Defining our pension ambitions
Today the DWP has published what those in the pensions world are calling the defined ambition paper, though its proper title is Reshaping workplace pensions for future generations.
This is a rich and welcome contribution to debate, although some of its suggestions on the future of DB pensions will be opposed by unions.
There is an important debate to be had about how we design pensions, particularly as auto-enrolment brings millions into pensions savings for the first time. But it is also highly technical, so what I want to do in what will probably become a rather extended post is to explain the issues. For insomniacs there is even more in the Touchstone paper I wrote with Craig Berry. The pensions experts that are always quoted in media reports all have particular backgrounds or represent commercial interests, but these issues are too important to be left to the closed pensions world.
Most pensions in the UK are of two types:
- Defined benefit (DB) pensions: where what you get as pension income is guaranteed by your employer and is based on how many years you contribute to the scheme, your pay and the rules of the particular scheme. These once dominated private sector retirement provision but have been in decline for many years.
- Defined contribution (DC) pensions: where you and your employer contribute to a savings account to build up what is usually referred to as a pension pot. When you retire, you turn this pot into a regular income, usually by buying an annuity. This is an insurance product that guarantees to pay a regular income in return for handing over your pension pot. DC now dominates the private sector, and a huge majority of those who have been auto-enrolled into pension schemes, or are about to be, will be in DC pensions.
The key to understanding how different types of pension work is to ask who bears the various risks:
- investment risk: Contributions made while you work will be invested. While you can predict returns, you cannot guarantee them. In a DB pension the employer absorbs this risk. If returns are less than expected then they will have to make up the difference (though will probably raise contributions as a result). In a DC scheme each individual bears the risk – so if investment returns are better than expected your pension pot will gain and if worse you lose. There is a further complication that annuity rates – how much regular income you can buy for a particular size of pot also depend on investment returns. At present they are particularly low as investment returns in the kind of assets insurers need to back annuities are poor. In general if you are prepared to take more risk you expect better returns but more volatile outcomes. But DC savers generally do not want to lose out from volatility so investments tend to get moved into safe but low return investments as they approach retirement.
- longevity risk: No-one knows how long they will live, but there are two types of longevity risk to think about in pension schemes – individual variability and changes in average longevity over time. If you know that a particular group of people will on average die when they are 80 it is reasonably easy to work out how much individual pension to pay each person. But if that average changes over time it becomes much harder. Longevity has increased faster than most have expected, though for the first time in many years it has just gone down. In a DB scheme the employer bears the longevity risk. In a DC scheme the annuity provider bears the longevity risk but only at the point of retirement. Growing longevity is one reason behind declining annuity rates and the growing cost of DB schemes.
- inflation risk: Again there are two types of inflation risk. Inflation can reduce the value of investments while people are saving, but the real inflation issue is whether pensions in payment keep up with the cost of living. This risk is shared between employer and individual in most DB schemes: payments are indexed up to a point bu if inflation is higher the individual loses out. In DC schemes people can choose how to handle the inflation risk by buying an annuity that either provides a lower initial income but which will be linked in some way to inflation in the future or a higher but fixed income that will not be linked to inflation.
There are many ways of arranging, sharing, managing and insuring these risks other than the ways that they are done in traditional DB and traditional DC schemes in the UK. And today’s DWP paper is a comprehensive exploration of possible alternatives.
Steve Webb, the pensions minister, deserves praise for encouraging this debate, and it is fair to say that as it has proceeded there has been a real journey with ideas that initially looked attractive fading away, while others have come up the outside.
But there is also a problem with it that flows from the two initial spurs to the debate:
- There has been a long campaign by pensions professionals to deregulate DB pensions. Their argument is that if you can reduce the risks and volatility that employers have to bear they will keep schemes open. They also argue that there are employers who would be prepared to accept some risk in new schemes if the law made this easier – though less than that in a full DB scheme. This part of the debate is often know as DB minus.
- There has been a separate debate about the future of DC schemes. Many think that the risk that investments will shrink (as they can do sharply during a stock market crash) will deter new savers, and have been interested in whether savers can be guaranteed they get their savings back. Steve Webb has certainly been known to speak favourably of this approach. There are also other ideas about reforming and collectivising DC and these policies are often called DC plus.
Putting these debates into the same process and consultation paper can be seen as either a comprehensive assessment of pension models or an unhelpful confusion of two different issues. The choice is yours.
In Third Time Lucky we say that we are extremely sceptical that DB minus will achieve anything other than poorer pension outcomes for savers. Unfortunately most employers have already closed DB schemes to new entrants or shut them to all contributions. Those that through their own commitment or because of strong union organisation want to keep DB already have plenty of ways of controlling cost and unions have normally been prepared to negotiate changes. In particular we oppose the abolition of the need to index pensions in payment. This simply means that pensions fall in value every year, as DC savers who have bought fixed rate annuities discover rather quickly.
We are also rather sceptical that there are many employers who want to take on risk and we worry that the hybrid schemes that result are often hard to understand. That is not to say that they are a bad idea or should not be explored, but it may be better for employers who want to do the right thing or unions wanting to negotiate a better deal to simply secure a higher contribution from the employer.
This is a much more interesting debate, although it is worth stressing that not all pure DC schemes are the same. There has been real innovation in DC with smarter investment strategies that manage risk and volatility well, more care about annuitisation, better governance and much lower charges. NEST is a good example (disclaimer: I am a Trustee Member), but not the only one.
Some of the enthusiasm for guarantees seems to have faded since this process begun. The problem with guarantees is that if you insure them that carries a cost. That will increase charges and thus lower pension income. Promising people their cash back over their working life (with no allowance for inflation) might be cheap to insure, but that is because it is incredibly unlikely. Guaranteeing that there would never be a year on year fall however would require an investment strategy guaranteed to produce lousy returns – often less than inflation.
The paper has some interesting technical suggestions for better managing risk and annuity volatility which are too complex to explain here but are certainly worth exploring.
But there is another way into the DC plus debate, and that is what is known as collective DC and is practiced in the Netherlands and Denmark.
The traditional DB DC risk spectrum has the employer bearing nearly all the risk at the DB end with the individual saver bearing nearly all the risk at the DC end.
Collective DC does not expect the employer to take on risk (unrealistic for the vast majority of the workforce), but instead shares risks among members. There is no need for members to move all their investments into less risky but poor performing investments as they approach retirement. More radically, schemes can pay pensions direct rather than requiring members to buy annuities. Such pensions may not be guaranteed to increase in line with inflation every year, but as an important new paper from the RSA, also published this week, shows offer the prospect of pensions 33 per cent better than traditional DC.
While collective DC offers the best approach it is also the most challenging. There are difficult questions about how to share risk across different age groups. Collective DC also requires large scale pensions with governance that is both expert and also run solely in the interests of members. Risk sharing within commercial companies in with-profits products went badly wrong as this was not the case.
But large scale and good governance are in any case the keys to making DC work. Public policy does not need to demand collective DC now, but acting in the interests of savers by shaking up existing DC is both a good thing in itself and a precondition for the extra prize of collective DC.
But while we want every pound saved in a pension to work as hard as possible, you still need to save enough in the first place.
Improving pensions structures is well worth doing, and has taken a major step forwards this week in the DC section of the DWP paper and the RSA paper.
But it cannot be a substitute for getting higher contributions into pensions in the first place.