Treasury Committee 

Oral evidence: Autumn Statement 2016, HC 837

Tuesday 29 November 2016

Ordered by the House of Commons to be published on 30 November 2016.

Watch the meeting 

Members present: Mr Andrew Tyrie (Chair); Helen Goodman; Stephen Hammond; George Kerevan; Kit Malthouse; Chris Philp; Mr Jacob Rees‑Mogg; Rachel Reeves; Wes Streeting

Questions 1-85

Witnesses

I. Paul Johnson, Director, Institute for Fiscal Studies, and Carl Emmerson, Deputy Director, Institute for Fiscal Studies.

Examination of Witnesses

Witnesses: Paul Johnson and Carl Emmerson.

Q1                Chair: Thank you very much for coming to give evidence to us this morning.  This is the first of a number of hearings we are going to be having examining the Autumn Statement in some detail, and then we will report on it at the end of the period.  I would just like to begin, first of all, by asking you, Mr Johnson, how much value you attach to the latest guidance we are being given—targets, I do not know which word you fancy, rule—that we now have on fiscal policy and the direction of fiscal policy.  This is the latest of at least a dozen that I have counted up since 1997.  In fact, I got to 13.  What about you?  How many have you spotted and how much value do you think we should attach to these rules?

Paul Johnson: We have come up with 12, so you may have found one more than we have and we reckon that 10 of those have been broken or abandoned at some point.  That said, I am not wholly cynical about the usefulness of fiscal rules.  They provide a degree of guidance about what the Government’s intention is.  Now, clearly you can make that guidance more specific or less specific, and you need to get a balance between that.  Obviously the more specific it is, the more you are constrained and the more difficult it is to respond to economic uncertainty.  Of course, we are in a period of some economic uncertainty going forward.

Q2                Chair: It is very difficult to be subject to uncertainty going backwards, but anyway, apart from that point, which is very pernickety, presumably when you use the word “some” you really mean a great deal, because there is much more than usual. 

Paul Johnson: There are known unknowns, as it were.  As it turned out, we were subject to vast uncertainty in 2007 that we just did not know about.  Actually, even in 2010, most forecasts were thinking the economy was going to grow much more strongly than it turned out to do.  If you look back, you will see that there were periods when outturns were vastly different from expectations. 

What is slightly different now is that we know that there are certain decision points, for example whether we stay in the single market or not, and indeed uncertainties about how the economy will respond over the next couple of years, which we know are uncertain.  I do not think it is within most people’s band of uncertainty to have anything close to what happened in 2008.  It is just that we know that there are some changes coming along, but we do not really know exactly how the economy will respond to them. 

Q3                Chair: Is this new guidance that we are being given of more value or less value than the average of the last dozen?

Paul Johnson: It is of some value.  I am not sure whether it is of more or less than what we have had before.  What it is doing, first of all, is giving the Chancellor a reasonable amount of wriggle room.  To get to a deficit of 2% of national income, which is what he is targeting by the end of this Parliament, we are aiming at less than 1% at the moment, so there is a reasonable amount of headroom there.  It allows him to be flexible in the face of some small‑ish economic shocks but, if there is a big one, if we lose 2% of national income or thereabouts, we are going to start to hit that headroom. 

With the existence of this target—and this is where we get into the question of how valuable it is—if the economy does that much worse, will he really try to rein in spending or increase taxes to achieve it?  I do not know.  In addition, of course, he has let himself a trapdoor here, which is to say that, if in the view of the Treasury this time—not in the view of the OBR, but in the view of the Treasury or the Chancellor—there has been a significant negative economic shock, then he can suspend them.  What this is doing is giving us a sense that, if the economy goes along broadly as expected with some wriggle room either way, this is what we are going to aim for but, fairly clearly, if it does not, then we are going to try something else.

Chair: The short answer to my question is, as you put it, you do not know, but it is has a non‑negligible value, even though it is extremely low.

Paul Johnson: It may even be not extremely low.  If you look at previous fiscal rules, even where they have eventually been broken, they have clearly constrained behaviour while they have been there, even to the extent that they may have been gamed.

Chair: That is the key point.  That is very helpful.  Chris Philp.

Q4                Chris Philp: Thank you, Chairman.  I was going to ask about fiscal rules as well.  The Chairman has already alluded to the large number of fiscal rules that have been implemented and then discarded when they have been inconvenient.  I have heard the answer you just gave but, given that successive Governments have essentially discarded fiscal rules or rewritten fiscal rules, as soon as they have become inconvenient or unachievable, do they really have any value, notwithstanding Paul’s comments a moment ago?

Carl Emmerson: I think they do.  It is important to remember that we do not have an optimal rule that you should obey at all periods of time.  In practical terms that means that we should not be surprised if, occasionally, the right thing to do is to get rid of your rule.  In 2008, the Government were right to get rid of its fiscal rule.

Chris Philp: It is not occasionally, is it?  It is 12 or 13 times.

Carl Emmerson: Occasionally it would be right to do it.  On other occasions, we have just had bad rules and we should not have had those rules in the first place.  In 2008, the Government were right to get rid of its “golden rule” and it was the same with the investment rule.  I believe the Chancellor was right to abandon the target to get rid of the deficit in 2019‑20, given the uncertainty that is ahead of us.  What we need to have is a) some rules that give you some flexibility and b) an appreciation that something bad might come along that means the right thing to do is to get rid of the rule.  That does not mean that the rule has no value.

The two rules that were obeyed or kept to, which I think have a lot of attractions, were Mr Osborne’s rules during the coalition years, where he said, “I will aim to balance the current budget X years out.”  Critics of that do not like the fact that X years out is always X years away; you never actually get there.  A nice feature of that rule is that, if some shock comes along, it gives you a bit of time to adjust.  You do not have to do immediate tax rises or spending cuts.  You can plan and adjust gradually over time, and that is very attractive.  It is perhaps not an accident that they were the two rules that were not breached while they were in operation.

Chris Philp: It was because they were the loosest.

Carl Emmerson: They were not the loosest.  There were ones that just allowed you more scope and flexibility.  As I say, if you have a rule that says, for example, we will eliminate the deficit in 2019‑20, that means you could start that financial year on course to meet it.  You could get to Christmas and suddenly realise you are off course.  You have to announce an emergency budget if you take that rule seriously, which I do not think is credible. 

Q5                Chris Philp: Of course, as you said, we did have the target of reaching fiscal breakeven or eliminating the deficit by 2019‑20.  That date has now gone and there is no specific date replacing it.  The Chancellor said, “at some point in the next Parliament”, which potentially covers five years.  The range is not a short period of time; it is five years.  Do you think that that lack of specificity is problematic?

Carl Emmerson: A better way forwards would be to say, supposing we do want to get to budget balance, we will set policy consistent with being there in five, six, seven or however many years you think it is appropriate to adjust, given the amount of uncertainty there is and where we are starting from.  Fixing the number of years, but always having it in a forward‑looking way would be the right thing to do.  What I think is particularly unattractive about saying “at some point in the next Parliament” is it is determine by factors other than pure economics and fiscal policy.  We do not know when the next election will happen and we do not know when the election after that will happen, so we might end up having to balance the budget by 2020, if there were two elections in quick succession, which again I do not think is credible.

Q6                Chris Philp: Implicit in his comment was the assumption that there would be an election in 2020 and an election in 2025.  Mr Johnson, do you think the UK budget currently suffers from excessive hypothecation, so forward commitment, in certain areas?  For example in the tax base, we have committed not to change to increase VAT and income tax.  We have committed to raise the thresholds to £12,500 and £50,000 respectively.  We have a glide path for corporation tax coming down to 17% and, on the spending side, huge portions of the spending budget are ring‑fenced: the schools budget, the NHS budget, the defence budget at 2% of GDP and the overseas aid budget at 0.7% of GNI.  Do you think that successive Chancellors, by adding these levels of commitment, both on the tax side and on the spending side, have actually given them very little room for manoeuvre when it comes to changing the fiscal outcome?

Paul Johnson: There is a lack of room for manoeuvre there.  The three biggest taxes, as you said—NI, income tax and VAT—are fixed in terms of their main rates.  If you did want to raise significant additional revenue going forward, then you have to fiddle around with either some of the minor taxes or some particular aspects of those, and that may not be the most economically efficient way of raising tax. 

That said, clearly there have always historically been political constraints around extending the VAT base or raising the main rates of income tax.  It has not happened ever so often in the past.  The formal constraint and the political constraint may not be ever so different.  If you are looking at the spending side of things, again you are right: there are ring‑fences around large fractions of spending, particularly if you include state pension spending within that.  Again on things like the NHS the ring‑fence is not generous.  It is not at a level where you are thinking, “The NHS is getting huge amounts of money relative to other departments.”  Again, whether the formal ring‑fence makes a big difference to outcomes I do not know. 

The big choices here are about what fraction of national income you want to take in tax.  If we do not want to take a much bigger fraction than we do at the moment, then the impact of those ring‑fences is probably relatively small.  Equally, if we want to be on the overall spending glide path that we are on, other than the effective ring‑fence around state pensions and possibly overseas aid, I do not think the impact is terribly big.

Carl Emmerson: One additional point on that is that one thing we are balancing off is whether providing some certainty is helpful.  You could imagine for corporation tax announcing in advance what you are going to do, if you are sure you are going to do it, might make the policy even better, because it allows companies to invest with the certainty of what that environment is going to be like.  It is not that complete flexibility is necessarily the right thing to do.  There may be some areas where some certainty is a good thing to provide.  Certainty might not keep things exactly as they are at the moment.  It might be that, for example, we are going to cut the corporation tax rate.

Q7                Chris Philp: You have mentioned a couple of areas where savings could be achieved by breaking current ring‑fences.  You specifically mentioned the triple lock on pensions and overseas aid.  If you were going to look for spending savings in such a way that it minimised the effect on public services and on the economy more widely, which areas would you recommend the Chancellor looks at first?

Paul Johnson: There are obviously a whole series of political choices in there about where you would want to make savings, but you have said to protect public services, so that excludes all of public service spending.  That essentially only leaves spending on benefits, which are pensioner benefits and benefits for those at working age.  We know that benefits for those of working age are already looking to be cut reasonably significantly over the next four or five years, whereas those for pension age, on a whole, have been left untouched.  I think it is a political decision.  If you are excluding public services and you know that a significant chunk of welfare has been cut, there is not very much left.

Q8                Chris Philp: Are you surprised that the UK Government has stuck by its 0.7% of GNI commitment on overseas aid, when other developed countries spend much less?  For example, the USA is at 0.2%; France and Germany, at 0.3%.  As you know, the OECD average is 0.28% and the EU average is 0.45%. 

Paul Johnson: Again, it is a political choice for the Government how much they want to spend on that.

Chris Philp: I asked if you are surprised.  Are you surprised?

Paul Johnson: It was in the manifesto.  Actually, I am not so cynical that I am surprised when Governments take their manifestos seriously because, in general, they do tend to take their manifestos reasonably seriously.

Q9                Chris Philp: There is a technical change that has been made to the definition of public sector net debt, which is part of the reason why the forecast has bumped up to around 90% in 2019‑20.  When you look at the Bank of England’s monetary stimulus, it looks like we are including the liabilities that have been created, but excluding the assets, loans to banks and so on, because they are allegedly illiquid, despite the fact that they are of high grade.  Do you agree with the ONS’s definitional adjustment?

Carl Emmerson: The ONS here will be following international accounting rules about what counts as a liquid asset and what does not, so I am not going to dispute that they have made the right decision.  However, in terms of the economics of it, if you are creating some liabilities in order to make a loan to the financial sector for just four years and you reasonably expect to get that money back, the economics say you should think about the value of that asset that you have just got as well.  It would be more sensible to consider a measure of debt that excludes both the liabilities and the assets, or includes both.  Either way, you essentially get to the same picture. 

It is an unfortunate thing that the Government’s target looks at the measure of debt, which includes the Bank of England effect.  It would be better to exclude it for two reasons.  First, they have picked a target that should fall in 2020‑21, which happens to be just when the financial sector will hopefully—touch wood—be paying back those loans, so the measure of debt will naturally fall very sharply in that year, so it is quite an easy target to meet.  Secondly, it is odd to include a target for something that can be affected by things; supposing the MPC decided to do more of this in 2019, say.  It would make this target very difficult to meet for reasons that are not really about the total indebtedness of the public sector.  It would be better to target a measure of debt that does not have this feature. 

Q10            Chris Philp: Briefly before finishing, I was on Sky News the other evening discussing your report on the outlook for wage growth.  Despite the fact that, actually, wage growth has been quite strong recently, particularly for low earners, you had produced a very gloomy forecast suggesting that real wages are going to be lower in 2020 than they were in 2010.  That flies in the face of what has actually happened in the last two or three years, does it not?

Paul Johnson: The forecast is actually the OBR’s forecast.  Their forecast for real wages in 2020 are that they will be slightly below where they were in 2008.  We are at a position now where median wages are, I think, 6% or 7% lower than they were in 2008.  That is where we are now and that is a pretty sad state of affairs really, compared to where earnings tend to go over time.  That is still the cost of the financial crisis in 2008.  That is despite a year or two of reasonably decent wage growth, because there was such an initial shock.  The OBR figures are now for very modest growth actually, over the next few years.  The next few years look better than the last six or seven on those numbers, but it is still not enough to make up the losses that we have seen over the last seven or eight years.

Chris Philp: They occurred particularly in 2008‑09, you say.

Paul Johnson: There was a big drop immediately after the financial crisis.  What is particularly surprising about this period of recovery relative to most previous periods of recovery is that wage growth has been very, very slow, certainly up until a year or so.

Q11            Chris Philp: Finally, is there a growing divergence between the median and the mean?

Paul Johnson: No, there is not actually.  I cannot remember the exact numbers but, in this period, we have not seen a significant growth in wage inequality since 2008.  In particular at the bottom, because the minimum wage has been protected, it has done better than both median and mean earnings.  Certainly going forward the National Living Wage will do an awful lot better.  Over a large chunk of the distribution, at least you have seen a compression, rather than the reverse.

Chair: We are going to go into that in more detail later on.

Q12            Kit Malthouse: Good morning.  I just want to ask you some questions about the OBR forecast itself.  My reading of it is that they are forecasting a sudden drop in GDP growth over the next two quarters, down to the point at which Article 50 is likely to be triggered in March, and then a sudden rise in GDP growth over the next three quarters.  Effectively, by the time we get to 2018‑19, things are back to normal.  Given the pattern of GDP growth over the previous three quarters, i.e. the three quarters that took in the vote of the referendum itself, does that seem rational to you?

Paul Johnson: I do not know whether this is a correct forecast.  It is broadly in line with what most other forecasters are saying and it reflects a view that the increased levels of uncertainty will have some impact on behaviour.  It is only when we really get some hard data, credible data on things like investment, that we are going to be able to answer that.  The short answer, in a sense, is that you need to really cross‑examine the OBR on what they have done.  I do not have a strong view.  I really do not have a strong view either way on whether what they are forecasting over the next couple of quarters is better or less good than what other people are forecasting.

Q13            Kit Malthouse: I am just asking you whether your professional opinion of their assumptions is that they are doing the right thing.  What they seem to be implying in the forecast is that the point of maximum uncertainty, and therefore minimum GDP growth, is the technical point at which Article 50 is triggered.  I would have thought the point of maximum uncertainty would have been the point of the referendum.

Carl Emmerson: It is not just that though.  There is also the depreciation of sterling pushing up prices, which is bearing down on consumer spending.  That clearly happens with a little bit of a lag after sterling falls, because it takes a little bit of time before it gets passed through into higher consumer prices.  It is not just uncertainty; there is also an effect on consumer spending.

Kit Malthouse: You think that, post Article 50 being triggered, the pound will start to rise.

Carl Emmerson: No.

Q14            Kit Malthouse: It seems just a very odd point.  The point I am trying to make is that it strikes me as an odd point to pick, this technical point at which a piece of paper is passed between two Governments.  That is the point of maximum pain, if you like, if that is what they are implying.

Paul Johnson: I do not know, but I doubt that their forecast is based on that.  We know that that is the point of the triggering and, therefore, we can assume this amount of GDP growth.  My guess is that this reflects over things that they think about the direction of the economy, but I would not claim to be an expert in exactly how they have come to that conclusion.

Kit Malthouse: It could just be coincidental is what you are saying.

Paul Johnson: It may well be that they would have had exactly the same forecast, even if the Prime Minister had said that it would be triggered in January or July.  I do not know. 

Q15            Kit Malthouse: They are saying that in quarter four this year, i.e. the quarter that we are now two thirds of the way through, GDP growth is going to drop by one tenth of a percentage point down to 3.6% from 0.5%.  From the indicators that are coming through so far in the quarter, would you hazard a view that it does not feel like that?

Paul Johnson: Honestly, I would not hazard any view about exactly what the current numbers are saying about this quarter’s GDP growth.

Q16            Kit Malthouse: In the opinions you are giving on the spending and all the other bits and pieces that are coming through, you are discounting the forecast, it sounds to me.  What purpose does the forecast perform for you?

Paul Johnson: The overall economic forecasts, first of all, are necessary for having any kind of public finance forecast.  Public finances depend on the path of the economy.  The OBR overall forecasts, over the next few years, are broadly similar to those of other forecasters including the Bank, if anything a little bit more optimistic.  The focus of the public finances is much less on what is the exact path over that period, and more what is the broad rate of growth over that period.  A key point, which we have discussed earlier, is that we know that there is a great deal of uncertainty over that.  The OBR has to, by law, provide a single central forecast over that period but, as they say and we certainly make clear in everything we say, there is a great deal of uncertainty about whether things will turn out somewhat better or somewhat worse than that central forecast.

Carl Emmerson: To quantify that uncertainty, the Government have about a two‑thirds chance of meeting its target to have a structural deficit of no more than 2% of GDP in 2020‑21.  There is a one‑in‑four chance that, to meet that target, you would need to do more tax rises or spending cuts in the current Parliament but, equivalently, the uncertainty is so big that there is a one‑in‑three chance that we will actually have a headline surplus in 2019‑20, the target that the Conservative Party manifesto committed to.  The uncertainty is so big that we should hope that things will turn out much better than the OBR expects, but we have to operate policy on the basis that we do not know what the future is going to look like.  We are going to have to be a bit nimble. 

Paul Johnson: That measure of uncertainty is based on previous forecast areas.  As we say, there have been periods of much greater ex post uncertainty, but this is a period where there is some ex ante uncertainty.

Q17            Kit Malthouse: There seems to be a general assumption in the economic forecasting profession that there is this uncertainty.  You are not quite sure when it is going to arrive, but it is definitely there somehow.  It did not arrive during the referendum period and yet this assumption still persists.  Do you think that says something about the forecasting community, as much as the forecast itself?

Paul Johnson: Things take a little while to happen.  We know that some measures of uncertainty have risen.  Clearly firms’ investment decisions and so on take a while to come through.  To be absolutely clear, the set of numbers to look for, over the next two years, in terms of understanding the impact on uncertainty are going to be private sector investment numbers.  When we get credible versions of those, we will know a lot more about what is actually happening.

Q18            Kit Malthouse: This is just the final thing on this.  It could be that we get to the end of this quarter and private sector investment numbers look pretty good.  We have had some big announcements of private sector investment—the forecast investment from Nissan, Google, JLR and all the rest of them that are planning to invest.  It looks pretty good.  This dip may not materialise.

Paul Johnson: As we said, there is uncertainty about it.  It may not materialise.  Let us hope so.

Kit Malthouse: Presumably the purpose of the forecast is to allow the Chancellor to do something about it.

Paul Johnson: He needs an operating assumption based on the best forecast that you can make, for making all of his policies and, indeed, to prepare responses if necessary.

Q19            Kit Malthouse: Could I just ask you some questions about inflation as well?  Do you think that the drop in interest rate earlier this year contributed to a drop in the exchange rate and, therefore, this implied inflation that is going to come through in import prices?

Paul Johnson: I do not know how important the drop in the interest rate was.  We know that there was an immediate sharp depreciation on the morning of 24 June.  The combination of that effect and any effect that the interest rate may have had has been that the pound is whatever it is, 15% below where it was six months ago.  Simple arithmetic suggests that that will increase the price level, in the medium run, by 2%.

Kit Malthouse: In old money, if you cut your interest rate, it generally had an impact on your exchange rate as well.

Paul Johnson: Indeed.

Kit Malthouse: It would be fair to assume that part of the fall in the pound might be caused by the cut in the interest rate.

Paul Johnson: Quite possibly.  Again, we are probably not the right people to ask about the macroeconomics of monetary policy.  Clearly the Bank did that for a reason and you need to think about what the counterfactual would be.

Q20            Kit Malthouse: The other thing I was going to ask you about was whether you have done any work on or had a look at the effect of the money supply on the inflation equation.

Paul Johnson: The answer to that is no, we have not.

Chair: That is because you do not consider yourself a macroeconomic forecaster.

Paul Johnson: No, and we are certainly not monetary economists, I am afraid. 

Q21            Stephen Hammond: Can we turn to the National Productivity Investment Fund?  The Chancellor set out that raising productivity was a “central challenge”.  I have three questions on it that stem from that fairly immediately.  One is whether that is the right central challenge and if that is a sensible measure of economic success.  Secondly, if that is the central challenge, has the Chancellor positioned the Autumn Statement correctly?  Probably first, although conversely I have asked it last, do you as an institution think there is a strong link between infrastructure spending, productivity and growth?

Paul Johnson: There is certainly a link between infrastructure, productivity and growth, and it is a long‑run link.  It is not a link that says, if we spend £5 billion more this year, productivity and growth will with certainty be significantly bigger next year, but we do know, for example, that particularly for some of the projects that the Chancellor announced, in terms of relatively small, well focused road building and so on, the economics benefits of those are reasonably significant.  Over the long run, I believe that the Government can make a difference to productivity through well targeted investment growth and all sorts of other things—the way that they regulate competition, the way that planning rules work, the way that the education system works and the way that the tax system works. 

All of these things have real impacts on productivity and growth, in the long run.  Governments can make quite a significant difference to productivity and growth in the long run.  Part of that is infrastructure.  An important part of that are things like research and development, and the sorts of road‑building‑type schemes.  I could not comment on any of the specific ones, but the Chancellor announced those sorts of things.  As part of a programme of policy to support productivity over the medium to long run that feels to me like a sensible element of such a strategy.

Carl Emmerson: In addition to that, often in these kinds of statements we hear announcements about capital spending, which to be honest do not really turn out to be that big.  I think the Government’s capital spending plans now, if they are delivered and the money is well spent, are reasonably significant.  They are looking to get public sector net investment up to about 2.25% of GDP.  That, for example, is around the level that the Labour Government were aiming for prior to the financial crisis.  If spending is got up to that level and maintained there, it will be far bigger than the amount we have tended to invest over the last 30 years, so it is a difference from the past. 

Q22            Stephen Hammond: To pick up what you have just said, Mr Emmerson, can we dig into the mix, because the mix is interesting?  By 2021‑22, 49% of the fund will have been allocated to housing, 18% to transport and 4% to digital communications.  Actually, a lot of that housing is effectively front‑end‑loaded, so it beggars the question, firstly, if that is the right sort of mix for the productivity challenge that the Government have set itself.  Secondly, the Government set themselves a project of finding projects that demonstrate a clear and strong contribution to economic growth.  Within that mix, are they likely to achieve those aims?

Paul Johnson: It is a reasonable mix.  Whether it is precisely the right mix, I do not know.  Research and development is a significant part of the mix here.  We still as a country spend significantly less than the G7/G20 average on research and development, both by government and by companies.  Again, the macro evidence is that that matters for growth in the long run.  We clearly have housing constraints in the faster‑growing bits of the UK.  Again, more investment there is an important part of growth in the long run.  Certainly for the type of road schemes, again we know that problems on the road create issues.  Is this exactly the right range of projects?  I do not know.  Is it a sensible mix?  Yes.

Stephen Hammond: Allocating half the fund to housing is a sensible mix.  It is 49% for 2021‑22.

Paul Johnson: I do not know; it is roughly that.  It is a reasonable shot.

Q23            Stephen Hammond: There are a number of questions from your response.  The first question is: if it is true that we are significantly behind in R&D spending, as indeed we are behind Germany, France, USA, Japan, etc, might not some targeted measures on patient capital, EIS and encouraging universities to establish accelerators as well as incubators be equally as effective as anything on Government spending?

Paul Johnson: I do not know how this Government spending is going to be spent but, yes, all of the things that you suggest would be sensible parts of a strategy.

Q24            Stephen Hammond: Are you concerned, as part of the productivity strategy, about the spending on skills and raising the skills mix, because you have seen over the last three years that a real decline in spending in DoE versus DfT has been significant?  Would you question what the Government should be doing in terms of skills?

Paul Johnson: Clearly skills is an absolutely crucial part of productivity going forward.  There are two issues there. One is what is the set of things that we should be doing and how we should be doing them.  The other is how much we should be spending on them.  Both of those are very big questions.  There are questions throughout the education system there, literally from whether we are teaching the right things in schools to whether we have the right routes through from 16 to 18, for both those going on to academic study and, in particular, those going on into apprenticeships and vocational education.  Do we have the right mix of things that people are doing at universities?  We know that there are quite a lot of people and quite a lot of courses at universities, who go on really not to do terribly well in the labour market, even if an average people are doing very well. 

For a hundred years, we have had a problem with vocational skills, the structure of apprenticeships and the institutions around those.  I do not pretend to have answers to those questions, but absolutely a focus on that will be completely crucial to where we go, going forward.  I have to say, of all the things that we talk about, in terms of what the set of levers is that Government can pull to improve productivity, we pretty much know, broadly speaking, what you need to do with roads and infrastructure, the tax system, competition, planning and all those sorts of things.  There are trade‑offs, and sometimes we do them and sometimes we do not.  Genuinely, we do not know exactly what it is that you need to do in practical terms to make the education system work better.

Q25            Stephen Hammond: There must be a question that if we cut, in real terms, the spending in particular targeted areas of raising skills, it is unlikely to be helpful. 

Paul Johnson: The spending on 16‑18 and further education has been cut really quite significantly, over the last few years, in a way that spending on schools and total spending on universities have not.

Stephen Hammond: That is likely to be a productivity issue in the future.

Paul Johnson: It could well be.

Q26            Stephen Hammond: Futurologists would say that the most important things of infrastructure spending ought to be the commitment to digital communications and upgrading that across the country, so that you concentrate on some of the areas of competitive advantage that we undoubtedly have in the UK, but also a redistribution of activity across the UK.  Do you think the level at 4% is the right sort of level? 

Paul Johnson: Personally I do not have a view on that at all.  I cannot answer that.  Sorry, I just do not know.

Chair: Whenever I hear about futurologists, my first thoughts are normally Mystic Meg, but you had something more serious. 

Q27            Stephen Hammond: You could well argue that, if we spread digital communications to rural areas, it could have quite a significant impact on the ability for businesses to be creative in those areas.  It would have a significant impact on whether or not you could put transport or housing into those areas.  It is a really important issue and there is considerable concern that, although the Government are committed to doing it, the reality of what is happening is that it is very slow and this is really too small. 

I have two final questions.  Sorry, I realise that colleagues want to get in.  You mentioned that the OBR had said that the National Productivity Investment Fund will lift PSNI to 2.3%.  OECD estimates that annual infrastructure investment ought to be 3.5% of GDP in developed economies.  What would you as the IFS judge as the right level of PSNI?

Paul Johnson: With all these things, there is not a right level.  There is a judgment about how much you need.

Stephen Hammond: What level would you consider appropriate for the UK economy facing the challenges it does at the moment?

Paul Johnson: For example, I do not know what the OECD counts as infrastructure investment, because there is quite a lot of infrastructure investment that happens in the private sector in the UK, which might happen in the public sector elsewhere.  The second point to say is that what Mr Hammond did last week was balancing what is actually feasible to deliver in the short run, with where he may want to get to in the long run.  What we do know from all of the last 20 years is that Government really struggle to ramp up infrastructure spending at all quickly.  Indeed, you will see the OBR figures assume that, even with this level of planned spending, the expectation is that we will underspend relative to plans.  I really do not know whether the right answer is 2%, 2.5% or 3% of national income. 

Q28            Stephen Hammond: My final question is we have spoken a number of times about the international comparability of productivity.  Do you think you are using the right definitions of workforce internationally or are they so different internationally that international comparisons are worthless?

Paul Johnson: You get different numbers according to whether you look at productivity per hour worked, productivity per person in work or productivity per person of working age.  Now, we do a lot better on productivity per person of working age than we do on productivity per hour worked, because we just have a lot more people in work.  That fact is rather important.  You at least need to show the two things together, because where you have countries that have a lot more people not in work then the average productivity over everyone is lower than on the other measure and that really matters.  At the very least, you need to keep both those measures in mind.

Q29            Chair: We had a brief word about this and I forewarned you about the likelihood that you would be asked questions in this field before this meeting.  I suggested then and I will repeat now, and hope to get on the record, that I think it would be extremely useful if the IFS takes a look at measures of productivity to see how credible these bald statements that our performance is shocking compared to other countries, and has actually been for decades, really are, how valuable these statistics are in measuring productivity and how useful productivity is as a guide for policy.

Paul Johnson: I am aware of work that has been done that tries to work out how the differences in the numbers are made up, what bits of it are real and what bits are statistical.

Q30            Chair: Perhaps you would like to make us aware of it all and do a piece of work for us.  We would be extremely interested to take a look at what you think the quality of productivity statistics is and how much weight we should attach the numbers and their composition.  Would you do that?

Paul Johnson: We can have a look, yes.

Chair: There are measurement issues on the labour inputs, which is what Stephen just referred to, and you have responded.  Also, in international comparisons there is considerable complexity in dealing with exchange rate fluctuations.  Then there are big cultural differences.  Some societies like a couple of months’ holiday a year, if they can get away with it, and some societies think a week will do.  This means that international comparisons are made much more complex, but we will put the ball in your court and await something back from you. 

Rachel Reeves: I believe Rupert Harrison used to do the work for the IFS on productivity.  Maybe you could see if you could get him back. 

Paul Johnson: I do not think we could afford him anymore.

Q31            Rachel Reeves: You said that, not me.  I wanted to ask you both about something that was not mentioned in the Autumn Statement, which was the National Health Service and spending on social care.  Mr Johnson, you said on Thursday that “Strikingly [the Chancellor] responded not at all to calls for more money for either the NHS or social care.  I’m going to stick my neck out and suggest he won’t be able to do that for much longer.”  Can you explain what you mean by that?

Paul Johnson: Yes.  There are two parts to that.  Spending on health over the next three years is due to rise by a very small amount, 1% to 2% over that period, relative to an average of 3% to 4% a year over the period of the NHS’s existence.  That is following a period of much smaller than average increases in health spending, at a period when the population is both growing and aging.  The pressures there, just looking from a top‑down point of view, look like they are going to be reasonably significant as demand grows. 

My sense from talking to people in the Health Service is that that is tough, but the thing that they are really struggling with is the fact that spending on social care has been cut, and I forget the precise numbers, but significantly because of the squeeze on spending in local government.  This is anecdote, but that is anecdotally putting significant additional pressure on the NHS.  This is an area of public service where demand is growing and where spending has not kept up, either with that demand or with the usual increases in spending over time.  Now, I am not saying for sure that that is going to result in a movement in policy, but the pressures there are, without question, significant.

Q32            Rachel Reeves: To keep up with rising demand, how much do you think NHS spending would need to increase by, per year?

Paul Johnson: We have history, which tells us that the number looks like 3% or 4%.  I am afraid I do not have the population numbers in my head, but we did some work a year or two ago and, if you look at just what has happened to the age and size of the population between 2010 and 2020, it implies that the overall growth in spending per age‑adjusted person is very small, 1% or 2% over that period.  Do not quote me on that, but it is a very small growth over that period.  On top of that, clearly there are the usual costs associated with demand, which appears to increase year by year irrespective of the size and age of the population, and the costs of drugs, technology and all those kinds of things.  I do not know what the number is, but it is clearly a service that will be under some stress by 2020, relative to where it was in 2010.

Q33            Rachel Reeves: Would you suggest that about 3% to 4% per year would be needed to keep up with rising demand?

Paul Johnson: That is certainly been the historical experience.  There is clearly an issue here, which is that you cannot increase spending on health faster than the rate of growth of the economy forever, otherwise it ends up taking the entire economy.  There has to be a point at which that level of increase has to tail off.  Where that point is I do not know. 

Q34            Rachel Reeves: Many people are saying, and you have just suggested as well, Mr Johnson, that spending on social care reduces costs for the National Health Service.  Has the IFS done any analysis on the savings to the NHS through spending on social care?

Paul Johnson: We are currently doing that.  In terms of getting a solid answer to that question, it is a long‑term research project and we are currently embarking on such a project, but we are not going to have an answer within the next several months, I am afraid.

Q35            Rachel Reeves: We will look forward to seeing that.  You said, as I remarked at the beginning, it was a surprise that the Chancellor did not announce any additional money for the NHS or social care.  Indeed, there was no mention of the NHS or social care in the 72‑page document.  Do you also think it was perhaps a surprise to people who had voted to leave the European Union that there was not an extra £350 million a week for the National Health Service?

Paul Johnson: We are not out of the European Union yet.  To the extent that we will be saving money from what we spend at the moment that will not come until at least 2019.

Q36            Rachel Reeves: Do you think people should look forward to a £350 million windfall for the National Health Service in March 2019?

Paul Johnson: It is in the Chancellor’s court as to how much he decides to spend on the NHS.  As we have said, the simple fact of leaving the European Union is not going to free up anything like £350 million a week to spend on anything.

Rachel Reeves: You think it is unlikely that, in March 2019, leaving the European Union would free up £350 million a week for the National Health Service.

Paul Johnson: Leaving the European Union in itself will not free up £350 million a week.  That does not mean that the Chancellor cannot find that within his new fiscal headroom.  He can find out within his new fiscal headroom, also he is able to spend that money, but he will not be able to spend that because we will have left the European Union.

Q37            Rachel Reeves: The Health Select Committee report said that the £10 billion given to the NHS was, in real terms, less than £10 billion for a variety of reasons: money going to Health Education England, money going to public health, etc.  Do you think that the Health Select Committee is right that the £10 billion that the Government say they have given to the NHS is, in fact, less than £10 billion, Mr Emmerson?

Carl Emmerson: This is about differences in definitions really.  The Government have increased the NHS budget by the amounts that it pledged to do in the manifesto, and I believe £10 billion is the number, although that is expressed in 2020 prices, which clearly makes it sound a little bit bigger than it would be if you put it in today’s prices.  Essentially, the firm commitment there is on course to be met. 

A broader measure of health spending, for example including what the Department of Health does outside of the NHS budget, is actually not going up as fast, so essentially what the Government have done is cut some things outside of the NHS, but within the health budget, to pay for those increases in the NHS budget.  The bigger issue here is really about how much money we are putting into the NHS.  As Paul says, there is a very small increase over the next few years.  It may meet the Government’s manifesto pledge, but it is much, much smaller than the NHS has been used to in its history and it is not enough to keep up with the growth in the population and the aging of that population.  Therefore, either you have to see a big increase in productivity or we should not be surprised if the NHS struggles.

Q38            Rachel Reeves: That is the point that Mr Johnson said in answer to my previous question.  Historically, NHS spending has gone up by 3% to 4% a year to meet rising.  Specifically, this £10 billion number is quite important.  In real terms, you have said that money has been moved from one budget to another, rather than there being additional money.  What additional money, in real terms, do you think there has been, Mr Emmerson?

Carl Emmerson: The Department of Health budget overall in increasing by a little over 1% between this current financial year and 2019‑20.

Q39            Rachel Reeves: What is that in pounds?

Carl Emmerson: I am afraid I do not have the number off the top of my head.  It is certainly a lot smaller than £10 billion.

Q40            Rachel Reeves: The Health Select Committee suggests that increases in health funding are closer to £6 billion.  Does that sound about right to you? 

Carl Emmerson: It sounds quite plausible.  It may even be smaller than that. 

Paul Johnson: There is a difference between NHS spending, of course, and health spending.  The key point is, as far as we can tell, the Government are meeting its commitment to increase NHS spending by a particular amount, but some of that money is coming from other bits of the health budget. 

Q41            George Kerevan: Good morning.  The Chancellor was remarkably candid when he said that he was increasing the Insurance Premium Tax in order to fund credits made in the Autumn Statement.  Why did he pick on IPT, do you think?

Paul Johnson: I will try to speculate as to why he chose that one.  It is not one of the taxes that the manifesto stops him from raising and, as we discussed, he is very much constrained in terms of what he can do, in terms of increasing some of the big taxes.  It is also a tax that has been raised a number of times in recent years.  It was initially introduced at 2% or 2.5%.  It was at 2.5% and has been increased several times since 2010.

Q42            George Kerevan: It was three times.  I noticed that, in the OBR discussion of revenue that would be forthcoming from the rise, which is about £1 billion over the review period, there is a caveat that says that the extra revenue might be offset by increased tax planning activity by insurance companies.  Would you be able to enlighten us as to what that might mean or have we caught you out on that one?

Paul Johnson: Not off the top of my head anyway.

Q43            George Kerevan: Has the Institute done any kind of analysis about the kind of revenue streams that we might get from IPT and what constrains them?

Carl Emmerson: No, but when looking at IPT, what we have said is that we would like to have VAT on insurance.  The value added for insurance is the difference between what you pay for insurance and the premiums that are paid out.  If we do not have VAT, you may wish to have an insurance premium tax to try to proxy for that, but setting it at a rate of 12% is far in excess of what it should be, given you have a VAT rate of 20%.  Most of what you pay in premiums to your insurance company will be paid out by them, so the value added is nowhere near great enough to justify an IPT rate of 12%.  It simply should be set at a lower rate.

Q44            George Kerevan: I am interested in this because if, as we suspect, he is raising IPT simply because he can do it and he is constrained elsewhere, rather than do it ad hoc, we should try to look at some of the aspects behind it.  Going in your direction, why do we not just put VAT on insurance as a service?  Why is there a separate tax regime?

Carl Emmerson: It is very difficult to try to measure exactly what the value added is with any financial service.  If you decide that that is too complicated and you do not go down that route, you might decide that you have an IPT as a proxy.  If you do go down that route, which is not crazy, you should have a rate that is much lower than 12%.  It has just got too high now. 

Paul Johnson: I think insurance fits within the definition of financial services that are constrained by European Union law, which says that financial services have to be exempt from VAT.  I am not 100% sure of this but, if insurance fits under that definition, then we are not actually able to put VAT on it at the moment.

Q45            George Kerevan: That is at the moment.  Let us go back.  If we say, therefore, we are constrained from a VAT approach, surely there is therefore an anomaly in the way IPT is being applied.  It is simply an indirect taxation on business, whereas a VAT approach at least would be deductible.  How would we assess the kind of complication that that introduces?

Paul Johnson: In the long run, it will reduce the amount of business that is highly dependent on buying insurance, because it is a non‑reclaimable tax.  It is not a tax like a tax on petrol, which is there to meet an externality.  It is not paying a social cost.  You can scale it.  You can say how much the cost to business is.  The scale of the impact on the structure of the economy is very hard to estimate but it is, as with all of these intermediate taxes, an inefficient tax from an economic view. 

George Kerevan: It causes a distortion in the market.

Paul Johnson: How big that distortion is I could not tell you.

Q46            George Kerevan: Do you know of anybody who has done any kind of analysis of how big that distortion might be?

Paul Johnson: I am not aware of it, but we could look.

Carl Emmerson: We would need to have a look.  Clearly it was a distortion even when IPT was at 2.5%, but you just think it is probably quite a small one.  You worry a lot more when you get up to an IPT of up to 12%.

Q47            George Kerevan: Going up to 12%, the Chancellor seems to figure, is a good way of raising revenue.  He was quite happy to admit he was just doing it for revenue.  Could you just clarify the point you were making earlier?  The Chancellor has made the comparison to VAT and said the standard rate of IPT, even at 12%, is below the standard rate of VAT, but you are saying that that is not a valid comparison.  Could you just explain that again?

Paul Johnson: The valid comparison would be VAT on the value added from insurance, which is essentially the difference between the amount we all pay for insurance in premia and the amount the insurance companies pay out again at the other end.  Now, I do not know what that is, but it is going to be a relatively small fraction of the total premium.  The value comparison would be 20% of that difference.  It is certainly not the case to say that, as was sort of implied in what the Chancellor said, the appropriate level of IPT would be 20%.

Q48            George Kerevan: Thank you.  Given that, and given the fact that you think that on a fiscally efficient approach IPT should be lower, what would you pitch it at?

Paul Johnson: My recollection from work that we did some considerable time ago, is that we thought that the initial rate of somewhere around 2.5% was not far off, but that is a recollection from previous work.  It may have gone up a bit, but it is clearly going to be below 12%. 

Q49            George Kerevan: IPT is still only applied to certain forms of insurance product.  It is not applied to reinsurance or to life insurance or to insurance on certain large industrial products, like ships.  Would that leave the Chancellor in a position that he could actually extend the tax base, if he wanted to go further down the route of IPT as a revenue raiser?

Paul Johnson: Those are very sensible exemptions from this, because of course the things that you are talking about—big industrial aircraft or what have you—are things that are essentially only going to be bought by business.  Life assurance of course is essentially a savings product, rather than a standard insurance product, so I would hope that he does not move in that direction, because that would be to introduce significant additional distortions in relatively new areas.

George Kerevan: Any further thoughts you have on IPT to guide us would be useful.  Thank you. 

Q50            Helen Goodman: Good morning.  Last week, you produced some graphs showing that old‑age pensioners, people on the state retirement pension, will get better off over the forecast period, as will people on the National Living Wage or the National Minimum Wage, but that people who are unemployed or chronically sick and disabled, living on benefits, would get significantly worse off and that people on average wages will also not be better off by 2020‑21, as they were in 2008.  Commenting on the path for real wages, you said, “One cannot stress enough how dreadful that is.”  I do not know if you want to elaborate on that point at all. 

Paul Johnson: Essentially I would repeat what we said.  We know, because this has already happened, that average real wages today are below where they were in 2008 and the projections are that they will not have recovered to their 2008 levels by 2020.  Now, they might scrape ahead or they might be not below but, whether they have scraped ahead or not, we will have had more than a decade with essentially no real earnings growth. 

We cannot find anything in history, certainly going back at least to the 1920s, where you see anything quite similar.  There seems to have been some deliberate misunderstanding of what we said to suggest that we were saying that people would be as badly off now as they were in the 1920s.  Of course that is not true.  Actually, what we have seen in every decade since them is at least some growth in real earnings over time and, in general, quite big growth in real earnings over a decade because, in general, earnings rise 1% to 2% a year in real terms, as the economy becomes more productive and so on.  That is something that people have got used to.  Real living standards rise because productivity rises and that drives up real earnings.  Having a decade where you have no real earnings growth, on average, is pretty unprecedented, certainly in pretty much anybody’s lifetime.

Q51            Helen Goodman: The pushback that you had from Number 10 was that real household disposable income will be increasing.  This comes out of table 3.7 in the OBR, which shows real household disposable income stagnating in 2017 and then rising to 1.9% in 2021.  Overall, there is an increase in people’s real household disposable income over the forecast period.  Does that include tax and benefits?  I think it does not include changes to tax and benefits, but I would be really grateful if you could clarify that point.

Paul Johnson: There are several things going on there.  In that measure, real household disposable income does include it, as I understand it.  It is after taxes and after benefits, so that has all been included.  First of all, it obviously includes all incomes, including the incomes of pensioners and so on.  It also includes, if I recall correctly, a measure for the value of owner occupation.  I think that is right; I am not sure.  Anyway, it is a different measure of income. 

It is not just earnings is the key thing.  It does not negate anything about what we understand about what we understand is happening certainly to median earnings over that period.  We know that average incomes as a whole, since 2008, have done a bit better than earnings and that is largely because we have seen really quite significant growth in pensioner incomes, for example, so average incomes have done better than earnings going backwards.  Going forwards, even on the newest projections real earnings are projected to grow a bit.  With average earnings growing a bit and incomes from pensions and so on growing a bit more, you would expect the average income across the population to grow, even if those in work are still doing relatively poorly.

Q52            Helen Goodman: Okay, in your presentation or what I was able to download of your presentation that you gave last Thursday, you set out the impact on expected net income for a part‑time lone parent in 2021, showing the impact of tax and benefit changes, the impact of the National Living Wage and the changes to the forecast.  She loses about £1,500 over this period.  Now, obviously this is a very important group of people, but it is a relatively small group of people taking the population as a whole.  It would be really nice if you could possibly give us some impacts for other groups of people and tell us how many of them there are—so a pensioner or a person on average incomes—so that we could actually see.  The distribution charts you have given us look at the changes with respect to tax and benefits.  Wes Streeting is going to ask you about that in a minute.  I do not want to go into that territory.  I am interested in the overall impact on people, putting together the wages data and the benefits data.  I do not think, apart from this one lone parent example, we really have that, so some more worked examples of perhaps some slightly more typical households would be really interesting for us to look at.

Carl Emmerson: We do have work underway that is going to project how the income distribution has evolved up to today how we think it might evolve over the next few years, if the OBR’s forecasts are right, and taking into account tax changes and benefit changes that the Government has planned.  That will be able to show us what is happening to incomes across the distribution through to the end of this Parliament, under this forecast, and we are going to publish that before the Budget.  It just takes a little bit of time to crank through.  We cannot do it overnight after the Autumn Statement.

Q53            Helen Goodman: Could you crank through some other examples like the lone parent one more speedily?

Carl Emmerson: We can, but what we are going to do is do it for a representative sample of the whole population, so we will do the whole distribution of incomes.  Rather than rely on Example A, Example B or Example C, we can do it for a sample of real people.

Q54            Helen Goodman: That is very helpful.  Will you send it to the Committee, so that we do not inadvertently miss it?

Carl Emmerson: Absolutely.  It is expected in February.

Q55            Helen Goodman: Thank you very much indeed.  One of the reasons that the forecast showed that people are not going to get as better off as we thought in March is because inflation is up.  Now, we know that food prices in particular are going to go up and we know that people on low incomes spend a higher proportion of their income on food than people on high incomes.  At one time, some work was done on having differential inflation indices for different groups of people.  I wondered if you had done anything on that, because I am concerned that the numbers that you have produced have underestimated the real falls in income for people on low incomes, while overestimating the benefit to people on high incomes.

Paul Johnson: We have actually done exactly that.  I can send you the link to what we published two or three weeks ago now, looking at the distribution impact specifically on inflation created by a falling exchange rate.  Because we know what set of goods people buy and what proportion of the value added there comes from imports, we can, in a fairly broad‑brush sense, look at what the income across the distribution is.  Broadly speaking it is flat. 

Whilst you are right that there is a lot of imported food, which rises in price, people who are further up the income distribution buy a lot of things like petrol, which has gone up quite a lot in price, and white goods and so on, which are mostly imported.  Actually, we do not see much in the way of differential inflation.  If anything, there is perhaps a little bit more of an impact on those higher up the distribution than those lower down, just because of the way the spending is distributed.  Yes, higher food prices hit people towards the bottom, but higher prices in other areas hit people further up the distribution.

Helen Goodman: That is interesting.  That has changed from maybe how it was 20 years ago, by the sound of things.

Paul Johnson: We looked at work on differential inflation from two or three years ago and found that, over the period of the recession, particularly because mortgage rates came down so much, if you include that in your inflation measure, some headline inflation measures were understating the impact on those towards the bottom and overstating the impact further up.  That reversed a bit over the last couple of years and, certainly from what we can see just looking at the impact of sterling depreciation, it is fairly flat across the income distribution.

Q56            Helen Goodman: You said you did that two or three weeks ago.  That was before the Autumn Statement.  Was there anything in the Autumn Statement or the OBR report that would make you revise what you said?

Paul Johnson: Not on that specifically, because that was just looking at the flow‑through of the depreciation in sterling, so that does not change any of that.

Q57            Helen Goodman: When I talk to my children, who are in their early 20s, they complain to me constantly that their standard of living is lower than mine was at their age.  Moreover, they will never have the standard of living that I have now, whatever they do.  It is completely out of their control.  Do you think that this is really a bad time to be a young person?

Paul Johnson: Certainly incomes among people in their 20s now are below where incomes of people in their 20s were back in 2006‑07.

Helen Goodman: I was not that young in 2006.  Maybe you can go back to 1986.

Paul Johnson: I am going to carry on.  What is more concerning is, first of all, that rates of owner occupation are half of what they were 20 years ago and still below where they were 30 or more years ago for that age group.  Possibly more worrying still for the long run is there is essentially zero coverage of defined benefit occupational pension schemes, in the private sector, for people in their 20s and early 30s, relative to something like 40%, 30 or 40 years ago.  The problems that are being faced are a combination of incomes that are much the same now as they were, say, 20 years ago, real problems getting into owner occupation and real problems saving substantial amounts through pensions.  Of course, negative real interest rates make it very hard to save significant money over time.  There is more difficulty for the younger generation than there has been for some considerable period.

I would add one thing to that.  At some point, you will presumably leave what you have for your children.  The amount that you or I can leave to our children will be very important in how well off they are in the long term, more so than has been the case for many decades.  The importance of parental support, either in life or afterwards, because everything else is becoming so much more difficult, is going to become an increasingly important part of the long‑term welfare of the younger generation.

Q58            Helen Goodman: Of course, and that has massive geographic region distributional implications, does it not?  In my constituency in County Durham, where people’s assets are worth much less and the gap is much bigger than people who have houses in London, this regional gap is going to increase.  Have you done anything or do you know of any work on it not just being a problem for young people now, as they are going to carry this problem through their lives, over the next 30 years, because of the particular conjunction of problems that they are facing now?

Paul Johnson: We have not provided anything like projections of what that will look like but, in the way that I have described it, we have taken it at least that far.

Q59            Wes Streeting: The Prime Minister pledged when she took office that the Government would work for every one of us.  How do you think this statement compares to the distributional analysis of the Autumn Statement?

Paul Johnson: The Autumn Statement did almost nothing in terms of distribution.  If you look at what we have done, there are very small gains from very small changes across the distribution, so the Autumn Statement itself was an appropriately low‑profile fiscal event from the point of view of making changes to that.

Q60            Wes Streeting: With that in mind, looking at the distribution analysis between now and 2020, the second‑lowest decile are forecast to lose £1,700 per year between 2015 and 2020 despite, as you said, a very modest boost from the Autumn Statement.  Is that a reflection of just how punitive the 2015 welfare changes were?

Paul Johnson: That is a reflection of a set of welfare changes that add up to about £12 billion in savings by 2023.  In particular, the biggest of these, let us not forget, is the freezing of most working‑age benefits, such that their real value will fall over time and will now fall by more than expected, because we expect inflation or price levels to be somewhat higher.  On top of that, there were some of the changes to Universal Credit, which will come in at that point.  Yes, what you see in that chart reflects substantial cuts to working‑age benefits and, because working‑age benefits of course go to people on relatively low incomes, then the impact will be on that group.

Q61            Wes Streeting: Based on your own distributional analysis and the Treasury’s, how would you describe the winners and losers from budget decisions over the course of this Parliament?

Paul Johnson: Essentially, there have not been very many winners.  By far the most substantive losers have been those who are on means‑tested benefits or benefits among those of working age.

Wes Streeting: They are the lowest‑income households.

Paul Johnson: Lower‑income working‑age households lose when you chop £12 billion off the welfare budget.

Q62            Wes Streeting: One of the new terms that has crept into everyday political use is “just about managing”, the so‑called JAMs.  Do we have any definition of those who might be just about managing?

Paul Johnson: I take it to mean people on relatively low incomes.  I do not know whether it is intended only to mean those in work but no, I do have a definition, I am afraid. 

Q63            Wes Streeting: Is it something the IFS might do some work on?

Paul Johnson: I do not think we can define JAMs, if that is what you mean.

Q64            Wes Streeting: It is a pretty meaningless term, is it not?

Carl Emmerson: We would tend to split the effects by different household types—pensioner, working age, perhaps split by whether someone is in work or not or whether they have children in the household.  We would tend to split by their position in the overall income distribution, because they are fairly obvious things to do.  In other analysis, we might say whether they are a renter or an owner occupier.  In our data, we do not ask people whether they are not managing, just about managing or well above just about managing. 

Q65            Wes Streeting: Sticking with the point of methodology for a moment, you will recall that it was largely thanks to the efforts of people around this table that we managed to get the distributional analysis reintroduced.  The previous Chancellor decided to abandon the practice and, to the new Chancellor’s credit, he listened to the arguments and has moved.  One of the things the Chancellor said to this Committee in a letter to our Chairman is that, consistent with all previous distributional analysis published by the Treasury, they are including the benefit to households from public services, “As it is my view that it is vital to ensure the analysis also reflects decisions on services, such as childcare and the NHS, which families also value.”  Having looked at the distributional analysis published alongside the Autumn Statement, is it your view that the methodology used in this distributional analysis is consistent with the methodology used for previous distributional analyses alongside Budgets and Autumn Statements?

Carl Emmerson: You would want to do both.  It is of interest to look at what is happening to household incomes from tax and benefit changes, and how that varies by household income.  It is also true that you would want to try, as much as possible, to incorporate the value of public services and that people are getting to enjoy how that is changing and how that looks across the distribution of total resources.  The Treasury analysis should be seen as complementary.  It is much harder to do; there are some methodological issues, but I think it is a worthwhile exercise.  I do not think you would want to do it instead of looking at what happens to household incomes.

Q66            Wes Streeting: I agree with that point.  From what you have seen, has the methodology used by the Treasury changed for their distributional analysis changed at all between the methodology used for this Autumn Statement and the methodology used for previous Budget and Autumn Statements or have you not looked that far into it?

Paul Johnson: I do not know, I am afraid.

Q67            Wes Streeting: Okay, fair enough.  I agree with the point you are making that it is useful or a case can be made to look at wider public service provision, alongside tax and benefit changes, in terms of household incomes.  That would make sense in terms of cash value of things like childcare, but the Chancellor has cited the NHS.  I am struggling to imagine how the provision of a GP or A&E provision is having an impact on pockets.

Carl Emmerson: You can say that these services have a value to people.  If they do not, we should not be providing them.  One way of doing that is to say how much we spend on them.  How much do we spend on, for example, the NHS and how does it vary across the population?  We know that different age groups, for example, receive different amounts of the spending, so that is one way of trying to attribute the spending, as it is to say that, actually, households with children or pension households tend to use the NHS more.  We will attribute the spending that we do on the NHS to those groups and those households.  I do not think that that is an unreasonable thing to do. 

Wes Streeting: In understanding family finances it is pretty meaningless. 

Carl Emmerson: You want to do it alongside looking at incomes.  I agree that childcare might directly help family finances, in a way that helping an ill pensioner or an ill child might not, but it is also the case that the NHS is providing a service to those households that you would want to try to value as well.

Q68            Wes Streeting: I am struggling, because the argument really centres on cash benefit.  For example, I do not have any children.  If I were a woman MP, I would be described as a “childless MP”, but being a man no one has bothered.  The fact is that I do not benefit from school provision for children, for example, but if I did have children who were attending school, it does not really make any difference to my income.

Carl Emmerson: It does relative to a world where the state does not provide any school education to your children.

Paul Johnson: You would presumably pay for it.

Q69            Wes Streeting: I am paying for other people’s children to go to school and I think that is money well spent.  What we are really trying to do with understanding the distributional analysis is what is actually happening to family finances and to people’s pockets, particularly at the moment, are we not? 

Paul Johnson: The point is, as Carl said, that there are two things that you might want to do.  One is exactly the thing that you describe.  If it is that, then you would not look at education, NHS spending and so on.  The other is the value of the things that the state provides to households.  Now, that is not the same as a cash transfer but, for sure, the education, health and so on that people have access to are worth something and that is genuinely a worthwhile thing to look at.  As a fraction of your income, this is worth an awful lot more to lower‑income people than it is to higher‑income people.  Clearly, we would be immensely worse off if we did not have these things.  There is a value in producing these things, but there is absolutely additional value and arguably more value in producing something that says, “This is the impact on cash incomes.”  They are two different things, but they are both valuable. 

Q70            Wes Streeting: I think we are splitting hairs a bit.  I agree fundamentally with the point you are making, so I want to just move on with the limited time we have left to look at tax and benefit reforms from the previous general elections to 2020.  Clearly the cuts are being felt hardest by those who are most in need.  Particularly given the benefit cap has tightened—and I have certainly, as an outer London MP, seen the consequence of that with my case work—I wonder if the IFS has done any work looking at what the impact of tax and benefit changes has been by region as well, on the ability for families to cope and exist in parts of the country where living costs are high, but the support provided is being decreased, year on year.

Carl Emmerson: We have not done that recently.  Either in the report I mentioned in February or in a follow‑up one later this year, we will do the regional split.  We are committed to doing it once this year.  I am not saying that it is not interesting, but there are other splits we also want to do, so we will be looking at that at some point over the next 12 months.  The benefit cap will not make a big difference here of course, because it is fiscally very small.  It has a very large effect on an incredibly small number of households and so, in terms of the big picture, it is not going to drive changes in incomes.  Even in terms of our decile chart it will not be a very important measure.  That is not to say that it is not having a big effect on some households; it is just a very small number of households.

Q71            Wes Streeting: That is the point I wanted to raise really, is it not?  Many, particularly the most recent, welfare changes, as you said, will not feature on the big‑picture analyses done, but for the families that are affected by these changes the impact is huge.  We will look at the evidence base from food bank usage to people presenting themselves to services for support with self‑harm, attempted suicide and unfortunately successful suicide cases.  We are putting a great deal of burden on the people with the smallest shoulders, are we not?

Paul Johnson: Just to repeat the point, it is a relatively small number of people in the context of the number of people who get social security benefits, but the impact on those people is probably bigger than any of the other changes that are made.  There are some very substantial losses for those who either have significant numbers of children or high housing costs.  It is a big impact on those people.

Q72            Wes Streeting: For the huge impact on those families, what is the impact on the scorecard? 

Carl Emmerson: It is tiny.

Paul Johnson: It is a few hundred million.

Carl Emmerson: A very large proportion of the direct savings have actually been handed back to those households through discretionary payments, so the overall saving for the Exchequer is really very small.

Q73            Wes Streeting: Effectively we are creating more bureaucracy through the need for discretionary payments.  We are putting a great deal of burden on people with the narrowest shoulders and the biggest inability to take the burden, and we are not even saving a lot of money for the taxpayer.  Is that a fair summary? 

Carl Emmerson: All of those statements are correct.

Q74            Mr Rees-Mogg: Following on from that, there must be some sense of fairness that people who are working should be better off in principle than people who are not.  There should be some limit on the benefits that are available.  That is not an unfair principle.

Carl Emmerson: It is not.  The reason that these households are losing very large sums of money is because they were getting very large sums of money from the state in benefits.

Mr Rees-Mogg: Some of these, prior to the early cuts, were quite extraordinary, though very small numbers were getting £70,000 or £80,000 in housing benefit.  I accept that they were rare, but this was considered to be a great unfairness.

Paul Johnson: The rules around housing benefit itself have been tightened up significantly now, so that would not have happened now, in any case.  You are right; this has been taken down to a total level of benefit of £20,000 a year which, after tax, represents a relatively significant salary.

Q75            Mr Rees-Mogg: Thank you.  I wanted to go back very briefly to the decade‑without‑real‑earnings‑growth point.  That is essentially still the aftereffect of the 2008 crisis. 

Paul Johnson: Correct.

Q76            Mr Rees-Mogg: There has been a slow improvement from 2012‑13, which is set to continue, depending on the forecast.  Your forecasts are entirely based on the OBR.  Is that correct?

Paul Johnson: Yes.

Q77            Mr Rees-Mogg: Do you think that that is the right starting point, when the OBR is so constrained by statute as to what it is allowed to consider?

Paul Johnson: It is the best starting point that we have.  I am not sure that the statutory constraints necessarily significantly impact what the OBR is forecasting, in terms of earnings growth.  To repeat the point from earlier, even with the relatively robust earnings growth that the OBR was forecasting back in March, you still only get to earnings in 2020 a tiny bit above where they were in 2008 so, qualitatively, the picture does not change.  There is a statistically politically interesting fact that it remains a little bit below where it was in 2008. 

Mr Rees-Mogg: The assumptions that they make for the overall economy are crucial to what feeds through to earnings.

Paul Johnson: Yes.

Q78            Mr Rees-Mogg: They assume, which I think they have to assume—I do not think they were wrong to assume this—that on leaving the European Union the UK will maintain tariffs against the rest of the world at exactly the same rates as they currently are.  Now, you would not have to make that assumption.  You could make an assumption of various options of what Government policy that has not been announced may be, but the OBR cannot do that.

Paul Johnson: Correct.

Q79            Mr Rees-Mogg: I just wonder why you do not go beyond the limitations of the OBR.  It seems to me it is right for the OBR to follow that, because that is what it was set up to do, but the IFS is there to give a broader picture.  What do you add if you just follow slavishly the OBR forecasts?

Paul Johnson: I honestly do not think we could say much more than this.  There is a world in which we could say that, if things turn out differently and better than the OBR assumes and the numbers look like this, or they turn out differently and worse than the OBR assumes and they look like that, but I am also not sure that that adds a great deal to the sum of human knowledge.  My sense is that, where the OBR is, they are very constrained in what they can say. 

Even what they are assuming about whether we maintain membership of the single market is actually slightly obscure.  That is almost the bigger point, in terms of where we are in 2020.  You could make one set of assumptions, which is that we are outside of the single market and, by 2020, we have not managed to change our tariffs and relationship with the rest of the world, in which these case these numbers would probably look worse; or you could make an assumption in which we are inside the single market and we have effective deals with the rest of the world.  In which case, perhaps they would look a little bit better. 

Q80            Mr Rees-Mogg: Is that not useful work for you to do?  Is it not sensible for you to say that “The Government have not declared their policy.  These are the policy options and this is what we think might happen under each option”?  Otherwise, the OBR has already done it and the IFS has to go looking for that.  

Paul Johnson: Broadly speaking, we can and have said that a world in which we remain within the single market is likely to be better for our national income than one in which we move outside.

Q81            Mr Rees-Mogg: What have you said about the abolition of all tariffs?

Paul Johnson: There is a strong case for abolishing tariffs over time.  That will clearly have at least a short‑run impact on the structure of the economy.  You could say that there may be a long‑term benefit from that, but there is going to be a short‑term restructuring cost.  Now, we are not in a position to forecast exactly who is going to bear the brunt of that cost or indeed where the long‑run benefits might lie.

Carl Emmerson: In addition, lowering the cost of trade would be very beneficial to living standards in the long run, but it is not just about tariffs.  Non‑tariff barriers can be very important too.  A world where we are outside the customs union and we move towards lower tariffs could help increase trade with the rest of the world, but you would also have to worry about if we have increased costs of trading with the countries with which we do the most trade at the moment and what effect that might have.  It is not just about a gain from trade from the rest of the world; you have to think very carefully about the costs of trade with the European Union.

Q82            Mr Rees-Mogg: There would be an immediate gain if we took tariffs off food.  Mrs Goodman has mentioned the issues with food inflation and the effects they have on the poorest in society’s budgets.  We have a 70% tariff on beef, for example.  If we were to remove these tariffs, the potential for an increase in living standards and a cut in inflation, where inflation is one of the major downsides of the OBR’s new forecast, would be fundamentally changed.

Paul Johnson: Clearly, if you get rid of tariffs on a set of things on which we are currently paying tariffs that will reduce prices on those things.  Equally, as Carl said, first that will have an impact on the structure of the economy.  They will be at least transition costs.  Secondly, if we are outside the customs union, even without tariffs, it will increase the costs of stuff coming in from the rest of the European Union.

Q83            Mr Rees-Mogg: This is all extremely interesting and it seems to me something that the IFS could usefully be doing.  Whilst you are restricting yourself just to following the OBR, the OBR is strictly limited in what it can do.  It has a clear mandate that it can say what Government policy is, but it cannot guess.  It says quite clearly that the Government have been very unhelpful in enlightening it as to what Government policy will be and that is for very good reasons, because the Government cannot whisper secretly to the OBR its whole Brexit strategy, whilst not telling anybody else. It would inevitably be reverse‑calculable from the forecasts the OBR was making, but I cannot see why you have not done that to see if you can come up with a range of things that may then inform policy, because the Government could then see which way things might go.  Slavishly following the OBR does not really take the debate very much further.  You are both doing the same thing, at this point.

Paul Johnson: We can think about whether there is something useful we can do in that arena.  I confess I cannot immediately see what it is, but we will think about that. 

Mr Rees-Mogg: This leads me on to Andrew Haldane’s recent speech, which I am sure you have read.

Paul Johnson: I am afraid I have not.

Q84            Mr Rees-Mogg: You have not.  It was a couple of weeks ago and it was the Shackle Biennial Memorial Lecture on 10 November.  He looks at what he calls the “failures of models, methodologies and mono‑cultures in economic forecasting” from 2008 onwards.  He says that the “pre‑crisis forecasts were very tightly bunched in a range of one percentage point” and that the forecasters saw “a continuation of the gentle undulations in the economy”.  The third point is that forecasts were not only wrong, but “spectacularly” so.  He goes on to say, which I think is very important, “While forecasting performance has improved in the period since, there has been a continued string of serially correlated errors, with the speed of the recovery consistently over-estimated. The average forecast error one year ahead has been consistently negative, averaging 0.5 percentage points per year. The average error two years ahead has been over one percentage point per year.” Are you concerned that forecasters all using the same models are all coming to the wrong answers and that this is very unhelpful for public policy, because we are effectively sticking our fingers in the wind and guessing which way the wind is blowing?

Paul Johnson: It is one of the reasons why we are not forecasters.  The point is clearly an important one.  In the aftermath of the crisis, we had a long period in which the long‑run impact of the crisis was, as you say, serially underestimated by all forecasters, which is why we have had to have, particularly in 2011 and 2012, additional years of austerity imposed relative to what was expected, because the economy has turned out to be a lot less good than expected. 

Now, the trouble is, going forward, we may well still be getting the answer wrong, but we do not know in which direction.  In one sense, the forecasts from the OBR are still relatively positive.  These are increases of—whatever it is—1.6%, 1.7% or 2.1% over the next three years.  If they are making the same kinds of mistakes they were making four or five years ago, things will turn out considerably worse.  On the other hand, for whatever reason, it could be that we have reached the flicking point at which forecasters are being too gloomy.  Rather like fiscal rules, these things might turn out to be wrong, thrown away or whatever, but they are the best we have to go on at the moment.  They are the best that the Chancellor has to make his plans and I do not think that there is any very good reason for thinking that they are particularly biased in one direction today, and in particular that the bias has moved from positive to negative.

Mr Rees-Mogg: You accept that forecasts are normally wrong.

Paul Johnson: I think everybody accepts that.

Q85            Mr Rees-Mogg: Everybody does.  The final question is that, from these forecasts that you know are wrong, you then make some very bold assertions.  You say, “One cannot stress enough how dreadful that is – more than a decade without real earnings growth.”  Actually, the last five years of that is not yet there, so this is a bold assertion on a forecast that will inevitably be inaccurate.  Do you think that it might be worth being a little more cautious in the boldness of your assertions, as we know that the evidence for those assertions is not going to turn out to be fully founded?

Paul Johnson: On that particular issue, no actually.  We have already had seven or eight years without it and we are still—whatever it is—7% below where we were.  It is almost inconceivable that we will not have a decade with no earnings growth, which we are in the middle of at the moment.  The best we can expect for 2020 is that earnings will be a little above where they were in 2008.  Actually, because that is referring to a decade that we are either seven years through or five years through, depending on which decade you are talking about, and we are 7% down on where we were, it is a pretty safe forecast that the best we are going to manage is just about back where we were at the beginning of the decade.  In all history that is still dreadful.

Mr Rees-Mogg: You will be authoritative on backward‑looking statements, but not forward‑looking ones, which sounds very prudent.

Carl Emmerson: Can we and can everyone do better in trying to stress the uncertainty around forecasts?  Yes, I think there is room there.  I did earlier try to stress that the Chancellor is on course to meet his forecast, by which I mean he has a two‑in‑three chance.  I do not know if you fancy those odds or not.  Actually, there is still a one‑in‑three chance that we will deliver a headline surplus in 2020.  I said that about an hour ago.  Trying hard to get across that it is very uncertain is important.

Mr Rees-Mogg: I thought that was a very important statement and I think it is worth emphasising the fundamental uncertainty of forecasts, because they are viewed as if they are fact and what are going to happen, when they are in fact honest opinion.  I do at least call them honest opinion.

Chair: We have asked you to take a look at productivity and, if you think it would be helpful to have conversations outside this hearing about what that could consist of, by all means come to us.  In the meantime, thank you very much for having given evidence to us this morning, and we will pursue our inquiries tomorrow with the OBR.