Examinatin of Witnesses (Questions 204-219)
LORD MOSER
KCB
14 JUNE 2006
Q204 Chairman: Lord Moser, can I welcome
you to the Sub-Committee this afternoon, and thank you very much
for assisting with our inquiry. I believe that you may wish to
make a short opening statement to us—or not?
Lord Moser: I do not think it
is necessary.
Q205 Chairman: Perhaps I could begin
then. You were quoted in the press—but you may have been
misrepresented—as describing the Government's proposals
as "deeply flawed". What would you identify as the chief
flaw?
Lord Moser: I do not remember
saying that, but you must presumably be referring to a quote.
I approve of the Government's proposals, of the Chancellor's proposals,
in principle. I think that the details in the consultation paper
need quite a lot of work on them. I cannot remember saying that
they are deeply flawed, but you never know.
Q206 Chairman: What are the areas
you think do need more work?
Lord Moser: The main issues which
certainly need more work are the obvious ones, which I think almost
every witness who has come before you has talked about—so
you must be a bit bored. Number one, that the independence—which
I do approve of—must in some way or other apply to the entire
system of statistics. This is the key area. It is also the area
on which I will focus in the Lords tomorrow. This is not something
that can just be confined to ONS. In fact, in my view ONS is the
best part of the system. It least needs any kind of reform. I
also have always taken the view, ever since I was in charge and
even before, that it is a single system. That is my greatest concern,
therefore. The next greatest concern, which links with that, is
that the proposed board, which I do approve of, must be non-executive—not
executive. Then there are other issues. Those, I think, would
be my two main problems. Along the line there is the issue of
pre-release, which again everybody has talked to you about and
I will certainly focus on in whatever else I say.
Chairman: I am sure my colleagues will
have questions on all those.
Q207 Peter Viggers: When I heard
an announcement, which I heard simplistically as that the control
of statistics would be made independent, I was pleased and really
quite excited. However, there are two areas of distinction, are
there not? That is, between departmental statistics, which as
I understand it will, in the present proposals, remain within
the departments themselves, and another distinction, which is
that between statistics that are National Statistics and statistics
that are not National Statistics. I fail to understand how such
a system could be effective. I do not know if you are able to
help me understand this.
Lord Moser: Thank you for that
question. Can I take them in the reverse order? The issue of national
statistics did not exist in my day, when I was the head of the
whole thing. That was introduced, if I remember correctly, in
the last major reforms—the framework reforms. The idea was
that you had one group of statistics, the so-called national statistics,
which are subject to serious quality control from the centre,
even if they come out of departments. Then you have another group
of statistics which are, so to speak, departmental statistics.
At the moment, the minister of a department—say, Education—decides
which are to be departmental and which are to be national. The
view I take—which again I will refer to in passing in the
Lords tomorrow—is very straightforward. If this system remains,
which is what is proposed in the Treasury document, if we retain
a system of National Statistics versus departmental statistics,
then it is crucial in my view that the decision as to which is
which rests not with the minister, as now, but with a board—the
new board. In my view, however, that system should be abolished
anyhow. I see no merit whatever in the distinction between national
and departmental statistics. I see only demerit, because it gives
the minister the chance of saying, "This is the group which
is entirely under me". With waiting lists at the moment,
some waiting list statistics are departmental; some are national.
This increases public distrust. So I would abolish the distinction.
That is on your second question. The first question is really
the one that is at the centre of my concerns. I have always taken
the view—which is the view of every head statistician throughout
the world—that government statistics, which should simply
be called "official statistics" not "national statistics",
are a single entity. They are all linked together, not only within
the national accounts but also to each other. Therefore, they
must organisationally be treated as a single entity—which
is why almost every other country has them all in one department.
In my day, we discussed endlessly whether we should go to a single
department. I discussed it with the three Prime Ministers I served.
We always ended up by saying, "No, let us keep to our decentralised
system, because it ensures the relevance of statistics to policies".
However, it must be run and led as a single system from the centre—and
that means the National Statistician, as she is now called, and
the board. It must be treated as a single system.
Q208 Peter Viggers: You have expressed
a concern about lack of confidence in statistics, even distrust
in some circles. If the proposals as currently put forward were
to be implemented, what do you think they would do in terms of
confidence in statistics?
Lord Moser: The whole trust issue,
as the Committee of course knows only too well, is a very complex
issue. One part of it is that figures go wrong—like the
1988, I think it was, national accounts which underestimated growth,
and the Chancellor of the time blamed policy decisions on that
mistake. So occasionally mistakes will be made. Quite often they
are actually revisions, which are inevitable. That does not help
trust. In my view, however, it is a very minor aspect of the trust
problem. The real trust problem is the way the statistics are
used, above all by ministers, and the fact that ministers are
able to get at the figures so far ahead of time—which was
not allowed in my day. It is the fact that they can spin what
they put alongside the figures—done more for comment—and
all that kind of thing. The public do not so much distrust figures:
they distrust the people who use the figures and the institutions,
and all of that. Also, this is the only country in which there
is a major trust problem. I have discussed this endlessly with
other official statisticians. It does not exist in Canada, Australia
or Sweden. It is a feature of the fact that our public distrust
politicians, distrust authority, and do not like figures. We have
a national problem.
Q209 Mr Newmark: Perhaps I could
comment on that. I think that part of the problem, as you have
identified, is lack of a joined-up single system, a lack of independence,
and probably the fact that politicians are delivering the message,
not an independent body. Those are the issues that you have identified.
What powers should the National Statistician and the independent
board have over the statistical system, and how should these powers
be divided between the two?
Lord Moser: This is genuinely
difficult. Perhaps I may answer in terms of how it was done in
my day, because I think it was done reasonably well—not
because of me, but also my successors. Number one, clearly the
National Statistician, as I will continue to call her and as she
should be called—she should not be called Chief Statistician,
that is a mistake—obviously has direct executive control
over the Office for National Statistics. I do not think that is
the problem. The problem is what authority or power does she have
over, let us say, the Department of Health statistics. It is a
very difficult issue. The way it was done in the old days—and
I think it is still meant to be done—is, number one, the
National Statistician is the head of profession. So she appoints
or helps to appoint the key statisticians in the Department of
Health, in consultation with the minister, et cetera. She
is a reporting line for the statisticians in each department.
That is the first thing, and it is very important. Secondly, ultimately
the buck stops with her—as it did with me. If there is a
major problem, say with waiting list statistics, she tries to
influence—I tried to influence my person in that ministry,
my statistician, to do things differently. Quite often one succeeds,
because maybe there is a better way and he has to persuade his
minister that there is a better way. If I do not succeed with
him, I have a go at the permanent secretary and say, "Look,
the way the waiting lists are collected or analysed is misleading.
We need to have a change". Sometimes that works. Sometimes
there is a committee which deals with all social statistics with
lots of departments, and I try there. If the worst comes to the
worst and I did not succeed through influence—it is not
power, it is influence—through the statistical networks,
then I went to my boss. This is important. My boss was the Prime
Minister. It is crucial that the board, or whatever ministers
are still involved in the whole business—I do not believe
that there will be no ministers involved, because it does not
work—ultimately, if there is a real problem, they are the
end of the solution line, and then usually it did get solved.
In terms of legislation, therefore, the challenge that you must
be grappling with is, if the legislation is to cover the entire
system—which I regard as absolutely essential—a way
has to be found of defining, first, the powers of the National
Statistician and, secondly, the powers of the board in relation
to the departments, obviously in consultation with the minister
involved and the permanent secretary. It is not easy but it can
be done.
Q210 Mr Newmark: Do you favour the
establishment of a non-executive board with oversight only? Is
that what you are saying, or would you retain control through
an external audit body, such as the Statistics Commission? I am
curious as to what powers you would give this board.
Lord Moser: The powers derived
from legislation.
Q211 Mr Newmark: But the powers that
they have—should they be one of oversight?
Lord Moser: If one needs one word,
it is scrutiny; it is advising the public, advising Parliament.
Q212 Mr Newmark: So it is an external
audit body effectively? It has more of an audit function. That
is really what you are saying.
Lord Moser: It is not a body that
can decide on how to collect particular figures. It is not an
executive body. That must be the National Statistician. I do not
believe in committees running organisations anyhow. It will by
definition be powerful, because of its composition, chairmanship,
and so on, but also because it is backed by statute.
Q213 Mr Newmark: Do you think that
ministers should play any role at all in appointing board members?
Lord Moser: If it was left to
me, I would have the Prime Minister being given that responsibility.
Q214 Mr Newmark: You do not think
that it should be an independent board of people, some sort of
independent committee, that appoints this?
Lord Moser: Let me make myself
clear. Of course the board is, above all, not only non-executive
but totally independent. However, the board has to be brought
about. The document proposes that the Treasury, the Chancellor,
appoints the board members. I do not have strong views about that.
Q215 Mr Newmark: You do not think
that it should be an independent panel deciding, though, i.e.
removing it one step from the Chancellor or Prime Minister in
deciding who should actually be on that board? An appointments
committee—call it what you will.
Lord Moser: I do not feel strongly
about that. I would totally trust the Chancellor or the Prime
Minister in appointing the members of the board. It has to be
very high-powered. I think the structure that is proposed makes
a lot of sense.
Q216 Peter Viggers: When I came fresh
to this subject, I did not realise that there was a difference
between National Statistics and other statistics, and that quarterly
waiting lists in the National Health Service are National Statistics
and monthly waiting lists are not national statistics. Should
there be one standard for official statistics? Should there be
a code to which they must comply, which would make them all official
National Statistics?
Lord Moser: As I say, if it was
left to me I would abolish the concept of national statistics.
I would replace it by what we have always had and what every other
country has: official statistics—call them what you like.
The great mistake is to have a category of statistics which are
left totally to the ministers' hands. It is a formula for lack
of trust, because anybody who looks into it can see that the minister
has decided that those particular things do not go anywhere near
ONS; they are totally for him or her to decide on. That, to me,
is a very basic flaw. However, I missed the first part of your
question.
Q217 Peter Viggers: It was merely
a preliminary comment, that I had not realised the difference
between national and other statistics. I think that you have answered
the question, Lord Moser.
Lord Moser: The problem is that
nobody out there knows the distinction. What they do know is that
there are a lot of statistics in the departments which the minister
has total control over. To answer an earlier question, to my mind
the Chancellor's initiative is totally welcome. The concept of
independence is totally welcome. It has to be coupled with some
subtle way that it covers the entire system. As soon as you do
that, you make real progress; but it does not solve the trust
problem totally. For example, pre-release issues have to be solved.
Q218 Mr Newmark: You keep referring
to the lack of trust in using statistics and you keep using the
health service as an example, on waiting lists. Does it not come
down to a problem of definition and clarification: that one person's
definition of a waiting list itself is different than might be
used elsewhere? It is this consistency that you want throughout
the whole system: that if we use the phrase "waiting list",
it means the same thing to the person who runs the department
as well as the Chief Statistician.
Lord Moser: I do not think that
the lack of trust problem is as subtle as that. The lack of trust
problem is the public conception that politicians, ministers,
and others play around with these figures to suit their policies.
Q219 Mr Newmark: So you feel that
you can actually manipulate figures without necessarily manipulating
definitions?
Lord Moser: I would not use that
word, sir. I would never use the word "manipulate".
I could quote you a lot of examples from my own personal experience
where ministers, far short of manipulating or wanting to manipulate,
put what would nowadays be called a slight spin. It is not easy
to interpret figures, and it should really be left to the statisticians.
In my 10 years there were two occasions when I actually resigned
my post because ministers did try to manipulate figures. I am
not going to say which they were if I am asked, but there were
two. It happens occasionally, but I do not think that it happens
nowadays because there is too much attention in the press.
|
