Memorandum submitted by the Society of
Business Economists
1. MAIN POINTS
1.1 The Office for National Statistics (ONS)
should be independent as a Non-Ministerial Department and, as
far as possible, for its funding, though not necessarily constrained
by a distinction between funding for ongoing and new work.
1.2 It is important that statistics are
adequately funded and it would be extremely damaging if the new
arrangements resulted in any further reduction in spending.
1.3 ONS responsibilities for statistical
accountability, quality and integrity should be clear and, as
far as possible, over-riding.
1.4 The proposals for the composition, appointment,
role and duties of the proposed governing board are unsatisfactory;
appointments by Ministers, rather than with their recommendation
or approval, are inconsistent with "independence". The
expectation that the board can satisfactorily perform the dual
functions of data and information provision, as well as being
critical scrutineers is impractical and undesirable.
1.5 ONS should report and be accountable
to Parliament, though the resource and time that Parliament can
devote to such a single and relatively narrow, though vital, operational
area is likely to be limited.
1.6 Ensuring the quality and integrity of
all Official statistics (ie not just "National Statistics")
should be a sine qua non of the purpose of the proposals.
Continuing to make an exception of the RPI is anomalous and should
be rescinded.
1.7 Despite the fact that public opinions
on the performance of the Statistics Commission are not universally
favourable, the recommendation that it should be wound up (and
its functions subsumed into those of the governing board) is inconsistent
with the Commission's own reports on legislation and trust.
1.8 The definition of "National Statistics"
has from their outset been fuzzy, difficult to pin down and identify,
and inexplicably and quirkily arbitrary; finding and identifying
them has also been difficult. This needs to be reformed; such
a distinction is not made elsewhere round the world.
1.9 Levels of public trust in official statistics
measured in recent ONS/Statistics Commission surveys are but a
benchmark, there is no similar earlier yardstick, nor any suggestion
that such measurement may need to be regularly repeated.
1.10 UK data are accepted as being generally
of high quality, but rules relating to statistical pre-releases
need overhauling as they appear to have led to some of the causes
of poor public trust; though errors and major revisions, some
definitional, have not helped.
1.11 Given the Government's desire to maintain
separate departmental and devolved administrations' statistical
roles and functions, it is desirable that the best and strongest
arrangements should be made to ensure that the National Statistician
(and the staff of ONS) are able to facilitate a well-co-ordinated
system which, whilst allowing separate data series to exist where
necessary, ensures the existence and accessibility of truly UK
data. Federal states, such as the US or Germany, do have not such
disadvantaged systems.
1.12 The role of users, professional and
other, needs to be more explicitly recognised and provided for,
both in accessibility and structural presence.
2. INTRODUCTION
2.1 The Society of Business Economists
(previously founded as the Business Economists' Group in 1953,
name changed 1969) currently has over 600 members. It is the leading
organisation serving business economists in the UK, as well as
having a small number working overseas.
3. GENERAL POINTS
3.1 The Statistics Commission (ref
paragraph 1.7 above). Although the Treasury consultation document
(paragraph 3.11) cites the Statistics Commission's Report No 18
Legislation to build trust in statistics, its three possible
models for reform, and (paragraph 3.12) recommending the third
model (a new statutory commission, directly responsible to Parliament),
this seems to have been transmuted, without explanation, into
somehow making the proposed ONS Board responsible inter alia
for acting as a "watchdog" on the ONS's own activities
as well as having some responsibilities for non-ONS functions
(ie relating to the quality of relevant work in departments and
devolved administrations). This model appears to more nearly resemble
the model of an "internal audit" department, than of
an independent external auditor/regulator.
3.2 National Statistics (paragraph
1.8 above). The difficulties of navigating the ONS website, which
is currently under renovation (and has been for some time), allied
to the fact that not all departmental websites make statistics
directly available, let alone identifying obviously whether or
not a table, publication or release is sanctified as "National
Statistics", is well-known to regular users. This concept
and its applications have been relatively incomprehensible since
their introduction and need to be thoroughly reviewed. Perhaps
attempting to include the term in legislation may be counter-productive.
3.3 Measuring National Statistics.
In the course of collecting evidence to help understanding of
the subject, a physical count of the number of National Statistics
"products" currently available has been compiled from
the ONS website. This is shown in the Annex.
3.4 UK-wide data (paragraph 1.11
above). The availability and accessibility of UK-wide statistics
is a major concern for many users and the Statistics Commission
has been reviewing it but has not yet published its findings.
At a recent well-attended meeting on the consultation document,
sponsored by the Commission, the Financial Times and the
Royal Statistical Society, John Pullinger (House of Commons Librarian,
formerly at ONS) quoted a telling example of the existence of
separate and different "official" data with similar
coverage for Scotland, produced by ONS and the Scottish Executive.
Oh dear!
3.5 The implementation of the recommendations
of the Allsopp reports (Review of Statistics for Policy Making,
December 2003 and March 2004) to improve regional and more local
data further underline the importance of maintaining coherence
of the system.
3.6 Minimising the business burden of
statistics (consultation document, paragraphs 4.24-4.26).
It is clearly desirable that the net burden on business of statistical
form-filling should be minimised. However, it should be recognised
that frequently there are conflicting views within the same business
since the individuals who complete statistical forms are not the
same as those who benefit from the resultant statistics. [6]
3.7 Compliance costs of individual
enquiries have been compiled for some years and presented as sterling
totals to impress or frighten. If such compliance costs were additionally
presented to include a "cost per respondent organisation",
the resultant perspective might reveal how relatively small most
such costs are compared with those associated with corporate reporting
or taxation.
3.8 Accelerating the sharing of data
between departments and/or administrative data is greatly
to be encouraged, subject to the standard confidentiality requirements.
3.9 Trust (paragraph 1.9 above).
Trust by all users, particularly the general public, is important
and central to the effectiveness of official statistical outputs.
Although the word "trust"; appears fleetingly (consultation
document, paragraph 4.3, 5th bullet point), it is a pity that
neither the Commission's Report No 24 Official Statistics:
Perceptions and Trust (2005) nor its Perceptions and Trust,
Internal Report, February 2006 are included in the Bibliography.
Nor are any of the findings of these discussed.
3.10 Awareness of trust. Most government
statistical data are derived from surveys. It is perhaps, surprising
that a survey on trust has not been undertaken previously. Many
commercial organisations have well-established, regular programmes
to measure their "brand image" as a matter of course
to maintain their "health" and could be useful exemplars
for a regular "Trustmeter" dipstick for official statistics.
3.11 National Statistician (consultation
document paragrsaph 4.32). It should be possible to find a more
appropriate title than the proposed "Chief Statistician",
given the long history of such nomenclature being used for the
lowest level of Government Statistical Service Senior Management.
Annex
NATIONAL STATISTICS: A COUNT OF INCLUSIONS/EXCLUSIONS
BY DEPARTMENT
Source: www.statistics.gov.uk May 2006
| Number of products* included as
| | | |
| | |
| |
| National |
| | |
| Statistics | |
| | |
| Other | |
| | |
| Total | NS |
| | |
| percentage | |
| | |
| | |
| |
| | |
| |
| Departments, Agencies and Devolved Administrations
| 1,597 | 257
| 1,854 | 86
|
| Cabinet Office | 2 |
| 2 | 100 |
| Department for Constitutional Affairs | 6
| 4 | 10 | 60
|
| Department for Culture, Media and Sport |
3 | 2 | 5 | 60
|
| Department for Education and Skills | 75
| 1 | 76 | 99
|
| Department for Environment, Food and Rural Affairs
| 78 | | 78
| 100 |
| Department for International Development |
1 | 2 | 3 | 33
|
| Department for Work and Pensions | 43
| 1 | 44 | 98
|
| Department of Health | 4 |
7 | 11 | 36 |
| Department of Trade and Industry | 20
| 21 | 41 | 49
|
| Department for Transport | 36
| | 36 | 100
|
| Forestry Commission | 7 |
6 | 13 | 54 |
| General Register Office for Scotland | 60
| 7 | 67 | 90
|
| Government Actuary's Department | 8
| 4 | 12 | 67
|
| Health and Safety Executive | 16
| 14 | 30 | 53
|
| Her Majesty's Revenue and Customs | 79
| 3 | 82 | 96
|
| Her Majesty's Treasury | 3 |
| 3 | 100 |
| Higher Education Funding Council for England and Wales
| | 5 | 5 |
0 |
| Higher Education Statistics Agency | 3
| | 3 | 100
|
| Home Office | 31 | 7
| 38 | 82 |
| Learning and Skills Council | 2
| | 2 | 100
|
| Ministry of Defence | 21 |
4 | 25 | 84 |
| National Assembly for Wales | 1481
| 462 | 194 | 76
|
| Northern Ireland Department | 95
| 46 | 141 | 67
|
| Office for National Statistics | 650
| 29 | 679 | 96
|
| Office of Fair Trading | |
3 | 3 | 0 |
| Department for Communities and Local Government (formerly Office of the Deputy Prime Minister)
| 27 | 21 | 48
| 56 |
| Scottish Executive | 144 |
5 | 149 | 97 |
| The Information Centre for Health and Social Care
| 35 | 14 | 49
| 71 |
| | |
| |
| Other Organisations |
| | | |
| British Geology Survey | |
5 | 5 | 0 |
| | |
| |
| Northern Ireland Departments |
95 | 46 | 141
| 67 |
| Education | 6 | 1
| 7 | 86 |
| Health, Social Service and Public Safety |
5 | 10 | 15 |
33 |
| Employment and Learning | 3
| | 3 | 100
|
| Enterprise, Trade and Investment | 21
| 1 | 22 | 95
|
| Environment | 2 |
| 2 | 100 |
| Regional Development | 2 |
| 2 | 100 |
| Social Development | 9 |
2 | 11 | 82 |
| Court Service | 2 | 1
| 3 | 67 |
| Economic Research Centre | 1
| | 1 | 100
|
| Housing Executive | | 1
| 1 | 0 |
| Statistics and Research Agency | 25
| 20 | 45 | 56
|
| Police Service | 3 |
| 3 | 100 |
| Tourist Board | | 5
| 5 | 0 |
| Agriculture and Rural Development | 16
| 5 | 21 | 76
|
| | |
| |
| Scotland | 144
| 5 | 149 |
97 |
| Scottish Executive | 91 |
| 91 | 100 |
| NHS Scotland | 37 | 53
| 42 | 88 |
| Visitscotland | 2 |
| 2 | 100 |
| Communities Scotland | 14 |
| 14 | 100 |
| | |
| |
| Office for National Statistics by Division/Business Area
| 650 | 29 |
679 | 96 |
| Business Prices and Sales | 13
| | 13 | 100
|
| Statistical Framework/Business Register |
36 | | 36 |
100 |
| Consumer Prices and General Inflation | 5
| 44 | 9 | 56
|
| Economic Analysis and Satellite Accounts |
6 | 5 | 11 |
55 |
| Employment, Earnings and Productivity | 4
| | 4 | 100
|
| Financial and Accounting Surveys | 47
| 2 | 49 | 96
|
| Health and Care | 22 | 6
| 28 | 79 |
| Labour Market | 17 |
| 17 | 100 |
| Methodology | 25 |
| 25 | 100 |
| National Accounts | 48 |
5 | 53 | 91 |
| National Expenditure and Income | 6
| | 6 | 100
|
| Population and Demography | 256
| 5 | 261 | 98
|
| Regional and Local | 3 |
| 3 | 100 |
| Social and Vital | 105 |
2 | 107 | 98 |
| Social Analysis and reporting | 43
| | 43 | 100
|
| Statistical Framework | 14 |
| 14 | 100 |
| | |
| |
*What constitutes a product varies widely, ranging from reference
compendia such as the Annual Abstract of Statistics or Social
Trends to individual topic series. Thus this is something of an
"Apples and Pears" analysis. However, its purpose is
to indicate the extent of Departments' NS/non NS products and
their quantity; it does not reflect volume of data.
1 includes two experimental.
2 which pre-date NS.
3 one awaiting decision.
4 one documentation; three experimental.
May 2006
6
Net burden = cost to enterprises of collecting statistics (gross
burden, commonly described as "compliance cost") minus
the value to enterprises of uses of the statistics (the benefits)
Andrew Machin Reducing Statistical Burdens on Business (GSS
Methodology Series no 9, 1998, page 7). Back
| 
