Letter - MSED to CSSP re Draft Statistics and Census (Jersey) Amendment Law 202- 7 June 2024

19-21 Broad Street | St Helier Jersey | JE2 3RR

By email

7th June 2024

Dear Chair

Draft Statistics and Census (Jersey) Amendment Law 202- P.29/2024

I am writing further to the Statistics Users' Group's (SUG) letter to the Panel of 31 May 2024 in which they provided feedback on the proposed amendments that have been lodged to the Statistics and Census (Jersey) Law 2018 by the Chief Minister.

The Chief Minister has asked me to act as the rapporteur for this debate.

The SUG overall support for the draft Statistics and Census (Jersey) Amendment Law 202- (the draft Law) is welcomed.  I am grateful to the Chair of the SUG and other members who have given up a significant amount of time over several years to develop this Amendment Law

In addition to expressing their support, the SUG highlighted certain specific aspects of the draft Law to the Panel. It is considered that these issues have been addressed in the policy report that accompanies the draft Law.

Nevertheless, to assist the Panel in its scrutiny of the draft Law, the Government has set out where it disagrees with the SUG's interpretation of the effect of the current or draft Law. These areas have been noted in the Annex to this letter. Further matters of contention have also been noted in the Annex.

I hope that this assists the Panel's work, and I would be pleased for officers to provide any further information or briefing on the draft Law as you may require.

Yours sincerely

Deputy Kirsten Morel

Minister for Sustainable Economic Development

Annex: Government Response to the Statistics Users' Group's letter to the Corporate Services Scrutiny Panel of 31 May 2024 regarding the draft Statistics and Census (Amendment) Law (P.29/2024)

Comments have been provided on some of the statements made in the SUG's letter to the Panel of 31 May, set out below.

1. "More engagement and consultation with SUG on the preparation of law drafting instructions would have been helpful."

• The SUG were engaged throughout the process of developing the Law and the underlying policy, as is evidenced by the Annex to the Chief Statistician's letter to the Panel of 31 May.

• It is noted that the process was resource and time sensitive for the SUG, given that they are unpaid. We do not consider it possible to have engaged them more, particularly as they work on a voluntary basis.

2. "Overall, the Draft Law reduces the oversight function and scrutiny authority of the SUG in the Current Law while adding formal engagement and representation responsibilities for statistics users."

• It should be noted that the Statistics Council takes on ostensibly the same role

as the Statistics Users Group. It has just been clarified to ensure that it is a

more effective provision i.e. it defines what "to oversee" means in this context. A comparison of their primary function in the current Law and the amended Law is produced below:

• Moreover, the Statistics Council has taken on an extensive series of new powers and functions which are not in the old Law i.e. its powers and functions have been extended:

As noted in the policy report at paragraph 54:

"the Statistics Council's role will be enhanced by taking on further specific responsibilities under the draft Law, including:

• advising the Chief Minister on the appointment of the Chief Statistician and being informed of their suspension or dismissal as soon as is appropriate (Schedule 1);

• advising the Chief Statistician on their annual report, the plan for Statistics Jersey and the strategy for the JSS (Article 7(4));

• presenting recommendations to the Chief Statistician, the Chief Minister and the States Assembly on statistics produced by the JSS as the Council sees fit (Article 7E(2)(a));

• raising concerns and issuing public comment on any statistics which are produced by public authorities and which describe on a representative basis the economic, demographic, social and environmental phenomena of Jersey (Article 7E(2)(a));

• advising the States Assembly where it has concerns in relation to the funding, resources or independence of the Office of the Chief Statistician or the JSS (Article 7F(3)(a));

• advising the Chief Statistician on the contents of and any revisions to the Code of Practice (Article 7I(2));

• advising the Chief Statistician and Chief Minister when the Council believes that independent expert reviews of tier 1 statistics are required (Article 7I(8) and (9));

• nominating statistics to be tier 1 statistics and advising the Chief Statistician on the contents of the list of tier 1 statistics (Article 7J(1) and (2));

• advising the Chief Statistician on the process for suspending statistical reports from the list of tier 1 statistics (Article 7J(6));

• advising the Chief Statistician on the programme of and criteria for reviews of tier 1 statistics (Article 7J(9)); and

• advising the Chief Minister on whether statistical reports should be removed from the list of tier 1 statistics (Article 7K(2))."

3. "The Draft Law transfers SUG's current independent standard setting and monitoring responsibilities to the Chief Statistician who is not required to consult with the Statistics Council except in specific limited circumstances."

• The Statistics Council retains its role, except in respect of publishing the Code of Practice. This is necessary for two reasons:

a) The UNECE's best practice Guidance makes it clear that this should be the role of the Chief Statistician (UNECE Guidance, paragraph 109, pp.27-28)

b) The Statistics Council is an advisory body designed to represent statistics users. It is voluntary and does not have the expertise or capacity to develop a Code. In practice, Statistics Jersey developed the current Code, not the SUG.

• As noted above, there are now many consultation requirements that are not in the current Law. In the current Law, the Chief Statistician is only required to seek the SUG's advice when setting the policies and priorities of Statistics Jersey and when discontinuing or starting statistical reports.

• As noted above, the Statistics Council retains the SUG's role of overseeing the statistical system. It now has enhanced powers to report concerns to the Assembly via its annual report.

4. "The Current Law thus reflects a quasi-regulatory role for SUG where it is the independent standard setter (Code of Practice) and also the entity that monitors the producers of statistics (Chief Statistician and Statistics Jersey) for compliance

– a good governance structure without the costs needed for formal regulation."

• The Law establishes the SUG as a statistical advisory council – it is not a regulator. The SUG has never had the expertise or resources to act as a regulator and nor has it had the powers to do so under the Law.

• It is clear from reviewing the archives at the Legislative Drafting Office that there was no intention to provide the SUG with a regulatory role.  

• Only Malta and the UK have statistics regulators. They are not common, and it would be disproportionately expensive to have one in a Jersey context.

• It is not usual for a volunteer advisory council to set a professional Code of Practice, nor is it practical or in accordance with UN guidance.

• The SUG has never had a role in monitoring the producers of statistics – the SUG monitors the statistical outputs produced by Statistics Jersey and other public

authorities. The Statistics Council retains this role.

5. "The Draft Law transfers SUG's current independent standard setting and monitoring responsibilities to the Chief Statistician who is not required to consult with the Statistics Council except in specific limited circumstances."

• The only roles that have been transferred to the Chief Statistician from the SUG/Statistics Council are to devise the Code of Practice and to develop the list of tier 1 statistics.

• The rationale for this is threefold:

• It accords with the UNECE Guidance

• These are technical documents that should be devised by a professional statistician and the SUG does not have the level of expertise required to develop them

• In practice, the SUG did not develop the current Statistics Code of Practice for Official Statistics – Statistics Jersey drafted it. In fact, some of the key documents that the SUG were required to produce under the Law (five-year plan and list of official statistics) were never published by the SUG, despite the Law coming into force in 2018

• Under the current Law, as noted above, the Chief Statistician is only required to seek the SUG's advice when setting the policies and priorities of Statistics Jersey and when discontinuing or starting statistical reports. They must now consult the Statistics Council on a far wider range of issues.

6. "This weakens the overall governance of the statistical system by limiting the independent oversight function, leaving the Statistics Council with primarily nuclear options (annual report, Chief Minister, States Assembly) to raise concerns."

• This is inaccurate.  The provisions enabling the SUG and the Statistics Council to comment on statistics are identical in the old and amended versions of the Law. The amendment Law provides them with an additional formal means to raise concerns via its annual report, and requirements to be consulted by both the Chief Minister and the Chief Statistician. All informal powers to raise concerns quietly behind the scenes, pursuant to its functions, are wholly available to the Statistics Council and have not been limited by the Amendment.

• It should also be noted that an annual report is not "a nuclear option"; this is designed to allow the Statistics Council to report to the States Assembly in the same way that the Chief Statistician is required to report to the States Assembly. The Statistics Council's annual report provides an opportunity for it to report on its activities that year, report on its plans, and raise any concerns. This improves the ability of the States Assembly to provide oversight.

7. "In the UK, the oversight and regulatory elements are included within the statistical system – in Jersey's case, the Draft Law should accordingly ensure that the Statistics Council is considered to be a key element of Jersey's statistical system and not excluded from it."

• There is a clear policy position, as set out at paragraph 70 of the policy report and agreed by the Statistics Legislation Steering Group that there should not be a statistics regulator in Jersey. As set out in paragraph 70 of the policy report:

"The Group considered whether the Law should establish a body to regulate statistics. It found that very few countries have a statistical regulator – only the UK and Malta have been identified. In addition, international guidance does not state that a statistical regulator is appropriate or necessary. Given the relatively small scale of Jersey's Statistical System by international standards, the Group considered that the costs of introducing a regulator far outweighed any benefits which might accrue for statistics users."

• As the SUG notes, the role of a Statistics Council/Statistical Advisory Council is to sit outside the statistical system and to oversee and monitor its statistical outputs.

• This is made clear in the UNECE Guidance. It is a policy point and not a substantive issue for the Law, however.

8. "To further strengthen independence in the production of core Tier 1 statistics, the Draft Law should ensure that the Chief Statistician, following consultation with the Statistics Council, has the authority to compel the minister responsible for the relevant public authority to collect data and produce a Tier 1 statistic – the current

Draft Law allows the minister to refuse the Chief Statistician's direction without publishing written justification for the decision [see Article 7J(3) of the Draft Law]."

• As set out at paragraph 74 of the policy report:

"The need to uphold the professional independence of the Chief Statistician must be balanced with the principles of democratic accountability, which dictate that it would not be appropriate for an unelected official to hold the responsibility for determining the government's entire statistical output and that this power is untrammelled, even by elected ministers. To provide the Chief Statistician with unrestrained powers to determine what are and are not tier 1 statistics may be as problematic as it would be to provide the power, unrestrained, to ministers."

9. "In summary, the body corporate model is more robust from a governance perspective but does require more resource to implement effectively – if the additional resource is not deemed to be justifiable, a number of the key limitations of the corporation sole model can be effectively addressed by addressing the observations noted in Sections 2, 3, 4 and 6."

• As noted in the policy report, extensive consideration was given to whether a body corporate would be a more effective model. However, it was agreed that it would not be effective because:

• It potentially compromised the Chief Statistician's ability to take decisions and so contravened the UNECE Guidance

• These are expensive to run and, if purely a governance board on the lines of the C&AG Board (i.e. not a corporate body in the classic sense of a company board taking key decisions), offered limited value compared with the expense of running it

• Instead, governance for the Chief Statistician is under the Public Finances Law, the Chief Minister and the States Assembly.

• Again, the SUG seem not to appreciate that it does not provide a governance role – it is an oversight body to represent the views of statistics users across the statistical system. It examines statistical outputs and plays no role in ensuring good governance of the Chief Statistician. This role is and will continue to be performed via Government audit, the Chief Minister and the States Assembly.

10. "Article 7I of the Draft Law might be strengthened in this respect by requiring the Chief Statistician to seek approval from the Statistics Council for the Code of Practice and any proposed changes to it so that the standard setter is not appearing to "mark their own homework."

• This would not accord with the UNECE Guidance which is clear that the Chief Statistician must assess statistical standards and methodologies.

• Nor would it be practicable, given that the Statistics Council would not have the expertise to overrule the Chief Statistician.

• Finally, the Code of Practice must still accord with the key statistical standards in the Law – the Code interprets these in more detail, but the Chief Statistician cannot override the Law.

11. "A key gap is the requirement for all government departments and public authorities to cooperate and prioritise the collection of appropriate data for the production of core Tier 1 statistics – too much time is currently spent on the negotiation of priorities and drafting of data sharing agreements, putting Jersey as a whole at risk of not being able to have access to key data sets and statistics to

make more timely and well-informed public policy decisions."

• This is not a gap in the Statistics and Census Law. The collection and publication of statistics by government departments is a decision for ministers which is independent of the Statistics and Census Law. The new reporting arrangements provide the opportunity for greater transparency about any data gaps.

• The reason that data sharing agreements are required is a matter of Data Protection Law. Regarding ministerial departments, the Government of Jersey is not a single legal entity – there are multiple departments and, therefore, multiple data controllers within Government. This falls outside the remit of the Statistics Law to resolve.

12. "Under Paragraph 1(3) of Schedule 1 to the Draft Law, the maximum term length for an individual to serve as Chief Statistician is 9 years. We note that the specific wording of the transition arrangements in Article 17 of the Draft Law will result in a tenure exceeding 9 years for the incumbent Chief Statistician. If the intention is to ensure that the incumbent continues on a permanent basis and is not limited to a specific term, we believe it is important to be transparent about this intention. For the avoidance of any doubt, SUG is supportive of the incumbent's continuation in the role."

• There has been no attempt to conceal this transitional arrangement.

• The current Chief Statistician is employed on a permanent employment contract. If the 9-year term limit did apply to him then, while he would be required to step down from his role as Chief Statistician under the Law, as a matter of employment law, he would continue to be employed unless there was some other reason to dismiss him. This would subvert the purpose of the statutory term limit.

13. "The Draft Law is currently expected to be effective immediately upon its approval. Although work has already been started to define a list of qualifying Tier 1 statistics produced by Statistics Jersey, this work would also need to be done by other public authorities and will require significant engagement and resource to achieve. Consideration of a defined transition period to allow public authorities to plan for this work and be able to demonstrate compliance by a certain point may be helpful."

• There is a transitional arrangement to deal with this under new Article 22(5). Work is underway to prepare an initial list, restricted to Tier 1 statistics produced by Statistics Jersey and has been planned for months. Addition of further Tier 1 statistics can take place over time in line with the provisions of the law amendments. There is, therefore, no need to slow the passage of the Law for this issue.