This amendment and amendments 46, 51, 52 and 53 replace references to a required probability level with references to a specified range. Here, clause 20(2)(c) is amended so that regulations may require initial targets in relation to collective benefits to be set at a level which ensures that the probability of meeting the targets falls within a specified range.
As has been our custom throughout, Mrs Riordan, I hope that you will allow me to explain the purpose of clause 20 and set out why we have chosen to amend it since the publication of the Bill. When my teenage children get a new book of fiction, they turn immediately to the final page to see whodunnit. So as not to keep the Committee unnecessarily in suspense, I will explain that the Bill specifies that we have to set a hard, single-figure probability for CDC, but the amendments will allow us to set a probability range. That is the dénouement of our exciting journey, but I hope that Members will stay with me throughout that journey.
In my explanation of clause 20, I will refer to clauses 26 and 28, because the other amendments in the group are consequential amendments to those clauses and have been grouped for rational reasons. Clause 20 talks about the targets and the probability of achieving them. Clause 26 talks about valuation reports to manage that process. Clause 28 is about what schemes must do when they underachieve or overachieve against those targets. The amendments will substitute probability ranges for probabilities in each clause.
Throughout the process, we have been keen to continue talking to potential providers, employers and members of schemes. Feedback on the Bill has indicated that to have a single probability—for example, to specify that there must be a 97.5% chance of funding a scheme up to the requisite level—is too rigid. These numbers are illustrative and picked at random, but if we said instead in regulations that a scheme could have a 90% to 95% chance, the managers of a scheme would have to take action only if the probability of reaching its targets fell below 90% or rose above 95%. Within that range, the scheme would have a margin of flexibility.
If schemes have annual valuations, we do not want, to use a technical term, to keep fiddling. If we go back to my 97.5% example, we would not want the managers of a scheme to discover in a valuation that the scheme was at 97.4% or 97.6% and to decide as a result to scrap indexation or to double pensions in payment. We do not want to keep mucking about with schemes. Having as the target a range of probabilities will allow a bit of ebb and flow between annual valuations. Scheme managers will monitor what is going on but will not be required to make a series of short-term changes.
It will become apparent that two things are going on here: a probability and a funding ratio. For example, a CDC scheme might have to have a 97.5% chance of reaching its target, but its funding level could be, say, 110% of the target benefit. We are talking about two different numbers: the probability of reaching the target, and the funding level. For the avoidance of doubt, the New Brunswick hospital scheme that I mentioned earlier has to meet both the probability level and the funding ratio in managing its plans.
Clause 20 provides that regulations may be made
“to require the trustees or managers of a pension scheme to set targets in relation to any collective benefits that may be provided by the scheme.”
The regulations may
“(a) impose requirements about the way that targets are expressed;
(b) impose requirements about the recording and publication of targets;
(c) require the trustees or managers to set initial targets at a level which ensures that the probability of meeting the targets is equal to or higher than a level of probability specified in the regulations”.
The regulations may specify that that must be certified by an actuary. As I said a moment ago, that target is unenforceable. That does not mean that it is unimportant or that there is no set of rules, regulations and expectations, but it is ultimately a target. No one can say at the end, “You told me this was what I would get. You haven’t delivered. I’m going to take you to court for not delivering the targets.” As long as the processes set out in the Bill have been followed, that is what the scheme has to do and a target cannot be enforced.
Targets are key for good governance. They ensure a meaningful relationship between the assets available for the provision of collective benefits and the target level of benefits that the scheme seeks to achieve. It is important to have a clear governance regime in place for collective benefits because there is not such a direct link between the contributions a member pays in and the benefits they may receive from the scheme as there is for money purchase benefits. In an ordinary individual money purchase DC pension, it is pretty obvious what has been put in. That money has been invested and charges have come out; it is that person’s pot. It is ring-fenced; it is theirs. In a collective scheme, however, a share of the assets is simply accrued according to the scheme rules. We definitely need good governance to make it absolutely clear what people can expect.
Targets are also essential for transparency for members. In relation to any collective benefits provided by a scheme, targets are necessary to ensure that members have sufficient clarity about what they might get. This is a statement of the blindingly obvious, but people want to know what they will get. It is not a promise or a certainty but a target. A target must apply to any collective benefits provided by the scheme. All the assets that relate to the provision of collective benefits must be allocated to member benefits transparently. We want members to be clear about the benefits being targeted for them by the scheme in all circumstances.
Subsection (2)(a) gives us the power to ensure that targets are expressed clearly and can be compared meaningfully. We state that we can
“impose requirements about the way that targets are expressed”.
That power may be used to ensure that schemes provide clear information about targets—for example, the benefits being targeted and the probability level. It also allows us to set appropriate regulations for different types of target. We could use the powers in subsection 2(a) and (b) to ensure a meaningful comparison between the collective benefits offered by different schemes where that was offered as a commercial product. Under clause 34, we can require that information to be published. That is an important point. This is complicated stuff if we are not careful; being able to compare one CDC scheme with another would be helpful. We can try to ensure that that information is provided.
We will only require schemes to meet a probability in relation to the target, whereas the New Brunswick scheme has both funding and probability requirements. We are taking a different approach, and that is important. People sometimes say, “The Netherlands is different. Canada is different. Denmark is different. Will it work here?” We are not cutting and pasting those models into our system; we are coming up with our own version. We intend to consult on the detailed figures in relation to the probability requirement. The probability required by regulations will have to set at a level that takes into account a number of different matters.
Giving members some confidence about what they can expect to receive in retirement may encourage people to engage with pension saving, enable them to plan for retirement and promote confidence in the scheme. In setting the probability standard, we must balance the desire to provide confidence to members about what they might get with the desire to allow schemes to adopt a range of different investment strategies that seek to deliver a good return for members on their pension contributions. I can tell the Committee that I intend us to consult fully on the regulations made under the clause.
Targets are relevant to clauses 26 and 28 because, once the schemes are up and running, there will be a governance process to ensure that they remain able to pay out the targeted benefits to the required probability. That is set out at various points in part 3. Under clause 26, we can require the trustees or managers of schemes to obtain regular valuation reports from an actuary to assess the probability of being able to pay out the targeted benefits. That will provide clarity about how robust the targeted benefits are to future changes, and I will discuss that clause in more detail later.
Where a valuation report shows that a scheme is over or under the required probability—in other words, a surplus or deficit—regulations may require the trustees or managers to follow that policy to address the deviation from the required probability range. Similarly, clause 28 may require the trustees or managers of a pension scheme to set out a policy in advance about how they will react to a valuation report that shows the scheme is over or under the required probability. Again, that policy may have to deal with a surplus or deficit in the probability in one or more of a range of ways set out in regulations. Again, I will come back to that later.
Clauses 20, 26 and 28 as a package provide a mechanism for ensuring that trustees and managers will monitor the probability of meeting the targets and take appropriate action if necessary. In combination, the clauses will ensure that there is a tenable link between contributions paid into the scheme, the investments held by it and the target benefit to be provided to members.
Finally, to clarify the amendments, in Government amendment 43, we have made the judgment that a probability range would be more effective than a probability level. Setting a single probability level would have been effective in providing clarity for members, but if it were used to trigger action, it would have the effect of requiring trustees or managers to implement their deficit or surplus policies at almost every valuation. We want to require schemes that offer collective benefits to take prompt action to restore a deficit or surplus, but we also want them to have the flexibility to take longer-term decisions, so that they do not have to react every year to every deviation from a single probability level. The main amendment is Government amendment 43, and the consequential changes from a probability level to a probability range are covered in amendments 46, 51, 52 and 53.
The hon. Gentleman makes a good point. We will consult on that when introducing regulations. As he said, if there is a huge probability range, practically any outcome will be within the probability range and although the probability range will always be hit, people would not often be paid the pension they were expecting. The whole point is to reduce the volatility of people’s pension outcomes, so it would be at odds with our goal to have an absurdly large probability range.
Will the Minister be more specific about the impact? I understand the reasons for flexibility, but does that not require the scheme to carry more surplus in case it underperforms? He said the pension will be valued regularly, but what does that mean and how often will it have to be valued?
We envisage an annual valuation. Obviously, people who run pension schemes are monitoring things every day, but we envisage a formal annual valuation. We are trying to strike a balance. We all know that there are year-to-year fluctuations and we do not want to require schemes to change their fundamental features just because of short-term fluctuations. That is one reason for having a probability range. The hon. Gentleman refers to schemes being required to carry a surplus. They will be required to act in a way that satisfies a set probability of achieving the target benefits.
Some Canadian examples have quite high probability numbers at 97.5%, but in one example, more than 40% of the assets are in equities and 20% are in property and other assets of that sort. It has not followed in practice that having quite a tough probability test has required schemes to be very cautious. The two can sit together. We are trying to give people targets, annual statements and so on that mean something to them and have some credibility, without being too rigid and expecting schemes to adjust too much to short-term fluctuations.
I apologise for going on at some length, but we are getting into some fairly novel concepts in British pension schemes, and I hope that it has been helpful to set out what we mean by a target and why we want to amend the clause.
I have several observations about these amendments and what the Minister described as novel concepts. It strikes me from reading part 3 that we are putting a lot of authority and responsibility in the hands of actuaries. That comes across clearly in the clauses. Actuarial science is a noble profession in this country and others, but it is worth reflecting on the responsibility being placed in actuaries’ hands. Actuaries have a lot of responsibility anyway, and certainly in defined-benefit pension schemes. It is inevitable that actuaries will play a significant role in any form of pension scheme other than defined contribution, but how would the Minister describe the role that actuaries will play?
This relates to a previous theme of our discussions, which is that this form of pension scheme requires savers to place significant trust in experts. It strikes me that it is a little like a doctor-patient relationship. We know that patients put a lot of trust in doctors to get them the best outcome for their health problems. Such are the complexities of medicine that someone with expertise and training must make such fundamental decisions. The actuary’s role in these schemes feels a little like that, and it brings the wider issue of governance into focus. We will undoubtedly discuss that more when we come to other amendments, but my initial observation is that this form of pension scheme saving puts a lot of emphasis on actuaries making good decisions. As I said, actuarial science is a noble profession, but actuaries have not always got things right in the past, which casts doubt on the scientific aspect in some respects.
The Minister described this as complicated stuff, and he is right. We must consider how much complexity a pension scheme can bear when communicating with its members. Complexity is an issue with pensions generally, and communication with scheme members is always a challenge. That brings into view how to explain to scheme members how collective defined-contribution schemes might work. He had a stab at explaining it, manfully—and probably, to experts outside this room, successfully—but the Government’s approach is to put into legislation a necessity for the probability that a target will be met, not the probability that the target will be met plus a funding ratio requirement, as in New Brunswick, for example.
It would be useful if the Minister elaborated on that in terms that might be understood more widely. It is difficult; I am asking him to do something that I am not prepared to try myself—that is the luxury of being in opposition—but I think that if we are to communicate the advantages of that type of pension scheme outside this place and to people other than actuaries, pension consultants and the like, he must try.
The Minister’s explanation of these amendments brought again into purview the issue of placing lots of decisions into secondary legislation. There might well be good reasons for that, but it is a feature of recent pensions Bills and perhaps less recent ones, too. Will he explain the rationale for that approach? On that note, I will be delighted to hear what other Committee members have to say and specifically the Minister’s response to those points.
I was looking forward to hearing what everyone else on the Committee thought, but let me give it a quick go. Am I spending too much time with actuaries, as I have been accused of doing? It is important to understand what the role of actuaries is and is not. The trustees or scheme managers—I will say “trustees” as shorthand—act on behalf of scheme members. They are expected to ensure that the scheme follows its rules. They obviously must take action if the data presented to them show that the scheme is not on course to hit its probability. That is what the trustees do. They take expert actuarial advice. The actuary might also say, “You are going to be outside your probability range, but there is a range of things you could do.” They could say that increasing employee contributions would have a certain effect on probability, or changing the investment mix to reduce risk would have another. The trustees will make the decisions, having been informed by expert advice on the different options and how they would feed through into the probability. That seems a sensible division of labour. The experts will crunch the numbers and tell the trustees how things stand and the implications of different choices, but the trustees will make the decision.
I hope that it is a Freudian slip that the Minister has repeatedly mentioned trustees but not managers. That brings into view how important good governance will be. Is it just about trustees? I assume from what the Minister has said previously and from the Bill that not just trustees but managers will interact with actuaries.
I think that the record will show that I said trustees or managers, but, for brevity, let us say trustees. Clearly, managers could manage CDC schemes, as we discussed earlier. A Government amendment sets out more precisely the duties of managers with respect to the interests of members. I am simply trying to avoid saying trustees or managers all the time; there is no great Freudian or other significance in that.
To clarify the point about actuaries, they have a key role to play because they are independent. They are not members of the scheme or employers; they are giving independent expert advice. We have powers to require actuaries to meet specified requirements and hold specified qualifications under clause 27. I accept that actuaries, like politicians, do not always get things right, but they do have a crucial role to play.
Yes, the buck stops with the trustees or managers, but they will depend on expert and independent advice. The actuary assessing the scheme’s funding position, probabilities and options for change must not have a vested interest. It is right to have independent expert advice.
There is the wider issue of trusting in experts for our pensions, in the way that we might see our doctors. Most people would think that trusting a doctor to give expert advice on health or lifestyle that the patient chooses whether to follow is not a bad model. Compared with what? Compared with an individual DC. Without the clause, most of the people we are talking about will probably end up with pensions that are individual DC. In that case, someone also has to trust experts such as investment companies and managers to invest the money wisely and get a good return. However, in that world nobody is trying to give people a target or any expectation; they just get what they get.
I suspect that the hon. Gentleman would agree with our argument that, although it is more complicated to have targets and probabilities, it is probably better for a scheme member, although there is a lot going on under the bonnet. That relates to his point about complexity. I jump in my car and do not have the faintest idea of what is going on under the bonnet, but I know that it gets me there.
There is a balance to be struck in member communication between giving people enough information to make informed choices but not the drains-up version, which will be a complete turn off and not mean anything to anybody. I do not think that is unduly paternalistic; it is realistic. The trustees—the people in charge of the scheme—need to know what is going on at a sophisticated level, but the scheme member does not need too much information. All the complexity needs to be available for those who are interested, but it does not need to be imposed on scheme members.
I certainly was not suggesting that people should reflect on the finer points of the investment strategy and press for change, or, at a fundamental level, on whether or not to be in the scheme. Under auto-enrolment, they will be placed in a scheme that they do not have to stay in, and if they are not satisfied with the governance or with the targets, they may take their money elsewhere. That is the type of thing that individuals will be choosing.
Collectively, for example, trade unions might want to scrutinise things on behalf of employees in the work force. If they feel that the CDC scheme to which the work force are being automatically enrolled is not as good as another CDC scheme, they could make representations to the employer to choose a different scheme. So it is horses for courses. All the information has to be available, but we do not impose complicated stuff on people who are not well placed to deal with it. That is the balance we are trying to strike.
The hon. Gentleman probed further on targeting and having a probability range as against a funding target. In our judgment, a funding figure is backward looking. The year comes to an end and the clever people do their clever sums. They come up with a number and tell someone that they were 110% funded last year. The probability is forward looking: as we go into any year, based on the best assumptions about what will happen in future, what is the probability of delivering what was targeted? In our judgment, the most intuitive and best way—not the only way—is for the people in charge of the scheme to go into a year planning to achieve their targets with a set of probability, or in a set probability range, rather than potentially reacting with considerable lag to events in terms of the funding outcomes. I do not think that schemes that do that differently are doing it wrong. I just think that it is probably clearer to scheme members to have a forward-looking model, rather than a backward-looking model. That is the judgement we have come to.
Finally, the hon. Gentleman asked about the balance between primary and secondary legislation, which is a familiar theme. In dealing with new forms of pensions—new models and new providers—we can speculate on what they might look like. We talk to people, and we amend the Bill when they tell us that we have not got it quite right, so the dialogue is ongoing. However, on trying to hard-wire particular features of the model—such as limits on probability ranges—in primary legislation, I do not think we are in that place yet. We do not anticipate that any of this will go live until post-April 2016. As we have got time, we want to set out the overall framework and go on talking to potential providers, employers and members about what they want, and then we can frame the regulations in the light of that, rather than second-guessing them at this stage. That is why we have chosen this balance. I hope that that is helpful. I commend the amendment to the Committee.