Robert Chote: the full transcript
Last week I interviewed Robert Chote, the boss of the Treasury watchdog/forecaster the Office for Budget Responsibility (and also a former economics correspondent of The Independent – it’s been all downhill since).
Here’s the write-up.
Chote told me that George Osborne’s intransigence will almost certainly mean the OBR will not be costing the spending proposals of opposition parties at the next election (story here).
But for fiscal geeks there were some other interesting, if less newsworthy, observations so I’m publishing the full transcript below.
* Thanks to the OBR the Treasury can no longer “nudge” spreadsheets to deliver helpful forecasts.
* Chote says the OBR’s output gap estimate is likely to continue to “evolve” in the coming years, meaning the fiscal hole could shrink and the envelope for spending could increase.
* Chote has not yet applied for a new term (his present one ends next year) but if he does, he wants a decision to be made on his re-appointment before the general election and with cross-party backing.
* And….Chote doesn’t do hillwalking
Have you enjoyed your three and a half years in the job?
Yes I have very much. I’ve enjoyed it enormously. The group of people we’ve ended up with is terrific. I didn’t know what to expect moving from the Institute for Fiscal Studies but I’ve just been delighted by the quality of those people and the spread of skills. I think that the process, which is unusual by international standards, having the necessary conversation with officials and policymakers prior to the Budget…that’s worked well. The important thing there was sort of being clear before you embark on forecasts: what’s the timetable? For it all to work well we need to get information from Treasury and DWP [Department for Work and Pension] and HMRC [Her Majesty’s Revenue & customs] in good time. That’s worked out very well and that’s down to civil service chiefs of staff. We have our meetings with the Chancellor in the run up to each forecast and he asks us questions about how we’re going to respond to particular measure’s he’s announcing and that’s worked well. The mechanics of it have been good.
So nothing’s been sprung on you at the last minute?
No, this Budget there was the decision on extending childcare support to everyone in universal credit. It was basically decided late on and we only heard that on the Monday before the Budget by which stage the documents had gone to the printers. So we highlighted that we hadn’t had a chance to include that. We pointed out what the Treasury’s estimate of how much that was going cost but said we hadn’t had a chance to scrutinise this. From our point of view it’s not good enough to say: “Oh we’ll come up with something else that offsets the cost”.
Is that bad practice from the Treasury – or is it inevitable?
It’s not inevitable – it’s not happened that often. We’ve had one other measure where we thought we didn’t have enough time and information to reach a judgement. It was a relatively small measure a few forecasts ago and to be honest it was a measure that was very difficult to come up with a fully articulated costing for anyway. Clearly, in an ideal world when you’re publishing a forecast you want to include every stated government policy at the time they stand up. However, it’s not for us to say the government can’t announce any policy between giving us the final thing. It’s a choice they have to make.
So you don’t say “please don’t let this happen again”?
What I don’t think is an excuse is to say “we’ve announced something that costs some money, but don’t worry about not including it because we’re going to announce something else that we haven’t thought of yet in the future that will offset this.”
Have you ever said “no- we’re not going to sign off on that” to the Treasury”?
No, it’s an iterative process. They present us with the draft scorecard….When we see that initial list we sit down with the Treasury and HMRC and we tell them which of them we think we’ll need to devote most time to and probably want to have meetings about. But there will be other ones where it will be relatively straight forward and you can sign it off. But inevitably as this process goes on in the six to eight weeks running up to the Budget some things will drop off that list, and other things will come on. We’ve never had a situation where we’ve looked at a policy, we’ve suggested changes to the costing, the Treasury has not accepted those costings and we’ve had to say: “The Treasury thinks this will cost £200m and that’s what they’ve put in their score card [but] we think it will actually cost £500m and that’s what we’ve put in our forecast.” We’ve never had a situation like that. Sometimes the initial list of measures will include three things that are kind of alternatives. You might scrutinise all of them and depending on what you say about each one, that may affect their choice about which of them to do
But is there a debate about costings? We think X and they say half X etc?
No, I would have expected that before I came in to be the case. Outside people sometimes assume that the relationship is an inherently bilateral one where we’re on one side of the table, the government is on the other side of the table, they have a line that says this costs £500m, and we just kind of fight it out to a standstill. Which isn’t really how it works. For a start it’s a triangular relationship. We’re around the table, so is the Treasury and so then is HMRC if it’s a tax measure, DWP if it’s most sorts of welfare measures, DEC [Department of Energy & Climate Change] if it’s energy.
That sounds like a pentagon…
It’s generally only three: there’ll be one responsible department in addition to the Treasury. And HMRC if it’s tax will bring a lengthy paper explaining the tax base, what’s the behavioural change expected to be etc and we will discuss that and at the end of the day we have to decide what to put in our forecast whatever the Treasury puts in that table. There’s obviously an incentive on everyone’s side. You don’t want to a have a completely different set of numbers so generally speaking when we think 10 per cent of people will respond this way etc normally our view is taken on board. A lot of it is very uncertain. If at the end of the day we were unhappy we would say publicly what our alternative costs was. That situation has not arisen yet.
The OBR was set up to restore public trust in the public accounts. In the 2012 Autumn Statement George Osborne delayed payments to various international bodies solely to allow him to say that the deficit was falling slightly year on year. Is that not something you, as guardians of integrity in the official accounts should have blown the whistle on?
In a sense quite the opposite. The reason the Treasury was devoting so much time to going round departments and persuading them not to spend very much in the end was precisely because we have the forecast. In the old days you’d look at your forecast of the budget deficit and if it’s not going down a bit in that year you just [mimes elbowing], give the spreadsheet a good nudge, or let’s just change the assumptions about this or that tax, let’s just shove nominal GDP up a bit. And you’d have managed to smooth that out. That isn’t possible under the [new] process. If the Treasury wants to keep the deficit falling year in year out they’ve got to do some policy to get it to happen. And indeed in that instance it was striking that they put a lot of effort into insuring that spending in the last few months of that fiscal year would be lower. Now part of it, which we highlighted transparently in the way we wrote it up, was indeed things like taking some spending and shifting it into following year – but also genuinely getting departments to underspend the spending limits they’d been given by what now, looking back, is clearly a historically very large amount relative to previous years and indeed to what we’re assuming this year. There’s a separate question which is: does it really matter very much if the cash value of the deficit is falling when the figures are going to be hugely revised over the subsequent months anyway, so why devote effort to ensuring that you’re one million one side of the line (triumph!), one million the other side (disaster)? Does that make a lot of sense? At the end of the day that’s for the Chancellor to decide how much that matters. The striking thing is that in the old days that would have been fairly straightforward: a little bit of spreadsheet magic would have sorted that out very nicely. That is no longer available.
So it’s a sign of success then?
Indeed, very much so.
And you’re not worried you’re being gamed?
At the end of the day the Treasury can decide where it’s going to be. It is for them to decide where they want to end up by persuading us that the measures they’ve announced are sufficient to do that. You may want to ponder whether it’s a sensible use of anybody’s time to make those kind of precisely graduated changes – but that’s exactly how it should work. What the system doesn’t do, very sensibly, is put you in a world in which we publicly say something that comes as a surprise. The Treasury’s got targets that it needs to hit. We need to tell them what will happen if you don’t do anything and what you need to do in order to hit the target. And if they decide to do it with a £100m margin that’s not gaming the system. They’ve taken the decision to do what is necessary to move that side of the line. That’s the process working properly.
The rumour is that you were sceptical of the value of forecasting before you took this job. Is that true?
Indeed, I was a consumer of it doing your job for many years. One of the striking things is if I think back to 1990 to 1994 and if you look at the outturn estimates of GDP were over the period coming out of the early 1990s recession: it’s completely revised. And three years after the pre-crisis peak, the numbers that I was writing about showed that GDP was still 2 per cent below the peak. The data now show that three years after it was 2 per cent above the peak. So four percentage points difference. So one thing I’m very conscious of is that we’re dealing with very early drafts of economic history. In the more recent context we have a very sharp contrast between GDP being flat until early 2013 and then taking off quite sharply. I’d offer you a modest bet that in 10 years’ time when we look at the data that distinction will not be quite so clear as it looks now. That said, for the time being we have to take the data as the best judgement we have. We don’t do what the Bank does, which is to forecast the re-writing of history. That was one reason to be cautious about the forecasting process. The other one is obviously the range of views about which reasonable people can disagree, particularly around turning points. Go back to the March 2008 Budget. [In] the compilation of forecasts that was published one of 38 forecasters predicted there would be a fall in calendar year GDP in either 2008 or 2009. It ended up falling by 6-7 per cent. Peter Warburton was the only person who predicted very small falls in both years, which was still way off. No sensible person bets the farm on any particular macroeconomic forecast being right. The key question is if you have particular objectives for the public finances in mind is: by how much does this forecast have to be wrong, and in what sort of ways does it have to be wrong, for this to matter. At the end of the day we only do macro forecasting because we have to get to a fiscal forecast. We don’t know anything more or any less than anyone else because it’s all in the public domain. Where we do have additional information, where we do add value, is getting to a forecast for the public finances, which has always been the black box of the fiscal forecasting process. When I was at the IFS in the early 2000s, when we were expressing concerns then that that we didn’t think the public finances were going to improve at the rate expected, which turned out to be the case, it wasn’t particularly because we though the economy forecasts were wrong, it’s that we thought that the assumptions in the fiscal forecast conditioned on those economic forecasts were a bit optimistic, in particular the pick-up in corporation tax.
So your value seems to be as a forecaster in a steady state economy, when there aren’t external shocks. Is that fair?
If you had imagined that, for example, the OBR had been in existence in the early 2000s and that we had been publishing fiscal forecasts then, I suspect they would have been less optimistic than the ones being produced by the Treasury at the time. However, we would not have spotted any more than anyone else that we were going to have a colossal drop in nominal GDP and that’s obviously what has driven most of the deterioration in the public finances. You’re right, there are errors that arise in relatively normal times and those errors tend to be in getting from the economy to the fiscal forecast. Then errors at turning points, like missed recessions, is basically an error in the economic forecast rather than in the fiscal forecast. We tease this out directly. The thing we produce called the Forecast Evaluation Report we go back to the earlier forecasts and we decompose the error into: How much of it is because you got the economy wrong? How much of it is because you got the fiscal wrong? How much of it is because the ONS reclassified? etc. That’s our least read and perhaps most interesting publication because we can go back and look at what went on and it highlights how data revisions cast fresh light on what has happened. Having an exercise [like that] once a year is actually very healthy.
So your goal is to get the fiscal error part down?
That would certainly be nice. But the key thing is to be more transparent. The main problem with a lot of that stuff is that it’s just very black box. Certainly from the IFS perspective there were some bits of the fiscal forecast that as an outsider you can second guess pretty easily. With the IFS micro simulation model you can have as good a guess at what income tax or VAT is going to do over the medium term as somebody in house. But there’s an awful lot of the forecast where we at the IFS had to say: we’ll just have to take the Treasury’s word. Whereas what we do is bottom up. I guess one of the things that’s striking about moving to this job is realising how small a proportion of what was going on the public finances as an outsider you were ever able to get your hand on. That’s not to say that it still isn’t very important for people to do that work
What about the output gap. Every 1 per cent variation means around £11bn more or less in future cuts. Do you feel a sense of responsibility about that?
You do think about those things – this has consequences for the policy choices that people are making. The key thing I hope that policymakers take when they consume our stuff is not simply to look at the central forecast but to look at the uncertainty around that. Our job is basically top come up with our best judgement of the median outcome because we’re being tasked to say whether there’s a greater than 50 per cent chance. Although in fact we don’t do a sort of subjective probability distribution, it’s all focused on whether the 50 per cent chance is higher or lower. For the policymakers there’s an additional layer of complexity and decision making which is: what are the relative costs to them of being wrong in particular directions? So, on the fiscal consolidation, there’s obviously a wide debate (in which thankfully we’re not required to participate) on the relative risks of going quicker or going slower given you don’t know how much consolidation is eventually going to have to be done. That’s a judgement that is going to have to evolve over time. On the output gap stuff it’s very uncertain, it’s always been very uncertain. I’m struck now that the range of outside estimates is minus 5 per cent to plus 1 and you think crikey we really don’t know what this is to within 6 percentage points of potential GDP. But Ken Clarke had his panel of independent experts look at this in the summer 1996 and as I dimly recall the difference between the highest and lowest estimate was 6 percentage points! You’d think there’s a lot more reason for this to be uncertain now than in 1996 given the scale of the departure from the previous trend that we’ve seen. But [also] you’re doing the forecast over a five year period, so if the outlook evolves, policy can evolve, so it’s not like you’re locked into a policy as it will effect policy and tax in five years’ time. We’ll get new information about that over time – it will evolve. If you look back over the last three years the hole has appeared to be larger and policy has responded. Maybe looking forward the hole will sometimes appear smaller and policy can respond. Or it may appear larger again. It’s a constantly evolving process.
What about the fact that the Government pushing ever more cuts out into the next Parliament? Is that legitimate?
We judge them against the targets they’ve set themselves. On that one they’ve chosen it to be a rolling five year ahead target. In the days when we had a fixed period through which the meeting of the golden rule had to be judged there were anxieties about that because it mattered so much where your start and end point was. We have a forward looking target in part because of the concerns expressed by people when we didn’t move. So its horses for courses to some degree. But the key point is that we’re not just saying: are you meeting it in that particular year. We are providing the full forecast through that period. We’re explaining whether as the mandate year moves forward we still tell people whether the target would have been met or missed in the earlier years so that’s all transparent. We give people all the information to know whether the year five which is now year three [for example] you’re hitting it or missing it.
Are the spending cuts achievable?
If that underspending finished and we ended up in a situation where they were finding it hard to meet their spending objectives in the short term we might have to look again at whether to make some sort of a judgement over the longer term. This is ultimately a political choice. Other countries have shares of public spending higher and lower than we do and for us to say it can’t be done here on that basis…It’s not unreasonable for other people to do so. The IFS have said “don’t you think it’s more likely there will be tax increases?” – it’s a perfectly reasonable debate.
If you were asked to advise the Chancellor on how the remit of the OBR should change what would you say?
I’m happy. The overarching duty gives us a broad enough remit to be transparent about things. There’s a specific issue [about] costing policy options. The Congressional Budget Office in the US thinks up policy options of its own to look at: “If you want to reduce the budget deficit here are your ways you might do it”. The [OBR’s] remit is pretty well laid out. The things that have been added on since we started – looking at the devolved Scottish tax receipts. The other new thing is looking at the welfare cap. There will be an interesting addition. The Government has a 2 per cent margin. One consequence that will have is that we will need to be explicit about what we think. When the forecast [for welfare spending] changes do we think it will be [a result of] policy [or changes in the economy]?
In general do you have the resources you need?
Yes I think so. We will get a bit more resource to do the welfare job. I think that it’s very important that we have a budget that is set out for a number of years, as a safeguard for independence. In other countries where governments have tried to put pressure on their fiscal watchdogs the budget is obviously quite an effect way to do that and so being quite transparent about the amount of money we’ve got, recognising that if the remit changes you might need more or less.
Have you got that at the moment?
Yes, it’s set out over the period of the spending review. And then we’re looking to get another multi-year settlement. Basically we have agreement on what the next multi-year period is going to be. That will take us two years beyond [2015-16]. When that’ll be included in the Treasury Budget I’m not entirely sure. It’ll be a formal thing. You don’t want to be in a situation where every year you’re having a debate about how much money you’re going to get to run the organisation.
You’ve got a five year term. Are you going to apply for a new one?
It depends whether the Chancellor wants me to do it. I don’t come up for renewal until October 2015. Basically, I obviously don’t want to be in a situation where there has to be a changeover around the next general election time. But I haven’t had the conversation yet with the Chancellor yet about whether he wants me to carry on and I would be very happy to.
But it might be a different Chancellor after next May?
The renewal doesn’t necessarily happen right at the last minute. Because otherwise if you left it right until the last minute and they said “no” then I’ve got kids to feed! So [fellow executive] Graham Parker has just been renewed but his term doesn’t actually run out until the autumn. But I would suspect and very much hope that it did come to wanting to replace me that’s something that the Chancellor would talk to the other parties about. You wouldn’t want to do it if the opposition parties didn’t want you to do the job.
So providing that agreement’s there would you like to do another five years in this job?
I haven’t thought about that yet. What I don’t think would be very helpful is to go around the election time because then you’ve got the potential then if there’s a new government and if they want a fiscal statement relatively early to have someone new in this job trying to get to grips with that. There’s certainly some attraction in going on at least a little bit beyond the election. But as I say I haven’t had the conversation. I’m enjoying it very. I’m not desperate for early release!
But 10 years is a very long time!
Graham Parker and Steve Nickell have been renewed and both have said they’re likely to serve another three. So it’s not five or nothing. I don’t have to leave in 2015 or 2020. There’s slightly greater flexibility.
Given your rather turbulent history as a journalist with the previous government could you work with Labour?
Absolutely. I’ve been a hack and doing the IFS and other things since the early 1990s which has involved saying things whoever was in government at the time sometimes didn’t want to read. That was true under the Tories as well as under Labour. I had meetings with Ken Clarke when he was Chancellor and he had the stuff that I had written sitting in front of him. He probably hadn’t read it quite as closely as Gordon [Brown]! But no, no, it’s fine. I see Ed Balls, erm, not frequently but from time to time – obviously when he’s been thinking about our role as regards the opposition party. I’d have no difficulty working with the other parties. Ed has been arguing that we should have an extended role rather than a narrower one. That’s nice. Whatever the merits of doing these things [costing opposition manifestos] now it’s nice that people think the organisation is doing a good enough job that they want to expand it, rather than shutting it down, or getting it to do a lot less. But I haven’t had a broader conversation about what other plans anybody else might have for us.
Have you said to the Chancellor: let’s get on with and let us cost the proposals of other parties?
I said it to the Treasury Committee when I was in the appointment process and obviously to the Chancellor going through that process that I wanted Parliament to give us a clear steer as to whether they wanted us to do that job or not. And they did give us a clear steer back in 2010/11 that they didn’t want us to do that job. But it’s entirely reasonable that some people want us to think about looking at the issue again now. As I’ve said, I think this close to the end of the Parliament, getting us to do it now would be quite a big ask because of the need to get, in the first instance, agreement in principle, and then agreement on the rules of the game in place.
Are we past the point when it’s a good idea?
We would want to be in a position in early summer to start looking for the extra people we were likely to need, to look for the extra floor space, to put together the detailed rules of the game. And in order to do that you would need not only political agreement on the principle, but political agreement on the substance of how this would work by the early summer. That’s not in my hands.
If nothing’s happened in three months?
By that stage that would be too late. You’d need first of all an agreement on principle, which there isn’t at the moment. The Chancellor perfectly reasonably has said he doesn’t think this is the right time to do this. The reasons he has cited are it’s the first general election we’ve existed – you don’t want to throw the OBR as a relatively young body into a politically contested territory now. And the issue which I raised in the letter – this does have some interesting implications for the role of civil servants in their interaction with…what sort of work can civil servants in effect do with opposition parties that the government of the day doesn’t have access to? The civil service is basically about serving the government of the day. I’m sure it’s not an insuperable thing – but it’s something that does have broader implications than just for us. It’s something that matters particularly in this regime because the whole OBR model is to have a relatively small watchdog, but with legal and formal and informal right of access. If we were the Congressional Budget Office with 250 people we could probably just get on with and do it ourselves – but we’re not in that world.
Just to be clear if nothing has been done by summer in principle it’s not a good idea? Is there a window which is shut?
My current expectation is that window is likely to shut before there is agreement simply on the basis of where the parties are at the moment. If there is a change in view and they all think it’s a good idea and they can get together and be fairly clear about what role they want us to play…. this has fairly big implications for parties, how early they come up with manifesto proposals – a lot earlier than would be the case normally. That’s certainly been the lesson from the Netherlands. It does extend the whole pre-election process.
I’ve seen your hobbies listed as “hillwalking and photography”. Is that right?
I’ve no idea where that came from! Maybe from the IFS where I used to have a lot of pictures on my wall of camera snaps I’d taken of the coast road in Cornwall. I think somebody was trying desperately to find something …I did get a few emails from people saying: “Hillwalking!” We have young family which takes up a lot of leisure time. I certainly wouldn’t say hillwalking.
Your wife works for the Treasury as a senior civil servant. Do you spend a lot of time talking about the public finances over dinner?
No, it’s normally who’s picking the kids up. We don’t have any substantive professional interaction. I think there was one meeting we’ve both been present at in a sort of work capacity. That was the formal meeting with the Chancellor – there’s a whole selection of Treasury officials who come along to that.
Do you choose not to talk about work because it would be unwise?
Not really. It doesn’t arise very much as an issue because we’re [the OBR] trying to be transparent. So it’s not as though I would have any great secrets anyway. The flow of information is to go quite well. Inevitably it’s quite nice that we each know what life is like and we can sympathise accordingly. But we don’t have the need to have conversations on the work stuff.
Is it nice to get a break from spreadsheets in the evenings. Or do you relish firing up Excel?
I’m ashamed that even after eight years at the IFS that my Excel technique is now what it could be. I basically bring the same attitude to this job as I did when I was a hack and when I was at the IFS. It’s about trying to explain to people what’s going on fundamentally. You’re doing it in a different way and in a different format…That’s what I thought I was doing at the Indy, what I thought I was doing at the FT, at the IFS and the same thing here. I occasionally get the twitch of thinking “goodness me that’s a good story” and then remembering that I don’t have anyone to write it for.
You could always tell me!
Perhaps not! [But] I’m always conscious whenever we’re writing our stuff, what would I be thinking of as a consumer of this back on the papers, on IFS. Have we actually explained this, is there not a really good graph…I’ve got the great advantage in this job as in previous ones of working in an organisation with a lot of people who are a lot smarter economists than I am – what you’re bringing is a degree of external perspective. Why is what they’re saying about this now different from what they were saying about it nine months ago, or three months ago? The most important thing we do in the reports is to try to answer those questions. We now have a detailed breakdown. I hope there’s vastly more material. Almost all of that is of great interest to some people – no one, with the possible exception of my former colleagues, is deeply interested in all of it. You also want to be able to tell people why things have changed over time. That’s why there’s a lot of effort that means regarding comparisons from year to year – the Royal Mail, the APF, classification changes – you want to strip that out as well. The key thing is for people to be able to say: why is the picture of the public finances moving from year to year and why is it changing from forecast to forecast? And how transparent can you be about those two things? As a consumer at the IFS, whenever you see a particular sort of tax receipt rising significantly as a share of GDP alarm bells ought to ring. Hang on, you expect taxes to rise with the overall economy – why are we expecting to get so much more from this in five years’ time? Which is why we emphasised what is happening with stamp duties and inheritance tax.
And your transparency has put pressure on the Bank to be more transparent….
Transparency is good for accountability. It’s good for informing public debate. I think it’s also good as an internal disciple. If you know that you’re going to have to publish the details of the judgements that you make you can’t just slip in some convenient judgement that makes the numbers easier to explain just because you know no one will ever know what that judgement was. You’ve got to be able to explain why you’ve done it.
Is that something that the Bank was guilty of?
I don’t know. They were doing most of the forecast activity to support the deliberations of the MPC. It’s a different sort of organisation with different goals and different needs. Mervyn’s [King’s] views was excessively precise forecasts underplayed the uncertainty. You occasionally get: “you publish all this detail but you don’t know whether the growth rate is going to be here or there”. But I think that is different. I think from our point of view we need to show our working. We need to be quantitatively precise. The way we highlight the importance of uncertainty is in the fan charts, the sensitivity analysis. If you get more accurate fan charts out of it that’s nice and a bonus – you hope that would be the case – but fundamentally it’s about transparency.
You’re having an external review?
Yes the former head of the Canadian fiscal body, Kevin Page [the former Canadian Parliamentary Budget Officer] will be doing a review of the quality of our outputs and the process that goes into it – over the summer. It was required by our legislation – the non-executive directors had to produce it. Kevin had a proven track record of being very forthright and very independent. It had to be done within the first five years. The summer after the election would have been the last [available] time. It’s very good to have external scrutiny. Having something that’s sort of independent and peer review means it’s less likely to be used as a source of indirect political pressure than if it was a process within government.Tagged in: Office for Budget Responsibility, Robert Chote
Latest from Independent journalists on Twitter