Under clause 5(6), discretions to vary the benefit which are only capable of being used for reasons related to a member’s individual circumstances are disregarded when establishing whether or not there is a full pensions promise. This amendment allows the Secretary of State to specify in regulations other requirements that must be met in order for these discretions to be disregarded.
I shall explain what clause 5 does, and I hope that we will not need a separate stand part debate. The clause’s heading—“Meaning of ‘pensions promise’ etc”—is a central concept in the definitions. It will astonish the Committee if I say that, in this section, “promise” carries its ordinary meaning. In other words, when a scheme contains a promise, it contains a commitment to the member that they will receive a certain pension from the scheme.
Our existing legislation operates to ensure that scheme funding and capital adequacy requirements will apply, so that such commitments are appropriately backed. There is the possibility, of course, of a funding shortfall, such as where the employer becomes insolvent in a DB scheme. In those cases, members may not receive the full amount that was promised to them and would instead have compensation from the PPF. It does not make it any less of a promise, even though there can be circumstances when that promise did not end up being delivered. So the word “promise” has been chosen with great care in drafting the Bill to convey that a commitment has been made that goes beyond an intention or an expectation on the part of the person or scheme making it. Although legislative protections are in place to minimise the risk of the promise not being met, it does not in itself provide a complete cast-iron guarantee to the member. I hope that what I mean is clear.
Subsection (1) specifies that for the purpose of a DB scheme definition there is a “full pensions promise” provided to members if
“at all times before the benefit comes into payment,” there is a promise
“about the level of the benefit” that will be received. I stress that
“the level of the benefit is to be determined wholly by reference to that promise in all circumstances.”
Subsection (2) states that for the purpose of shared risk in DC scheme definitions a pensions promise is provided if there is a promise to members at some point during the accumulation phase about the level of benefit that will be received. The level is the rate of income or the amount of the lump sum.
To be clear about what we mean by “pensions promise”, subsection (3) specifies what a reference to a promise about the level of a retirement pension does and does not include. Subsection (3)(a)
“ includes a promise about factors, other than longevity, that will be used to calculate the level of” retirement benefit—for example, how long someone has been in the scheme or whether it is linked to a member’s salary—but the promise does not include longevity factors, because the effect would be to allow longevity risk sharing in a DB scheme, which is not our intention. To be clear, there are occupational pension schemes at the moment—big household name employers—where the pension age at which someone will draw the pension they are building up varies according to the latest data on life expectancy. That can change, obviously, as new data come forward. So we would regard that as a risk-sharing scheme, not a DB scheme, because of that element of risk sharing on longevity improvements between the employer and employee.
Subsection (4) makes provision for a promise if the scheme sets out the promise, or it is obtained from a third party. For example, an employer might pay an insurance company to underpin some of the scheme benefits. This enables a pension scheme to be defined on the basis of a pensions promise regardless of who pays for it, whether the accountability sits with the scheme itself, the employer, or a third party. The scheme categorisation therefore depends on what the scheme offers to members, not on whether individual members take that up.
Subsection (6) deals with what happens when there is discretion to vary the benefit. This is where amendment 41 comes in. Subsection (6) sets out certain discretions that may be present in schemes without affecting whether the scheme contains a full pensions promise. For example, many schemes allow ill health early retirement. That is a discretion, so it can mean that someone gets something different from what was expected at the start, but we do not regard that as undermining its being a DB scheme with a hard pensions promise. So the fact that the scheme has discretion to pay the money early would not be regarded as undermining its being a DB scheme. An example would be to make provision for ill health early retirement. Because those discretions are exercised only on an individual basis, they are different from discretions that are applied at the scheme level. We have therefore set out that the discretion for individual circumstances does not constitute a failure of the test in subsection (1)(a). In other words, it could still be a DB scheme.
I will say a word on Government amendment 41, and I hope that the Committee can see why I had to explain clause 5 before turning to it. The amendment will add a regulation-making power under which discretions that are capable of being used only for reasons relating to a member’s individual circumstance would have to meet any other requirements that might be specified in regulations. We have tabled the amendment as a result of conversations we have had with the industry on the definitions and how various scheme designs would be categorised. We will use the regulation-making power to prevent abuse of the discretion exclusion. It will allow us to prescribe in regulations that the exclusion of such discretions may apply only if certain other requirements are met. I hope that that clarifies the purpose of Government amendment 41 and clause 5.
This is the nub of the question about how defined-ambition or shared-risk schemes work. I am trying to work out exactly how a scheme member knows the difference between a promise, a target, an intention, an aspiration, a hope and perhaps, in some cases, a dream. I might think that I have been given a promise that I will receive two thirds of my final salary, when I have actually been given a target of two thirds of my final salary. If that turns out not to be affordable, I might only receive 60%, and the employer might take some of the extra risk.
There may be many scenarios in which an employee has been given a promise through a defined benefit scheme, but if the promise looks to be wholly unaffordable for the employer, the scheme rules might be changed following consultation with members, who all agree, for example, to retire a year later or to receive a lower percentage of their final salary. At that point, what would happen? Would the employee still be in a defined-benefits scheme but with a new promise, or would they somehow have moved into a shared-risk scheme? It strikes me that the former would be true, because they would still have a promise; it would just not be the promise that they had had before. If, however, a scheme has to keep redoing that process and saying, “Sorry, that thing that we did last year is still unaffordable, so we need to do it again,” could it accidentally become a shared-risk scheme because the promise keeps changing? Or would an employee in such a situation always be in a defined-benefits scheme?
The other area of confusion is where defined contribution becomes shared risk. Last week, I received an annual pension statement that told me that if I kept going at the same rate, I would have £20,000 in my pot and my annual income would be, if I remember correctly, £536. That implies that I have some indication of what my annual income will be from that pot, but that is clearly not a promise or a target; it is simply a guide to what that £20,000 would provide. The problem is that if I did not know that, I might think that it was some kind of promise or target, and I might think that I had something.
If I were in a shared-risk scheme, what more would I have? I assume that I would have a set of complicated rules that would tell me that that £536 was not quite a promise but was more than a hope and a bit more than an expectation; it was what the scheme would try to deliver to me, and as long as the scheme could roughly afford it, I ought to get that. Perhaps the range on which I would not get that would be slightly protected, so if the scheme could only afford £500, I might still get £536. If the scheme could only afford £250, however, there might be a wholesale reconsideration.
It strikes me that there are real problems about how the system will work in practice. How will people know, in clear language, exactly what kind of scheme they are in? From the language used in the annual statement, it is quite hard to work out what is a promise and what is not, or what is a target and what is an indication. That must be clear to make the whole thing useful. I think that the regulator will have to put a lot of effort into working out exactly what flexibility shared-risk schemes are allowed when they give a promise that is not a promise, and exactly what ranges they will be allowed to operate under. Where would they be expected to meet a promise that is not a promise? Where would the pension scheme trustees or governance committee or whatever else be expected to say, “This promise we are making is not realistic. Even if it was a range of promise; even if we say you will get £450 to £550, we know that we cannot meet that now. Should we move the range down at this point?”
We really need to think through how much flexibility we will allow shared-risk schemes in working out what their promise is, how they communicate it, when they change and what assumptions they make. Otherwise I suspect we will end up in a horrible mess. We will not have any kind of promise in a shared-risk scheme. There will be a hope or a dream that will just fly around every year with changes and annuity rates or changes in stock market years or whatever else. It will not be quite as useful as we all hope this will be.
I think what we want is, “Look, Mr Mills, you will get two thirds of your final salary. We are working to get to that. The employer will fund the scheme on that basis. In reality it is only if we really, really cannot afford to get anywhere near that two thirds of your final salary that we will have to change that. Here are the situations where we would do that and this is how we would go about doing it.” I think that is what we are hoping for, but I am not totally sure that we quite achieve the clarity we need in these definitions.
I am grateful to my hon. Friend, who raises understandable concerns about promises that are not promises and whether they are really worth the paper they are written on. I hope that I can offer him some reassurance, but I want to take him back to my three columns and my two rows. Shared risk, which was a phrase he used a lot, is defined ambition. It is the one in the middle where there is a promise on an element of the provision. The example I have given is a big supermarket which has done a cash balance scheme and has promised that for each year an employee works they will get some percentage of their salary as a cash amount when they retire. The supermarket cannot change its mind about that. It is backed by law. It has to fund for it; it is a promise.
It is shared risk because nobody knows what pension someone will get for that pot of money; nobody knows how long they will live, what inflation will be in retirement and what annuity rates will be. There is some risk for the individual, but in general terms there is a bit that is promised and then there is the bit that is subject to uncertainty as faced by the individual. The shared risk space—the column in the middle—is exactly as we would expect it to be. The scheme is clear what it is promising and what it is not promising. My hon. Friend rightly said that good governance and good communication are vital to all of that. Scheme members need to know what it is they are being promised.
My strong suspicion is that someone in that sort of pension scheme—I described a cash balance, but that is just an example—will know what they have got. They will have a guaranteed pot that they have been promised. They will get a statement each year and what has already been promised cannot be taken away from them. It is there. I stress again that there are plenty of legal protections for pension promises which have already been made. The scheme can change things for the future but it cannot just say, “We promised you that last year but we have changed our mind.”
I am not familiar with the scheme the Minister talks about, but I assume that for that promise to be meaningful annual contributions must go up by inflation. To be deliverable that promise must rely on investment returns keeping pace with what was expected so that when someone reaches retirement age their pot will be worth x. If investment returns fell through the floor the scheme might have to say, “Well, actually we cannot get near the cash target that we thought you would get” and so would have some flexibility in changing.
No. The risk of that bit of the pension promise is on the employer. They have made a promise. They have promised a pot of money. They cannot come back a few years later and say, “Well, the stock market has not done very well so we have changed our mind.” It is a pension promise. Under the Bill, that sort of thing is a promise which has to be honoured. They can say, “Guys, we cannot afford this scheme for the future and therefore we will make a less generous promise or no promise at all for the future.” However, hard pension promises have to be kept. I think the sort of scheme my hon. Friend was more worried about was a collective defined-contributions scheme, which is my third column and the collective bit.
I think that I have dealt with that point. We use the word “promise” in the Bill rather than “guarantee”, because if it was guaranteed, it would have to happen and, as the hon. Gentleman says, the one circumstance in which it does not happen is when a company goes into insolvency, because there is then no one to fulfil that promise. That is the only distinction. Funds go in, a promise is made and, for as long as the company exists, it is expected to honour its promise. It seems reasonable that promises can be kept only by things that exist, as it were.
However, CDCs are a different model. First, there are no promises. It is not that there is a promise that is not a promise; there ain’t no promise in the first place. There is, however, a target that the scheme will have a probability of reaching. I will come back to probabilities and ranges later on, but a scheme, for example, might fund with a 97.5% probability of meeting its target. Therefore, in one in 40 years it would not work, but it would do in 39 of 40 years. In general, the target will be met based on the funding levels going in but, in extremis, it might go outside that probability range or funding target. In such circumstances, a process will be in place in which either the firm or individual puts more money in, or the target or the pensions in payment are reduced. There will be a lot of flex when things go wrong, but no promise to begin with.
My hon. Friend the Member for Amber Valley is right that we must make clear to members of the scheme what is being promised, what is just a target and what happens when a target is not met. For example, if a target will not be met, will it be the workers or the retired members who bear that, or both, and what is the process for doing all of that? Arguably, what went wrong with the Dutch scheme was that while the employers knew—or their lawyers did—that a promise was not a hard promise but a target, the scheme members thought that they had a promise. That is why there has been conflict in the Netherlands: the members never had a promise, but that was not properly communicated to them. As we are starting this from scratch, we need to establish on day one what is a promise and what is not.
To reiterate, on the extreme example of DB that my hon. Friend gave, a DB pension promise cannot be ripped up. If the scheme has promised a DB, whether that is two thirds of final salary or whatever, that must be met; it cannot decide to pay, for example, 60% for past service. The scheme has to be funded and the regulator will require a recovery plan. Therefore, a promise really is a promise, but some schemes in the Bill will not offer promises and the members need to be told about that.
There will be targets, which are better than nothing. My hon. Friend’s DC pension will not have a target. It will have a guess, described as a forecast or an estimate, but next year he might get a piece of paper with a very different number on it. That is how DC works. With CDC, however, there will be a target, which people will be told about, and a specific effort will be made to aim for that target. I hope that that is a useful distinction.