Uncorrected transcript from the Select Committee on Economic Affairs inquiry into The use of RPI - October 9

Witness: Dr Ben Broadbent, Deputy Governor of the Bank of England.

Examination of witness

Dr Ben Broadbent.

Q61            The Chairman: Dr Broadbent, welcome to the Economic Affairs Select Committee. We are hoping that you will help us to deal with this very knotty question. I begin by inviting you to tell us what you think. Both the statistical authorities and the Government have told the Committee that the RPI is an adequate measure of inflation. To what extent is CPI a good measure of general household inflation?

Dr Ben Broadbent: I think it is. In the phrasing of the question, are you suggesting that general household inflation is qualitatively different from a consumer price index? I do not know whether that is behind your question. I do not think it is. I think there is also a single best measure of consumer prices and, in that respect, the CPI is definitely superior to the RPI.

The Chairman: Could you elaborate on why you do not think it is? We have had evidence in which people have argued that you can have a rate of inflation that gives a kind of macroeconomic view of the economy, which is not necessarily the same as what Mrs Snooks discovers when she goes shopping and in her household bills.

Dr Ben Broadbent: I do not see that in some ways. It is true that the ONS is developing what are called household cost indices. Those include some outgoings that households might recognise as costs, in particular, say, interest payments. For me, that comes down to whether you call that consumption or income. There are very good reasons for this, which relate to something which is also a problem with the RPI. Say you are a landlord and you get rent, or income on your deposit account: that is income. In the national accounts, interest payments are deductions from income, not consumption. So if you accept that those things are not consumption, then you should not be measuring them in a consumer price index. It is not quite clear whether you should measure them in something called a household cost index. For some people they are deductions from income; for others who get interest on their deposits they are additions to income. It is not quite clear to me what is meant by “as households experience it”, and whether that is justification for constructing an index of something other than consumption.

The Chairman: I am not expressing a view, but, for example, John Pullinger, the National Statistician, when he appeared before the Committee in July, said: “Over the last 10 years … for inflation, we have been trying to capture two rather different things. There is an economic concept that we are trying to capture, which we have based on consumption. It is consistent with the measure used in other European countries and has been the basis of the inflation targeting that successive Chancellors have asked the Governor of the Bank of England to look at. It has consistency and coherence”. Also, the Royal Statistical Society said in written evidence: “Consumer price indices are used for many different purposes and no one index can meet the whole variety of needs. In particular, there are two different primary purposes which require different indices: macroeconomic needs, such as inflation targeting and international comparison, for which the Consumer Prices Index (CPI) and similar indices were designed; and understanding how inflation affects households, which was the RPI’s original aim”. The Bank appears to have a completely different view from the National Statistician and the Royal Statistical Society. 

Dr Ben Broadbent: No. It is partly a question of language. They are both macroeconomic. It is not that one is suitable for macroeconomic purposes and the other is not. I would draw a distinction between two of the differences. It is perfectly legitimate to ask oneself the question: “What is the experience of different groups of households?” Some may spend more on some things than others and you may want to work out the differing rates of change of those particular price indices.

The Chairman: Forgive me, but that is not the point that they were making.

Dr Ben Broadbent: I know, but it is partly. They said “as experienced by different households”, but I am not quite sure what is meant by “the experience of households”. Let me give you an example, if I may. The weight that is attached to mortgage interest payments is a gross one. It says: “How much do gross mortgage interest payments make out of income?” Nowhere in there is the interest of receipts on deposits but, conceptually, that is exactly the same thing, only with a different sign. So what does one mean by cost, as opposed to income? In the national accounts both those things—and I think this makes perfect sense—are defined as parts of income, one with a positive sign, one with a negative. It is a matter of semantics whether you call one of them costs and one income. They are more naturally both defined in the same way, because one is just the counterpart of the other. One is the other with a negative sign. It is not at all obvious what the difference is between a consumer price index and “the experience of households”. I do not think that calling it “the experience of households” drives you in one particular direction. It is not clear to me what is meant by it.

Lord Turnbull: A difficulty is that people who largely pay mortgage interest are a different age group from people who are paying the interest. Netted off, you may capture the experience of no one. But that, I suppose, is true of the whole thing: pensioners have different budgets from millennials and so on. In that sense, netting off is the correct thing to do, but in some ways it is more interesting to look at the different groups.

Dr Ben Broadbent: I can see that there may be an interest in something—I would not necessarily call it “cost of living”—which is a mixture or blend of income and consumption, but it is tricky to do consistently, as you say. You say that it applies to no one, but I would have thought that there were many households who have both a deposit and a mortgage, to varying degrees. It is true that none may be exactly the same weight as the country as a whole. All I am saying is that to call something a household experience does not easily lead to some particular construction of either bits of income or bits of costs. The virtue of the CPI is that you know what you are getting: it is the price of consumption. 

Q62            Lord Sharkey: I am going to talk about how we reflect owner-occupier costs in any index. Is it your view that CPIH, for example, reflects owner-occupier costs better than the RPI? In parentheses, would it be very difficult to fix the RPI, which has the merit of being a long-standing index, so that it more accurately represented owner-occupier costs?

Dr Ben Broadbent: In principle, yes. I think it is true that the CPIH captures them better. Some of these issues relate to what we were talking about a moment ago. It is not clear to me that the cost of a mortgage is the best way to measure the cost of housing. There are 27 million households in the UK; there are fewer than 10 million owner-occupier mortgages. Is the cost of a house to someone who happens already to have paid off his or her mortgage really zero? I do not think it is because, in some kind of opportunity-cost sense, they could move out of the house and let it. What they are really doing, in an economic sense, is acting as both landlord and tenant of the same property. That is the basis on which the CPIH measures housing costs. I think it is the right one. 

The RPI has two elements to it: one is the cost of a mortgage, and the average mortgage is far smaller than the average value of a house. As I said, fewer than half the people have one; the rest are in the process of paying it off. They also have the term of depreciation, which is just an index of house prices. Certainly depreciation has nothing to do with the cost of buying a house. It depends on the cost of a builder and that is not the way that the RPI measure is constructed.

Lord Sharkey: It seems to me, listening to what you are saying, that you acknowledge that there is something necessarily arbitrary about the construction of any index, and no index can claim to—

Dr Ben Broadbent: No. I think there is something definitely superior about the way owner-occupied housing costs are measured in the CPIH, relative to those in the RPI. Economically, it is a much more defensible way of measuring those costs.

Lord Sharkey: My question was going to be: why would you choose one over the other? Is it because you see a technical flaw in the RPI, in the sense that it does not represent something, or is there some other intrinsic merit in the CPI?

Dr Ben Broadbent: It is a bit of both. In other words, if you were to ask me the best way of measuring owner-occupied housing costs, I would have to say that it was the imputed rent version that is in the CPIH. For the same reasons I can identify specific flaws in the way the RPI measures these things. 

You also asked whether it would be possible for the ONS to improve that particular aspect of the RPI. In principle, yes of course it could choose to say: “Now we are going to measure owner-occupied housing costs in the same way”. That gets us into a completely new area of the consequences of that, of which the ONS may well be aware. Whether it is that change or, indeed, any other, there are material consequences. As a matter of technical principle, it is entirely possible for the ONS to redefine things in that way. 

The ONS has made plain its concerns with the RPI. Perhaps I can put the point the other way around: when it was constructing CPIH it had the capacity to choose various different methods, including the interest-cost approach in the RPI of measuring the cost of owner occupation. For very good reasons it said no: the better way to do it is imputed rent and rental equivalence. That is why it adopted that approach for CPIH, where it was free to choose any of them. 

Q63            Lord Lamont of Lerwick: Following on from that, you explained why you prefer rental equivalence and the imputation system. Is the rental market in this country large and representative enough to make rental equivalence possible?

Dr Ben Broadbent: That is a good question. It has certainly grown. Very roughly, say one-third of those 27 million households are now renters. Many of them might be renting social housing. It is also the case that, because each house will differ from the other, it is not an entirely straightforward task to construct rental equivalence series for owner-occupied housing. In principle, you should go through a fine level of detail and it may be that the stock of rental properties is not identical in its quality, respects or location to the stock of owner-occupied ones and you have to do some work to translate one into the other. What you are trying to do is say: were the existing stock of owner-occupied properties to be rented, what would be the average price at which they were rented. There is some work in getting from one to the other.

My view is that it is big enough, but it is not likely to be an absolutely perfect measure. It is still sufficiently superior to the interest-cost approach, the use of which does not make nearly as much sense to me as an economist. There were some teething problems in getting at the rental equivalent in the early days of CPIH. It has now settled down reasonably well. The ONS was right to redesignate CPIH as a national statistic some time last year.

Lord Lamont of Lerwick: When the Governor came before the Committee, we discussed the idea of changing the inflation target to CPIH. He said: “Not at this stage, because it does not have the track record in place”. How long do you think it would have to be before you judged that to be in place?

Dr Ben Broadbent: Ultimately, this is a matter for the Government. It is hard to say exactly when. I think I am right in saying that the ONS will have access to all the raw data that it gets from the Valuation Office. At the moment, it gets what it needs to construct the owner-occupied housing bit of CPIH. It does not have the complete, cell-by-cell data, but I think it is gaining access to that in the new year. I cannot give you an exact date, but as time goes on one is more reassured that this thing has bedded down without any problem. It becomes less and less likely that we will find any problems with it.

Lord Lamont of Lerwick: How important is it to have a measure that is compatible, or informally harmonised, with that of other countries? Is that important, from the point of view of inflation targeting? 

Dr Ben Broadbent: Not particularly. It is helpful, not for inflation targeting—

Lord Lamont of Lerwick: A number of people have said that that was one of the motives for doing it originally.

Dr Ben Broadbent: The interesting thing, of course, is that in the EU it took a long time—and has not happened yet—for people to agree the precisely right way to measure owner-occupied housing costs. We did not land on a common definition. The CPIH is not used in exactly the same way in other countries. I do not think that is a bar to its becoming the target for the MPC and for the inflation target. We should think about whatever the best measure is. Nor does the answer to that question depend on whether we are in or out of the EU. It will not change that assessment. We could have switched earlier. As I understand it, it was these teething problems that dissuaded the Government from switching.

Lord Lamont of Lerwick: Am I right in saying that Ireland bases its housing cost index on interest payments? 

Dr Ben Broadbent: I would have to check that. It may be the case. It would not surprise me, given the links between our two countries and how it is measured in the RPI, but I would have to check it.  

Q64            Lord Kerr of Kinlochard: Dr Broadbent, you are a member of the RPI committee. What does it do?

Dr Ben Broadbent: In the Bank, that is true. It has status only in respect of the Bank’s obligations in the 2007 Act. We have a part to play in the process of determining or passing judgment on potential changes in the RPI.

Lord Kerr of Kinlochard: Of which there have not been many. Your committee in the Bank is not concerned with reform—

Dr Ben Broadbent: No. Sorry, I have interrupted.

Lord Kerr of Kinlochard: Your committee in the Bank does not consider potential reforms to the RPI. 

Dr Ben Broadbent: Were they to be submitted to us, we would. There are changes to the RPI of some kind all the time, as there are with any index. The weights change and the surveys on which the weights are based change. Things are submitted to us from time to time on that basis. Our job is very simple: it is to ask whether the proposed change is a fundamental one, which we take to mean something other than routine changes in weights based on new survey evidence—a methodological change—and whether it is “materially detrimental” to holders of bonds. In other words, would it tend to depress the rate of inflation of the RPI? That is our job. 

Lord Kerr of Kinlochard: I understand the importance of fundamental change in relation to the bond market and so on. What puzzles me is this: I imagine that in the past, pre-CPI, there was fairly regular re-examination, modification and adjustment of the RPI. Yet we are now told by all our witnesses that it contains known errors, but nobody wants to put them right.

Dr Ben Broadbent: I do not know what the exact position would have been before the 2007 Act, or how many changes there were. Equally, whether or not we had exactly this sequential procedure, such has been the depth and volume of contracts—including, most obviously those that the Government have—for index-linked gilts and some pensions that are contingent on the RPI, whether you want to move away from it, and how you do, is a very big question for the Government.  

To put it another way, I do not think it is the fault of sequencing, which may be imperfect. The ONS thinks about a change and sends it to us. We deem it to be fundamental and materially detrimental, or both, or neither, and then it is pushed off to the Government. Whatever the precise sequencing, however this decision is taken, the Government’s role is central. Only the Government can take certain decisions, and there are many consequences that follow.

So, in answer to your question, I do not know for a fact that there were no such changes but if, for example, the ONS had proposed, even before 2007 or before there was the CPI, doing away with the Carli weighting—which added so significantly to our PI inflation in 2010 with the clothing price collection and more generally has done for ever relative to CPI—I think there would have been consequences for the Government to think about, even if they had not had this procedure, even if there had not also been something called the CPI and even without 2007.

The Chairman: What is the procedure and sequencing you are talking about? Is there a set procedure? The evidence that we have had from the Chancellor is that we have not asked him to make a change, and the ONS has said that it is not going to ask the Chancellor because it is pointless. But you are implying that there is some kind of sequence.

Dr Ben Broadbent: The way that it would work is written down; that is what I mean. I am saying not that there is a sequence of events that has happened but that, in principle, the law sets up a sequence for the decision-making.

The Chairman: But it is clearly not working

Dr Ben Broadbent: It depends what you mean; my point is that it is not the sequencing. Let me describe what I mean by sequencing. In principle, for any decision—whether changes to the RPI or whatever the source of proposed changes to the RPI—the ONS proposes something, and we have a committee at the Bank of England that sits down and asks, “Is it fundamental or is it materially detrimental?” Then it passes to the Government. The Chancellor in particular has to say, “I will take that into account. Do I or do I not make the change?” It is still his or her decision as Chancellor. These days, the ONS says, “We kind of know that the Bank is going to say it is detrimental and that the Government will than say they will not do it, so what is the point?” At the same time, the ONS makes very plain its opinion of the RPI: it is no longer a national statistic and it does not even appear in the press release for consumer prices. It is an odd situation.

The only point I was trying to make earlier to Lord Kerr is that I do not think that changing that sequencing somehow would in and of itself make the decision easier. It is a very difficult decision. If the ONS is right that the Government are reluctant to do this—it is quite understandable that the Government are reluctant—there would be big consequences were they to turn round and say either that the RPI is not going to exist or that they are happy to rename what we currently call the CPIH and call it the RPI. 

The Chairman: Sorry, Lord Kerr, I interrupted your question. 

Lord Kerr of Kinlochard: It all seems academic to me because it has not happened in Chancellor Hammond’s time and it did not happen in Chancellor Osborne’s time. You did not send to the Chancellor any ONS recommendation with your view that this is a fundamental change; therefore, the Chancellor never had to exercise that judgment. 

Dr Ben Broadbent: I understand.

Lord Kerr of Kinlochard: So my point is that, before this system was set up, the RPI was, presumably, kept up to date—or people tried to keep it up to date. That process seems to have stalled.

Dr Ben Broadbent: That may overcharacterise the situation. I am trying to remember. I think that there was, maybe before my time as DG, an occasion when we said a change was fundamental. I would have to check that, but I think that is the case. I do not know how one would measure this, but it may be that the pace of changes has slowed. Maybe it has not; I do not know whether that is true. Either way, my sense is that it would be a caricature to say that, before CPI existed, RPI was changed willy-nilly and all the time and was imperfect in some sense, and that afterwards it has not been. 

Let me make it plain: the main problem with RPI that the ONS has pointed out is a problem that has been there from the beginning and has never been changed—nor would anyone think of changing it—which is the way that inflation rates are weighted. That has been there, and no one thought of changing it before. Whatever the procedure and whatever the sequencing of decision-making, that had not happened. It so happened that there was a change in the methodology of collecting clothes prices that made this problem even bigger; it was highlighted with the so-called Carli measure. But, let us be clear, the main problem with it was not addressed before or after the CPI was there.

Lord Kerr of Kinlochard: It seems to me that the motivation for nothing coming up this sequencing pipeline is that the people at the start all believe that the RPI should wither on the vine. It seems to me that if we are stuck with the RPI for a time, it is very strange to be using something that has known defects, which are thought to be increasing. I would have thought that you need somehow to break that. Letting it wither on the vine seems to me to be a perfectly good idea, but in no way with making some improvements and remedying the defects in this thing while it still exists.

Dr Ben Broadbent: I understand. My only observation is that the biggest defect, the one glaring defect, which was accentuated by the change in the collection of clothes prices, is a very discrete change. It is not a small thing and the consequences are significant. This is a decision that is difficult for the Government and can only be taken by the Government. Whatever the particular arrangements for making reforms, I do not think it is just the ONS’s timidity or the structure of the process that means it had not happened for ages. As to the faults in the RPI and its index, one of my economists handed me a pamphlet by Irving Fisher from 1920 which pointed out how poor the Carli weighting measure was. This is not a problem that suddenly emerged and might have been cured much earlier. It is not clear to me that the system before was perfect and allowed all these things to change, because the biggest single problem with this index was there throughout. 

The Chairman: It is a problem that we were encouraged to look at by your boss, the Governor. 

Q65            Lord Turnbull: Is the problem not that you are being asked the wrong question? It is not whether something is fundamental and detrimental. The question should be: is it detrimental, but is it justified in terms of best statistical practice? It seems to me that you are frozen into something where you can see it is glaring but it needs quite significant change. Whether it is the Bank or the Government, simply saying that it is fundamental and detrimental should not mean that no change is made, because something that is absolutely supported by best statistical advice may be saying that you should make this change. The ONS has a duty to provide the best-quality statistics that it can. It cannot do that if it is saying that this is detrimental and therefore it cannot make it. We have to allow inferior measures. 

Dr Ben Broadbent: As I say, I really do not think the fault lies there. Let me paint a different world, where we had the same arrangements for having reforms, or proposed reforms, but where the RPI was entirely inconsequential. No private sector contracts were linked to it and no debt linked to it. Of course this would have happened. The problem is not the structure of the arrangements by which reforms are suggested, it is because of the consequences that follow from a change. That is the fundamental issue and it is why only the Government can grasp this nettle and have sufficient power to co-ordinate things, if indeed they decide that that is the best thing to do. When you say that the ONS has a duty to measure things, it has fulfilled that duty on one view: it has relegated this from national statistics data.

Lord Turnbull: It is not fulfilling that duty. It is saying quite clearly, “We are going to allow a substandard index to carry on. We are going to calculate it. We are going to publish it. We are going to remove some status from it”. So it is not following the practice of maintaining at all times the best-quality statistics that it could provide.

Dr Ben Broadbent: Well, let us not forget that the Act says that it has to produce the RPI, so the ONS certainly cannot take that particular decision on its own and say that it is going to stop producing it—it is required by law to do so. Let us take the most dramatic version: it could have said, “Yes, we are going to carry on producing it but we are simply going to reform so that it is identical to the CPIH, which we think is the best measure of consumer prices”. I suppose it could propose that, but it would still land at the door of the Chancellor, who would have this big decision, with lots of consequences for account holders. In the end it will be a government process, co-ordinated with the ONS.

Lord Turnbull: He cannot take that decision if people do not write down their decision and put it in front of him. 

Dr Ben Broadbent: I think he could take it anyway, let me put it that way. He does not have to wait. 

Lord Turnbull: May I clarify one thing? In the CPI that is targeted by the monetary committee, how much housing cost is in there?

Dr Ben Broadbent: Not much for owner-occupier. The CPIH differs in that respect. In practice, because the real price of rents has not moved around that much—in other words, rental inflation relative to CPI—the two rates of inflation, CPI versus CPIH, have not differed that markedly. We will target what we are asked to target. To put it this way: the switch from RPIX to CPI back in 2004-05 was in some ways bigger than this one would be.

Lord Sharkey: I was going to observe that, when we discussed all this, the Chancellor told us that he had never been asked to make a resolution. The fact is that the system, or rather the ONS, is preventing the Government from examining this question. 

Dr Ben Broadbent: I am not sure that I agree 100%. The Government are free to look at this with or without the ONS’s recommendation. 

Lord Sharkey: In the abstract sense, the Government are of course free to look at anything they like, but in the real-world sense, the ONS, by not acting, is preventing the Government from making a judgment about this.

Dr Ben Broadbent: “Preventing” is too strong a word. I get your point, but I think the Government are perfectly free to look at it as they wish. It is a very big question with lots of consequences, so that is more material than the procedural arrangements that we have. What I am pointing out is that, long before we had this procedure, there was still what I view as a pretty significant error—or at least a second-best way of measuring prices in the RPI. It was there from 1947 and no one challenged it, so it is a bit odd now to say that the current procedure is the thing that has prevented it from being changed. 

Lord Turnbull: We had one piece of evidence that was rather surprising. You referred to Carli, which has the feature relating to something going down and bouncing back to where it was. Simon Briscoe, who is part of the RSS, said that that is not actually how Carli is used. He said that it uses levels rather than price changes. Therefore, you are blaming Carli for something that it is not to blame for.

Dr Ben Broadbent: No, I do not think that is right; I think it is clearly to blame. Let me give an example of the result of this. You have the same sample of raw data—item by item; price collection—used for CPI and RPI. That property of things, measuring average inflation rates, means thereby having a situation where you have a range of things that go up and then come down. Therefore they are not returning and not having zero average inflation across the two periods, which is worse the greater the dispersion of the inflation rates. What happened in 2010, for various reasons, including the way clothes price changes were collected, was that there was a great deal more dispersion in the inflation rates of the individual items. That is what suddenly pushed this up. It was very material: close to half a percentage point a year. 

If I look at inflation for clothes prices, at the cumulative change, CPI has gone up 15% since 2010 and RPI has gone up 100%. There are certain items within that where there are massive variations from month to month, because there are sales and then the sales end. A particular item that shows you how detailed this can get is women’s vests and strappy tops. It is 20% on the CPI and 500% on the RPI. That is purely as a result of the difference in methodology; there is absolutely no difference at all in the raw data. It is Carli, and you can show what it would be. The ONS has a series of different weightings showing what it would be. 

I take all the questions about the procedure and the point that it may be imperfect, but this is really about the consequences of the change. They are sufficiently significant that it has to be the Government who are the mover here, I think. 

Q66            Lord Burns: Could I move on to the question of what to do about this? As you point out, we have a lot of gilts that have very long maturity dates and which are tied to the RPI; we have pension obligations that are tied to the RPI; and we have this index, which, we have heard, is regarded as not very satisfactory. Should not the first step be to say that we will not use RPI on anything from now on, including gilts and all contracts, and to say that, as far as the Government are concerned, they will not link anything to the RPI going forward? Is that a minimum first step?

Dr Ben Broadbent: Broadly speaking, yes, I think so. It would be a necessary condition if you really wanted a serious move away from RPI. There are a couple of thingsthat, out of self-interest, if you were going to do that you would have to have something else. If you started to issue, for example, government debt linked to some other index, presumably you would want to link it to what you think is the best measure. It would also make sense, therefore, to have asked the Monetary Policy Committee to target that. I would have thought that that would also be a precondition. But, yes, broadly speaking that is right. Whatever the ultimate solution, any serious solution would involve saying, “Index contracts from now on will index to this rather than that”.

Lord Burns: Can you think of any solution to that fact that there are these very long contracts? When we discussed this with the Governor, he seemed to imply that, given sufficient warning to all parties concerned, it may be possible to fix some time between now and 60 or 70 years hence.

Dr Ben Broadbent: I think that in his evidence he suggested something sooner.

Lord Burns: Yes.

Dr Ben Broadbent: Broadly speaking, you can imagine a path out of this that is relatively gentle, where you say that all new debt is indexed to the CPI, encourage the private sector, to the extent that DB pensions are still being written, to say that they will be, even beyond retirement; and accept the fact that the existing gilts and all the existing trust deeds that govern the indexing of pensions are as they are for as long as they exist. It would take many decades for them to—as someone said earlier—wither on the vine entirely. 

Then there is a more radical set of options, which involves cutting off the RPI or indeed having the ONS, as I suggested—though this is a decision that would have to be taken in consultation with the Government, who would be responsible for it in the end—saying that what we currently call the CPIH is now the RPI, so de facto cutting it off. Those are more radical things that would have legal consequences. With gilts, the difficulty is that they vary contract by contract—some specify precisely the RPI, some do it without saying exactly what the RPI means and so forth. But you are right that all of them would involve a move in new contracts, whether those are indexed gilts or pension funds, to indexing to something else. 

Lord Burns: What about this period over which you think we could do the transition, if the Government were minded to do that? What would be a transition period—

Dr Ben Broadbent: To moving to these new contracts?

Lord Burns: Yes, that would cause the least trouble—for existing contracts.

Dr Ben Broadbent: For existing contracts? I do not know. As you know, it is a continuum. Some of the gilts are very long-lived. We have been issuing them continually, obviously. The rate at which new pension contracts are being written is not that high. In the private sector at least, the rate at which DB pensions are being written has slowed a lot. As I say—this is just a selfish concern—I would have thought that the Government would want to ensure that, before they did anything in that respect, or at least as part of it, they say, “Let’s switch the inflation targets to whichever index we are now going to use”.

Lord Burns: We have noticed that the RPI has an awful lot of supporters. That may be partly tied to the fact that it grows faster than the CPI.

Dr Ben Broadbent: It may well be.

Lord Burns: The people who are recipients of things that are indexed to the RPI may be the reason it is so popular, but it has also been said to us that there are certain circumstances in which CPI and the weighting system could lead to underestimating and could have a downward bias in it. 

Dr Ben Broadbent: I do not think that is true

Lord Burns: Are you quite confident that there is no bias and the CPI weighting measure used is the one that creates the most accurate measure?

Dr Ben Broadbent: It is the most accurate measure; I do not think it leads to an underestimate. As I say, the failings of the Carli index have been evident for a century.

Q67            Baroness Bowles of Berkhamsted: I am trying to apply some fundamental logic to this. Is your mandate on your committee that looks at these things asymmetric? All you have to do is to see whether something is materially detrimental to gilt holders, whereas you do not have to pipe up and say when something is rather jolly good for them.

Dr Ben Broadbent: No. That is absolutely right, and it is by design. We are not asked to say, “This is materially beneficial”. But very clearly, this change in the way that clothes prices are collected led precisely to that and has added to the cost of the existing gilts at the cost of taxpayers and to the benefit of the gilts—it is tens of millions of pounds. 

Baroness Bowles of Berkhamsted: But now you cannot back off from it, because to correct it you have to give an opinion that says, although they have had a bonus, if you like, for a certain number of years, to rub that bonus out is detrimental. 

Dr Ben Broadbent: No. I understand. I come back to the issue that it is not so much that it is our role or the role of the ONS. It is a fundamental legal difficulty. I am no lawyer—some of the RPI-linked gilts have an explicit clause saying that the owner can go back and redeem them at par. Now that they are trading above par, maybe it would not be worth it. There is a broader question—I do not know the law on this and I do not consult lawyers. If you made a change that benefits them, they are obviously not going to come to you but, if you make a material change that costs the owner of a contract, they may say that you are no longer fulfilling the contract. So there is a clear asymmetry. 

The Chairman: On that point, does that not depend on whether the contract says that it is a measure of inflation or says something specific about the RPI?

Dr Ben Broadbent: Some of them are specific, some of them are not. Even those that are specific about the RPI do not say, “It has to be measured in this way”. I do not know the answer to this question, but they may none the less have grounds for thinking about or having a go at saying to the Government, “Hang on, you cannot just redefine the RPI as something that never grows and is just a fixed number. That is a breach of contract”. There is a clear asymmetry in the way that the law would apply.

Baroness Bowles of Berkhamsted: I understand that, but I think the opinion of the Bank of England is an important one, not least because you say it is up to the Chancellor and that he could do these things. Obviously a Chancellor fiddling around with RPI and so forth, even for the best and honest and proper reasons, maybe backed by recommendations from this Committee, is still nevertheless going to worry that he is open to accusations that he is fiddling the books. In general, they would probably prefer not to have that accusation. Is the Bank in the situation now that it can never recommend this correction to the clothing methodology in the RPI?

Dr Ben Broadbent: No. I come back to the point that the constraint on the Government is not that someone else has not recommended it. The most important constraint is that it is a very material decision with distributional consequences for owners of gilts and the future taxpayers that have to meet the costs of funding them and also for a host of other RPI-linked contracts that the Government write, on one side or the other of those contracts. I do not think that the overriding constraint is whether we have to say whether it is detrimental. This is a big political task for the Government, if they choose to engage with it. Tweaking with the procedure will not make it any easier or harder for the Government than it already is. 

Lord Sharkey: Just to be clear, your committee, as I understand it, as you describe it, does not in fact make a recommendation, it makes a judgment. 

Dr Ben Broadbent: It makes a judgment, quite so. That is it. Let us be clear: it is not up to us to make recommendations. The way it works at the moment, the ONS considers changes and is obliged under the law to pass them to us. We have to make these very particular judgments, and then it goes to the Chancellor. 

The Chairman: Am I missing something here? When the Chancellor came before the Committee, he said, “I cannot take a view on this because I have not been advised to do so”. You are saying, “This can’t be done unless the Chancellor decides it”. So am I right to conclude that what you and the ONS are doing is protecting the Chancellor from taking a difficult political decision?

Dr Ben Broadbent: No. I am simply pointing out that it is of that nature. The Chancellor, of course, is right in one way. The only thing I would say in response is that there is nothing to prevent him from considering the question. 

The Chairman: Except that you have not asked it. The ONS has not asked it.

Dr Ben Broadbent: He can consider any questions he wants.

The Chairman: The ONS has said to us, “We are not going to ask the Chancellor because we know the Chancellor will not do it”.

Dr Ben Broadbent: Clearly there is a co-ordination issue here. I make the point again: it is not the overriding problem. Suppose we had some perfect system, or suppose it was not this procedure, the Government would still face the big question of whether to do this. What are the potential risks? What do the contracts, which are very widespread, say? What are the distributional consequences? These are the material things.

The Chairman: Yes, we have that point.

Baroness Bowles of Berkhamsted: I am trying to move on to working out whose business it is to know what legal consequences are of changing the RPI. Is it the Government through the DMO? Should the Bank know what the effect should be or does the Bank have no role?

Dr Ben Broadbent: It is the Government; these are government contracts, let us be clear. It would have to be the Government who go through this. Unless it was the very gentle process that Lord Burns suggested, where you simply shift all new contracts to another index and allow the rest gradually to shrink, it would not be without legal risk. I say that without being a lawyer; I do not know, but I doubt it would be.

Baroness Bowles of Berkhamsted: I understand that but, against that, there is concern about, as the RPI withers on the vine, so to speak, what enthusiasm there is to keep on computing and updating it properly, not least when we have this problem that there is already something wrong with the methodology. 

Dr Ben Broadbent: That would happen. As I say, the ONS is currently legally obliged to do it and will carry on doing it as long as it is asked to do so, I am sure. 

Lord Tugendhat: I think we have exhausted the question. 

Q68            Lord Kerr of Kinlochard: On market fragmentation, our rivers were run by terrifying tales from the Debt Management Office about how, if you started issuing CPI-linked gilts while there were RPI-linked gilts in the market, as well as non-index-linked gilts, the market would fragment and prices would go shooting up. I was not very convinced by that. Could you give us two minutes on the risk of market fragmentation if we started issuing CPI-linked gilts right now?

Dr Ben Broadbent: It is an important question, and the DMO is right to think about this. There is an awful lot of government debt out there and it is important that we have a deep, liquid, functioning market that minimises the cost of issuing that debt. I cannot see as a matter of principle why it would become fragmented. I suspect, though I was not there exactly at the time, that there were similar concerns about issuing RPI-linked debt at the same time as conventional gilts and that somehow this would be an issue in quite a liquid market. I would also point out that there are private sector CPI-linked contracts. There is a swaps market in CPI and a couple of private sector institutions have issued CPI-linked debt of their own. There has been a demand for that, and I think that there would be a demand for this too. In some sense, as I say, it is the same co-ordination problem. Once you have done it, there would be other contracts—more contracts—linked to it and a liquid market would develop. So I can understand the misgivings. The DMO’s job is to try to operate in a deep liquid market. It may take time for that to develop, but I do not think it would be a permanent and costly fragmentation. 

Lord Kerr of Kinlochard: So you think that the demand would be there. We were also advised that, because so many index-linked gilts are held by pension funds whose liabilities are increasingly RPI-linked, they would not be very keen to buy CPI-linked.

Dr Ben Broadbent: Again, this is an equilibrium. It is not necessarily the best one but maybe over time you would have another one where they do not. I would also say that, for the purposes of revaluation, which is what has happened to the liability of a pension for someone, the employee, who is still working for a firm, many pension providers, many private companies have switched that to CPI. Then RPI kicks in once people retire, when it starts paying out linked to the RPI. So I would have thought, at least in the shorter term, they would actively want RPI to hedge against CPI-linked debt. There is a CPI exposure within many private sector firms and pension funds. 

Lord Kerr of Kinlochard: Yes. So you see no particular difficulty in starting to issue CPI-linked now. Could I take you a little further? Supposing that one stopped—here is Lord Burns’s suggestion—issuing RPI-linked, do you think that it would change the gilts market in any particularly worrying way?

Dr Ben Broadbent: It may be the more legitimate—there will continue to be these obligations for firms that have linked to RPI. The trustees these days take quite a strict view about trying to hedge risk wherever they can. So it would be a question of how quickly those obligations—those pension obligations, for example—went away and how quickly the stock of RPI-linked debt itself would dwindle. You have to make some comparison. It would still exist, obviously; the stock would be there.

Lord Kerr of Kinlochard: Yes. Thank you.

Baroness Bowles of Berkhamsted: Would there be a price increase in that stock when everybody suddenly wanted it?

Dr Ben Broadbent: Possibly, but I think there would be demand for the alternative as well. I do not think the one somehow cleans out the other or dominates. There would be a price that equilibrates them. 

Lord Burns: Presumably it would not take long before some clever person in the Bank of England calculated what the market’s view was of the expected difference between the two.

Dr Ben Broadbent: There is, as I say, a swaps market; it is not that liquid at the moment, but there is a swaps market. Even if you did not issue any new RPI-linked government debt, I have no doubt that the private sector would come up with some swap that priced the difference. 

The Chairman: Dr Broadbent, thank you very much. I know this is not a particularly easy subject. You have brought clarity to our thinking, for which we are very grateful. Thank you very much.

Dr Ben Broadbent: Thank you very much.