Cabinet handbook - Annex B

Standards and guidance for the management and publication of Official Statistics and Social Research

• All official statistics produced by, or on behalf of, the Welsh Government are covered by the Statistics and Registration Service Act 2007, supported by a statutory Code of Practice for Statistics, overseen by an independent regulator, and managed by the Chief Statistician who has decision-making responsibility for the content, methods and timing of statistical releases.
• The Code of Practice does not only relate to the regular statistical products produced by the Government, but also the way we use statistics elsewhere such as in speeches, strategies, Senedd questions.
• Government social and economic research is covered by the Government Social Research Code which includes a Publication Protocol for research. This is not statutory but has previously been adopted within the Welsh Government as part of five high level Principles for Research and Evaluation agreed with Ministers and published on the Welsh Government website.
• Access to statistics and research before they are published is strictly limited to a small number of officials and relevant Ministers and they cannot be quoted before publication. There is an expectation that any statistics quoted by Ministers are published transparently and equally available to all.
• Statistics and research standards in place are consistent with the Civil Service core values of integrity, honesty, objectivity and impartiality.
• The Chief Statistician, Stephanie Howarth, and Chief Social Research Officer, Steven Marshall, are available to provide advice to any Minister on statistics or research matters.

• Consistent adherence to professional standards and procedures is fundamental to building and sustaining trust in government. Rules on the handling of evidence and analysis apply to all civil servants not just members of the analytical professions.
• Adherence to these standards minimises both the actual misuse of statistics and research as well as the perception of misuse. This protects ministers from being exposed if data are inaccurate or used incorrectly and protects ministers from being accused of misusing or interfering in the production of statistics and research.
• Welsh Government analysts have 2 key roles:
  1. Providing support to ministers and policy officials in the normal course of business. This includes briefing, contributions to the policy process, monitoring of targets and performance, production of analysis and responses to Senedd questions.
  2. Publishing social research and statistics that meet the needs of society and support the transparency of Welsh Government decision making, in a manner that is open, trustworthy and trusted. These data are used for a variety of reasons including delivery (for example school self-evaluation); and holding the performance of local government, the public sector and the Welsh Government to account.

• The Code of Practice for Statistics is a statutory Code which applies to all staff within the Welsh Government and is subject to external assessment and comment by an independent regulator.
• In line with the code, all official statistical publications should be pre-announced, publicly stating when an output is planned to be published well in advance of this publication date. Any changes to the publication date must be agreed by the Chief Statistician and be solely for statistical reasons.
• Statistical outputs also come under the Pre-release Access to Official Statistics (Wales) Order 2009. This legislation sets out the rules governing the limited number of individuals who may be granted access to statistics prior to publication and under what circumstances. For all non-market sensitive statistics, the maximum period allowed is 5 working days prior to publication. In most cases, Ministers will receive briefing about statistical publications 5 working days prior to publication. The Ministerial Code requires Ministers to be mindful of the Code of Practice for Statistics and give due regard to pre-release access requirements.
• The pre-release period can be shorter than 5 days. This will include outputs with a tight production period, those produced on a monthly basis for example; and others where we rely on statistics provided by other government departments to produce our own analysis. Statistics produced by UK government departments have a 24-hour pre-release period, and the Office for National Statistics does not give pre-release to their statistics (or only in very exceptional circumstances).  
• Ministers must not use, quote or share pre-release statistics prior to their publication. This includes making any indication to others as to whether or not the statistics are favourable.
• Any statistics used by ministers, for example in ministerial speeches or debates, must be in the public domain. This is in line with expectations of transparency and equality of access to statistics. The Office for Statistics Regulation can and does intervene where it feels this principle has not been met.
• Ministers may also get access to a range of management information about their delivery responsibilities, such as NHS performance data. Where these data are published as official statistics, they cannot be used publicly ahead of release. Even where they are not routinely published as official statistics there is an expectation any public use of data or statistics includes the publication of the underlying source.
• he Office for Statistics Regulation oversees compliance with the Code of Practice for Statistics by government departments. They assess producers of official statistics to establish their compliance with the Code of Practice. They can also proactively intervene in cases where they are made aware that official statistics are used inappropriately or that the code has been breached.
• Ad-hoc statistical requests are published on a fortnightly basis. These are responses provided to external requests for statistics, where the specific detail is not available through published sources. The responses are published to meet expectations around the transparent and accessible release of statistics and data. On occasion the ad-hoc process is also used to put unpublished statistics into the public domain to support use by policy officials and/or ministers.
• The Chief Statistician is the Head of Profession for statistics in the Welsh Government. The Code of Practice states that the Chief Statistician has the sole responsibility for deciding on statistical methods, standards and procedures, and on the content, methods and timing of statistical releases.
• It has also been previously agreed that on matters relating to official statistics, public comment might be attributed to the Chief Statistician in order to reinforce the independence of the role. Statisticians have also undertaken several briefings for journalists, with Ministerial agreement, which aids interpretation of statistics. The ability for statisticians to publicly comment on statistics is a requirement of the Code of Practice.

• The term 'government research' applies equally to both economic research and social research.
• The Government Social Research Code applies across the 4 UK administrations as a voluntary code and is monitored by the central unit within HM Treasury which covers both the Social Research and Economist professions. A key part of the Code is the Publication Protocol and this details the scope of its application (principally that internal advice is outside the scope) and the basic mechanics of the publication process.
• Ministers in a previous administration put in place 5 high level Principles for Research and Evaluation. These are published on the Welsh Government website and include a commitment to the Research Publication Protocol:
  • Early consideration of evaluation - new policies and programmes will consider the needs of evaluation from the outset.
  • Effective planning of research and evaluation - the internal evidence planning process will focus on providing the evidence to support the government’s strategic priorities. An annual evidence plan will be published each year.
  • Research is published - we will publish research according to the Government Social Research publication protocol.
  • Research is acted upon - ministers will be briefed on a proposed response to the recommendations for every evaluation.
  • Value for money (VFM) - where it is possible to do so we will build in explicit VFM analysis in evaluations.
• There is a presumption towards publication as part of the Publication Protocol, and that publication should follow certain standards such as preannouncement. However, the practical arrangements are for individual administrations and departments to decide upon. A Forward Look covering all ongoing research is maintained which sets out expected publication dates and this is circulated to Ministerial Offices and Communications teams on a weekly basis The Forward Look also summarises the details of the publication arrangements put in place within Welsh Government.
  • Ministers are briefed on research 3 weeks prior to the planned publication date
  • The publication date is preannounced on the Welsh Government website 2 weeks prior to the planned publication date
  • The timings can change with the prior agreement of the Chief Social Research Officer.
• Ministers must not use or quote or share pre-release research findings prior to their publication. This includes making any indication to others as to whether or not the findings are favourable or making public decisions that are based on unpublished research.
• Any research findings used by ministers, for example in ministerial speeches or debates, must be in the public domain. This is in line with expectations of transparency and equality of access to social and economic research.
• Researchers have undertaken several briefings for journalists, with ministerial agreement, which aids interpretation of research findings.
• The Chief Social Research Officer and Chief Economist each have overall responsibility for Social Research and Economic Research respectively.