Examination of Witnesses (Questions 120
WEDNESDAY 1 MARCH 2000
JOHNSON MP, MR
120. Mr Grice, could I put my point to you in
a slightly different way? The Governor in November, and the Deputy
Governor yesterday, in front of the Committee, said that they
were still uncertain about the reliability of the Average Earnings
Index over a 12-month period because the 12 months begins in this
period of uncertainty. Does the Treasury, in its own work on forecasting,
share that view?
(Mr Grice) We share the view that the Average Earnings
Index, like, indeed, a number of macro-economic indicators and
statistics, are not perfect. Measurement error, and so on, occurs.
What we know, in the context of the Average Earnings Index, is
that some of the improvements which were required would take a
period of time to implement, and I think it is true that we are
not yet fully through that period of time. So to that extent we
would not say "These numbers are absolutely right, we know
they are absolutely right", but, actually, that is by no
means true of most statistics. We certainly would not put faith
in any single indicator.
121. Coming, then, to a particular point that
has occurred and which does give rise to concern, one of the crucial
periods of calculation of earnings is, of course, the quarter
that ends in December, because that is the quarter in which the
City bonuses are fired in, which have an effect upon the Index.
In fact, it would appear that these figures, which were released
in February, reached the market earlier than the time of formal
(Miss Johnson) I do not think that is absolutely clear,
by any means.
122. Could you clarify this.
(Miss Johnson) First of all, ONS has an extremely
good record on security arrangements. As you will appreciate because
I know you take a keen interest in this, an awful lot of statistics
are released and many of them are market sensitive and both Dr
Holt and the other senior managers in ONS, and indeed I and the
Treasury, take any implication of a lapse in security very seriously
indeed. In relation to the point you have raised about the AEI
figures, recently we did not find any evidence that was compelling
enough to warrant a formal inquiry. Despite that, the Director
of ONS has ordered a review of security procedures in order to
make sure that the highest standards of security continue to be
123. This is specifically as a result of this
incident in February?
(Miss Johnson) Yes.
124. You are giving the Committee an assurance
that despite the fact that the markets appeared to have this information
in advance of formal publicationYou do not believe that
to be the case?
(Miss Johnson) The thing is analysts are permanently
trying to second-guess what the figures will be. In this particular
case Dresdner Kleinwort Benson put out a figure of 5.5 per cent
which, as you know, was the figure in question on the Friday and
any analyst could have taken into account the same factors that
ONS are taking into account in arriving at that calculation. They
have sophisticated statistical capabilities themselves and could
have come up with the same figure. It would be surprising if a
figure came out that none of a whole range of analysts who were
actually estimating or guessing the figure had got right. So we
would expect some people to get this right on the basis of the
work that they will do to try and second guess what those statistics
are in the run-up to the time at which they are released.
125. Nonetheless the concern that there was
a sudden increase in trades that particular morning was sufficient
to cause you or the Director of ONS to at least consider whether
there had been a leak?
(Miss Johnson) Yes, but the markets themselves were
of the view that there had not been, that it was simply a rumour
in effect that had been put out on the wire an hour or so before
126. Could I have confirmation of that from
(Mr Goldsmith) Obviously we are very alert, we read
the papers as assiduously as everyone else, and that rumour was
around and because we take them so seriously we check them out,
we always have look and we looked at what happened in the market.
We did not do that with our own competence, we took advice on
that as well from the Treasury and elsewhere, and we were unable
to form a conclusive view that there had been any adverse or unusual
market movements that would confirm the allegation of a leak.
It is at that point in a sense technically and professionally
we could have stopped because rumours of leaks and stories in
markets are not unusual. As the Minister just said, although we
could not substantiate there had been any leaks, nonetheless we
take them sufficiently seriously that we are having a look at
our own internal processes. That is the confirmation we can offer
you professionally and technically that although in this case
the evidence is not there nonetheless ratcheting the belt and
braces, mixing my metaphors horribly, we are having another look,
but it is normal professional concern rather than absolute total
evidence or even significant evidence that there had been a leak.
(Miss Johnson) Can I just emphasise that we do take
it very, very seriously and we are very concerned to make sure
that the highest security standards are maintained on these things
so it is not treated in a light-hearted fashion, even the hint
that there may be such a problem, but we believe that it is not
127. Could I then come on to what has become
quite a hot topic, the question of vacancies, because you will
appreciate that Ministers have been concerned with the particular
issue of vacancies just at the moment. Now, as I understand it,
the vacancy figures that the Government is using to reach the
figure of one million approximately at the present time is achieved
by taking the number of recorded vacancies and multiplying that
by threeby two, so that 300,000 becomes a million.
(Miss Johnson) I am not quite following the drift
of the question at the moment. Perhaps, Mr Cousins, you could
clarify. You mentioned three and then two and then appeared to
multiply the figure by three actually.
128. Can I take you to two specific expressions
of this. One is in the Government's publication The Goal: Full
Employment which was published to a wider world though, sadly,
not actually to Parliament yesterday and it says, paragraph 4,
"Record levels of vacancies" is the title of the paragraph:
"Moreover for every vacancy at job centres there are around
two others in the wider economy, bringing the total number of
vacancies in Britain today to around one million." So you
are multiplying that by three. By sheer luck, I suppose, I had
asked a Parliamentary Question about vacancies and I got the answer
from Dr Holt last week and this is how it was expressed there.
It was expressed, I have to say, rather more cautiously. "Latest
estimates suggest that about a third of all vacancies nationally
are notified to job centres. This proportion may vary between
regions and between occupations as well as over time."
(Miss Johnson) That is certainly the information I
have on this subject, yes.
129. That is obviously rather more guardedly
suggested than in the more bold and striking terms in which it
appears in the document The Goal: Full Employment. I am
just a bit concerned about the soundness of taking all the vacancy
figures we know and multiplying them by three.
(Miss Johnson) As indeed happens in many other countries,
the administrative data actually provide the only official indicator
of vacancies. That data is, in fact, amongst other things, frequent
and timely and obviously fairly readily available. But what it
provides is some detail, for example, by occupation and local
area, and there are not actually any sampling errors in it because
it is straight data as supplied, but obviously then there is a
question about what level of vacancies are not supplied to the
job centres as vacancies, and we know that that is a very high
percentage of vacancies and the estimate is about a third are
notified to the job centres, as you said, and the remainder are
not advertised to job centres but because employers are not under
any obligation, as you know, to notify them, but are in fact nonetheless
available to the market, they are privately advertised and all
the rest of it in the normal way employers deal with vacancies.
It is that basic breakdown between those that are notified to
job centres and those which are not which provides the underlying
arithmetic which leads to the figures you have quoted earlier
on. That is the view of ONS and I am not here as a Government
Minister to question the basis on which the ONS operates. Alan
may wish to comment on that.
130. That is perfectly proper, I should address
my questions here to Mr Goldsmith, quite clearly. Are you satisfied
that it is sound to multiply for a period of ten years routinely
all vacancies everywhere by three?
(Mr Goldsmith) I think Dr Holt's answer gave the right
signal that the soundness is less than absolute; it is indicative.
What we know for a fact is the vacancy numbers, these are administrative
data, they exist. As the Minister has just said, we do not know
what vacancies are out there that are not notified to unemployment
agencies. There are suggestions as to why do we not try and count
advertisements in newspapers and people have tried to do that
but are unable to add to that particular exercise the right degree
of statistical integrity. We are as an office looking at whether
we can access the data we get through employer surveys as another
way of coming at it. The dilemma there for us is we are very conscious
that the more demands we make on people in surveysbut it
cannot be absolutely sound. That has to be the answer.
131. It cannot be absolutely sound. Let me put
the question another way then. Why multiply by three, why not
by two and a half or four?
(Mr Goldsmith) Because I think the indications are
in so far as statistical methodologies can be applied and a little
bit of thumb in the air, that is the order of magnitude, but it
is seen to be no more than that and there are appropriate health
warnings attached whenever those conclusions are drawn.
132. That is all very well but in The Independent
yesterday we get "Nearly a million vacancies are unfilled
in the United Kingdom". There is no health warning there.
(Mr Goldsmith) No.
(Miss Johnson) We are not responsible and neither
are ONS for exactly what goes into newspapers. If we were responsible
for it, it would be a different world. Whether it would be a happier
world or not is something individual Members will have their own
views about but, indeed, it would be a different world. We are
very happy to send you a note on the methodology that goes into
the factor of three that you are interested in in essence in this
133. We have had a long night
(Miss Johnson) I had forgotten about it, you should
not mention it.
Mr Cousins: People have various ways of passing
the time and mine was to look at vacancy statistics.
Mr Ruffley: You should get out more, Jim!
134. That is what my wife is always telling
me. She means of course with her! And so let me share a puzzle
with you. One of the sets of vacancy statistics which the Government
highlighted in this report was for the North East of England,
which is of course the area I represent, and it used the figure
60,000 vacancies, so I look up the vacancy statistics for the
North East of England. Let me share with you this puzzle and maybe
in the dark watches of night I missed something and you or Mr
Goldsmith will put me right straightaway. The number of recorded
vacancies in the North East runs at between 10,000 and 12,000
over a period of several years, within that range. In May 1999
there is a sudden increase in the North East of 2,000 in the number
of vacancies and the number of vacancies adds in the succeeding
months between 1,000 and 2,000 more each month until October when
the increase stops by which time the number of vacancies in the
North East has doubledbetween May and October. That does
not seem to me to be quite right, particularly when you are multiplying
this doubled figure by three on the sort of rule of thumb that
you have got. Mr Goldsmith, tell me, that does seem a bit of a
puzzle, does it not?
(Mr Goldsmith) It is a very significant increase in
percentage terms. I would have to come back to the methodology
135. If you could explain that I would be extremely
grateful because it has a knock-on effect which quite surprised
me. The North East is the smallest of the English regions. It
contains four per cent of the population of the United Kingdom
and at the start of 1999 it had four per cent of the vacancies,
but this doubling of the vacancies in the North East alone that
occurs between May and October 1999 accounts for a quarter of
the total increase in vacancies across the United Kingdom for
1999. Does not that seem a bit odd? You are nodding.
(Mr Goldsmith) That is a valid point you are making.
(Miss Johnson) You are raising an interesting question.
It is not one to which we can give you an answer this morning
especially not after a long night, as I think you described it.
But I am happy to write to you about the questions you have raised.
136. Finally then just because the Committee
has been advised about this by the Greater Manchester Low Pay
Unit, they went off and did a study of the vacancies that they
could find in the job centres and they went round and added up
all the vacancies they could find and they could only find 40
per cent of the vacancies that were attributed to the Manchester
job centre at that particular time. This is a bit worrying, is
(Miss Johnson) Can I comment perhaps initially on
that. I think ONS does recognise there are difficulties with coverage
of some of the local figures. For example, some local figures
have been affected by the introduction of central vacancies take
by the Employment Service whereby one office acts and takes all
the vacancies for an area notified by employers. So in fact, from
June of last year ONS ceased to publish the vacancy figures for
individual job centres in the regional press releases only and
so there are some issues there. I suspect the work done by the
Manchester Low Pay Unit may indeed be affected by these kinds
of issues. I do not know if you wish to comment further on that.
(Mr Goldsmith) No.
137. Perhaps the Committee could be informed
about all of this. Clearly there is something quite odd about
these vacancy figures both in terms of how they are calculated
at the national level, how they are built up at the regional level
and how they exist at the local level too.
(Miss Johnson) We will endeavour to meet your request
for more information on this.
138. Can we just come back to the average earnings
point just for a moment, Minister. You quoted the Dresdner Kleinwort
Benson figure of 5.5 which was the eventual figure but it is true
to say, is it not, that the general market forecast was around
5.0 to 5.1?
(Miss Johnson) It was certainly lower. I cannot comment
whether it was that figure but it was certainly lower.
139. So the rumour was 5.5 at 8.30 in the morning
and that turned out to be the correct figure?
(Miss Johnson) Wellyes. What the nature of
this rumour is is something that is not clear. That is why I hesitate
to say yes to that.