Examination of witnesses (Questions 260-279)|
MONDAY 26 FEBRUARY 2001
260. Do you know something we do not?
(Mrs Rosewell) I know absolutely nothing.
261. Is that something which is going around,
that there is an expectation the Chancellor may change the target
(Mrs Rosewell) I have not heard that, except some
journalists have been mentioning it. I have no knowledge whatsoever.
I merely offer it as a possibility.
262. It was an interesting speculation. Can
I ask some questions about data and what the January minutes said
about data? In paragraph 11 of the January minutes it refers to,
"Part of the puzzle presented by the consumption and investment
data might relate to mismeasurement ... it was ... possible that
the growth rate of real consumption was being overstated. It was
plausible too that investment was being understated." Then
it goes on in paragraph 21 to say, "It was possible that
the GDP deflator might have been mismeasured." Two quick
questions for each of you. One, in relative terms are the problems
over that kind of data mismeasurement better or worse now than
they have been historically? Secondly, how much of a difference
do those mismeasurements referred to in the January minutes make
to the present conduct of policy formation? Or are these references
to mismeasurement not desperately important things?
(Mr Bootle) I am not aware of any evidence to suggest
that the problems of mismeasurement are any worse than they were.
I think that the points you rightly pick up can be emphasised
rather because of the low levels of inflation we have got down
to and the fact these small differences actually are now quite
significant with regard to inflation performance, and also we
have got in the inflation report and the minutes someone picking
over them in great detail. After all, we have had plenty of occasions
in the past when GDP numbers themselves have been well out. In
1988 the first estimate of GDP growth for that year is now shown
to be 1 per cent off from the latest revised figure. That is one
of the reasons, after all, why the boom in the late 1980s got
as bad as it did.
(Mr Miles) I think that is absolutely right. Revisions
to data happen all the time and historically there have been lots
of huge revisions to GDP and to trade flows particularly. I have
not seen any evidence that the scale of revision is larger now
than in the past.
(Mrs Rosewell) I think it is right that the scale
of revision has always been a serious problem. It is very important
to remember that the revisions have now become relatively more
important because instead of dealing with the 15 per cents we
are down to the 2 or 3 per cents. But it could also be argued
that the difficulty of measuring the economy has increased over
the last few decades. In the old days you counted "things"
and most of the effort in statistics was put into the manufacturing
sector where it was also much more productive to put it. Now you
are trying to measure things which are in principle more difficult
to measure, and it is only in the last five or ten years that
we have had quarterly measures in the service sector at all, and
statisticians are still struggling with where the margins of error
are in those and therefore the adjustments they make to first
estimates. Secondly, the advent of computer systems and investment
in computer systems has also substantially confused where you
put consumption and investment as far as corporates are concerned.
Investment in software has traditionally been dealt with not as
investment but as consumption of servicesis that right,
is it not right, what do you with the sharply falling price of
computer supplies and how do you adjust for those within the construction
of different indices? So the scope for error in some of these
series has probably substantially increased over the last decade
or so, and it is not yet clear how you escape from that either.
(Professor Scott) Forgive me, I cannot remember the
discussionI normally avoid detailed discussions of deflatorsbut
I think this question is absolutely about IT. The US has changed
its way of measuring IT, software is now included as an investment.
You have difficulty dealing with commodities which have substantial
increases in quality and big falls in prices, and that is going
to cause all sorts of problems for consumption, investment and
GDP deflators. It is less of a problem for the UK because we do
not make so many computers but it does raise the issue of what
the trend rate of growth is in the economy. Sushil Wadhwani is
on record as saying he thinks we do not measure the economy very
well, PRS has said that if it adopted the methodology the US has
just adopted which boosted its growth rate, UK growth would be
much faster than it is now. It is a slightly two-edged sword because
if you are mismeasuring the economy then growth is actually faster
than we see, so if you are a supply-side structural change person
it is likely you are not actually observing properly the rate
263. Do any of those issues impact on Monetary
Policy Committee decision-making? Do they say, "It is a zero
gain" and just get on with it, or is any change likely to
take place as a result of Sushil Wadhwani's arguments? Is there
going to be a sea-change in the way we account for these mismeasurements
which, as you say, have historically always been there?
(Mrs Rosewell) I am not sure it does make any difference
to the way you view the economy. If you are used to saying, "The
economy as we measure it has grown by a 2.25, 2.5 per cent trend,
and that is how we expect it to go on in the future", then
the fact it actually grew by 3.25 per cent or whatever instead
is just an adjustment factor. The difficulty comes if you think
that factor is changing. If we have been mismeasuring it, we can
go on mismeasuring it. It is a bit like the CBI survey which asks,
"Are you more or less optimistic than you were four months
ago?" This is not actually the question which anybody answers.
Most answer the question, "Am I optimistic or not at the
moment?", but the question has been asked in the same way
for so long that you can adjust for that and the thing is not
to mess up the series by changing the question. That is actually
better than trying to correctly measure what it is you think you
want to measure. But if people are changing the way in which they
count output and that is getting more intense going forward, then
that would make a difficulty. The problem is of course we do not
know whether that is the case or not, but if you think the world
is getting more virtual, if you like, and more computer intensive,
harder to measure, then it will get worse going forward and that
does make a difference.
264. But not an immediate cause for concern?
(Mrs Rosewell) No more an immediate cause for concern
than it was five years ago. It is something you can continually
adjust your way out of.
(Professor Scott) Having sat in on a lot of macro-monetary
forecast meetingssign of a misspent youthone of
the key questions you have to look at is, is the data sensible
and is it being badly measured? If the data is wrong, then you
are not going to have the same relationship you have seen in the
past, so that question is important. Is it a reason to say, "I
do not believe the central forecast that comes out of the macro-monetary
model", and I think there are grounds to be more suspicious
than in the past. That is very germane to why inflation has been
under-shooting for so long. That is the key question.
(Mr Bootle) There is one issue of disagreement in
the stats which I think the Bank ought to be concerned about,
and it is one they do refer to and I mentioned it in the last
point of my note, namely the different picture given by different
measures of inflation. We are well-used to the HICP figure coming
out differently and the discrepancy has not altered a great deal
over the time we have been discussing it, but it is right to note
that the consumer expenditure deflator in the third quarter of
last year was only 0.6 per cent, the lowest for 40 years, and
the GDP deflator covering the economy as a whole was only 1.6.
This, again, relates to my earlier point about looking at the
evidence before you, imperfect though it is, and being guided
by the numbers. The fact those figures are as low as they are,
a good deal lower than the RPI, I should have thought would raise
a few questions, in my mind anyway, about the extent of inflationary
pressures in the economy at the moment.
265. Table B6 in the inflation report, which
has been referred to already, shows quite a wide range of views
as to where inflation will be in two years' time, and yet the
vote to lower interest rates was unanimous. How do you explain
(Mrs Rosewell) I explain it by you having two different
views as part of the same process. So there is probably a group
who would have wanted a bigger cut, could not get it, so converged
on the smaller cut and that became unanimous as you had the series
of votes going through the process.
266. But the implication of seeing inflation
higher would have been that the cut was not required.
(Professor Scott) This was pre-dating the recent vote.
I was not suggesting you should ignore current inflation. The
inflation report says there are three things you should worry
about: one is the current low level of inflation, one is strong
consumer bank lending, and the third is the labour market. We
have seen in the most recent months no sign that wage inflation
is rising, if anything it is calming or easing off, and for me
that is what has accounted for it. If you think there is supply
side growth which is very robust, you can still think the economy
will grow too fast and you can have an overheating of the labour
market. I think it was this very slight easing of the labour market
which converted that 5-4 vote, combined with the current low level
267. Any other views?
(Mr Miles) I think Bridget is right. If we had been
told what the result of a vote wasif indeed such a vote
took placewhere the proposal was to cut by 50 basis points,
I suspect very strongly you would not have seen 9-0.
(Mrs Rosewell) It was probably 5-4.
(Mr Miles) Quite possibly.
268. The earlier discussion has been questioning
the fan charts and there is the possibility that the central projection
in these fan charts for inflation can represent a minority position
on the MPC. Is there a form of presentation of these inflation
forecasts which would be rather more meaningful and transparent
than the present arrangement?
(Mrs Rosewell) I cannot think of one. That was my
answer to an earlier question to Mr Kidney. By disentangling the
fan charts, which is something to do with the general uncertainty
around a particular set of assumptions and what might be the uncertainty
around some alternative set of assumptions, we could end up with
nine different fan charts, one for each member of the Committee.
You could then superimpose those. There was a paper in the quarterly
bulletin 18 or so months ago talking about having different fan
charts, which did suggest there was something along those lines
going on. What they have seemed to try to do here is disentangle
it to some extent by having this Table B6 about the alternative
assumptions along with a fan chart and there is a fan chart to
go with those assumptions which might be a completely different
fan chart, so unless we want to go down the road of having nine
different fan charts or some group of fan charts I think we are
probably stuck with this as the best way of doing it. It is an
advance on having simply one line and nothing else.
(Mr Miles) There is another table which you could
in principle have, which is, without revealing who they are, simply
nine different estimates of the most likely rate of inflation,
say, two years ahead. You cannot infer anything like that from
this. What we do not know from the fan chart is whether the spread
of possible outcomes and skewness is a view everybody has or whether
there are some people who think inflation would be a bit above
2.5 and then another group who think it will be a long way beneath.
You cannot work anything like that out from this stuff, but we
could have nine different numbers.
(Mrs Rosewell) On unchanged interest rates.
(Mr Miles) Yes.
(Mr Bootle) Do we know enough about the process by
which the fan chart is generated? I do not mean the technical
stuff on computer models, I mean the to-ing and fro-ing, or rather
lack of it, between the Bank staff and the MPC members? Presumably
what happens is that at some stage or another a draft fan chart
of the likely future appears, but what actually happens thereafter?
This is a genuine question. I do not know what happens, I do not
know whether anyone in the public domain does know what happens,
but this seems to me the critical question. One could imagine
either extensive scope for to-ing and fro-ing and discussion about
what the central view of the fan chart is, and in another state
of the world it might be a rather cursory procedure under which
Committee members think, "What is the point of engaging in
a great argy-bargy about the central forecast because the real
issue is, having got a central forecast, what do we say about
what we think the risks are and which way we are going to vote?"
It is not clear to me that we know enough about how this central
view is made.
269. That is clearly a question we can ask the
MPC tomorrow morning.
(Professor Scott) Given the divergences amongst the
MPC members, the current fan chart does not seem to reflect the
spread of views of MPC members. The chart David is talking about
is a great chart and hopefully it should be incorporated into
this chart. This is a great way of translating a multi-personal,
multi-national conversation into one chart, but if there is very
strong editorial control it may not reflect the nine views.
270. Don Kohn suggested to us earlier that in
fact they had started by spending all their time on the building
blocks of the chart, the technical bit, but now they spend more
time on the chart at the end than they do on the building blocks.
That is what he said to us.
(Mrs Rosewell) It might be worth revisiting the description
of how the fan chart which was published was put together, and
then discovering how far it is now different. I think there is
some thought that as new members of the Committee have come on
it has, as you say, become less of a technical process and more
of a risk assessment process, and in that case what is the relationship
between the fan chart and table B6? Perhaps that is the key question
271. In Mr Bootle's note to us, it says, "One
is left with the suspicion that the central forecast is being
driven by hard line members of the Committee and that it does
not fairly reflect the balance of views on the Committee as a
whole." I wonder how that grabs the other three of you as
(Professor Scott) It is clear there are two forecasts
and the absolutely critical point is how you build uncertainty
around the forecasts. I think one part of the Committee is saying,
"Either the world has not changed, in which case we are right
in doing what we are doing, or the world has changed, in which
case inflation will be lower than the target, and that is a bias
we have to put up with." I think that camp is a strong one.
I do not think they are necessarily saying there is only one view
of the world, but saying, "This is how we are going to deal
with the uncertainty around the forecast." The other group
who feel the world has changed feel you can be much more relaxed
about financial controls. I think I would put it in a slightly
different way. You may want to use the same terminology but I
think we can put it in a slightly less oppositional way and more
rational decision-making way, how do you deal with risks around
the central forecast. If there are two different camps, it is
clear which side one of those camps is, and that may well be connected
with the idea that an under-shoot of inflation is better than
an over-shoot of inflation in the early years of forming credibility.
That is hard line but it may be a justifiable hard line. It is
not a pure inflation matter, there is a reason for it.
272. The other two of you?
(Mrs Rosewell) I think that there is a group of hard
liners who are characterised by a combination of thinking that
the world is the same sort of place as it used to be and as an
implication of that that inflation is around the corner all the
time, the historically determinist group perhaps, and that is
probably the one that is represented most strongly both on the
Committee and in life for that matter, and that you could argue
is behind the sort of fan chart we have here. The minority view
is then the view that you cannot rely on history to describe what
is going on at the moment, there is a change in behaviourof
which indeed the existence of the MPC is one instanceand
that inflation is not ready to re-ignite around every corner,
and that view is not fully represented in these fan charts. If
you want to convey that in terms of probabilities and risks you
can do so, but I doubt the individual members of the Committee
would do that as it happens. They have a coherent picture of the
way the economy works, and that is one coherent picture, and there
is an alternative coherent picture to the way the economy works
and they are alternatives. If that is a correct representation
of the schools of thought if you like within the Committee, but
that is perhaps something we do not actually know. If that is
the case, they are alternatives, and it is not a question of the
balance of probabilities. The balance of probabilities might be
used by some third group standing outside these two, to judge
between them on probabilistic grounds, but I do not think you
can use that as a sort of get-out within the Committee itself,
if it is the case that my characterisation of these two views
is the correct one.
(Mr Miles) I think it is possible to over-estimate
the degree of disagreement on the Monetary Policy Committee. I
suspect there is wide agreement thatand indeed the charts
show itinflation is very likely to be below 2.5 per
cent for most of the next two years. Nobody is dissenting from
the proposition that there is more chance that it will be under
2.5 per cent than above it two years from now. If there is
a disagreement, I tend to agree with what Andrew Scott said, which
is, that it is about attitudes to the risks and do you seem them
as symmetric? Do you consider inflation being at, say, 4 per cent
as being as bad as inflation being at 1.5 per cent? How bad
you think inflation being at 1.5 per cent is depends very
much on the wider economic environment. Inflation being at 2 or
1.5 per cent and staying there in an environment with unemployment
very low and growth continuing, is completely different from under-shooting
the inflation target in an environment of stagnant output and
rising unemployment. I suspect, and I have no strong evidence
for this, there are several members of the Committee who take
the view that keeping interest rates where they are is more likely
to lead to an under-shoot even two years ahead than it is to an
over-shoot, but given this is the first cycle we have gone through
with an independent central bank it would be a bad outcome to
end up at the end of the cycle, a couple of years from now, with
inflation above 2.5 per cent. They would rather have it a
bit beneath it especially if you were not paying a high price
in terms of unemployment or stagnation to have inflation there.
273. Who are the hard line members of the Committee?
Can we just infer it from the voting records or is it some rather
more special status in terms of driving the process forward?
(Mr Bootle) I think the voting records will tell you
quite a lot but what I am suggesting has to be seen in the context
of the comments I have just made about what I think is a considerable
degree of opacity about the fan charts. This is really precision
disguising opacity. Putting those two things together, the person
I think who needs to be questioned most intensely on this whole
process is none other than the Deputy Governor, Mervyn King, and
I am suggesting he is the person to whom some of the questions
that David Miles has been raising should be addressedlike
how much value you place upon under-shoot versus over-shoot. I
do think this is a fantastically important question that David
has raised, and it is not obvious to me that his description of
what is permissible within the existing regime is permissible.
I thought there were clear statements both by the Chancellor and
by the Governor and other Committee members to the effect that
the target was perfectly symmetrical. There is an issue as to
what costs are incurred by under and over-shoots but I think it
is quite clear in the remit that 1.5 per cent is to be regarded
as bad as 3.5 per cent. I very much agree with the thrust
of what David is saying, that given the youth of the whole regime
if one were, as it were, at all politically-minded, with a small
"p", sitting on this Committee and thinking about the
health of the economy and the sustainability of the regime going
forward, one might well interpret things somewhat differently.
One might be secretly rather pleased about the idea of under-shoot,
but my understanding of the remit is that it is not within the
remit of the MPC.
(Mrs Rosewell) I think that is absolutely right.
274. If we got to a situation in which the famous
letter had to be written because of an undershoot, what would
that do for market sentiment? What would the reaction of the markets
be to that?
(Mr Bootle) I do not think myself the markets' reaction
would be that significant. After all, writing a letter as opposed
to not writing a letter is a difference of 0.1 per cent. I think
technically it has to be 1.4 or otherwise it would be more than
0.1 per cent outside the target. I think its significance, dare
I say it, is rather more in this domain and with regard to this
whole question of the symmetry of the target and the suspicion
that perhaps, in fact, the operators on the Committee are not
regarding it symmetrically. It is really rather in that context
that I think it is going to be significant. I do not think the
markets are going to be much bothered by the fact that it is 1.4
(Professor Scott) I suspect they feel that it would
give more weight to the soft liners rather than the hard liners.
It would not have a big effect. The news would be in the inflation
numbers which they would know already, but they would probably
feel it would constrain the enthusiasm of the hard liners, so
it may alter their expectations.
(Mr Miles) I think it depends what is in the letter.
I believe it is an open letter that we all get to see.
Chairman: You will be able to have hearings
275. It would be rather exciting, a constitutional
(Mrs Rosewell) Excellent.
(Professor Scott) Yes.
276. Hearings on a letter.
(Mr Miles) It could be very short.
(Mr Bootle) In a sense I think how significant the
letter proves to be is partly, as it were, up to you as to how
you interpret it. It seems to me quite important that the Governor
has on numerous occasions said he does not think it is significant
at all, that he is surprised he has not had to write umpteen letters
already, he has got them stacked up ready to go, it does not really
much matter. I think that it might matter depending on the context
of inflation. The fact that we have been undershooting for the
best part of two years and then he would have to write a letter
is, I think, altogether different from writing a letter out of
the blue because you have suddenly had some oil shock which has
come along and upset you.
277. Professor Scott in his thoughts to the
Committee in italics is reassuring us that the undershoot has
been modest, well within the standard error bands and has not
been accompanied by weak growth. Then he reverts to ordinary type
in which a rather different thought emerges which is, "I
suspect," that is him, "that the economic climate of
the next 12-18 months will not be so forgivingan inflation
undershoot during this time period may well be accompanied by
below trend growth." I wonder if I could ask the other three
of you how that grabs you. Do you see that as being a sensible
thought, that if the economic climate is not so forgiving in the
next 12-18 months then the consequences of an undershoot may be
rather greater future in the future than it has been in the past?
(Mr Miles) I think that the implications of undershooting
the target by a certain amount, be it by 2 per cent or 0.5 per
cent, critically depend on what is happening in the rest of the
economy. It is one thing to undershoot by that amount consistently
when unemployment is relatively low, has been falling and growth
is okay. It is very different from an environment where unemployment
is really beginning to rise and output is very sluggish. I think
whether or not over the next year the real economy will be substantially
weaker than it has been over the last few years or not is still
a very open question. We could go through the list of indicators
that are still quite strongthe housing market still appears
to be quite strong and unemployment is low and all thatand
then there is a bunch of other indicators that are slightly more
worrying. It seems to me that looking at the thing in the round
it is still pretty evenly balanced and I would not say that there
is a strong weight of evidence that the economy is going to be
a lot slower over the next year than it has over the last couple
(Mrs Rosewell) I would say that the economy will be
slower over the next year than it has been over the last couple
of years, particularly over the last year, and we are already
seeing signs of that. We have certainly seen that unemployment
has stopped falling, or pretty much stopped falling, and in some
places, in quite a lot of places now, it has started to rise depending
on which measures you take and so on. I think that is going to
work through. Manufacturing is doing a bit better because the
exchange rate has fallen off the top. The general pattern of the
economy is down, so therefore that will intensify the undershooting
that we have already got, and indeed that is what we can see in
the fan charts that we have got. Inflation continues to move down
in the forecast, as does GDP growth. Then it all starts to accelerate
when we get through this year and into 2002. That is the reason
why we all think that there will be some continued cuts in interest
rates over the coming year. Quite when and quite how much there
is rather less agreement on, but the general path is you could
stick us into the MPC meetings and the outcome might well be the
278. Do you have any thoughts, Mr Bootle?
(Mr Bootle) No.
Mr Cousins: My last question really is
about this. It is described as a speech, so let us call it that.
It is the speech of Dr Wadhwani
Chairman: The Newcastle speech.
279. The Newcastle speech, yes. He was at a
more interesting meeting last Thursday in Newcastle than the one
I was at, but that is another story. In the course of his remarks
he says this: "Our econometric models tend to let us down
when there are large changes in confidence." He finishes
by saying: "It is going to be more important than ever to
have our ear close to the ground." Now, apart from the fact
that obviously going around with your ear close to the ground
is an inherently undignified posture, particularly if the ground
does not say anything to you, are we really in that kind of situation
in which we should place more reliance on the sort of touchy-feely
ear close to the ground kind of stuff than on the econometric
(Mrs Rosewell) I think we are always in that position.
That is why I do not run an econometric model any more.