I beg to move amendment 20, in schedule 16, page 85, line 17, leave out paragraph 1(4)(e).
Schedule 16 sets out some of the detail of the provision. The justification for the deletion is that, on our reading, the schedule is intended to give the Secretary of State the ability to exempt all existing pots from the automatic transfer mechanism. The Minister painted what is thought to be an alarming picture of enormous amounts of money—£0.75 trillion, I think—being brought into the stranded pot system, but that shows how big the problems are with the current pension system. It tells us, first, how many stranded pots there are, and, secondly, how little engagement there is with the current pension system—the Minister made great play with the word “engagement”. Surely an aggregate stranded pension pot of such size illuminates the significant problems with the current pensions market.
The amendment would remove the ability from the Secretary of State to exempt all existing pots from the automatic transfer system. If the Minister intends to proceed on that basis, it explains why the DWP thinks that by 2050 or 2060, 87% of people will still have multiple pots—a situation we would all deplore. This is a significant point of disagreement between the Minister and me. If his £0.75 trillion figure is correct, it is still not clear to me why that amount of money is better off stranded than being brought back into the system. On that basis, I commend the amendment to the Committee.
This comes back to the issue of walking before one can run. We envisage a massive potential problem of millions—and eventually tens of millions—of small pension pots. We want to get to a situation in which people’s pension savings are consolidated in many fewer places. But getting from here to there, given the scale of the problem described by the hon. Gentleman—£0.75 trillion of dormant pots by 2050 even under our plans, or without auto-transfers anyway—is a huge challenge. He should bear in mind that dormant pots can be DB as well as DC, so there are issues around DB automatic transfers, but there is no difference between us on fragmentation being a problem.
All we propose to do is initially set a date beyond which automatic transfers will apply. One option would be to say that all automatic enrolment pots would be auto-transferred, while another would be to say that all pots that become dormant after Royal Assent would be auto-transferred—in fact, the two options would cover many of the same pots.
Over time, once the infrastructure is in place and we are confident that we have dealt with things such as pension liberation fraud, transfers and identity—once the infrastructure is in place to deal with those sorts of things quickly, cleanly and cheaply—there might be huge potential for expanding the scope of all that. We might do so through raising the pot size limit or by bringing the date back in time; indeed there might in effect not be a date. The basic idea is that we start with something manageable. Even with our pot size limit and date proposed, we could be talking about 1 million dormant pots a year if we are not careful.
All we are trying to do through paragraph (1)(4)(e) of the schedule is to allow us to implement the policy gradually. The amendment would stop us from doing that; it is the “big bang” amendment. Having, for 13 years, not thought that there was a problem with small pots or that it was necessary to legislate for dormant pots at all, the hon. Gentleman now wants to amend the Bill so that every dormant pot in the entire British pension universe is within the scope of the policy. I hope that he will, on reflection, think that the amendment is absurd.
Clearly, we need to get the scheme up and running and to do so at a manageable scale. Just to provide the Committee with one example of some of the problems with the process, pension scheme record keeping is not universally brilliant. A year or two ago, I responded to an Adjournment debate about a pension scheme for Grimsby’s fishermen. I seem to recall that one of the problems with some of the pension records was that they said, “Captain Jones”, and that was about as much as it had. There was no national insurance number or date of birth; it just said, “Captain Jones’s pension”. I hope that we have come a long way since then.
If we use the date powers that the schedule will provide us to say, for example, that new auto-enrolment schemes are within scope, I think we can then have more confidence—although not total confidence—about data quality. If we simply said, as the amendment suggests, that every scheme ever would be within scope, there would be carnage, because many pension schemes out there have lousy record keeping.
The Pensions Regulator is working on the issue and trying to drive up standards, but to try to graft new legal duties on a system where the data are, shall we say, of variable quality would be a nightmare. Even in the context of auto-enrolment, someone told me the other day that a pension scheme had had trouble because certain provisions with auto-enrolment referred to the state pension age of the scheme member—one has to know whether a member is a man or a woman. Apparently that pension scheme did not know whether its members were men or women and was going down the first names to find out whether it was a Julie or a Jack and then which pension age applied. It is incredible to think that that is a modern pension scheme.
We have an awful journey to go on to ensure that pension schemes have high-quality data. The idea that we should not, in some way, restrict the universal pension schemes in the first instance to which automatic transfers apply, or should apply a legal duty to practically every scheme, all with variable standards of record keeping, and expect schemes to be able to ask other schemes for data that they might not have or to send pension schemes to the right place is, to put it gently, running before we can walk.
The ability to constrain what we are doing to the newest schemes—those set up under auto-enrolment and those that we are clearest about—seems to be entirely measured. Oppositions always want us to go fast, but we believe in getting the policy right, which is why we need the powers in the Bill.
May I pick up on a couple of points? The first is about our earlier discussion—the Minister mentioned it this afternoon—regarding the lack of a level playing field between new entrants and current players in the marketplace, if we have a situation where only pots stranded after Royal Assent are included in the automatic transfer mechanism. We need to think about that a little.
As things stand, the book of an insurance company that is a current player in the market equals existing contracts plus future contracts. Existing contracts will be exempt from the new system, while future contracts over £10,000 will be exempt. We are talking about the biggest players in the UK pensions market. If the Minister has his way, the bulk of existing providers’ business will be exempt from the new system.
New providers such as B&CE, NOW and NEST have few existing contracts—they have only just started—and most future ones will not have accumulated £10,000 by the time Royal Assent is granted. The new entrants into the marketplace, who, as the Minister would agree, are driving down charges and increasing quality, will therefore face a situation where the bulk of their business rightly has to meet the challenges of the automatic transfer mechanism. However, the bulk of the existing providers—the big players in the marketplace—will be exempted.
That is a significant competition issue and we need to probe that a little further. At the moment of automatic transfer in the pot follows member mechanism that the Minister proposes, the company with the current contract with the saver can try to persuade that customer to opt out of the automatic transfer and stay with the existing provider. In that case, the competition would be between the company with the current contract and the company to which the contract would automatically transfer under pot follows member.
On that basis, looking at it from a fair perspective of working out how matters will proceed, that will affect virtually all business of the new providers and little of the business of the insurance companies; they will have much more margin to play with and there will be a big cost difference.
More widely, if there are excess profits on the contracts written before pot follows member—those that will not be brought into that system that might exist at the moment, rather than after Royal Assent—there is a danger of cross-subsidy. Admittedly, that would be within a company, but it would still be cross-subsidy because if the Minister takes the view, rightly, that a transfer mechanism, whatever it be, is a way of driving forward quality and delivering more value for the saver, that probably involves reducing excess profits where those exist. However, the current providers will be able to retain those profits in that situation and use them to fund a retention offer for the saver.
The lack of a level playing field under the Minister’s preferred system is a big issue. He mentioned that he is aware of that, but that is significant for anybody who cares about competition and I am sure that he will consider that further. He suggested the possibility of a system that does not require transfers every year to deal with that problem, but then we would be again into the situation where things might need to take place to make his chosen mechanism work properly that are not laid out before us. We simply do not know. He indicated that he is interested in that, but, for us as a Bill Committee, there is nothing for us to look at. We do know, however, that under his system there will not be a level playing field between new entrants to the market and existing players.
We all believe that new entrants can be really effective and I think that the Minister would agree that bringing down charges in the early stages of auto-enrolment is important. NEST, B&CE and NOW have played a big role; they came into the marketplace and said, “We will charge this”, which was significantly lower than what had been the case and the other providers responded to that on price and hopefully on quality, too. That is a significant issue and something that further supports the Opposition’s view that giving the Secretary of State the ability to exempt all existing pots from the automatic transfer mechanism is not something that we can support.
There is obviously a lot of detail in schedule 16. It goes on for many pages, but it might be helpful if I explain how “pot follows member” might work and set out a couple of models. We produced what we call inside the Department the policy certainty document—I slightly overstate it there—in April 2013. I want to give the Committee a bit more clarity about how this might work in practice. The basic idea is that a person works for firm A and is auto-enrolled by firm A into pension scheme A. He moves to firm B and leaves behind a pot below the pot-size limit. Firm B auto-enrols him into pension scheme B, which then establishes that he has a dormant pension pot in scheme A and, unless he objects, pulls the old pension pot into the new scheme.
We are essentially looking at two ways of doing that. One is what might be called a central database model. Under the powers in schedule 16, the scheme in which money has been left behind would have a duty to notify the central database that there was a dormant pot. So there would be a dormant pot alert. That is then stored on the central database. Then the new scheme that has auto-enrolled the employee checks the central database to see whether the new scheme member has any dormant pots. It finds one, contacts scheme A and pulls the money across. Clearly, that is an IT project. I keep telling my colleague in the Department that as Excel spreadsheets can have a million lines on them these days, surely all we need is a big Excel spread sheet. I am advised that we need more than that. However, the basic principle is obvious. The dormant pot is registered as what is called the ceding scheme and then the automatic transfer scheme pulls it across.
One can see the attractions of that kind of model. It could be a nucleus for a broader pensions database where perhaps information about all an employee’s pension schemes were held even if they did not end up physically consolidated. As was mentioned earlier, portability is important. This database could be used as a facilitator for pension transfers more generally. We certainly envisage that pension transfers have to be a lot cheaper and simpler than they are at the moment. That is absolutely clear. That would be one model. We envisage—we have powers in schedule 16 to do this—that ultimately the industry will pay for this. The Government will not in the long run pay for this.
Clearly, if there are set-up costs, there might be issues about whether the Government make it happen. We have indicated that there is a levy, for example, on pension schemes, so one option would be to use the levy to recoup the costs. However, the basic idea is that, as our impact assessment shows, the net present value of savings to the pensions industry of getting rid of millions upon millions of stranded pots runs to £6.4 billion. We think that asking the industry ultimately to pay the cost of the infrastructure to save them £6.4 billion is a pretty good bargain.
I thank the Minister for his elaboration on the cost of an automatic transfer mechanism. Would it be fair to say that the Government have not come to an agreement with the industry on the industry funding this? Is that a fair assessment of the position? Do we know precisely how much it will cost?
We certainly do not know precisely how much it would cost. It depends how it is done. For example, just to cite one of the things that we have to be careful about, we are increasingly aware of a growing phenomenon called pensions liberation fraud. It sounds rather too glamorous for something that is actually rather seedy and rips people off. There is a tendency for people with pensions to be approached and encouraged to transfer them into decidedly shady pension schemes. They then find that they have breached Inland Revenue rules by early access and have a hefty tax penalty, that the new schemes have very high charges, and they can end up losing all their money.
Clearly, we do not want to facilitate that sort of thing through automatic transfers, so we will need an infrastructure just to make sure that money goes from proper pension scheme to proper pension scheme. That kind of thing needs doing. I should add that it would need doing in the hon. Gentleman’s automatic aggregator model as well. One would have to be quite clear where the money was going and about the bona fides of the people the money was being sent to. That is one model, an IT-based central database and so on, which is from where I instinctively came to this.
There is another way of doing it, which would, in a sense, be more paper based. We refer to it as a pension information transfer document, or a pensions P45. The basic idea would be that, when someone leaves a firm, the reference number of the scheme that they have come from is on the P45 or similar. When they are auto-enrolled on a new scheme and it asks to see their pensions P45, it would have the reference number of the scheme they came from. The new scheme could then simply directly contact the old scheme and ask if there was a dormant pot. The old scheme would say yes and send the money across.
At one level, that is attractive, because it avoids the need for a big database and all the rest. The downside is that we all lose bits of paper. How many times have people started a new job and been asked for their P45 from their previous employer? Then it is: “What did I do with it? I have lost it”. We know there would be significant drop-out if we went down that route. So we might get 70% of dormant pots going across and the other 30% never supplying the information.
We are trying to strike a balance. There is a relatively cheap and simple paper-based way of doing it—it need not be paper based, but a direct way of doing it. We also want to think about the burden on employers. Would that put a burden on firms so that every time somebody left them, they had to put on the P45 the reference number of the scheme of which the individual was a member? If the firm had only one pension scheme, that would be blindingly obvious. However, many firms have multiple schemes, multiple sections and so on. We therefore want to be a bit careful about the burden on firms.
We are still talking to IT people, the pensions industry, consultants, employers and many others about how best to deliver this. That is why the Bill gives us flexibility. Schedule 16 sets out a set of general rules, such as the duty of a scheme someone is leaving to report dormant pots, and the duty of the scheme someone joins to obtain that pot. There are rules about the right to opt out, because this is an automated process, and the powers for us to raise money to pay for all that. There is obviously a lot more detail in schedule 16, but I hope that this gives the Committee a flavour of how we plan to use the powers in it, and I commend it.
I thank the Minister for that run through schedule 16. He is right. There is more detail in schedule 16, but it is fair to say that the Government’s desire automatically to transfer pots via “pot follows member” is a long way from fruition.
The Minister mentioned a couple of options: a centralised IT database or perhaps a paper-based system, but he pointed out the downside of bits of paper. He mentioned enthusiastically something Aviva had said about small pots and the pot follows member system. Is he aware of the proposal from, I believe, John Lawson, of Aviva and the ubiquitous Tom McPhail at Hargreaves Lansdowne? According to the pensions press and things that I have seen, they have come up with a proposal for perhaps making this work. I would be interested in the Minister’s view.
The Minister mentioned that, in the end, the industry will have to pay for this. That does not surprise one. He also mentioned a levy. For other members of the Committee who might not be aware of it, will he say more about how that might work and what levy he refers to?
I am grateful to the hon. Gentleman. He asked about the alternative proposition that John Lawson and Tom McPhail had put forward, which, to Labour party ears might be pronounced “OMOP”. That is, one member, one pot. Their basic proposition is that, when people change jobs, they leave the pot behind and the new employer has to put money into the pot that they started with.
Their interesting argument for that is that it avoids moving money around. Money would probably stay in the first scheme someone joined. Someone would be free to say to a new employer, “No, put money into my new pot,” but inertia being such a dynamic force, the chances are that large numbers of people would end up with their pension with the employer they started with. That potentially introduces an element of lottery; the pension scheme of where someone first worked will accumulate all that person’s pension. That is an interesting notion.
I have met John Lawson and Tom McPhail to talk about that, as I am always hungry for new ideas in a remorseless way. One of my worries is that an employer could end up having to send pension contributions for each employee to each of the pots they first started with. If the employer were a supermarket or similar with tens or even hundreds of thousands of workers, it would have to log and register each preferred pension pot. It would not, on the whole, be putting money into the supermarket pension scheme, with perhaps having workplace presentations about the merits of that scheme and advisers coming in. Instead there would be great fragmentation.
I can see the attraction of not moving the money around. However, on the other hand, employers would be less engaged in workplace pensions because they would not be putting money into their own pension scheme, but perhaps into that of a rival supermarket. While it is not an absurd idea and I have looked at and thought about it, it raises a raft of different issues. It also does not gel with automatic enrolment. If someone works for supermarket A and goes to supermarket B, does the latter automatically enrol that person into its scheme, and then by default all the money goes into the first scheme? It does not conceptually fit well with auto-enrolment. While it is an interesting idea, we have a number of reservations about it and do not think it deals with the issue.
The hon. Gentleman asked about the levy. There is a levy on pension schemes separate from the pension protection fund levy, which is a DB thing. All pension schemes pay a levy related to their size, to fund things such as the Pensions Regulator. If I remember rightly, the ombudsman is funded through the levy. It is not a huge sum of money, but it enables us to run the infrastructure of pensions regulation. That is the power that we could use there.