Memorandum submitted by Ian Maclean[5]
SUMMARY
1.1 The definition of National Statistics
in the Consultative Document (CD) is far too narrow, production
based, not demand driven by needs of civil society to play an
informed role in the democratic process, actively participating
in debates on key public issues and empowered to asses the performance
of the government of the day. The test is "what statistics
are needed" to fulfil the citizens requirement to relevant
as well as reliable official statistics (see 2.1, 2.2 below).
1.2 The EU Council is ahead of us with its
Plan D—Democracy, Dialogue and Debate. (see 2.3-26 below.)
1.3 The proposed Board will be unique in
its concentration of power. Structures to ensure that the needs
of users are identified, evaluated and implemented are essential
if public trust is to be gained and maintained (see 2.7, 2.8 and
2.10).
1.4 The appointment procedure in the CD
is confusing. The first part of the paragraph describes an open
selection procedure, the second part states Ministers will appoint
(see 2.9).
2. Scope of National Statistics (NS)
2.1 The Consultation Document (CD) Paragraph
1.3 accurately defines the role of NS, but then goes on to describe
NS purely in production terms, as meeting strict integrity and
reliability standards, losing sight totally of the reason why
the concept of NS is so important. A recent book on the English
Civil War describes in 1649 England was declared a "Commonwealth"—a
free state where power flowed upwards from the people, not downwards
from a monarch. The reality was very different, but there are
echoes of this ideal in the current move to transparent government
and the Freedom of Information Act, of which statistics should
have been a part. Involving the public in debates on key issues
in crime, education, health and the performance of the public
sector is the essence of democracy in the 21st century. Statistics
are an important element in achieving these goals, but to succeed
NS must be demand driven. What official statistics does society
need to participate fully in the democratic process? The Framework
Document has the uplifting phrase—a window on the work of
government. As described in the CD, NS are a very dark window.
2.2 Public disillusionment with the political
process has many fall outs, including voter apathy and direct
action. Involving the public in evidence based debate is one of
the elements in the mix needed to reverse these trends. Statistics
are not valued for their own sake, but because they are an essential
element of decision making. So let us approach the scope of NS
by identifying decisions that need to be underpinned by statistics.
2.3 Britain is not an island. Jokes about
fog in the channel, Europe cut off apart the UK is in the slow
stream when we discuss the role of official statistics in advancing
democratic debate. The European Union through its various channels
has, for several years, been actively advancing the role of official
statistics. The integrity of official statistics was guaranteed
as long ago as 1997 by clause 285 of the Amsterdam Treaty. A report
on the proposal for an EU Citizens Charter in October 2002 listed
the key issues for discussion as:
— The feeling that more must be done
to encourage the EU citizen or voter to engage in the democratic
process at all levels.
— The acceptance that, if citizens
are to be in a position to judge the performance of those they
elect, they have a right to access reliable and up to date statistics
that make this possible.
The concept of an Information Society for all,
based on access to statistics held by public sector bodies.
2.4 Then early in 2006 a White paper on
an EU Communication policy stressed the need for institutions
and governments to pay more attention to consulting the public
and feeding their views into the policy making processes. The
White paper is a complement to the Commission's Plan D—Democracy,
Dialogue and Debate. The DGINS Conference, Copenhagen, May 2005,
also discussed the challenge of communicating statistics.
A proposal for an upgraded EU Statistics Advisory
Council has just been prepared for ratification by the European
Parliament.
2.5 The Statistics Advisory Committee shall
deliver opinions and present reports on user requirements in the
production and dissemination of Community Statistics to the European
Parliament, the Council and the Commission and the Commission
shall report to the Committee on a yearly basis on how it has
taken account of the Committees opinions.
2.6 So the EU statistics law will explicitly
recognise the need to service users outside Government. Are we
going to be left behind again? There are two main options A Statistics
Advisory Council separate from the Board but reporting to it,
or build on paragraph 4.31 of the CD which states the Board Chair
is responsible for "determining the structure of its committees".
If the former, the Advisory Council should be recognised and authorised
by the legislation. If the latter the legislation should include
a clause that one of the committees should be responsible for
identifying and evaluating user needs.
Power Boarding
2.7 The attempt to operate with an external
guardian—the Statistics Commission—was not a success.
The National Statistician resented what was seen as a confrontational
rather than a co-operative relationship and the Commission lacked
teeth. The positioning of the Board inside the statistical system
resolves this problem but at the expense of weakening the perceived
independence of that system. An exchange, provided certain safeguards
are in place, worth making for greater operational effectiveness
and a unified assault on public mistrust in Statistics per
se, a much wider and deeper problem than just re-branding
selected official statistics as trustworthy. On their own, statistics
are just numbers, they need sympathetic presentation to facilitate
their use by the public in democratic debate. The ONS has regarded
itself as a quarry to which you bring your own spade and shovel.
The Board is in a strong position to start a product development
programme to bring "statistics into focus" (the title
of an EU series), converting a valuable raw material into essential
aids for decisions. The Norwegian initiative "statistics
for a general election" shows the way forward. Provided certain
safeguards are in place is an important qualification. The problem
of the invisible user still remains. The Board High Level Objectives—paragraph
4.17—repeats the production bias and introduces a disturbing
concept, meeting user needs is mentioned, but in Freudian tones
they are described as "key users" ie the Bank of England.
Recommendation, replace "key" by "agreed".
Under the Board structure there is an assessment function (4.33)
which will assess National Statistics and the output of the ONS
against the code through a unit that although located in the ONS
will report directly to the Board.
2.8 Another recommendation, add a user needs
identification and evaluation function to the Board structure.
If it is not in place in the legislation it will be a long hard
struggle to get it there.
2.9 Hail Caesar is the thought that crosses
the mind when reading the brief for the chair, more a Lord King
or a Tiny Rowlands than a figurehead and arguably a full time
rather than a part time job. It is a crucial appointment, the
top job in National Statistics and a very powerful national figure.
The selection process for the chair we are told in several places
in the "consultation document", will be by "open
and fair competition", but paragraph 4.34 on the "Independent
Selection Process", is confusing. The first part confirms
this view, but the second part states baldly that "the Government
proposes that the Chair should be appointed by the Crown on the
advice of Ministers". The Government expects that Ministers
will appoint the other non-executive members of the Board. This
apparent contradiction presumably arises from a subtle distinction
between appointment and selection, but the omens from the first
selection/appointment process for the Statistics Commission are
not good—over 1,000 applications, leading after Treasury
sifting (what an expressive term) to a single interview session
with nine candidates for seven places.
Lessons from other countries
2.10 I reviewed the practice of some 40
countries in a report for the Statistics Commission in 2002. The
main observations are that apart from a handful of countries legislation
is actively pursued but there is no agreement on content, long,
short; detailed or just principles, you take your choice. There
is, however, universal enthusiasm for a Statistics Council, ranging
widely in size (eight in Ireland, 160 in France) and composition.
The importance of wide ranging user representation is increasingly
recognised with the Dutch model (27 designated user sectors) the
most admired, The Chief Statistician is almost invariably a member
and frequently the Chair. Ministers are also members, but the
Councils are mostly advisory not executive. With its concentration
of power the proposed UK Board will be unique.
May 2006
5 Chairman, Dissemination Committee EU Statistics
Advisory Council and past Chairman, Statistics Users Council,
1989-2004. Back
| 
