Our second item of business is to take evidence on the United Kingdom economic and fiscal outlook. I welcome to the meeting Robert Chote, chairman of the Office for Budget Responsibility. Before we move to questions, I invite him to make an opening statement.
Good morning, convener, and thank you for the invitation. It is a pleasure to be back. I should preface what I say today by warning you that we are now in what is formally the purdah period as a result of the UK general election, which means that I am supposed to be even more opaque and incomprehensible in my answers than I normally would be. If you think that you have understood what I have said, then it is not what I meant to say. [Laughter.]
I will start off by saying a little about the broad highlights of the latest economic and fiscal forecast that we published alongside the UK budget. In doing so, I will compare it with what we published at the end of last year at the time of the autumn statement.
Thinking about what changed between December and March, there were a number of factors that were both positive and negative for the economy. Most dramatically, there was a further sharp fall in the oil price—the spot price fell by about 27 per cent and the medium-term level implied by the first couple of years of the futures curve fell by about 17 per cent between the two forecasts.
We also saw a further substantial rise in net inward migration into the UK—about 298,000 in the year—and that is significantly higher than we had been anticipating for the levels over the next few years. We have now plugged into our forecast an assumption that net inward migration will decline to about 165,000 a year, rather than 105,000 a year. We have to choose from a variety of population projections that are presented by the Office for National Statistics.
We have seen a downward movement in market interest rate expectations. In one way, that can be seen as a positive stimulus; on the other hand, it may be reflecting the fact that people are gloomier about growth prospects. It could be read either way.
We had a series of downward revisions to past gross domestic product growth estimates through 2013 and 2014, which were very slightly reversed in the data that was published yesterday.
We had another disappointing quarter for productivity growth—worker output per hour continued to show weak productivity by historical standards—and we also had a weaker global outlook, judging from the forecasts published by the International Monetary Fund, the Organisation for Economic Co-operation and Development and the like.
We judged that all of those factors taken together would have a relatively modest net effect on our GDP forecast and on the budget deficit. We have edged up our growth forecast slightly in the near term, reflecting the fact that the lower oil price reduces inflation—primarily via lower fuel prices—and temporarily pushes up real incomes, leading us to assume a boost from consumer spending.
Slightly further into the future, growth is nudged down in 2017 because of the assumption that oil production will be weaker throughout the forecast as a result of the lower oil price. Right at the end of the forecast there is a slight upward nudge because of the assumed higher levels of net inward migration. Over the forecast as a whole, there is growth of about 2.5 per cent a year over the five years.
The other main difference in our latest forecast is lower inflation in the near term than we were anticipating before, with inflation near zero and presumably negative in some months on the consumer prices index measure.
I will turn now to the position of the public finances. We have revised down both receipts and spending somewhat over the course of the forecast. On the spending side in particular, the combination of lower market interest rates and lower inflation means that servicing index-linked gilts is less expensive. Welfare bills are also less expensive, because of the presumption that benefit values will not be uprated as quickly as they otherwise would have been.
The impact on the public finances is offset somewhat. On the basis of the overall approach in the Government’s medium-term profile for public spending, which it has changed, the fact is that money saved on debt interest does not necessarily feed through to a lower forecast for Government borrowing but reduces the previously implied squeeze on public services.
There are a number of reasons why we have pulled the estimates for receipts down: the oil receipts, lower interest rates, which will reduce the income the Government gets from its assets, and changes in various tax measures.
The starting point for the impact of the budget policy measures is what might be thought of as the menu with prices in the Treasury’s red book—the list of individual tax and spending measures and their costs. The 2015 budget is another fiscal event pretty much in keeping with every one subsequent to the—quote unquote—emergency budget in 2010, in that the giveaways broadly match the takeaways over the five years and, indeed, in most individual years.
The budget policy measures do not therefore make a great deal of difference to the outlook for borrowing. To summarise, the Government is bringing in some more money from the banks and by reducing the value of pension relief for relatively high earners, and it is spending that money on things such as higher income tax allowance, the subsidy for first-time buyers and measures to help the oil sector.
What has made more difference is that the Government has yet to say what it wants us to assume is its policy on choices over expenditure on public services and capital spending. At the moment, we have detailed plans for public expenditure department by department set out only for 2015-16. That requires us to make an assumption for the remaining four years of the forecast.
We asked the Government what it wants us to assume, and we have been given an assumption that we are assured is the agreed view of the coalition as a whole, although both parties would say that, if they were governing alone, they would be doing some things differently.
The Government has chosen to tell us directly not how much it wants to spend on public services but how much it wants to spend in total. The forecast that we produce for items such as debt interest and welfare can be subtracted from that figure, leaving an implied envelope left over for public services spending.
There have been some significant changes to that envelope. The Government has slightly increased the squeeze on total public spending that it wants us to assume through to 2018-19, but it has dropped the idea of cutting total public spending as a share of GDP in 2019-20, the final year of the forecast.
In consequence, the implied profile for the change in spending on public services over the five years of the next Parliament looks somewhat uneven. We refer to the profile as a rollercoaster in our report. There will be a sharper real cut in public spending on public services in 2016-17 and 2017-18 than anything we have seen over this Parliament, but that is put into reverse in 2019-20.
On rationale, it is obviously for the Government to say why it has chosen those figures, but there are a number of things that can be said about the shape of the public finance forecast as a whole that the Government has achieved—for want of a better word—and the consequences that that has had. It has ensured that our forecasts for borrowing are lower in every year to 2018-19 than they were in December. That has been achieved by tightening the squeeze in the middle. It is no longer the case—as it was in the December forecast—that public spending is set to fall to its lowest share of GDP since the 1930s: the additional cut in spending in the final year has been dropped.
The Government is also still achieving, with some room for manoeuvre, the fiscal targets that it set itself with the rolling three-year target for a particular measure of the budget deficit. That is one reason why the Government has continued to pencil in the particularly sharp squeeze on spending in years 2 and 3 of the next Parliament.
Finally, the Government has also managed to see debt as a share of GDP fall in 2015-16, which is a year earlier than in our December forecast. It has achieved that by announcing additional asset sales: the Northern Rock securitisation vehicle, Granite, and more sales of Lloyds Bank shares. It is important to point out that, if you sell an asset for roughly the present value of the future flow of income that you will get from it, that does not make the public finances better off—you are changing the flow of money that is coming in. It is worth noting that the sales bring in an additional £20 billion to reduce debt in 2015-16 but, in effect, reduce the Government’s income in the remaining years of the forecast by about £10 billion.
On the face of it, we might look at the profile of public services spending and ask why, if we had a blank sheet of paper, we would design it like that. However, we have to bear it in mind that there are various objectives for what the Government wanted the public finances forecast to look like, and that is what drops out as a consequence.
Secondly, it is the agreed policy of the coalition. Both the Conservatives and the Liberal Democrats have said that they would do different things if they were governing alone. For example, the Conservatives have said that they would not need to squeeze public services as much in the middle of the next Parliament, because they would find more money from welfare cuts and measures to tackle tax avoidance. The UK Parliament has instructed us not to look at the alternative policies of different parties, so I merely note that rather than comment on it.
Thank you very much for a fascinating introductory statement and for the documentation that has been provided to committee members.
I will start with the asset sale. You said that, if the £20 billion of assets had not been sold, the Government would have expected to raise about £10 million of income as a result, if the asset had been retained. What level of interest do you estimate that it would have had to pay on the £20 billion debt over the years?
The effect of the asset sale is to reduce net debt by about £20 billion in total in 2015-16. When it comes to asset sales, we set a relatively high hurdle before we are willing to include them in our forecast. We need to have a reasonable degree of certainty about what exactly the Government intends to sell and, crucially, when it is likely to sell it—in which financial years that is likely to take effect. We then reach a judgment on the amount of money that the sale is likely to bring in. We pressed the Government for details, because if it wished us to include the asset sale in the forecast, we needed to have a reasonable amount of certainty on those things.
There always remains uncertainty about the impact of such asset sales, because we obviously do not know what market conditions will be like in 2015-16 and whether the Government will proceed with the sales as currently described. However, we think that the Government has been up front enough about the sales and about the amount of money that will be involved to include them in the forecast.
As I say, there is an up-front gain to the public finances of £20 billion-ish in 2015-16, but one consequence of that is that the Government will not get the income that it would have done from those assets over the remainder of the forecast—that is about £10 billion over the remaining years of the forecast, and it would stretch on further into the future. If we assume that the Government sells something for roughly what it is worth, that comes out in the wash at the end. The asset sale temporarily reduces public sector net debt rather than reducing it permanently; it is a change in the profile.
The point that I was trying to get at is whether the interest that would have had to be paid on the £20 billion would be more or less than the £10 billion that the Government would otherwise have received in income from the asset. I know that, given market conditions, the value of the asset can possibly fall to £15 billion or go up to £30 billion—who knows? However, on current trends, how would that work out?
We will have taken that into account in the forecast. I am not sure—given the amounts involved and the level of interest rates—that, in respect of the health of the public finances, the sums would be very large in aggregate terms. I can check whether we have a firmer number for that and get back to you.
That would be interesting.
You said that there are detailed expenditure plans for 2015-16 but nothing beyond that and that you had to do some extrapolations. You talked about the implied envelope. Is that the £12 billion that has been mooted?
The £12 billion—there are lots of £12 billions around.
There are indeed.
09:45
We have plugged into our forecast the number that the coalition has signed off together, but both parties have said that that number does not reflect what they would do if they were governing alone.
The Conservatives have said that they would like to find an additional £12 billion from welfare spending, which means that they would not necessarily need to squeeze public services by as much as is implied in the central forecast that we have at the moment. They have not said where that £12 billion would come from. We have not put it in the forecast for two reasons. The first is that it is not Government policy; it is Conservative policy. The second reason is that, even if it was Government policy, we would only include the £12 billion once it had been explained to us how it would be found, because that obviously affects the whole shape of the forecast.
If that is the £12 billion that you are referring to, it is essentially one of the ways in which the Conservatives say they would depart from what the coalition has managed to agree together for its spending path.
Okay—thank you.
Now, I have asked you this question in previous years, and I asked the Institute for Fiscal Studies more or less the same question this year and, indeed, last year. It is about productivity. You talked about another disappointing quarter for productivity growth, about downward revisions to estimates for economic growth in 2014 and about the world economy—you touched on them in your opening statement. You also talked about productivity falling on an hourly basis in the final quarter, falling short of the forecast once again. This is my $64,000 question: why are we continuing to see this difficulty, with productivity stubbornly refusing to rise? Obviously, it is having a serious impact on economic growth overall, and it is affecting the economy and competitiveness.
I would love to say that we have cracked it since the last time you asked the question.
That is exactly what the IFS said. I was hoping that you would be the fount of all knowledge on this occasion and get one over on the IFS. I realise that it is not in your paper, but you must have some thoughts about why it should be.
I think that the main explanation to which most of us cling more than any other—although, given the scale of the shortfall from historical experience, we have to assume that there must be a number of explanations adding together—is the presumption that the difficulties in the financial system, including credit conditions and the fact that the largest banks are not functioning as they would do in normal times, are preventing the efficient reallocation of capital away from firms that would ordinarily shut under these circumstances, to relatively young, potentially rapidly growing, innovative and productive firms. That hinders productivity growth and also the ability to take advantage of falls in the exchange rate and in export markets.
One half of the argument is the so-called zombie firms argument: that some relatively unproductive firms do not go out of business because of the combined effect of relatively low interest rates, relatively low wage growth and banks being reluctant to pull the plug on them. On the other hand—and I think that my main macroeconomist colleague on the committee would place more emphasis on this than on the zombie firms—there is the inability of potentially rapidly growing, innovative firms to get access to enough working capital and the ability to expand.
That remains the straw to which we all cling as the main explanation. The Bank of England has devoted thousands of person-hours to the problem and it has come up with a long list, from which I think that that would be one of the main explanations that it would cite.
It is a very important assumption—and one that we continue to assume—that over time productivity growth and, consequently, the potential output of the economy will return to the sorts of growth rates that we have seen historically. That is partly an act of faith—there does not seem to be a firm evidence base for saying that that is not going to be the case. However, there are different views. Some people would say that we can now expect productivity growth to rebound relatively rapidly—they are more confident than we are of that. Other people argue that we have had three industrial revolutions and that is our lot—we are not going to see productivity growth reviving.
Not in this forecast but in the previous one, we had some alternative scenarios of what the forecast would look like. There was work on the assumption that, for example, rather than productivity growth gradually recovering to the historical trend rate, there was the sort of improvement that there was in the 1980s. There was also work on the assumption that recent history continued and productivity growth remained very depressed. In short, summarising the bad scenario produces a bad outcome and summarising the good scenario produces a good outcome.
A weaker outlook for potential growth in GDP is crucial in constraining the outlook for the fiscal position. If you kick a hole in the potential of the economy, you create an additional element of the budget deficit that is presumed to be structural. It will not disappear as the economy gets back to its normal level consistent with the Bank of England achieving the inflation target over time.
Perhaps by the next time I am here we will have cracked it, but that is as good as we have at the moment.
One hopes that that will be the case.
I understand that normal productivity growth is about 1.4 per cent, and I assume that it would have been projected to increase, certainly in the public sector, given the significant reduction in people in the sector and the continued rising demands on it. Is much of the projected growth in the private sector? Surely having 298,000 migrants, most of whom will be young and dynamic working age people, should in itself boost productivity.
What is the GDP per capita growth? You said that growth was forecast at 3 per cent and is now at 2.6 per cent, but that is economic as opposed to per capita growth. I take it that per capita growth is now lower, because, if the population is increasing but the economy is not growing as fast as it should have been, GDP per capita growth will be significantly less.
Let us start with the impact of the migration change on per capita GDP. We have shifted to what the ONS calls its principal population projection rather than the low migration population projection, and that assumes a figure of 165,000 per year rather than 105,000.
The consequence of higher migration for the potential aggregate growth rate of GDP over the course of the forecast is an increase of about 0.6 percentage points. The increase in population accounts for 0.5 of that 0.6 and does not have an effect on per capita GDP. The remaining 0.1 of that 0.6 reflects that the employment rate is presumed to be higher as a consequence of net inward migration, and that relates to the fact that migrants are more likely to be of working age than is the population in general.
We make the assumption that net inward migrants have the same productivity and the same age and gender-specific employment rates as the native population. The consequence of that assumption is that an increase in the employment rate increases per capita GDP. It does not, however, increase GDP per worker: it is just that there are more workers per chunk of the population because of net inward migration than there would be otherwise.
You are now going to have to remind me about your initial question on productivity.
The point I was trying to make was that surely—
Oh yes, the public sector—
Yes, as productivity has presumably increased in the public sector because of the huge decrease in the numbers, most of the decline or stagnation in productivity would be in the private sector.
It is interesting. We had been surprised by the fact that cutting public services spending did not have more of a direct drag on GDP via the bit of GDP that is the Government’s consumption of goods and services. It took us longer than perhaps it should have done to catch up with the impact.
Although overall growth has generally been weaker than we anticipated, we overestimated the size of the direct drag from the public spending cuts because of the way in which output and therefore productivity within the public sector are measured. For a reasonably large chunk of public services, the output is measured directly but relatively simply through, for example, the number of children being educated or number of operations taking place.
Let us take education for example. Because the Government has not responded to an X per cent cut in education spending by saying that it will educate X per cent fewer children but, rather, has changed the overall quality and quantity of educational services, and because that has not shown up in the measure of direct output, it has not dragged output down and, as you say, it shows up in productivity and the price mix instead.
It took us a while to realise that that would be the case. A few forecasts back, we assumed that the cuts would be a greater drag from that source than they tended to be in the measured numbers.
That is a different question from whether the overall fiscal consolidation is having the impact on the economy that we would anticipate. That is a broader question, which brings in what is going on in the private sector as well as what is going on in the public sector. However, you are right to highlight the fact that the link has not been straightforward and shows up in a relatively good productivity performance.
I will not ask any more about that, but one would have thought that advancing technology alone would help to boost productivity, which it has done over previous years and generations. Perhaps a further leap in technology will take things forward.
In the end, it depends a lot on how we measure the output. It is not to deny that the cuts might have broader effects but if, for the purposes of the national accounts, we measure the output of education as the number of people being educated, many things simply do not show up.
Paragraph 1.43 of the executive summary of the “Economic and fiscal outlook” says:
“There is considerable uncertainty around our central forecast.”
I want to move on to talk about Scotland in that regard.
The Scottish rate of income tax will be introduced a year from today. The forecasts for it have changed markedly over the years. For example, the forecast for 2016-17 has fallen by 18 per cent since your March 2012 forecast, which is significant. The forecast is now that the receipts will go from £4.379 billion to £5.748 billion, which is a 30 per cent increase over five years.
There have also been significant changes to other forecasts, such as a 15 per cent reduction since December in your predictions for land and buildings transaction tax for 2016-17 from £600 million to £510 million. However, over the period 2014-15 to 2019-20, you predict a 90 per cent increase for receipts from LBTT and, overall, you predict a total increase in tax receipts for the devolved taxes of 35 per cent over five years, which seems to me significant in relation to the background economic growth of 2.5 or 3 per cent.
Some of the explanation is in the documentation but, for the record apart from anything else, will you explain why your forecasts have changed significantly over the past three years and, indeed, in the past quarter with regard to LBTT? Will you also explain why you are so optimistic about the tax forecasts? We would all be delighted if they came true, but a 35 per cent increase in tax receipts in Scotland in five years seems pretty optimistic.
In both cases, most of the changes over time are the consequence of changes in the UK forecast as distinct from changes in the Scottish share if you take the SRIT forecast, for example.
The income tax story comes back in part to the productivity puzzle that we were talking about a moment ago. What you have seen is a fall in the effective tax rate. It is possible to be wrong about the amount of income tax that you will bring in because you are wrong about the amount of income, but you can also be wrong about the income tax forecast because you are wrong about the amount of income tax that you will bring in per pound of wages and salaries.
I realise that the UK changes to personal allowance have had an impact on previous forecasts.
10:00
Yes, that is true. The policy change is subsequent to that, and it is of course an important driver of what is going on with the Scottish share, looking forward.
Income tax forecasts have disappointed partly because we have continued to be surprised that wage growth has not picked up; it has been weaker than expected while employment growth has been stronger. That is another manifestation of the productivity puzzle.
Generally, if labour income—wages and salaries—rises by £10 billion, the Exchequer gets more bang for that buck if that is a consequence of everybody’s wages going up by 10 per cent than if it is a consequence of the amount of employment going up by 10 per cent. The reason for that is that, if wage growth is relatively strong throughout the income distribution, that pulls more of people’s incomes into higher tax brackets, which is a process called fiscal drag. We normally rely on that to raise the average tax rate year in, year out, which is an important reason why, if we look into the future, we see the income tax numbers rising. However, that has not happened, partly because of the change that I have described. If we have employment growth rather than wage growth, we may end up with a larger number of people on relatively low incomes not paying very much more tax, rather than helping to drive people up through higher tax rates.
That is one reason why the forecasts have disappointed. In addition, employment growth, which has been stronger than expected, has not been as tax rich as one would have anticipated. Some of the growth has been in self-employment, and the evidence suggests that there has been an increase in self-employed jobs that pay less than self-employed jobs normally do on average. For that reason, the average tax rate, or the effective tax rate, is weaker. That is another reason for the changes.
Why are the figures expected to pick up? Partly, that comes back to our assumption that we will return to more normal rates of productivity growth and, associated with that, more normal rates of wage growth. That will return us to fiscal drag and help to raise the average tax rate as we go forward.
Also, because inflation is lower than anticipated, the uprating of allowances and thresholds is less than it otherwise would be, which means that people do not have to travel as far to get into the higher rates. That is another reason for forecasting a relatively robust growth rate for receipts, looking forward. However, the key judgment on which the issue depends is about whether we will get back to the sorts of rates of productivity and earnings growth that we have seen in the past.
As I said, most of the story has been to do with the UK forecast. The Scottish share has been relatively stable over quite a period, at a little over 3 per cent, but it has declined and we assume that it will continue to decline in future. As you say, that is primarily a consequence of the policy changes that have been announced. We have had revenue-raising measures at the top—the additional rate of income tax and the withdrawal of the personal allowance at relatively high rates—while, at the bottom, the personal allowance has been raised, which narrows the tax base at that level. Because that has made the income tax system more reliant on the incomes of the people towards the top and because of the difference in income distribution between Scotland and the rest of the UK, that basically helps to explain the figures on the share.
On LBTT, one reason why we have moved our forecast down is that we have revised down our assumption about the normal level of transactions—the frequency with which people move and buy and sell houses—relative to what we had before. Again, that is true for the rest of the UK, with the stamp duty land tax system, and for the Scottish system. Basically, we look at a number of years and try to derive a historical average to which we return. We looked back and thought that some of the years that we were averaging over were not particularly representative, so we shifted that.
With LBTT, because the thresholds are assumed to be fixed—there has been no announcement about indexing them—we again have the issue of fiscal drag. If house prices increase, more transactions will be pulled into relatively higher bands. It may also be assumed that, in addition to the recovery of house prices, transactions will recover back to the assumed, albeit lower, historical norm. Indeed, it is not unusual for stamp duty or LBTT-type taxes to have quite big ups and downs because they are not driven merely by what is happening to prices but by what is happening to transactions.
An additional twist is that the consequence of moving to the sliced systems on both sides of the border is that we expect to get more fiscal drag out of the new systems than we got out of the old system, because the average tax rate goes on rising at the top rather than being capped at the highest slab rate. Therefore, as the value of the transaction gets higher and higher, the average tax rate is asymptoting towards the highest marginal rate.
It poses an interesting issue for us, as well as for the Scottish Fiscal Commission. The new structure will mean that the amount of receipts from LBTT and from the similar structure of SDLT in the rest of the UK, will become increasingly dependent on the number of transactions for relatively highly priced properties. That is always a hard adjustment to make because it is not simply a question of coming up with a macroeconomic forecast that provides an assumption on what average house prices will do. Indeed, in the UK context, we always have to worry about whether London is doing something different from the rest of the UK and then we have to adjust for that difference.
Looking back, we see that we have been through a period when London outperformed the rest of the country, as we assumed it would. That seems to have dissipated, but we are not assuming that that will be the case looking forward. However, even when forecasting for LBTT, you would have to be concerned not merely about what the average performance of house prices will be across Scotland, but about what will happen in areas that have relatively high house prices.
It is striking that, in the housing market as a whole, Scotland, in terms of the movement in the average house price, does not look very different from the rest of the UK excluding London. Indeed, it looks more similar to the rest of the UK excluding London than, for example, Wales, Northern Ireland and north-east England do.
I suspect that when the Scottish Government and the Scottish Fiscal Commission look at the LBTT forecasts over time—we will do that, too—they will wonder whether there are reasons why more highly priced properties perform differently from the rest, and those reasons will be quite an important consideration. That may not be entirely down to what is going on in the macro forecast. I am not an expert in that area, but my guess would be that the number, the movement or the behaviour of relatively highly priced properties in the Aberdeen area has as much to do with what is going on in the oil sector as it has to do with the overall macro picture.
I understand that you are unable to produce Scottish macroeconomic forecasts because macroeconomic forecasts and economic determinants are generally not available or are available only after a long lag. Is that still the case?
Yes, that is right.
What kind of lag are we talking about before Scottish figures can be assessed?
It varies from statistic to statistic. I am sure that Scottish Fiscal Commission colleagues would be more up to speed on that issue.
Clearly, if you wanted to take a different view about what was going to happen to the path of labour income and the average house price, a macroeconomic forecast would shed light on that. However, that would not get you away from the other judgments that you must make. For example, on the income tax forecast, once you have a macroeconomic forecast, you would have to worry about whether policy measures are resulting in forestalling and whether something particularly unusual is going on with the distribution and size of bonuses in and outside the financial sector. In the LBTT and SDLT forecast, you would worry about whether something different is going on with relatively highly priced properties versus other properties. If you are looking at landfill tax, policy questions are as important as macro questions. For example, if you come up with a macro forecast, that would give you some sense of what is happening to landfill tax.
If you had a completely and fully articulated, up-to-date Scottish macro forecast, that would be a useful input, but there would still be a lot of awkward things on which to take a judgment on a forecast-by-forecast basis, which means that even that would not get you the whole way.
Thank you very much for that. My first colleague to ask questions will be Malcolm Chisholm.
Mr Chote, I realise that you have to be even less political than usual—you never are political, of course—but we tend to be pulled in the opposite direction.
One of your most quoted soundbites has been the “rollercoaster profile”, which sparked the accusation that spending levels were returning to 1930s levels. You probably cannot comment on that, but I am interested in the fact that the squeeze is increased slightly until 2018. Is that a significant change from previously? I am not quite clear what the reason for it is, unless it is to create a rollercoaster.
It is not an enormous change. The Government has chosen to tell us what to assume about public services spending by coming up with a rule, or assumption, that describes what happens to overall spending over the period and then backing out the implied profile for public services spending. That rule has now become quite complicated. Back in 2011, the Government gave us an assumption for what should happen to public spending beyond the period for which there were detailed plans and we could explain it in our document in 29 words. The rule has now become sufficiently complicated that it takes us 428 words to explain it. That gives some sense of how much effort has gone into getting the profile that the Government wants.
In practice, that has meant that, because of what has happened to inflation and interest rates in the forecast, the debt interest payment forecast has come down quite a lot. Essentially, in the way that the Government has expressed the assumption to us, it has said that it wants to bank some of that in a way that reduces the budget deficit so that it is, perhaps not uncoincidentally, slightly lower than it was in the forecast that we produced in December, and the rest of it effectively loosens the implied squeeze on public services spending. Over that period, the rollercoaster is somewhat less rollercoastery than it would otherwise have been.
If you are looking for the explanation for why the pattern looks as it does, it is a combination of wanting to have borrowing lower in most years than in the previous forecast; wanting to achieve a particular amount of fiscal consolidation by 2017-18; and not wanting to have public spending hitting a post-war low at the end of the process. Once all those things have been achieved, through the 428 words explaining the overall profile for spending, the rollercoaster is what drops out as the consequence for public services spending. It is not that the Government has said, “This is the path of real changes in spending on public services that we think makes sense for the planning of public expenditure.”
As I said, both coalition parties say that they would do different things that mean that the pattern would not look like that if they were governing alone. The key point is that that pattern is now the baseline for whichever party or parties come into power after the election. If they do not want it to look like that, they will have to explain what they would do differently, and then we will come back and tell you what the path looks like as a consequence.
The working assumption is that at some point in the autumn of this year the new Government will have to set out some detailed plans, or the envelope for the detailed plans, over some years into the future. What we do not know is how many years into the future, which may of course depend in part on what the Government looks like after the election.
My next quote is not quite such a striking soundbite as the previous one, but it has been used by various commentators. You stated:
“The Coalition Government’s policy decisions in this Budget are not expected to have a material impact on the economy.”
That might be thought to be surprising, given the purpose of a budget. The Government has been lucky in that there have been lower oil prices and lower inflation although, obviously, that is not lucky for Scotland. You also say that the economy ended 2014 in a weaker state than expected. Yesterday some figures for January suggested that it might be even weaker. Can you comment on any of that? Looking forward, are you still fairly optimistic about the growth of the economy?
10:15
We have not updated the central forecast since the budget. Although the latest numbers on GDP show higher growth on average over 2014, we had the downward revisions that have been partially reversed and the main reason for the change during the calendar year is because there was a small upward revision to GDP in the first quarter of 2014.
I am not sure that the numbers that came out yesterday tell us anything terribly different about the momentum at the moment. The fourth quarter GDP growth rate was revised up a tad, but the third quarter was revised down and the index of services was perhaps slightly weaker than anticipated in January. I do not think that someone would look at that and say that there was a different momentum story arising from those revisions.
Our judgment that the policies are unlikely to have a material effect is partly a reflection of the fact that this is yet another package in which the aggregate size of the giveaways broadly offsets the aggregate size of the takeaways. It is quite hard to produce a policy package that has a dramatic effect on the economy for which that is also the case. Therefore, we have not made any dramatic changes to the growth profile as a consequence.
We can consider individual policies; for example, one could argue that the subsidy for first-time buyers is likely to push up house prices—that is to be expected as it is increasing demand and not doing very much for supply. However, we have made the assumption that the magnitude of that policy change is not large enough to materially change the forecast.
One obvious difference—this is not a policy change, although policy partially reverses it—is that the decline in oil production that is assumed through the forecast makes a difference in terms of the overall GDP path. Basically, we have assumed that the tax measures that were announced in the budget will reverse or ameliorate about a third of the decline in production that we would have otherwise anticipated from what has gone on with the oil price over that period. For the oil sector, that is material, but in terms of the macro picture it is not necessarily so.
The fallout from the fiscal mandate has been much discussed in Scotland, in particular the first part, which is the aim to achieve
“cyclically adjusted current balance by the end of the rolling, five year fiscal period”.
From your report it is obvious that you are confident that you understand what that means, but other people in the debate in the House of Commons, which has also been much quoted in Scotland, said that it was all rather obscure, particularly the terms “cyclically adjusted” and “rolling”. Are you confident that you know what it means?
I think so. We have certainly produced a forecast and a judgment on whether the Government will hit the target or not on that basis.
What is being targeted? Another way of putting it is to say that we want the Government to raise enough in revenue to cover its spending on things other than capital investment—that is what balancing the current budget means. “Cyclically adjusted” means that we would want that to be the case if the economy was running at a Goldilocks level that was neither too hot nor too cold. We are assuming that there will still be a little bit of spare capacity in the economy in 2017, which would marginally make the underlying position look worse. However, we do not have much spare capacity, so the difference between cyclically adjusted and non-cyclically adjusted is becoming smaller as that gap closes.
On the rolling nature of the forecast, in the new framework, as in the old one, there is one target that is at a fixed date. Until the end of last year, the Government wanted the debt to GDP ratio to fall in 2015-16—it does not matter when we do the forecast, it was always 2015-16—but it has now changed that target date to 2016-17, which is a fixed date. The aim of achieving the cyclically adjusted current budget balance is rolling in the sense that, every time we do a forecast, the Government wants that to be true in the third year of the forecast. Every year, we roll the forecast forward another year and that target then rolls forward into the future.
The rationale behind that is that if the target is stuck at a particular date we could end up having to make some very dramatic policy changes with a year to go, and it is better to keep our eyes focused on the medium term. The Government has decided to shrink the horizon over which it is looking from five years to three. Some would argue that it would have been better off sticking with five while others would prefer a fixed date, but that is what it has chosen to do and that is what we have to stick to.
That is presumably one of the reasons why there is still a relatively tight squeeze in years 2 and 3 of the next Parliament—that is what helps to determine the deficit in 2016-17. However, when we roll the forecast forward another year, the target year will move to 2017-18. Obviously, at one level, you could say, “Well, that’s fine. You are always saying, ‘Jam tomorrow’, but you never deliver the jam today.” However, we explain what was going on with the old versions of the targets as well as the new so, in our analysis, it will be entirely transparent as to whether the promised land is simply being shifted one year further into the future or whether the Government is continuing to deliver it in the earlier years.
The Government has changed the fiscal mandate in that it now relates to the current budget, whereas previously it related to more than that. I do not know whether you can comment on this, but do you feel that, in order to meet that mandate, we need to have the squeeze that we having? Given that the mandate now relates to the current budget, a related question is how much capital expenditure is financed by current expenditure. I presume that that is relevant to whether the target is hit, given that capital expenditure could now be exempted from the fiscal mandate.
The Government’s target has not changed. It related to the current budget before, and it relates to the current budget now. The only thing that has changed is that the target has moved from year 5 to year 3.
We are basing our forecasts on the national accounts definition of what constitutes capital expenditure versus what constitutes current expenditure. Local authorities, for example, have rules on whether they can use current income to finance capital spending, but we are looking at the total magnitude of non-investment spending and receipts and comparing them on a national accounts basis. As a result, there should not be any scope for that sort of gaming.
There is, of course, a debate on whether the distinction between current and capital spending is the right sort of thing to be worrying about. One might be happy to borrow for things that will benefit future generations, because future generations can help pay for it by servicing the debt that has been raised, but others might ask why we are happier to do that for certain areas of capital spending that might not deliver great value for money and less happy to do it for things that have longer-term benefits, such as training teachers and doctors. That is all about how the rule is designed, and is therefore not a question for us to comment on. However, people have looked at the matter, and the Government has chosen to set the rule in this way.
That is also common to the previous Labour Government’s golden rule. The rationale at the time was that, if there needed to be a squeeze, capital spending should not necessarily bear an undue proportion of it; in practice, however, when times get tough, Governments often cut what they can, not necessarily what they should, and those are the consequences. I am not saying that that is happening in this particular case, but it is one of the shared rationales for making that distinction. People might say that that might or might not be the right thing to target, but we have to police the Government against the targets that it has chosen to be policed against.
Just to be clear, are you saying that, if we spend current or resource money on capital, you would classify that as capital expenditure?
The capital depends on what you are spending it on—
This is important. We are very familiar with the issue in the Scottish budget, but I wonder how much resource is spent on capital across the UK, because it is relevant to whether or not the fiscal mandate is met. If we borrowed for that capital, we would obviously free up that resource or current expenditure for current services.
In the national accounts sense, we take a step back from that and say that the Government sets out detailed plans for what it is going to spend on capital and what it is going to spend on current and then it basically has a set of plans to bring in revenue, which is spent on both of those things and on servicing debt and paying for the welfare system et cetera. In policing the issue using the national accounts, we do not look at what particular bits of revenue are paying for particular bits of expenditure. A whole lot of revenue comes in and the Government has spending plans, some of which are plans to spend on capital and some of which are plans to spend on current. In terms of the fiscal rule, we are looking at whether the Government is raising enough in revenue to pay for the current resource bit.
Yes, but it would make a difference. If we spend the resource on teachers and nurses et cetera and borrowed, we would then have the same amount of capital expenditure, or are you saying that we could not get round it that way?
The Government in effect decides what it wants to spend on capital and what it wants to spend on current, and it brings in some money simultaneously. We are obviously comparing both so, on the aggregate budget deficit, we consider whether the Government is bringing in enough revenue to pay for both of those things and, in judging the current budget, we consider whether the Government is bringing in enough revenue to pay for the non-investment bit.
Okay. Thanks.
I want to ask about a couple of points. One is inflation. You state:
“We expect CPI inflation to return to the Government’s 2 per cent target relatively slowly”.
You also state:
“We expect inflation to remain below 1½ per cent until the end of 2016”.
I had a gut feeling that inflation would jump up again at some stage, but I have obviously got that wrong. Inflation seems to be quite low at the moment. Would it matter if inflation went negative—say -0.1 per cent or some tiny amount? Would that have a psychological impact? Would it have any real impact? Have you looked at that at all?
Our working assumption, which seems to be being borne out in the data at the moment, is that, on consumer prices index inflation, basically we are heading to averaging fractionally above zero on a quarterly basis, which is entirely consistent with it being negative year on year in some months.
I think that the consensus view is that, if we end up with a bit of deflation of that sort, it is of the relatively benign kind that reflects the fact that, in particular, the oil price has declined quite a lot. That reduces the amount that households have to spend on fuel, which means that the rest of their income goes further, which temporarily boosts consumption and is a positive for economic activity, albeit temporarily. Looking further into the future, the path of consumer spending depends much more on whether we get the productivity growth and wage growth that are a sustainable source of that.
There would be much more worry if we got into a situation in which, psychologically, people assumed that inflation was going to remain negative and prices were going to fall over time, which would encourage them to say that they will not spend today because it will be cheaper to spend tomorrow. That becomes a vicious cycle.
I think that people’s presumption at the moment is that that is not likely to be the case; it is not what we have in our forecast. One reason for that is that the oil price path that we plug into the forecast is essentially the spot price today and what the futures curve is telling us about prices over the next couple of years, and then we hold it constant at that point thereafter, basically because the IMF and others suggest that the futures market is not liquid enough to be providing much useful information beyond that point.
In the implied movement of the oil price, we have a sharp fall relative to our December forecast, but one that is partially reversed over the next couple of years. When that is mapped through the forecast that we have for consumer price inflation, it drops down sharply to zero, but after a year the base effect of the lower oil price and lower fuel price drops out and is no longer pulling the inflation rate down, so that inflation kicks back up again relatively quickly to about 1 per cent from zero.
10:30Then, as you say, we have inflation moving relatively slowly back towards the Government’s, and the Bank of England’s, 2 per cent target by the end of the forecast. The bank may have inflation moving back more quickly, but the difference is really nothing to get excited about—the rate is back to about 1.8 per cent by 2018. The lagged effect of movements in the exchange rate is one reason why we do not expect it to snap straight back.
A key reason why most people would see that as the benign form of deflation is that there is an assumed one-off temporary gain from the reduction in the oil price, which will not go on and on falling. The oil price and the futures curve are slightly lower now than they were when we did the forecast, but not dramatically so. For that reason, we are not projecting that inflation will be negative for an extended period of time in the sort of way that central bankers and others would worry about as being a deflationary spiral.
In what you have said and in the report, you mentioned that inflation is linked to sterling’s value. Presumably a lot of that is outwith our control, as it depends on the economies of China and the United States and that kind of thing. I guess that that is quite hard to predict, too.
We assume that the exchange rate moves in line with the relationship between expected levels of interest rates in different countries and that there is a response there. We are not making any particular judgments on the exchange rate—our view comes from where the exchange rate is now and where the expected levels of interest rates are going to be in different countries.
Different measures of inflation matter in different parts of the fiscal forecast. For example, consumer price inflation will matter in the uprating of income tax bands, and retail price inflation will matter for the cost of index-linked gilts and for the uprating of excise duty. The mapping from what is going on with inflation to what goes on in the fiscal forecast is not entirely straightforward.
You mentioned oil prices, which is the other point that I want to touch on. As far as I can see, I do not think that anybody predicted the fall before it happened. People were arguing about what the price would be and then it seemed to fall below everybody’s expectation. Have we learnt from that, or were there lessons to learn, about predicting the oil price, or does it just continue to be incredibly difficult and, again, probably linked, say, to the Chinese economy and how much they use?
It is very difficult—if I could predict the oil price accurately, I would be doing something much more remunerative.
In the past, particularly when we have discussed the oil receipts forecast, I have said that it is a highly uncertain and very mobile forecast and there is a very volatile path for receipts. The lesson that I would draw from that is that the forecast could have been wrong in the opposite direction and one has to make plans on the basis that the forecast is affected by the oil price, by what is going on with production and by what is going on with the level of investment in the industry.
It is the most volatile of the receipts paths by some distance and there is a whole variety of reasons for that. The swings can be quite large over time—we have seen oil receipts drop from roughly £12 billion to roughly £1 billion a year, then go back up to roughly £12 billion and then drop back to probably about £1 billion a year. The oil price is driven by global demand and supply, by particular geopolitical factors and by what is going on with particular suppliers. We could try to second-guess the futures market, but I think that this is probably the best that we have to go on.
As you know, when we do our longer-term projections we go to quite a lot of effort to explain what the differences would be if there were different assumptions, looking at a range of forecasts for the oil price and seeing what impact that would have. It is a very volatile receipt stream and the forecast areas are enormous. The lesson to draw from that is that we are unlikely to move to a point to where it is trivial to forecast in the future, and whoever is getting those receipts must plan on that basis.
Okay. Thank you.
First of all, I want to ask about the devolved taxes, particularly landfill tax, which has not been touched on so far. I had a lot of questions ready about this, but then I saw that you had published an addendum—or a slight change—to the original forecast. Can you tell us in broad terms, based on the change that you have made, what is happening to landfill tax in Scotland over the next couple of years?
We have the landfill tax dropping from about £103 million in 2014-15 to about £90 million and then picking up a little bit after that, ending up at £99 million by the end of the forecast. That is a somewhat more optimistic view.
There was an error in the initial forecast on how the Department for Environment, Food and Rural Affairs projections for waste were being interpreted, and that has moved things back. Fundamentally, it is still a downward revision in December’s forecast, which reflects the fact that receipts in 2014-15 came in lower than expected. That pushes through the remainder of the forecast. There is also a lower projection for the proportion of local authority waste that will be sent to landfill, and that is based on the DEFRA projections, which suggest a steeper fall in the near term.
That pushes in one direction, but pushing in the other is our assumption that the tax rates are raised in line with RPI inflation; in other words, things are pushing in both directions. Towards the end of the forecast, we assume a flatter trend in waste sent to landfill. Again, because we are looking at the shares of receipts coming in, we are not basing the forecast on an explicit assumption that landfill moves in line with Government targets, here or anywhere else, which is what might be assumed in the Scottish Government forecast.
The forecast goes up to 2019-20, but over the longer term, do you see it as a tax that will eventually diminish over time? Presumably we will get to zero waste at some point, and the tax take will come pretty close to nil. Do you think that that will happen over time, or is it not something that you have looked at?
It depends on the policy setting thereafter. If the amount of landfill declines relatively steeply, there is a choice to be made about whether to raise the tax rate to offset that and how far you want to go down that route before you decide that the tax rate is too high. In the absence of a fully articulated policy in that respect, we assume that the rate rises with RPI inflation. If different choices are made, the decline will, I presume, slow as a consequence.
You have been asked a couple of questions about LBTT already, but I would like to return to it briefly. I note that your projections for commercial LBTT have been revised upwards slightly since your December forecast.
Yes. I suspect that that is on the basis of the latest information about prices and transactions, such as we have to go on.
However, the residential LBTT has been revised down. You highlight four reasons for that in table 3.3 of the “Economic and fiscal outlook: Devolved taxes forecast”; one is the “Change in LBTT rates”, which shows a fall of £47 million in 2015-16. Does that reflect the fact that the Scottish Government changed the thresholds post December?
Yes. The December forecast used the original thresholds that were announced at the time of the draft budget. We have now taken on board the new ones that Mr Swinney announced subsequently.
What does the heading “Modelling changes” mean?
It refers generally to the variety of ways of taking on board what the latest data is telling us as well as looking at the nature of the model and whether there are any particular wrinkles or whether better adjustments can be made. As you will see, the figure is relatively small.
You show property transactions down by £37 million. What is that based on?
It comes back to the question that the convener asked earlier. We have assumed that the rate of property transactions to which we return in the medium term is somewhat lower than we had previously assumed, because we looked at the years that we were averaging it over. Again, relative to the Scottish Government, we are assuming a long-term transactions rate of about 5 to 6 per cent. As a result, our forecast might be a bit lower than the Scottish Government’s, but not dramatically so.
In paragraph 3.15 of the devolved taxes forecast, you say that your
“forecast takes into account the bringing forward of some higher-priced transactions ... and some delayed transactions at the lower end.”
According to you, SDLT goes up by £11 million in 2014-15, while LBTT goes down by £20 million in 2015-16. Where in table 3.3 is that change of £20 million reflected? Is it a change in rates or property transactions? I presume it is not a modelling change. Is that £20 million split across different categories, or is it all lumped into one?
I assume that it falls within the “Change in LBTT rates” category; it is certainly not in the “Property transactions” or “House prices” categories. It might be in the “Modelling changes” category, but it probably comes under “Change in LBTT rates”.
This is a consequence of the rates having been announced some time before they were implemented. We know from painful experience of the changes in the higher rates of income tax that when people know about changes in tax rates well in advance and they can change the timing of when they pay the tax, they have an incentive to do just that.
That is fine. Have you been formally asked by Government to look at the taxes that will come to Scotland as a result of the Smith commission report, or is that something that has not officially crossed your path yet?
It has not crossed our path yet, but obviously I am waiting with keen interest to see what it is all going to look like. We want to make sure that we do the best that we can.
As for how all of this will evolve, the complexity of the job will depend a lot on the degree to which the devolution of taxes results in different policy. Scotland has had the right to move the income tax rate for a while now, but it has not done so. As a result, having to decide the assumptions for the behavioural response to a new rate has not arisen as a practical issue. The issue has arisen with LBTT, because a whole new system has been introduced, and because the system in the rest of the UK has changed, too. It is a question of considering not merely which new areas we might have to look at but whether any room for manoeuvre will in fact be used and, as a result, what sorts of issues will arise in forecasting.
You have already been asked a couple of questions about oil, but I want to return to the issue. According to table 4.12 on page 116 of the “Economic and fiscal outlook”, your December forecast for the 2015-16 figures was £2.2 billion, while your March forecast is £0.7 billion, which is a change of -£1.5 billion. Can you talk us through the “Pre-measures forecast changes” and the changes in the “Budget measures” in that table? With regard to the pre-measures changes, the -£1.1 billion change in oil prices reflects the drop in oil prices, and I suppose that there is a similar reason for the change in gas prices. How do those changes manifest themselves in the “Production”, “Expenditure” and “Modelling and outturn receipts” headings in table 4.12?
The assumption is that the lower oil price results in lower production. With regard to the outlook for production, we use the work of the Department for Energy and Climate Change, which has a model of how production is likely to evolve based on Oil & Gas UK’s survey of activity in the basin.
Given the scale of the price change, a judgment has to be made, first, about what would happen to production in the absence of the Government doing anything, and secondly, what difference it would make if the Government made any changes. Not entirely surprisingly, the production effect builds over time, as the assumption is that oil and gas production falls further below the line that was in place in the December forecast. To put it in a straightforward way, a reduction in production generates a reduction in receipts.
Expenditure moves in the opposite direction, because a lot of expenditure, such as capital expenditure, is allowable against tax. If there is, as we have seen in recent years, a burst of relatively high investment in the North Sea, the hope is that that will generate more production and thus more receipts in the future. However, in the near term, a dramatic increase in investment reduces receipts, because there is more money that the firms can offset against tax. The assumption is that the lower oil price not only encourages lower production but discourages investment, because fewer projects are likely to get over the hurdle rates; that lowers expenditure, which actually has positive rather than negative effect on receipts. Generally speaking, those two things always move in opposite directions.
10:45The “Modelling and outturn receipts” line in table 4.12 partly takes on board what the latest numbers through the year are showing. I am not sure whether it is specifically the case here, but modelling changes often relate to how Her Majesty’s Revenue and Customs looks at which fields are expected to produce what and then links that to whether the firms in question are in a likely position to be paying tax. Sometimes modelling changes can reflect a change in the view or an updating of understanding of field ownership.
So those are the pre-measures. With regard to budget measures, you forecast a -£0.2 billion static effect and no behavioural effect at all. Does that mean that the tax changes for 2015-16 will result in £200 million less?
That is right. Although the more generous treatment of tax costs money, the assumption is that, eventually, the amount of expenditure and production will increase, which is likely to have a positive effect. As a consequence of the oil price change on its own, we would, roughly speaking, have reduced expected production at the end of the forecast by about 30 per cent in the absence of any policy measures. However, taking the policy measures into account, we assume that the decline will be about 20 per cent. The policy measures are not assumed to be sufficient to completely outweigh the implications in the change in the price for production, but they partly offset them.
So we are hoping for greater production—or at least not as large a decline in production—as a consequence, and you make certain assumptions on that basis.
I should say that, given the size of the changes, there is uncertainty around both the pre-measures forecast and exactly how much impact the measures will have. We can make some relatively precisely calibrated judgments about how many projects might be shoved above the line, but there is a broader confidence issue about whether in the long term this is a sector to be investing in. Judging that for the purposes of the forecast is not straightforward, and I would not claim that we have done anything terribly scientific to that end.
So you are saying that it is harder to predict the outcomes, and that things can move in different directions.
Your forecast for the cost of the budget measures is -£0.2 billion in 2015-16, -£0.4 billion in 2016-17, -£0.3 billion in 2017-18 and so on. Are those costs slightly more predictable, or do they also rely on a range of factors?
They rely on a range of factors including the amount of activity to take advantage of the incentives. The key point to note is that the size of the specific identifiable cost to the Exchequer is dwarfed by the changes in the revenues implied by the movements in production and price. Exactly how that set of incentives is going to shape up in terms of cost is much less important than the fact that it is very hard to predict with any confidence what a change of this magnitude in the oil price is likely to generate and how much of that would be reversed by the policy measures. There is uncertainty around those numbers, but that is the least of our problems.
I am grateful for that. Thank you.
To touch a little bit more on oil and gas, I note the adjustments that you made to forecast oil receipts in the outlook for the 2015 budget compared with those in the outlook for the autumn statement. In the intervening period—between December and March—there was some upturn in the oil price. Most forecasters expect there to be a reasonable upturn across the rest of the year with some stabilisation across the piece because the view is that, other than Saudi Arabia, pretty much none of the countries in the Organization of the Petroleum Exporting Countries can sustain a significantly lower oil price than the one that we are experiencing at present. Given those factors, which are being widely commented on, why was there such a dramatic adjustment for future years in your forecasting? Will you elaborate on that?
In the assumptions that we make about the oil price that are factored into the forecast, we incorporate the change in the spot price in the near term, obviously, and then assume that the oil price moves in line with the futures curve over the next couple of years. Then, reflecting some research that was done at the International Monetary Fund, we would say that the market was too thin at that point to provide us with a great deal of additional useful information and, therefore, we assume that the oil price was constant thereafter.
If we look at the way in which things moved between December and the March forecast, we see that the spot price was about 27 per cent lower when we closed the March forecast. To do that forecast, we took an average over a number of days about a fortnight before we shut the forecast. However, as you say, the assumption in the markets was that some of that decrease would be reversed in the future. Therefore, the oil price that we assume for the end of the forecast is only 17 per cent lower, rather than 27 per cent lower, which takes into account the effect that you mentioned.
Clearly, there is a wide variety of other forecasts, some of which are relatively technical and data based and some of which draw prognostications about the world economy and geopolitical events. We have highlighted the matter in the past using the US Energy Information Administration’s projections, which give us an enormously wide range around any central forecast, but we do not think that there is a strong case for saying that we would be in a position to second-guess the futures market over that sort of time horizon.
That is what we have plugged in and that is why it shows exactly what you say: the oil price declined and has picked up a bit. I have not looked at it closely recently, but I think that the decline now as distinct from the position in March is because the spot price and the futures curve are lower still but not by a dramatic margin.
There are a number of international analyses of oil price that are not borne out in what we see now. What value do you attach to five-year forecasting given that, between December and March, you had to alter your five-year forecasting radically? When do you anticipate reconsidering that five-year forecast and making any necessary adjustments to it?
We update the forecast whenever there is a fiscal event. Under normal circumstances, that would mean that we would come back to it at the next autumn statement, which would be late November or December. Of course, that might be different this year because, if we have a new Government and it decides to have an additional budget relatively early on in the summer—as the incoming coalition did in 2010—we might return to it earlier than that. It would depend on whether an incoming Government wanted to announce a package of measures, whether that needed a new forecast or whether the Government was happy to use the basis that we had before. We would come back to it at that stage.
On the value of five-year forecasting, Governments have medium-term fiscal plans. There is merit in producing a five-year forecast and being able to set out a fully articulated view. However, I have done numerous talks to the effect that we need only look at the scale of the jumps up and down in oil receipts to see that we have been consistently overoptimistic about the level of them since we began—as you can see from all the blue lines lying above the black one, which has continued to fall, in chart 4.5 on page 114 of the outlook—although there were people around who said that we were undershooting. The lesson that I draw from that is that I would not expect anyone to get that line right very often.
You make some commentary on the fiscal mandate at paragraphs 1.13 and 1.14. It appears that the fiscal mandate now applies in year 3, rather than in year 5. Is the effect of that that the cuts in public spending that would have taken place over a five-year period are now expected to take place over a three-year period? Is that what is anticipated as a result?
No, it is not. In terms of the forecast of the deficit, the Government is not saying, “We want in year 3 to get to where the OBR was forecasting we would be in year 5.” If that were the case, as you say, we would be concertinaing the remaining fiscal consolidation into three years, rather than five years, so that the fiscal consolidation as set out and implied in particular by the medium-term path of public spending, would continue beyond the fiscal mandate date, and there would be another year of spending squeeze in 2018-19. Our forecast for that year has not moved dramatically since December.
Basically, you can think of it as the Government saying that, at a five-year horizon, it looked as though we would be overachieving this forecast by a significant margin, whereas in the third year we would be expecting to overachieve it by a smaller margin, so now let us aim for the third year instead. It is not trying to do five years’ work in three years but is saying that year 3 is the year on which we should focus people’s attention, so that we are not doing dramatically more consolidation by then.
That explains the comment in paragraph 1.14 that the Government is on track to meet the mandate with £16.8 billion to spare, compared with the previous expectation that it would meet it with £38.8 billion to spare. Is that what that refers to?
That is right. One would expect the budget balance to improve over time as the spending cuts go on and as the tax revenue picks up. The margin by which the Government would expect to hit it in year 5, if that were still the target, would now be significantly lower than it was in December, because the Government has made the decision to drop the additional year of spending cuts in 2019-20 as a share of GDP.
Is that the fiscal mandate that was essentially approved and voted on as part of the autumn statement?
That is right.
So it is a three-year period. To achieve that fiscal mandate—if one had signed up to it—according to what you are saying in paragraph 1.14, it will require there to be significant cuts in public spending either way.
I would not use the word “require”. The way in which Government policy is set out at the moment, in terms of the tax rates, the benefit rates and the overall path for total spending on public services and capital investment, means that, over that period, the current forecast implies that most of the action takes place in terms of cuts in public services spending. It is also true that, if you look over the full five years of the forecast and think about getting from a deficit of roughly 5 per cent of GDP now to a small surplus in 2019-20, about 70 per cent of that additional deficit reduction takes the form of implied cuts in public services spending as a share of GDP. That does not have to be the way it is done, but it is what is implied by the Government policy choices that we have taken on board at the moment.
Both the coalition parties would say that, if governing alone, that is not what they would want to see, and that, for different reasons, they would have less of a squeeze on public services. As I say, the Conservatives have talked about welfare and tax avoidance measures, and the Liberal Democrats have talked about tax measures and having a different degree of ambition on the overall borrowing figure.
It is “required” in the sense that that is what is required by the policies that we have been given in order to produce the forecast, but policies can always change and people can always do things in different ways or try to achieve something different.
It would necessitate spending reductions of some form. You mentioned welfare and some of us would argue that talking about welfare measures is a way of saying that cuts in public services expenditure can be avoided by cutting welfare, which may not be appropriate. Whatever the argument, it would necessitate cuts to take place somewhere in order for the fiscal mandate to be achieved.
11:00
With the room for manoeuvre or margin for error that is implied in our current forecast, that is what arises out of current tax rates, current welfare policy, the detailed spending plans for 2015-16 and the implied spending totals for subsequent years. Any of those things could be changed by a future Government. If it wished to do so, a future Government could do more on tax or more on welfare. It could choose to aim to achieve a different target or the same target with a different margin. Those are all policy choices, but we have to look on the basis of current policy.
You say that a Government could choose to alter the target. I presume that that would be different from the fiscal mandate that was agreed to in the autumn statement.
A future Government could choose to have a different target. It could also choose to try to achieve the target with a different margin for error. That would not necessarily mean having to change the targets. As we say in the “Economic and fiscal outlook”,
“On our central forecast, the Government is on track to meet its new fiscal mandate with £16.8 billion to spare.”
The choice could be made not to achieve that with £17 billion to spare, to have a different target or to have a different composition of measures to get there.
I have a brief question. The OBR’s forecast is that household debt will continue to rise. According to your papers, household debt is now far above the levels of 2008. What are your thoughts about the sustainability of that? Do you expect that to stabilise at any point?
There is a slightly less steep increase over the course of the forecast. For those of you who have the “Economic and fiscal outlook” in front of you, the forecast for household debt to income is on page 73. As members can see, the figures remain broadly flat until 2016 and pick up thereafter. Primarily, the debt to income ratio rises not because that is required to fuel consumer spending over the period—that is more dependent on what goes on with productivity and wages—but because it is assumed that house prices will grow more rapidly and transactions will recover. Therefore, quite a lot of that is secured lending on housing, which results in a corresponding change on the household assets side as well as on the debt side.
Things have been revised down since last time partly because the starting point in the latest data is lower in cash terms than it was. In addition, the growth in mortgage debt is less than it was in the previous forecast. That goes back to the point that has been made twice that fewer housing transactions are taking place and less debt is being generated as a consequence. There is also less accumulation of unsecured debt. That is because we have different forecasts for the amount that households are consuming and investing. Therefore, there is less of a push up there.
You are right about the overall picture: the debt to income ratio is rising. Given the rest of the forecast and the stance of monetary policy, that is not necessarily surprising or inconsistent, but we highlight that as one of the potential risks to the forecast. We are seeing household activity moving in that direction, and that could affect the path of the forecast.
In a sense, that forecast is a microcosm of a broader issue. On the face of it, when we started out talking about the GDP growth rates, the picture looked very stable over the next five years. Growth is chugging along at 2.5 per cent a year, inflation is falling in the near term but basically heading back to the target, and interest rates are not moving terribly far over the forecast implied by Government receipts.
However, there is quite a change in the composition of the forecast, because there is a substantial fiscal consolidation continuing throughout. A useful way to look at it is to consider the various balances of net lending by each individual economy in the rest of the world and how all those things have to add up to zero and, for that to be the case, there is an implication for household debt.
We have a relatively robust improvement in business investment and some improvement on the overseas balance, although the trade deficit is not really improving very much and remains a modest drag on GDP growth throughout. The income balance is improving. The latest figures suggest that, in 2014, the current account balance was at its biggest deficit since the 1800s.
We see all those things moving over time. On the one hand, there is a forecast that looks very stable at a headline level and, on the other, if you look at the changes in the household balances and other balances in the economy, there is quite a lot going on under the surface.
Finally, there has been a change in the labour market generally, with different employment patterns—we have more self-employment, a growth in zero-hours contracts and more of what feels like less stable employment practice—but how is all that factored in? How do you forecast such things?
As we discussed earlier, that is a key factor in terms of understanding and projecting what is going on with income tax receipts in particular. It is precisely one of the reasons why income tax receipts have been relatively disappointing, partly because headline employment has been performing better than we anticipated and earnings growth has not been picking up as we have anticipated in a series of forecasts. On top of that, there is the fact that some of the issues that you have highlighted mean that we have got lower receipts out of the increasing employment, because more of the self-employed appear to be on relatively low incomes than is normally the case when there is a rise in self-employment income.
That said, we do not do an extremely detailed breakdown of the labour market forecast in terms of exactly what sorts of employment and contracts are going on. Basically, we make a judgment on what the unemployment rate consistent with stable inflation is in the medium term. Our view on that is not very different from that of the Bank of England, which is that we assume that it gets back to about 5.5 per cent. We assume that the unemployment rate will not fall as fast over the remainder of the forecast as it has done over recent years, which is the flipside of saying that we think that the productivity growth that has been absent in recent years will hopefully return. As we have underlined here, that remains one of the biggest uncertainties in the forecast: if and when normal service will be resumed.
Thank you.
That has exhausted questions from other committee members, but I have a couple more before we finish.
On page 178 of the “Economic and fiscal outlook”, you say:
“the UK began the period with the second highest deficit (after the US) and ended with the second highest (after Japan)”.
You go on to say:
“The contribution of lower spending to that fall was the largest among these countries. The UK was the only country where the deficit has not been reduced by having revenue growing faster than national income.”
Going forward, is that how you see the relationship between the UK and other countries? Is there still a significant difference between how the UK is approaching the issue and how other countries are tackling it?
We have looked back at what has happened over the last few years in terms of making the international comparisons, but we do not do detailed forecasts for all the other countries. As you will see from the comparison that we have done, receipts have made less of a contribution in the UK than in most other countries. There are a number of reasons for that, which we have already covered.
The coalition Government has announced additional tax increases, notably the increase in VAT early on. However, about half of the gross tax increase of 3 and a bit per cent of GDP has been handed back in the form of other tax cuts, notably the increase in the income tax personal allowance and the reductions in the headline rate of corporation tax. The remainder of the gross tax increase has been swallowed up by the disappointment on the effective tax rates for income tax that we have just been talking about and the cut in oil receipts and so on.
Looking forward to deficit reduction over the next five years, we will get some more in from receipts and, hopefully, a return to earnings and real wage growth, and there will be some fiscal drag. On the other hand, the effect of past policy measures will move in the opposite direction and some recovery in the housing market will be combined with the rates on the new SDLT and LBTT. However, on this forecast, 70 per cent of the deficit reduction over the next five years will still be from reductions in implied public services spending.
Whereas it was 82 per cent in the current Parliament. Would that be right?
It sounds plausible. I do not have that number.
That is the IFS figure. Are you of roughly the same view?
Depending on precisely what measure you look at and over what time period, it is either that or somewhat higher. There was more of a contribution from capital spending cuts in the past than we will have in the future. The welfare changes are delivering some more of the consolidation looking forward, whereas in the past welfare spending had risen quite sharply as a share of GDP over the course of the crisis and the early recovery period, partly because inflation remained relatively high while earnings growth was relatively weak.
Indeed. The last issue that I want to touch on is the OBR itself. Edward Troup, who is the second permanent secretary at HMRC, told this committee on 21 January:
“We measure and forecast, and the published forecasts are signed off by the Office for Budget Responsibility, but we do most of the leg work on forecasting, and the analysis is done internally within HMRC.”
He continued:
“Although the OBR has been praised for its independence, from our perspective, the process feels very much the same as it was when the Treasury was doing the forecasting—we had the same conversations with colleagues in the Treasury, and the Treasury would make those forecasts. Both then and now, it is HMRC that provides the underlying data and the first cut of the forecasts for discussion.”—[Official Report, Finance Committee, 21 January 2015; c 43-45.]
Do you have any comment on that?
He is right in the sense of the “first cut”. For all the individual forecasts that we do, we basically provide HMRC with the economic forecast, different bits of which matter for different taxes. What is going on with labour income matters for income tax et cetera. HMRC cranks the handle on that and comes back to us, say in the weeks running up to a budget or an autumn statement, with what Edward Troup describes as a “first cut”.
What then happens is that we have very detailed discussions in which we tell HMRC how we want it to change those numbers. I suspect that in the old days, HMRC had conversations with the Treasury, which also told it to change the numbers that it came to in the first instance, but perhaps the Treasury told HMRC to change the numbers for different reasons from the reasons why we tell it to change them. That is the whole point of setting up the process in the first place.
I do not think that Edward Troup meant that HMRC basically comes to us with some numbers and we say, “Yeah, that looks fine,” toss them to one side and go off for tea. It is our forecast; we tell HMRC what the forecast is. If HMRC wants to have a new model to predict what a particular tax will generate, it has to come to us and tell us what it is intending and we say whether we think it is sensible or whether it should go back and think about it again or suggest that it double runs it for a while until we are happy with it. Then there are all the judgments about how you interpret recent history and the numbers as they come in during the year, which is obviously the administrative data that HMRC has, such as whether something is news or noise, whether it is something that you will want to push forward into the future years of the forecast or whether it is a one-off distortion that will come out, and what you want to assume about how much change and avoidance are going to take place.
The model for the OBR, compared with, say, the larger fiscal watchdogs, is that we have a small group of people and a legal right to the time, effort and assistance of HMRC and the DWP on the welfare side—they are the ones that matter most. However, the key point is that these are our forecasts. HMRC knows that these are our forecasts. That may well condition what sort of first cut they bring to us, as distinct from the sort of first cut that it might have brought to the politicians in the old days.
11:15We are very grateful for the work that HMRC does. There is a meaningful degree of arm’s length between it and the Treasury and Treasury ministers. When it brings us a first cut, it does not have the whiff of political interference about it. It may be something that we want to change a lot but, as I say, at the end of the day, it is our forecast, so we do it the way that we want to and we make the judgments. However, the fact that the information is brought to us by HMRC rather than by ministers’ direct representatives is symbolically and practically important, and it conditions the behaviour of everyone in the process. That is a useful feature of the system.
It also helps that HMRC has the ability to use taxpayer confidential information, because neither we nor the Treasury can see such detailed information. If a forecast was being done only out of the Treasury on, for example, corporation tax, that can matter quite a lot if you are getting a relatively large amount of revenue out of a relatively small number of taxpayers.
There is a different methodology—for the time being at least—on LBTT forecasting because we can use HMRC’s knowledge of the detailed micro data in a certain way. I do not think that the Scottish Government can see that micro data, although we would not see it either. Obviously, as you go forward, there is an issue about where the forecasting activity resides in relation to the Scottish Government, Revenue Scotland and the commission. I know that that issue was raised by the committee and in the paper that was put out last week. No one-size-fits-all model works for everyone.
Speaking personally on how we are doing the job, I take comfort from the fact that, at the end of the day, I am coming up with a central forecast and not judging whether I am willing to accept someone else’s forecast. I also take comfort from the fact that we have HMRC as a good, robust professional organisation providing us with material that, as I say, does not particularly have the whiff of politics about it.
Thank you very much for that comprehensive answer and, indeed, for your evidence. Do you want to make any other points to the committee before we wind up the session?
No, I think that we have covered everything pretty exhaustively.
I think that we more or less have. Thank you very much once again, Robert. I will be seeing you later, but we will have a break to allow for a changeover of witnesses.
11:17 Meeting suspended.