Public Sector Pension Reform - Report - 12 May 2014

CONTENTS

1. KEY FINDINGS AND RECOMMENDATIONS  3
2. THE PANEL'S REVIEW  5
3. REPORT OF BWCI CONSULTING LIMITED  15

1.  KEY FINDINGS AND RECOMMENDATIONS

Key Findings

1. Adoption of the draft Law would effectively provide in-principle approval for the move from a final salary scheme to a Career Average Revalued Earnings (CARE) scheme. The rationale for such a move is compelling and the debate on the draft Law should therefore proceed, albeit with the caveat that the details of how the scheme will operate will not be agreed until the draft Regulations have been finalised.
2. Further evidence is required in respect of the affordability of the proposed employer's contribution cap in the long term.
3. Further clarity is required regarding the manner in which Article 8(1) of the draft Law would be applied.
4. It would be beneficial for Regulations under the draft Law to include, in respect of the CARE scheme, provisions in relation to investment strategies; prudent and best estimate funding assumptions; and the declaration of conflicts of interest.
5. The aim to implement the proposed CARE scheme on 1st January 2015 means that the development and consideration of the draft Regulations will be undertaken within a very tight timescale. Sufficient time must be allowed for those Regulations to be considered.
6. Further information and analysis is required in respect of the cost comparison between PECRS and the proposed CARE scheme; the sensitivity of the results to the assumptions underlying the calculation of the anticipated contribution rates; and the quantification of risks of underfunding within the CARE scheme.
7. The concept of prudence within the funding assumptions to be used under the proposed CARE scheme should be clearly established.
8. Some of the protections which would be afforded to current members of PECRS in the move to a CARE scheme are essentially unfair. They appear to have been included for pragmatic reasons to ensure the proposed reforms would be acceptable to employees.
9. Appropriate provision needs to be made within the draft Regulations for the circumstances in which Admitted Bodies to the CARE scheme wished to leave the scheme.
10. It is not ideal for both the PECRS and CARE scheme Committee of Management and the States Employment Board (as employer) to receive actuarial advice on the choice and prudence of assumptions from the same actuarial firm, since there is a risk that the advice may not be, or seen to be, completely independent.

11. There needs to be clarity regarding the administration costs arising from implementation of the CARE scheme and confidence that the staff resources would be sufficient and adequately trained. These are matters which will be pursued during Phase 2 of the Scrutiny Review.
12. Notwithstanding the large amount of communication which has taken place, care should be taken to ensure that communication with members of PECRS is not inadvertently misleading about the status of the proposed reforms.

Recommendations

1. Prior to the debate on the draft Law (and the accompanying amendments), the States Employment Board should ensure the States Assembly is provided with sufficient evidence on the affordability of the proposed employer's contribution cap.
2. The States Employment Board should clarify the policy that would be followed in the application of Article 8(1) of the draft Law.
3. The States Employment Board should clarify whether and how provision will be made in Regulations for the matters identified by Scrutiny's expert advisor.
4. The States Employment Board should take appropriate steps to ensure that additional information and analysis identified by Scrutiny's expert advisor is made available before the draft Regulations are debated.
5. The States Employment Board should ensure that the Regulations underpinning the proposed CARE scheme incorporate the concept of prudence being used within the funding assumptions.
6. The States Employment Board should ensure that the draft Regulations make appropriate provision for the mechanism which would apply if one of the Admitted Bodies to the CARE scheme wished to leave the scheme.
7. The States Employment Board should take appropriate steps to ensure that, from 1st January 2015, actuarial advice to the Board and to the PECRS Committee of Management on the choice and prudence of assumptions is provided by separate actuarial firms.

2. THE PANEL'S REVIEW

Introduction

1. Draft Public Employees (Pensions) (Jersey) Law 201- (P.28/2014) was lodged by the States Employment Board (SEB) on 11th March 2014 and, at the time of this report's presentation, was due to be debated at the sitting of 13th May 2014. Its adoption would see the first legislative step towards reform of public sector pension provision in the Island; draft Regulations would subsequently be lodged for debate by the States Assembly. The proposed reforms would see the effective closure of the current scheme, the Public Employees Contributory Retirement Scheme (PECRS), to new members and the establishment of the Public Employees Pension Scheme (PEPS). PEPS would begin on 1st January 2015 and would be a Career Average Revalued Earnings (CARE) scheme, rather than a final salary scheme. We shall refer to it as the CARE scheme'.
2. Following the lodging of the draft Law, we agreed to undertake a Scrutiny Review of the proposed reforms. Given the demands of our work programme and the technical nature of this topic, we commissioned an expert advisor to review the proposals. BWCI Consulting Limited was appointed on 4th April 2014 and we are grateful for the work it has undertaken in reporting within a particularly tight timescale.
3. BWCI Consulting Limited has prepared a report, which we have appended. The report sets out the background to the types of pension scheme that might be introduced and to PECRS as it currently stands. Our advisor then proceeds to explore the proposals for the CARE scheme in relation to a number of areas, including the principles which are intended to underpin the proposed reforms: affordability, sustainability and fairness.
4. Our advisor has made 15 recommendations for us to consider, which we do below and we have drawn our own Key Findings and Recommendations. The information underlying our Key Findings and Recommendations is contained within our advisor's report, however, and we would urge people to read that report in order to gain a full understanding of the proposed reforms and the issues involved.
5. For this review the Corporate Services Scrutiny Panel comprised Senator S C Ferguson (Chairman); Deputy J G Reed (Vice-Chairman); and Deputy R J Rondel. Connétable D W Mezbourian , as a Member of the SEB, declared an interest and took no part in the proceedings. A copy of our Terms of Reference has been included in our advisor's report.

6. The written material provided to us and our advisor during our review is also listed within the advisor's  report.   In  addition,  we  held  three  public  hearings  and  received  two  written submissions, as follows:

Public Hearings

22nd April 2014  Chief Minister (as Chairman of the SEB), accompanied by the

Chief Executive, the Treasurer of the States, the Project Director of Pensions and the Human Resources Senior Advisor

23rd April 2014  Chairman of the Joint Negotiating Board, accompanied by two

other members of the Board (including its Secretary)

23rd April 2014  Chairman of the PECRS Committee of Management Written Submissions

14th April 2014  Chairman of the PECRS Committee of Management 17th April 2014  Joint Negotiating Board

The written submissions and the transcripts of the public hearings are available to read on the Scrutiny website (www.scrutiny.gov.je).

7. We agreed to undertake our review in two phases. Phase 1 (of which this report represents the culmination) was to last until the debate on the draft Law. Phase 2 would begin once (and if) the draft Law were adopted and would last until the debate on the draft Regulations. It is apparent that 5 of our advisor's recommendations are most relevant to Phase 1 of our review and therefore to the debate on the draft Law. We shall therefore consider those recommendations first before moving on to the 10 recommendations which relate more directly to Phase 2 (and which we will therefore follow up during that Phase). In each instance, we have provided the reference in our advisor's report where our advisor has considered the issues in question.

Recommendations relating to the Draft Law Scheme Design

8. Our advisor has recommended that we consider whether the arguments put forward for a CARE scheme are compelling. The rationale underlying that recommendation is covered in Section 4.4 of the advisor's report.
9. In  considering  this  recommendation,  we  have  noted  that  the  draft  Law  itself  does  not explicitly establish a CARE scheme. In any event, very little detail is contained within the draft  Law  and  the  proposed  reforms  would  not  be  implemented  unless  and  until  draft Regulations  were  debated  and  approved. Nevertheless,  it  is  clear  from  the  report accompanying the draft Law that it is intended as a step towards the introduction of a CARE scheme. Approval of the draft Law could therefore be seen as an in-principle approval of the introduction of a CARE scheme.

10. We have taken into account the fact that work on the proposed reforms has been underway for some time and that there has been recognition from all parties that PECRS needs to be reformed. We have also taken into account that the proposals have already been subject to negotiation between the employer (i.e. the SEB) and employee representatives and that the agreement to maintain a defined benefit scheme and to introduce a CARE scheme have formed part of those negotiations. We have also considered the argument that aligning Jersey's public sector provision with the provision that exists in the United Kingdom (UK) (where CARE schemes have already been introduced) is in the Island's best interests in terms of recruitment. Thus, whilst it is difficult to disagree with the assertion that the move to a CARE scheme will mean that public sector provision is still more generous than what one might find in the private sector, the rationale underlying such a move is indeed compelling and rejection of the draft Law would risk derailing the proposed reform process. As ever, the devil will be in the detail and it is when the draft Regulations are finalised that the detail will become truly apparent.

KEY FINDING

11. Adoption of the draft Law would effectively provide in-principle approval for the move from a final salary scheme to a CARE scheme. The rationale for such a move is compelling and the debate on the draft Law should therefore proceed, albeit with the caveat that the details of how the scheme will operate will not be agreed until the draft Regulations have been finalised.  

Affordability

12. Our advisor has recommended that we seek further evidence to satisfy ourselves on whether the 16% / 16.5% cost cap (for the employer) would be affordable for the tax-payer in the long-term. This is covered in Section 6.4 of the advisor's report.
13. The need for the Assembly to consider this question has been raised by the decision of the SEB to lodge an amendment to the draft Law which would place the employer's contribution cap of 16.5% within the draft Law (and the decision of Deputy E.J. Noel to lodge an amendment to that amendment which would see the cap limited to 16%). We understand

that it had initially been foreseen that the employer's contribution cap would be incorporated within  the  draft  Regulations.   If  the  Assembly  is  to  be  asked  to  debate  whether  the employer's contribution cap should be incorporated within primary legislation, it should be provided with sufficient information on the implications of that cap.

KEY FINDING

14. Further evidence is required in respect of the affordability of the proposed employer's contribution cap in the long term.

RECOMMENDATION

15. Prior to the debate on the draft Law (and the accompanying amendments), the States Employment Board should ensure the States Assembly is provided with sufficient evidence on the affordability of the proposed employer's contribution cap.  

Draft Legislation

16. Our advisor has recommended that we consider the text of the draft Law and, in particular, the  risk  of  ambiguity  in  interpreting  "significant"  in  Article  8(1).   Our  advisor  has  also recommended  that  the  Chief  Minister  seek  appropriate  legal  and/or  actuarial  advice  in relation to that Article. The Article covers instances in which Regulations made under the draft Law make retrospective provision which appear to have either significant adverse effects' or which may be otherwise unfavourable'. In Section 10.4 of the report, our advisor has highlighted that these are subjective terms and that the manner of their interpretation is important. Clarity is required to be able to understand how Article 8 would be applied.
17. Such clarity might come from the provision of legal advice, as our advisor has highlighted. However, we are also reminded of the advice provided by HM Solicitor General to the Assembly during the recent debate on the Interim Population Policy: that it is preferable to have policies in place to ensure consistency of decision-making by Ministers. This might well be such an instance where a policy would be beneficial to ensure consistency of decision- making under Article 8 of the draft Law.

RECOMMENDATION

19. The States Employment Board should clarify the policy that would be followed in the application of Article 8(1) of the draft Law.
20. In Section 10.5 of the report, our advisor has also highlighted other areas which it would be beneficial to include within the scope of Regulations and has therefore recommended that we consider whether it would be helpful to expand the list of areas for potential regulation. The areas identified by our advisor are investments, investment strategy and returns; prudent and best estimate funding assumptions; and the declaration of conflicts of interest within the Committee of Management.

KEY FINDING

21. It would be beneficial for Regulations under the draft Law to include, in respect of the CARE  scheme,  provisions  in  relation  to  investment  strategies;  prudent  and  best estimate funding assumptions; and the declaration of conflicts of interest.

RECOMMENDATION

22. The States Employment Board should clarify whether and how provision will be made in Regulations for the matters identified by Scrutiny's expert advisor.

Legislative Process

23. In Section 10.7 of the advisor's report, it is recommended that we satisfy ourselves that the legislative timetable is sufficient to review what are expected to be complex Regulations.
24. The proposed CARE scheme is due to be implemented on 1st January 2015. Before that time, the draft Law needs to be debated by the States Assembly; the unions (should they choose to do so)  need to ballot  their members; and the draft Regulations need to be developed, consulted upon and debated by the States Assembly. In that latter regard, both the PECRS Committee of Management and the Joint Negotiating Group will need to consider the draft Regulations to ensure that from their perspectives they say what they should. The draft Regulations will also be subjected to Scrutiny. There is seemingly a desire from all parties to get through these processes to allow for implementation to occur in January 2015 but these are not insignificant stages through which the proposed reforms will need to pass.
25. This is happening in an election year when there is an even greater drive for States business to be completed before the summer recess. The current plan is therefore for the draft Regulations to be debated at a sitting of the Assembly in July 2014, ahead of which the process we have described above need to be completed. There is, we understand, the procedural  possibility  that  the  draft  Regulations  could  be  debated  at  the  first  sitting  in September 2014 (should such a move become necessary) but the desirability of doing so, given the proximity of the elections, is questionable. The development and implementation of the proposed CARE scheme is being undertaken within a tight timescale.

KEY FINDING

26. The aim to implement the proposed CARE scheme on 1st January 2015 means that the development and consideration of the draft Regulations will be undertaken within a very tight timescale. Sufficient time must be allowed for those Regulations to be considered.

Recommendations relating to the Draft Regulations Further Analysis and Information

27. Our advisor has made three recommendations that further analysis or information in relation to the CARE scheme is required. First, our advisor has suggested that additional information is required to enable the costs of PECRS and the new CARE Scheme to be compared on a consistent basis. The nature of the information required is described in Section 5.7 of our advisor's report. Secondly, our advisor has recommended that additional sensitivity testing is carried out to determine the funding contribution rates on a range of assumptions. This is also covered in Section 5.7. Finally, in Section 7.2.2 of the report our advisor has highlighted the need for additional modelling to be carried out to quantify the risk of underfunding even after benefit increases have been reduced to the minimum levels over the sustainability time horizon of 25 years.
28. In all three instances, our advisor has indicated the type of information or analysis which is required (analysis which we anticipate would be undertaken for the SEB by its actuarial advisors). We recognise that such information would assist the States Assembly when it comes to debate the draft Regulations and therefore support our advisor's recommendations.

KEY FINDING

29. Further  information  and  analysis  is  required  in  respect  of  the  cost  comparison between PECRS and the proposed CARE scheme; the sensitivity of the results to the assumptions underlying the calculation of the anticipated contribution rates; and the quantification of risks of underfunding within the CARE scheme.

RECOMMENDATION

30. The States Employment Board should take appropriate steps to ensure that additional information and analysis identified by Scrutiny's expert advisor is made available before the draft Regulations are debated.

Funding of New Arrangements

31. Our  advisor  has  recommended  that  the  intended  level  of  prudence  in  the  funding assumptions be agreed at the outset and written into the regulations. This is a matter covered in Section 7.2.1 of our advisor's report. The fact that, within the proposed CARE scheme, funding assumptions would be based upon prudence (rather than best estimates, as is the case for PECRS) was often stated during the review. It is a matter we raised with the Chief Minister and there was an undertaking on his part to consider how the concept of prudence could best be dealt with in the Regulations.[1]

KEY FINDING

32. The  concept  of  prudence  within  the  funding  assumptions  to  be  used  under  the proposed CARE scheme should be clearly established.

RECOMMENDATION

33. The States Employment Board should ensure that the Regulations underpinning the proposed CARE scheme incorporate the concept of prudence being used within the funding assumptions.  

Fairness

34. Our advisor has recommended that we consider whether the fairness of the protections for certain members should be challenged. This is covered Section 8.3.1 of our advisors report. The protections in question are that those within 7 years of normal retirement age would be able to elect to retain their current provision (i.e. the final salary scheme) rather than move to the CARE scheme. Similar protection would be afforded to those members of PECRS on a 1/45th accrual rate. As our advisor has highlighted, the protection for those within 7 years of normal retirement age means that those least affected by the move to a CARE scheme would receive the most protection.

35. We  have  noted  from  the  evidence  we  have  received,  and  from  our  advisor's recommendation,  that  such  protections  are  essentially  unfair  in  that  they  mean  certain members of PECRS would stand to benefit where others would not. Indeed, it was not a recommendation  of  the  Technical  Working  Group  that  such  protections  be  included. However, they were evidently a negotiating requirement of the Joint Negotiating Group and we recognise that the protections have been introduced following a negotiation process between that Group and the employer. They might therefore represent a pragmatic decision to compromise and thereby to make the proposed reforms acceptable. However, that does not alter the position that, in themselves, those protections are unfair.

KEY FINDING

36. Some of the protections which would be afforded to current members of PECRS in the move to a CARE scheme are essentially unfair. They appear to have been included for pragmatic  reasons  to  ensure  the  proposed  reforms  would  be  acceptable  to employees.

Risk-Sharing

37. From the information we have received to date, our advisor has highlighted that it is not clear what the mechanism would be if one of the Admitted Bodies to the CARE scheme wished to leave the scheme. Our advisor has therefore recommended that we pursue this in Phase 2 of our review to ensure that it is reflected in the draft Regulations. It is a matter we would expect to be covered in the Regulations and we would therefore hope that the SEB would make provision for such arrangements.

KEY FINDING

38. Appropriate  provision  needs  to  be  made  within  the  draft  Regulations  for  the circumstances in which Admitted Bodies to the CARE scheme wished to leave the scheme.

RECOMMENDATION

39. The  States  Employment  Board  should  ensure  that  the  draft  Regulations  make appropriate provision for the mechanism which would apply if one of the Admitted Bodies to the CARE scheme wished to leave the scheme.

Governance

40. In Section 11.5 of our advisor's report, consideration is given to the fact that both the PECRS Committee of Management and the SEB (as employer) receive actuarial advice from the same firm. Our advisor has highlighted the issues raised by this arrangement and has recommended  that  we  consider  whether  we  are  content  with  that  arrangement.   We understand that this position is not ideal and although our advisor has also suggested a compromise arrangement that could potentially be implemented, we believe it would be preferable  for  the  situation  to  be  addressed  outright.   The  intended  reforms  and  the introduction of the proposed CARE scheme would appear to present an ideal situation for that to happen.

Administration

43. Our advisor has made two recommendations in relation to the resource requirements of the proposed reforms. They are addressed in Sections 12.4 to 12.7 of our advisor's report. First,  our  advisor  has  recommended  that  we  consider  whether  the  additional  internal pensions  administration  costs  arising  from  the  CARE  scheme  should  be  identified  and managed  explicitly.   Secondly,  we  have  been  advised  to  seek  confirmation  that  the anticipated  additional  pensions  administration  staffing  resources  will  be  sufficient  and adequately trained prior to 1st January 2015.
44. The report accompanying the draft Law states that the resource implications of implementing the CARE scheme (i.e. the manpower and systems) will be met from the scheme itself. It is also stated that the draft Law does not in itself have any additional financial implications for the States. Even if the anticipated resources are to be funded from the scheme (and not from States revenue expenditure), we believe it important that there is full understanding of

what those resource requirements are and that there is confidence that appropriate systems and arrangements have been put in place for when the reforms are due to be implemented. We would expect such matters to have been subject to consideration already and for there to be an appropriate project plan. Nevertheless, these are matters we will pursue and seek to address in Phase 2 of our review.

KEY FINDING

45. There  needs  to  be  clarity  regarding  the  administration  costs  arising  from implementation of the CARE scheme and confidence that the staff resources would be sufficient and adequately trained. These are matters which will be pursued during Phase 2 of the Scrutiny Review.

Member Communications

46. We have noted the findings of our advisor that some of the communication with members of PECRS about the proposed reforms could be reviewed as misleading in that it has presented the reforms as a certainty rather than as proposals (which await approval by the Assembly). Not all of the communication falls into that category and we recognise that a good deal of communication with the members of PECRS has been undertaken. Nevertheless, it is important that the correct message is provided at all times.

KEY FINDING

47. Notwithstanding the large amount of communication which has taken place, care should  be  taken  to  ensure  that  communication  with  members  of  PECRS  is  not inadvertently misleading about the status of the proposed reforms.
48. In  considering  this  particular  recommendation,  we  have  recognised  that  communication about  the  proposed  reforms  has  been  undertaken  by  the  various  parties  involved:  the PECRS  Committee  of  Management;  the  employer  (i.e.  the  SEB);  and  employee representatives.   This  places  a  further  emphasis  on  the  need  for  care  in  undertaking communication to ensure that mixed messages are not provided. It also raises the question of which party should be primarily responsible for communication with the members of PECRS. In our view, it is the Committee of Management which should be clearly seen as primarily responsible.

3. REPORT OF BWCI CONSULTING LIMITED

Independent Review

of the Proposed Reforms to the Public Employees Contributory Retirement Scheme (PECRS)

Prepared for

The States of Jersey Corporate Services Scrutiny Panel

Prepared by

Mrs M Galpin, FIA

12 May 2014

Contents

1. Introduction  21
2. Types of Pension Scheme  24
3. Background to PECRS  26
4. Proposed benefit structure  30
5. Analysis of Costs  38
6. Affordability  45
7. Sustainability  48
8. Fairness  51
9. Risk-sharing arrangements  56
10. Draft primary legislation  59
11. Governance  63
12. Administrative arrangements  66
13. Member Communications  68
14. Conclusions  71 Appendix A Panel's Terms of Reference  75 Appendix B Information Considered  76 Appendix C Technical Details  78

Executive Summary

We have reviewed the proposed new CARE scheme arrangements and our principal conclusions are:

• Our analysis of Jersey average earnings increases suggests that the CARE benefits indexation is quite generous.
• Indexing the CARE benefits to uncapped RPI introduces a risk that benefits could increase rapidly and substantially in a high inflation environment before corrective action can be taken.
• If the costing assumptions are not sufficiently prudent then the 24% cost envelope will not be adequate to meet the funding costs.
• We have not seen any evidence that the 16.5% [or possibly 16%] cost cap would be affordable for the tax-payer in the long-term.
• The proposed risk-sharing arrangements are complex
• Our principal recommendations are that:
  • the Panel consider whether the arguments put forward by the TWG and the JNG for a CARE scheme are compelling (section 4.4)
  • the Panel consider whether to seek evidence that the 16.5% [or possibly 16%] cost cap would be affordable for the tax-payer in the long-term (section 6.4
  • the intended level of prudence in the funding assumptions is agreed at the outset and written into the regulations (section 7.2.1)
  • the Panel consider whether to seek additional information to enable the costs of PECRS and the new CARE scheme to be considered on a consistent basis (section 5.7)
  • additional sensitivity testing is carried out to determine the funding contribution rates on a range of assumptions (section 5.7)
  • additional  modelling  is  carried  out  to  quantify  the  risk  of  underfunding  over  the sustainability time horizon of 25 years, even after benefit increases have been reduced to the minimum levels (section 7.2.2)
  • the Panel consider whether the fairness of the protections for certain members should be challenged (section 8.3.1)
  • the Panel pursue what the mechanism would be if one of the Admitted Bodies wished to leave the CARE scheme (section 9.2)
  • the Panel consider whether the text of the draft law be reviewed, in particular in relation to Article 8(1) (section 10.4)
  • the Panel consider whether it would be helpful to expand the list of areas for potential regulation (section 10.5)
  • the Panel satisfy themselves that the legislative timetable is sufficient to review what are expected to be complex regulations adequately (section 10.7)

Executive Summary (continued)

• the Panel is content with the current arrangements, whereby actuarial advice to both the COM and SEB is provided by separate offices of the same company (section 11.7)
• the Panel consider whether the additional internal pensions administration costs arising from the CARE scheme should be identified and managed explicitly (section 12.7)
• the Panel verify that the anticipated additional pensions administration staffing resources will be sufficient and adequately trained prior to 1 January 2015 (section 12.7)
• that the Panel consider whether some of the online communication with members could be viewed as misleading and whether any action should be taken as a result (section 13.4.2)

1.  Introduction

1. Role of BWCI

BWCI Consulting Limited ("BWCI") has been engaged as an independent expert advisor by the Corporate Services Scrutiny Panel of the States of Jersey ("the Panel") to assist them with their review  of  the  proposed  reforms  to  the  Public  Employees  Contributory  Retirement  Scheme ("PECRS") and the related draft legislation. Details of the Panel's Terms of Reference for the review are set out in Appendix A of this report.

2. Structure of Review

The Panel's review will be broken down into two discrete phases. Phase 1 will run until the States' debate on the draft primary legislation on 13 May 2014. Phase 2 will then run from 14 May until the subsequent debate on the regulations, which is expected to be in July 2014.

3. Background

In  Strategic Plan 2012, the Council of Ministers stated that it would review the affordability, sustainability and fairness of public pension schemes and the proposed changes to PECRS follow from this review.

The review of PECRS sought to address a number of issues that had been identified:

• to determine the most appropriate structure for the public sector pension arrangements going forward
• to ensure that the future risk-sharing arrangements are as fair as possible
• to ensure that the solution put in place is both affordable for members, their employers and the taxpayers, as well as sustainable in the long-term
• to avoid any unnecessary cross-subsidy between the members
• the payment of the Pre-87 Debt (see section 3.3)

The proposals have had regard to the principles set out by Lord Hutton for the reform of UK public sector  pensions,  following  his  detailed  consideration  of  a  range  of  approaches  to  pension provision both in the UK and other jurisdictions. Lord Hutton concluded that a Career Average Revalued Earnings ("CARE") scheme would be the most appropriate structure for the public sector in the UK.

In November 2012, a report on the possible options for reform of PECRS was published by the Technical Working Group ("TWG"). Following the endorsement by the States Employment Board ("SEB") of the option to move to a CARE scheme, detailed proposals have been discussed with the Joint Negotiating Group (JNG). The States is now being asked to approve the reforms, for subsequent implementation on 1 January 2015. It is anticipated that will be done through a draft enabling law, followed by approval of draft regulations in which the details of the proposed reforms will be laid out.

4. Key Issues for the Panel

The Panel has identified the following key issues that it wishes to consider as part of its review of the proposed reforms:

1. How will the proposed structure for PECRS compare with the current structure in terms of sustainability, affordability and fairness?
2. What will be the financial impact of the proposed reforms for employers, employees and taxpayers?
3. What impact will the reforms have on benefits, risks and contribution rates, both from the employers' perspective and that of the employees?
4. Are the proposed reforms sustainable in the long term?
5. Are the proposed reforms fair for the various classes of members of PECRS?
6. What will be the future liability of the States within a reformed scheme?
7. What impact will adoption of the proposed reforms have on the repayment of the Pre 1987 Debt?

5. Scope of the Review

While the key issues outlined above form the central part of their review, the Panel has also asked that the scope of BWCI's report should be extended as necessary to cover other potential areas of risk that may be identified whilst considering the evidence presented. As a result, we have also commented upon the following:

• Governance arrangements
• Administration arrangements
• Member communications

We have not considered whether any changes might be required to contracts of employment, as a result of the proposed pension changes, since this is a legal issue and therefore is outside the scope of our report.

6. Evidence considered

In order to assist the Panel we have been provided with evidence from a number of sources. In addition, we have also drawn on additional information that is available online. The documents that have been considered are listed in Appendix B.

7. Abbreviations

There are a number of abbreviations and technical pension terms that are used throughout the evidence considered and for ease of reference, we have included the most frequently used terms in the box below.

2.  Types of Pension Scheme

1. Overview

Before reviewing the pension reform proposals, it may be helpful to summarise briefly the different types of pension scheme, how they operate and the relative advantages and disadvantages of each.

2. Types of Scheme

Essentially there are two types of pension arrangement:

• Defined contribution (DC or Money Purchase)
• Defined benefit (DB)

3. Defined Contribution Schemes

In a defined contribution scheme the employer and the members pay a specified contribution rate into a "pot" for each individual member. The retirement benefits available to a member will depend on the value of the funds in that individual member's pot and the cost of purchasing an annuity at retirement. This in turn depends on the contributions paid into the scheme and the investment return achieved. At retirement the member has the option to choose the type of pension that best suits their individual circumstances.

Defined contribution schemes are by far the most common type of open pension arrangement found in the private sector, both in Jersey and the UK. From an employer's perspective, the contribution rate is known and because the benefit is always equal to the value of a member's pot, there are no funding surpluses or deficits to be addressed.

From the member's perspective, defined contribution provides the flexibility to select the form of pension that best suits their personal circumstances, as well as usually providing some choice around  how  the  contributions  are  invested.   However,  the  members  bear  all  of  the  risks associated with poor investment returns and improving life expectancy.

4. Defined Benefit Schemes

The benefits payable from a defined benefit scheme are calculated by applying a set of rules. Normally the benefit is calculated as a proportion of salary for each year of service, according to the particular formula for an individual scheme. The benefits would normally be independent of both  the  investment  performance  of  the  scheme's assets  and  the  contributions  paid.   Both PECRS and the proposed new arrangements have most of the characteristics of a defined benefit schemes from the members' perspective. However, the risk-sharing arrangements described in section 9 mean that they are not fully defined benefit.

2.4.1  CARE Scheme

A Career Average Revalued Earnings ("CARE") scheme is a form of defined benefit scheme and in  many  ways  is  similar  to  a  traditional  defined  benefit  final  salary  scheme.   The  one  key difference is that the pension "earned" during each year of employment increases in the same way for all members, rather than in line with an individual's salary increases during their working life.

2. Types of Pension Scheme (continued)

At an individual level, a final salary pension arrangement will tend to favour "high flyers". The exact relationship between the benefits provided under a CARE and final salary scheme cannot be compared easily on a scheme wide basis, other than by considering the cost of providing each type  of  benefit  using  consistent  actuarial  assumptions.   We  have  requested  such  a  cost comparison  from  the  Treasurer,  and  while  we  have  received  some  additional  summarised information, we have some further points that we would like to be clarified before considering this further. We would recommend that the Panel pursue this in Phase 2 of their review (see section 5.7).

5. Actuarial valuations

The financial position of a defined benefit scheme needs to be reviewed on a regular basis by the scheme's actuary. Typically a formal actuarial valuation is carried out once every three years and involves the comparison of the market value of the scheme's assets with its liabilities at the date of the valuations. The value placed on the liabilities is calculated using a financial model that projects the expected amounts and timing of the benefit cashflows from the scheme, using a set of assumptions.

6. Valuation Assumptions

While the value of the assets is an objective value, based on the market value of the investments, the present value placed on the scheme's liabilities is much more subjective, since it is necessary to make assumptions about when and for how long the benefits are expected to be paid, as well as how they will increase.

As the cost of providing the benefits is to be partially met by the investment return on the scheme's  investments,  the  assumed  rate  of  return  on  the  investments  is  a  key  valuation assumption. This assumed rate of return is generally referred to as the discount rate. The discount rate, together with the assumption made for future Jersey inflation, are the key financial assumptions.   In  general,  it  is  the  difference  between  these  two  assumptions  that  is  more important than their absolute values.

3. Background to PECRS

1. Overview of PECRS

PECRS is a statutory scheme and, on the face of it, is a final salary defined benefit pension scheme. However, as Aon Hewitt point out in their report of 15 July 2011, it is hybrid between a defined benefit and a defined contribution scheme due to the unusual way in which the funding surplus and deficits are managed. The Committee of Management ("COM") has oversight of the operation of PECRS and operates in a broadly similar way to a trustee board for a pension arrangement established under trust.

The membership of PECRS comprises the employees of the States of Jersey, together with the employees of 24 other employers, known as the "Admitted Bodies". In total there were 13,250 members as at 31 December 2012. The majority of the membership are current public sector employees who are still accruing benefits. Around a quarter of the members are in receipt of pensions and the remainder are deferred pensioners who have left employment but who have not yet started to draw their pensions. The chart below illustrates the distribution of the membership.

PECRS Membership at 31 December 2012

Deferreds

Pensioners and Dependants Actives

Source: PECRS 2012 Annual Report

2. Benefits provided

3.2.1  Overview

Not all of the members have the same benefit structure. In particular, there are differences between the "uniformed" and "non-uniformed" staff, as well as some members with legacy benefit structures dating back to before the changes made to PECRS which gave rise to the Pre 1987 Debt (see section 3.3). The headline differences in benefit structure are around the accrual rates and normal retirement age.

The median pension provided by PECRS is understood to be £8,600 pa. However, there is a wide distribution of the size of pension which is illustrated in the chart below.

Distribution of Current PECRS Pension Amounts

Median = £8,600

1200 1000 800 600 400 200 0

< 5k > 80k 5k - 10k10k - 1515k k - 2020k k - 2525k k - 30k30k-35k35k - 40k40k- 45k45k - 5050k k - 5555k k - 6060k k - 6565k k - 7070k k - 7575k k - 80k

Size of pension (£ pa)

Source: Treasury and Resources presentation

3. Pension at retirement

The pension provided under PECRS at normal retirement age is calculated using the following formula:

Pension = Pensionable Service x Final Pensionable Earnings Accrual rate

4. Accrual rates

The lower the accrual rate the faster the rate, at which the pension builds up. For example, an 80ths accrual rate is equivalent to a pension of 1.25% of final pensionable earnings for each year of service. However, a 60ths accrual rate would provide a greater pension, as it is equivalent to the pension building up at a rate of 1.67% for each year of service. As illustrated in the pie chart below, the vast majority of the members have an accrual rate of 80ths.

Number of active members with each accrual rate

0.76% 15.68%

83.55%

Based on a total number of active members of 7,417.

Source: Treasury and Resources Presentation

NB: The pie chart illustrates the most recent information provided to the Panel on the numbers of members of PECRS with a particular accrual rate. This is why the figures for the total number of active members is different from the number in section 3.1 above. In addition, we understand that a very small number of new prison officers have an accrual rate of 1/70th which is not reflected in the pie chart.

5. Pension increases

PECRS does not currently provide guaranteed increases to pensions in payment for the majority of members. However, it is required to provide increases in line with the Jersey All Items Cost of Living Index provided that PECRS's financial position remains satisfactory.

Following the 2007 valuation the financial position of PECRS was found to be unsatisfactory and consequently pension increases were awarded at 0.3%pa below the rate of inflation. This is understood to be the first time that the COM had to take this action. The 2010 valuation indicated that subsequently the position had improved slightly and pension were re-adjusted to be 0.15% pa less than inflation.

3. Pre 1987 Debt

1. What is the Pre 1987 Debt?

Following changes made to Jersey's public sector pension arrangements in 1987, PECRS took on the liability for paying increases to pensions accrued in respect of service prior to the changes. Previously these increases were funded by the States on a "pay as you go" basis. While PECRS took on these additional liabilities, no additional contributions were paid into PECRS to cover this. This shortfall in the funding of the pension increases is how the "Pre 1987 Debt" arose.

It was not until 2003 that an agreement was reached between the COM and the Policy and Resources Committee on the framework for dealing with this debt and the detailed process was set out in a "10 point plan" and eventually enshrined in regulations in 2005.

2. How large is the Debt?

Under the 10 point plan the actuary to PECRS is required to certify the debt each year and a note of the amount is included in the States Accounts. The latest accounts available relate to 2012 and at 31 December 2012 the actuary certified that the debt was £250.45M, having reduced from £252.02M the previous year. However the note to the accounts also highlights the sensitivity of the value of the debt to small changes in the discount rate used to value it. In particular, a reduction of 0.1%pa in the assumed discount rate would increase the value of the debt by around £8M.

3. How is the debt to be eliminated?

It was agreed that from 1 January 2002 the employer's contribution rate of 15.6% of pensionable earnings would essentially be split into 13.6% to cover the cost of future benefits accruing and 2% would be allocated to paying off the Pre 1987 Debt. This approach was expected to eliminate the debt over an 82 year period.

According to the 2012 States Accounts, the payment towards the debt in 2012 was £4.1M and the expected payment in 2013 was £4.3M.

4. Eliminating the debt more quickly

It is recognised that the period over which the debt was initially planned to be eliminated is very long. The key problem with this is that the benefits to which the debt is linked would have had to have been paid out of PECRS before the debt is fully repaid. This would result in a weakening of the security of the benefits for younger members of PECRS until the full debt had been repaid. Consequently steps are being taken to accelerate the funding of the debt, in order to eliminate it over a shorter period. Section 9 of the written submission to the Panel dated 14 April 2014 from the Chairman of COM indicates that if the increased payments included in the States current Medium Term Plan (2013-2015) can be maintained, the repayment date of the debt is expected to be reduced by around 29 years to 2054. This is still a long period and the Treasurer has indicated that the debt will be repaid over a shorter period if it is possible to do so.

4. Proposed benefit structure

1. Introduction

Before commenting on the proposed fairness of the benefit structure, it is important to realise that the current proposals are slightly different from the initial proposition put forward by the TWG. Various  adjustments  have  been  made  as  a  result  of  discussions  with  the  JNG.

A summary of the changes that have been brought about by the negotiations are provided in Slide 11 of Part A of the Treasury team's presentation to the Panel.

2. Type of benefit structure

A key decision to be made is the type of benefit structure. The benefit design then follows from this. Essentially the choice is between a defined contribution arrangement and a form of defined benefit. Section 2 of this report provides a description of each type of scheme.

To help put this decision in context, virtually all private sector employers, both in Jersey and the UK have closed their defined benefit pension schemes in the last 20 years or so, replacing them with  a  defined  contribution  scheme.  Defined  contribution  schemes  are  seen  as  particularly attractive  to  private  sector  employers,  as  they  provide  complete  certainty  over  the  cost  of providing pension benefits and eliminate balance sheet volatility. Defined contribution schemes are not necessarily inferior to defined benefit schemes; it will depend on the level of contributions paid.

3. Reasons for proposed CARE Structure

In contrast to the position in the private sector, Lord Hutton's review of the UK public sector pension arrangements recommended a CARE benefit structure, which is a form of defined benefit scheme (see section 2.4.1).

The TWG considered a range of types of arrangements in their November 2012 report before proposing a CARE arrangement, citing various disadvantages of defined contribution schemes including:

the potential difficulty of recruiting specialist staff from the UK,

the lack of certainty of outcome of a defined contribution scheme, the administration costs and

the risk of poor decision making affecting members' benefit outcomes

However, the evidence that we have reviewed does not seem to entirely support these arguments and therefore the Panel may wish to question the robustness of the arguments put forward to support this central part of the proposed reforms.

1. Recruitment from the UK

In section 12.2 of the 2011 LCP Scheme Review for the COM it says "the current New Member Regulations benefit scales under PECRS are now much less valuable than typical public sector arrangements from which members might be transferring."

We  understand  that  these  New  Member  Regulations  apply  to  employees  joining  after  1 September 1989.

Section 12.4 of the report goes on to say that the volume of transfers was not thought to be insignificant, with more transfers in from the UK than back to the UK. We do not have any information about the current situation, but if it is still the case then the Panel might wish to consider  whether  this  would appear to  weaken the argument that a  less generous pension arrangement would act as an obstacle to recruitment.

While some staff with specialist skills will undoubtedly be required, does it necessarily follow that all public sector employees should be provided with a pension arrangement that is broadly in line with the UK public sector? If recruitment of key staff from the UK is an issue, a more targeted solution might be to provide cash-based compensation. A counter argument to this would be that it is treating certain categories of employees differently and goes against the fairness principle. Any potential difficulty of recruiting staff from the UK falls outside the scope of this report.

2. Uncertainty of benefit outcome and poor decision making

While there is inevitably some variation in the size of a defined contribution pot at retirement, these issues can be addressed by ensuring the scheme has a default investment strategy which automatically moves to lower risk investments as a member approaches their selected retirement age.

The limited number of annuity providers in the Channel Islands at the current time could also be a reason for not going down the defined contribution route. However, Jersey's tax legislation does already  provide for  an alternative form of retirement income known as "income drawdown", provided that an individual can demonstrate that they have a guaranteed income at least equal the Jersey single person's old age pension.

3. Administration costs

While the UK Government accepted the CARE benefit structure for the UK public sector, they have taken a different approach to the recently introduced auto enrolment legislation, which has

led to downward pressure in the UK on investment management costs associated with a defined contribution scheme.

4. Fairness to Taxpayers

Looking at the fairness criteria, most of the private sector employees in Jersey will either have no pension provided by their employer or will be in some form of defined contribution arrangement. Therefore it would appear reasonable to question if it is fair to taxpayers that they are funding a public sector defined benefit pension scheme, notwithstanding the effect on the pension amounts of the risk-sharing arrangements (see section 9).

4. Conclusions on benefit design

We have highlighted some of the main arguments for and against whether the public sector pension provision in Jersey should be on a CARE basis. Ultimately the solution needs to strike a balance  between  the  theoretical  best  outcome  for  each  of  the  stakeholders  and  what  is

This requires every UK employer to provide access to a workplace pension scheme and pay a minimum level of contributions into that scheme. The default provider for auto enrolment (NEST) provides pensions on a defined contribution basis.

achievable through negotiation in practice. In practice, there might be insurmountable resistance to the introduction of a defined contribution scheme. In particular, at the Panel hearing on 23 April, the JNG chairman said "I think one of our lines in the sand was that we would find totally unacceptable any move to a defined contribution scheme, ie away from the defined benefit scheme...". He then went on to say that a CARE scheme "gives certainty and we like that idea".

We  recommend  that  the  Panel  consider  whether  they  consider  whether  the  arguments  put forward by the TWG and the JNG for a CARE scheme are compelling.

5. Principal Changes

The key changes for members under the proposed new CARE arrangements are:

phased increases in members' contribution rates

the linkage of pension age to State Pension Age for non-uniformed staff

the increase in pension age for uniformed staff to 60

an accrual rate of 1/66th

introduction of CARE style benefits, revalued at a rate of 1% pa in excess of Jersey RPI guaranteed minimum pension increases of 50% of Jersey RPI (with 100% being the target) increase in death in service lump sum

inclusion of nominated co-habiting partner if move to CARE structure

protection from changes for specified categories of members (see Section 8.3.1)

the introduction of contribution caps for the employer and the employee

the introduction of risk-sharing in the event that the benefits are not fully funded

improved governance arrangements

6. Protection

While all members will be required to pay the higher contribution rates, some will have the option to continue to accrue benefits under the existing PECRS structure. The fairness of the protection arrangements is considered in section 8.3.1.

All members' benefits accrued prior to 1 January 2015 are to be fully protected. Therefore the changes only affect benefits in respect of service from 1 January 2015 onwards. Consequently the changes will have the least impact on those close to retirement (even before the effect of the protections discussed in section 8.3.1).

7. Comparison of Benefits

The table below compares the benefits which we understand are expected to be provided from 1 January 2015, if the proposed reforms are accepted:

8. Increases in members' contributions

It is proposed that all members' contributions will be increased, whether or not they opt to retain the existing benefit structure if they are entitled to protection or move to the CARE structure. This is fair since all members will be being treated in the same way. The affordability issue of a rise in contributions has also been mitigated by the phasing in of the increases for existing members. New employees from 1 January 2015 will however be required to pay the full rate on joining the scheme. Therefore not all members are being treated equally in the short term, but this is a trade off between fairness and affordability.

9. Normal Retirement Age

The linkage of normal retirement age with increases in SPA is one way of managing the cost of increasing life expectancy. The downward effect is twofold on the contribution rate; there is a longer period over which to pay the contributions and pensions are less costly to provide because they will be payable for a shorter period. Normally these two effects outweigh the additional cost of the extra benefit that accrues due to the longer period of pensionable service.

The proposal is for the non-uniformed staff to have their normal retirement age linked to SPA changes; this will have the effect of phasing in the effect so that it only applies to the younger staff. The increase will be gradually introduced between 2020 and 2031 and therefore will affect all employees who are aged 59 or less at the current time, with the full effect of an increase in normal retirement age to 67 for members under around age 50 at the current time.

For uniformed staff the increase is an immediate rise in normal retirement age from 55 to 60 from 1 January 2015. However, there is no agreed linkage with changes in SPA beyond this at the current time.

10. Accrual rate

The proposed CARE accrual rate of 1/66th is equivalent to benefits accruing at the rate of just over 1.5%pa. The vast majority of members in PECRS are on an 1/80ths accrual rate (or 1.25%) at the current time so this represents an improvement. However, for the members on 60ths it will be a reduction. Benefits are proposed to be protected for the members who currently have an accrual rate of 1/45ths[2] (2.22%). We consider the fairness of the protection arrangements in section 8.3.1.

11. Revaluation of CARE Benefits

The benefit accrued each year under a CARE scheme increases in some objective way between the year in which it is accrued until the date the employee leaves employment. This is generally

referred to as "indexation". Based on the information that was supplied to us by the Treasury on their analysis of the different types of CARE structures being negotiated within the UK public sector, the rates of indexation in the UK range from in line with UK CPI increases (in some cases plus an additional percentage of up to 1.6% pa) to in line with increases in average earnings.

1. Proposed Jersey indexation

The TWG's starting position was indexation in line with average earnings increases (assumed to be equivalent to RPI+1.5%pa for costing purposes). However, as part of the negotiations, the indexation rate was altered to RPI +1%pa in return for an improvement in the accrual rate from 1/70ths to 1/66ths, (ie a 6% improvement in the accrual rate). This change in indexation breaks the direct link with earnings growth.

In relation to this change, the Chairman of the COM commented to the Panel during the hearing on 23 April:

"one can understand why staff representatives would rather go for the certainty of RPI or RPI plus something, rather than rely on a link to earnings where they feel that earnings can be under significant employer influence."

2. Jersey Average Earnings Increases

The States Statistics Unit collates information on average earnings increases in Jersey and publishes data showing the historical increases since 1991.

This chart show the annual increases in average earnings together with the annual RPI increases. Over the period since 1990, average earnings in Jersey have increased by 4.5%pa, whereas RPI has increased by an average of 3.8%pa. Therefore over this period average earnings have grown at a rate of 0.7% pa in excess of inflation.

The  graph  also  illustrates  that  earnings  increases  have  been  quite  volatile,  with  the  lower observed increases from 2010 onwards reflecting the economic conditions generally.

Annual percentage change in Jersey average earnings

10

9

8

7

6

5

Annual Percentage Change 4

Jersey RPI

3

2

1

0

-1

1991 1993 1995 1997 1999 2001 2003 2005 2007 2009 2011 2013

Source: States of Jersey Statistics Unit

This second graph focuses on the real earnings increases, ie increases in excess of RPI. This illustrates more clearly how volatile real earnings growth can be. The dark blue line illustrates how accrued benefits would have been indexed in line with the RPI +1% indexation under the CARE scheme proposal.

Difference between annual changes in Jersey average earnings and the RPI

5 4.3 4.2

4 3.4

3

2.1 2 1.8

2 1.7 1.7

0.6 1.2 1.1 0.5 0.4 0.4 0.7

1

0

0

-0.2

-1

-1.2

-2 -1.5 -1.3 -1.7 -1.5

-2

-3

1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013

Source: States of Jersey Statistics Unit

3. Jersey Public Sector Average Earnings increases

Since 2002 the Statistics Unit has also been publishing the information separately for the private and public sectors.. The graph below illustrates changes in public sector average earnings increases, relative to RPI over the period. We have also added the red line at a rate of RPI +1% to illustrate what the CARE benefits indexation would have been had the structure been in place over that period.

Annual percentage change in Jersey public sector average earnings

10

8

6

4

Annual Percentage Change 2 Jersey RPI

0 Jersey RPI + 1%

-2

-4

2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013

Source: States of Jersey Statistics Unit

We understand that some of the volatility in the average earnings increases in the earlier years is as a result of the 2 year pay deals in the public sector. As a consequence, while earnings have been reviewed annually, increases were not always been applied annually. Over this 12 year period, average earnings in the public sector increased by 3.1% pa whereas RPI increased by 3.4%pa. While this has been a difficult period economically, over this period, the proposed CARE revaluations of RPI +1% pa look to be quite generous based on published historical experience.

4. Comments on indexation structure

We would highlight that by indexing the CARE benefits to uncapped RPI, the employer has no influence over the actual rate of revaluation. There is therefore a risk of very high inflation increasing the cost of the benefits substantially. With valuations being undertaken once every  three years and taking a year or so to complete, it might be some time before there is the opportunity to take any remedial action. This may then require a large adjustment to benefits through the risk-sharing mechanism.

From the information provided, it is also not clear what would happen if there were a period of deflation such that the RPI index fell by more than 1%. However, we have subsequently been advised by the Treasury, that benefits would not be cut back.

5. Analysis of Costs

1. Assumptions

For the avoidance of doubt, this section focuses on the assumptions for the CARE scheme. Different assumptions are used to place a value on the accrued PECRS benefits for funding purposes.

Aon Hewitt's Benefit Costings paper dated 11 April 2014 concludes that the proposed benefit structure is within the 24% cost envelope based on the assumptions used. However, their report makes it clear that further work would be required to confirm what funding methodology and assumptions would be appropriate for contribution-setting purposes.

The TWG report dated November 2012 states that, provided the investment strategy for the new scheme includes at least 50% growth assets, the Scheme Actuary's view was that the costing assumptions were a prudent assessment of the future at that time. However, this comment provides no indication as to the level of prudence incorporated into the assumptions. Prudent assumptions could mean anything from a 51% chance of success to a near 100% chance of success.

It is critical that the initial costing assumptions (subject to changes in market conditions) would be acceptable to all parties as prudent funding assumptions for future valuations. If the costing assumptions are not deemed to be sufficiently prudent then the proposed contribution rates will not be sufficient to meet the funding costs within the 24% cost envelope.

2. Financial Assumptions

The two key financial assumptions are the discount rate and the assumed inflation rate. In particular, it is the difference between these two assumptions, rather than their absolute values, which is critical to the value placed on the pension liabilities. Therefore it is important to consider both of these assumptions together.

3. Discount Rate

If a higher investment return is achieved on the assets held by a pension scheme then, all else being equal, lower contributions will be required. This is factored into the actuarial valuations through the discount rate. The discount rate can therefore be thought of as the investment return that is expected to be achieved on the assets held to meet the pension liabilities. In view of this, the discount rate for funding valuations is typically set based on a prudent assessment of the investment return expected to be achieved on the assets held. This means that, before an appropriately prudent discount rate can be determined, the relevant parties must have decided on a suitable investment strategy for the CARE scheme. We understand that this aspect of the new arrangements has not yet been considered in detail. However, we have been advised verbally by the Treasurer that it is anticipated in the short-term that the investment strategy will be the same as for PECRS. This is illustrated in the graph below, based on the benchmark strategy shown in PECRS's 2012 Annual Report. This shows that the strategy, after treating the expected income stream from the Pre 1987 Debt contributions as a salary-linked bond, is broadly 55% return seeking assets and 45% bond type assets. It is however recognised that as PECRS becomes more mature, the investment strategies of PECRS and the CARE benefits scheme might need to diverge.

Investment Strategy Pie Chart

Equities

Hedge Funds

Property

Bonds

Pre 87 Debt Repayments

Source: PECRS 2012 Annual Report

The discount rate for the Benefit Costings paper was set at 5% pa. This was derived as 3% pa above the UK Government's target for future CPI inflation. This follows the approach adopted in the UK for setting the discount rate for unfunded public sector pension schemes. We understand that initially this discount rate was adopted by Aon Hewitt to enable the proposed Jersey changes to  be  benchmarked  against  UK  public  sector  schemes.   However,  whilst  this  discount  rate derivation may be an appropriate methodology for largely unfunded UK public pension schemes, we cannot see a logical argument for why it would necessarily be a relevant methodology for determining funding costs for a funded Jersey pension arrangement. Section 7 of Aon Hewitt's paper of 15 July 2011 also considers what might be an appropriate discount rate to use to illustrate the cost of possible benefit scenarios. In particular it states quite rightly " it is not possible at this stage to select a discount rate with regard to the scheme's investments and we are required to use an alternative approach." This alternative approach of using assumptions set by the  UK Government to compare the cost  of UK public sector pension  arrangements by reference to expected growth in UK GDP, whilst helpful to compare the overall "generosity" of the Jersey CARE proposals with the UK arrangements, is not necessarily an appropriate basis on which to set the long term cost envelope.

While historically most funded private sector defined benefit pension schemes invested heavily in equities, more recent equity market volatility combined with the closure of most defined benefit schemes has led many pension schemes to de-risk their assets. Typically most defined benefit schemes now hold bond type investments to match their pensioner liabilities, whilst retaining equities to back the non-pensioner liabilities over the period until they reach retirement.

Ultimately it will be a decision for the COM and the SEB as to whether, as an open public sector scheme, the CARE scheme wishes to take a more aggressive investment strategy than the typical closed private sector defined benefit scheme. A more return seeking investment strategy will reduce the expected costs of the scheme. However, return-seeking assets produce volatile investment returns and this exposes the employer and members to the risk of benefit cuts and additional contributions if the anticipated investment returns do not materialise.

Once the investment strategy has been agreed, the level of prudence to be incorporated into the assumptions needs to be determined. It is intended that the assumptions for the CARE scheme will be determined on a prudent basis. However, as the degree of prudence can be subjective, we recommend that the intended level of prudence is agreed at the outset and written into the regulations. This will ensure consistency between valuations and avoid the risk of the level of prudence reducing over time, as has occurred in PECRS actuarial valuations in more recent years.

4. Inflation Assumption

The inflation assumption used for the Benefit Costings paper was 2.75% pa, set as the Bank of England's target for CPI (2% pa) plus an allowance for the expected long term difference between UK RPI and CPI (0.75%). We would highlight that:

• the  Bank  of  England's  target  for  CPI  is  only  a  target  and  actual  CPI  has  significantly exceeded this in recent years. Therefore, we would not classify this as a prudent CPI assumption.
• the long term difference between CPI and RPI is a difficult assumption to determine, but we suggest that Aon Hewitt are asked to provide their best estimate assumption and, if possible, the level of prudence incorporated in an assumed differential of 0.75%.

Jersey inflation has been assumed to be in line with UK inflation. There was a sustained period between the mid-1980s to the mid-2000s where Jersey inflation exceeded UK inflation. When asked for his opinion on this issue, Jersey's Economic Adviser said: "The difference between UK RPI and Jersey RPI is a tricky one. The long-term difference between the two is 0.25% (1950-2013). In recent years the differential has been low or in fact the other way. Over the very long-term it would be hard to argue that Jersey could persistently see higher inflation than the UK (especially given the starting point where this has persisted in the past) as that would push prices higher and higher relative to the UK. However, it is also difficult to completely ignore past trends that have persisted over the long-term. That suggests to me that using a range of a differential of 0-0.25% might be sensible to recognise the uncertainty".

• The COM may therefore wish to consider whether it would be prudent to make an allowance for Jersey RPI to exceed UK RPI in the valuation assumptions. Alternatively, the converse view is that Jersey would become uncompetitive if local inflation consistently exceeded UK inflation  over  the  long  term.   In  addition,  Aon  Hewitt  may  wish  to  factor  into  their determination  of  an  appropriate  Jersey  inflation  assumption  any  differences  in  the  RPI calculation methodology between the UK and Jersey. The graph below compares UK and Jersey RPI over the last 30 years.

UK RPI and Jersey RPI Comparison

12 10

8

6

Jersey RPI UK RPI

4

2

0 -2 -4

1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012

5. Demographic Assumptions

The Benefit Costings report is based on the same demographic assumptions as adopted for the 2010 valuation of PECRS. More recent mortality tables have been published since the valuation and we understand that the most recent costing figures include an allowance for updated mortality rates.

6. Funding Method

The Benefits Costings report notes that the calculations are based on the Projected Unit funding method with a 1 year control period. However, the paper states that Aon Hewitt are not advising on the funding methodology adopted. The Projected Unit method is most appropriate to use to assess the cost of future benefit accrual where the average age of a pension scheme's active membership is expected to be fairly stable. It might not be appropriate if a significant change in the membership were expected in future which might affect the age profile of the membership. We are not aware of any such change at the current time, but there may be a risk of this in the future if say any of the Admitted Bodies were to leave the scheme. We recommend that the Scheme Actuary is asked to comment on the suitability of the funding method and we have included this in the list of questions for Aon Hewitt in section 5.7 below.

7. Additional Calculations

As noted above, we believe that is it important that the benefit costings are calculated on a basis that is consistent with the basis that will be adopted for the regular valuations. In addition, calculating contribution rates for PECRS on this same basis will allow a direct comparison of the costs of the two benefit structures. We therefore recommend that the Panel consider whether the following additional information should be sought:

1. The COM are asked to provide details of the investment strategy that is expected to be implemented for the CARE benefits
2. The Scheme Actuary is asked to:

1. Determine  best-estimate  valuation  assumptions  based  on  the  proposed investment strategy
2. Determine  prudent  valuation  assumptions,  based  on  three  different  levels  of prudence (see table below)
3. Confirm that they believe the Projected Unit funding method to be an appropriate method for future funding valuations
4. Complete  the  table  below  summarising  the  future  service  contribution  rates (including expenses) for PECRS, CARE with the target level of pension increases and CARE with the minimum level of pension increases. This will then provide a better guide to the affordability and sustainability of the proposed arrangement and will also highlight comparative costings against the existing scheme.

8. Funding Risks

It is important to appreciate that actuarial valuation assumptions are only one set of assumptions about the future. There is no guarantee that they will be borne out in practice and inevitably actual experience will differ from the assumptions made. With hindsight, the actuarial valuations of the past under-estimated future life expectancy and over-estimated future investment returns. This has led most private sector defined benefit schemes to have significant funding shortfalls and to replace their defined benefit schemes with defined contribution arrangements for future service. Any defined benefit scheme remains exposed to these risks in the future.

9. Longevity Risk

Whilst  current  actuarial  valuations  are  making  a  greater  allowance  for  future  longevity improvements than has been the case in the past, there still remains the risk that the assumptions underestimate the future improvements. For example, further significant increases in longevity could occur as a result of a medical breakthrough. This would increase the cost of the benefits being provided.

10. Investment Risk

If the CARE scheme intends to invest in growth seeking assets such as equities then, over the long term, these are expected to produce higher returns than investment in bonds. However, there are historical periods where the reverse has occurred.

The main risk from equity investment is that volatile markets will lead to volatile funding levels. The graph below shows the variability in the FTSE100 equity index over the last 20 years. If the CARE scheme had been in place over this period then, depending on the dates of the actuarial valuations, the extreme equity swings could have pushed the funding level outside the proposed 95%-105% funding corridor, leading to benefit cuts when the markets fell.

FTSE 100

8,000 7,000 6,000 5,000 4,000 3,000 2,000 1,000 -

11. Inflation Risk

There appears to be no cap on the benefit increases awarded in times of high inflation (other than limiting the inflationary element to 50% as a result of funding reviews). Although we are currently in a relatively low inflation environment, there were periods in the 1970s when UK inflation exceeded 20%pa. There is a risk that future periods of very high Jersey inflation will substantially increase the CARE scheme's liabilities. This risk could be controlled by imposing a cap on the annual increase that will be applied to benefits. We understand from the treasury team that the possibility of including such a cap was considered as part of the refinement of the proposals but was subsequently discounted. However, we do not know the reason why it was decided not to pursue this..

A further risk is that during periods when the outlook for future inflation is very low, reducing increases to 50% of assumed inflation will have less of an impact on the funding level of the CARE scheme benefits than in times when the outlook for inflation is higher.

12. Contribution Cap

The proposed employer contribution cap will limit the risk to the States of Jersey. We understand that  this  cap  is  intended  to  apply  to  the  combined  contributions  required  for  future  service benefits, together with any contributions needed to meet past service deficits in relation to either PECRS or the CARE scheme benefits, to the extent that such deficits cannot be eliminated through reduced benefit increases.

Therefore if the future service, benefits together with the required deficit contributions, cannot be met within the contribution cap, future service benefits would be reduced. This means that new members would be paying contributions equal to the contribution cap but part of their contribution would be funding the shortfall in respect of past service for existing members. In other words, new members would be paying for more than the benefits they are accruing because they are being required to contribute towards the deficit in respect of existing members. The impact of this could be compounded if deficit contributions towards the PECRS benefits are also included within the contribution cap. This does not appear to be fair to new members.

6. Affordability

1. Stakeholders

The affordability of the proposals needs to be considered from the point of view of the three principal stakeholders:

• employees
• employers
• taxpayers

We consider each of these in turn, comparing the existing costs with the proposed changes.

2. Affordability for Employees

Membership  of  the  pension  arrangements  is  currently  a  condition  of  employment  for  all permanent staff and we are not aware of any plans to change this. Therefore all members must pay regular contributions, calculated as a percentage of their pensionable earnings.

From an employee's perspective, the affordability of the new pension arrangements will depend on the impact of the increase in the contributions on their disposable income, relative to the current position. We have focused our affordability considerations on the lower paid members, since they will have the lower disposable incomes and consequently they are those mostly likely to experience affordability issues.

85% of the membership pays contributions of 5% of their pensionable earnings. The other members pay either 6% or 6.25%.

Under the proposed changes non-uniformed staff would pay 8% of pensionable earnings and uniformed  staff  would  be  required  to  pay  10.1%.   Therefore  the  maximum  increase  in contributions would be:

• Non-uniformed staff:  3%
• Uniformed staff:  5.1%

To mitigate the financial impact of a jump in contributions, contribution increases will be phased in at a rate of 1% pa or so for most members. In addition, there are special provisions for non- uniformed staff with a basic salary of less than £30,000 pa. For these cases, the increase in contribution rate would be phased in over 4 years at 0.75% pa. For a non-uniformed employee with a basic salary of £30,000 pa or more the phasing in would be over a 3 year period. We understand that there are no corresponding special provisions for uniformed staff since there are no staff with a full time equivalent basic salary of less than £30,000 pa.

The  table  shows  the  gross  monthly  contributions  for  a  range  of  salaries  under  the  current structure and over the phasing-in period for non uniformed staff.

Essentially contractual pay and shift pay but excludes overtime payments or other fluctuating emoluments.

The table illustrates that for someone with pensionable earnings of £20,000 pa, the monthly increase  in contributions  in 2015,  before the effect of tax relief,  would be £12.50. For an employee with pensionable earnings of £80,000, the monthly increase in contributions would be £66.67.

In practice, the impact would be reduced due to the effect of tax relief. The vast majority of members will receive full tax-relief on their contributions. Under Income Tax legislation, tax-relief is phased out for those earning in excess of £150,000 pa, with no tax-relief being available for those earning £200,000 pa or more.

3. Affordability for Employer

1. 2013 Costs

The States Treasurer has advised that the pensionable earnings payroll for PECRS members in 2013 was around £240M pa. From the figures provided, it appears that on average, pensionable earnings represent around 95% of total pay. The percentage that pensionable earnings represent of total earnings will vary between different categories of member, depending on the extent to which their total pay is pensionable.

At the current time the employer is paying contributions to PECRS at the rate of 13.6% of pensionable earnings plus 2% pa for Pre 87 Debt contributions. Under the proposals the 13.6% will increase to 16% of pensionable earnings (subject to cap of 16% [or possibly 16.5%). The 2% of  pensionable  earnings  would  be  required  for  the  Pre  87  Debt  payments  would  continue. However, we have been advised by the Treasurer that these Pre 87 Debt contributions would be met from existing department budgets.

The employer contributions to PECRS in 2013 were:

£M

• Regular contributions of 13.6%  32
• Pre 87 Debt contributions   4 Total  36

2. Impact of contribution rate increase

Under the proposals, the employer's contribution rate would increase to 16% of pensionable earnings.

The phasing in period is reduced to 3 years if basic salary is £30,000 pa or more

If this had been applied in 2013 then the required PECRS contributions would have been:

£M

• Regular contributions of 16%  38
• Pre 87 Debt contributions   4 42

Therefore the proposals would have increased the 2013 contributions by £6M.

4. Affordability for Taxpayer

6.4.1  2013 Costs

While  the  cost  envelope  for  the  reformed  pension  arrangements  has  been  set  at  24%  of pensionable earnings, split in a 2:1 ratio between the employer and the member, we have not seen any evidence that the 16% [or possibly 16.5%] employer cost would be affordable for the tax-payer in the long-term and we recommend that the Panel consider whether they wish to seek further evidence to satisfy themselves on this point.

The risk-sharing arrangements, which form part of the package of reforms, are designed to limit the employer's contributions to 16% [or possibly 16.5%] of pensionable earnings. A cap of 16.5% is equivalent to an annual contribution of £39.6M, whereas a cap of 16% would be around £1.2M pa lower at £38.4M, based on a pensionable payroll of £240M.

7. Sustainability

1. What does sustainable mean?

Sustainability means the ability to be maintained. In a pensions context this is taken to mean a benefit structure that is not only affordable now but that is also expected to continue to be affordable in the future.

2. Measures put in place

Three  different  measures  are  proposed  with  a  view  to  making  the  CARE  arrangements sustainable for the target period of at least 25 years. These are as follows:

• Prudent valuation assumptions
• Risk-sharing
• Revised governance arrangements

1. Prudent Assumptions

The CARE benefits are to be funded on "prudent" actuarial assumptions. The most significant of these assumptions will be the assumed real rate of investment return (ie the nominal investment return less the assumption for future increases in Jersey RPI).

At the Panel hearing on 23 April, in response to the question as to how critical prudent funding assumptions are on the success and sustainability of the CARE benefits, the chairman of the COM said:

"I think that it is a very important factor in the sustainability of the new arrangements".  He went on to say:

"They would  not be sustainable if there was reluctance by the  States to pay an increased contribution. If a demand for increased contributions can be avoided then the scheme is much more likely to be sustainable. So there is almost a direct link there."

While he agreed that the concept of prudence should be reflected in the legislation, initially he did not feel able to say whether it should be in the Draft Law or the regulations. He subsequently indicated that the regulations should specify that the funding should be prudent.

We would recommend that the Panel push for the concept of prudent funding to be included in the regulations. If this is not done, then there is a very real risk that the degree of prudence in the assumptions drifts more towards "best estimate" over time. This might happen for example if there were pressure to have a funding level within the funding corridor of 95%-105%, perhaps to avoid the need to reduce benefits.

There is some evidence to suggest that prudency drift may have happened in PECRS historically. The Chairman of COM said at the hearing that when the current arrangements were set up in 1988 they were funded on prudent assumptions.

From the information that we have reviewed, it is not clear why the PECRS funding basis has moved to "best estimate" more recently. However, it appears to have at least been in part to manage the funding levels in the absence of increases in the contribution rates when deficits would have been revealed on more prudent assumptions.

To avoid this happening in the CARE scheme, we recommend that the Panel pursue the inclusion of  prudent  funding  assumptions  within  the  regulations. On  an  associated  point,  we  also recommend that the Panel consider whether the employer and the COM should be advised by separate actuarial firms. This is considered further in section 11.5

2. Risk-sharing arrangements

Within traditional defined benefit schemes it is the employer who picks up the balance of the cost if there is a funding shortfall. Over recent years the cost of providing pensions has increased and many employers have ceased providing defined benefit pension arrangements because they have become unaffordable.

This is not expected to occur in PECRS or the CARE scheme due to ability to reduce both past and future service benefits. If the target benefits are not fully funded then benefits would be reduced through lower benefit revaluations, thus improving the funding position. Consequently a large proportion of the underfunding risk is passed to members.

However, there is to be a minimum level of benefits that must be provided from PECRS and CARE scheme. If after reducing increases to their minimum levels the assets are still deemed insufficient to meet the liabilities then further contributions would be required. It is not clear that the States have committed to meeting any such shortfall.

Aon Hewitt have already produced some modelling to indicate the probability of pension increase cuts in the future, and some of these outcomes show pension increases at their minimum levels. This implies that underfunding remains a possibility even, once benefit increases have been cut to their minimum levels. We suggest that Aon Hewitt are asked to expand their modelling to calculate the probability of the funding level (on the minimum benefit structure) falling below 100% at future time points if joint contributions of 24% of pensionable earnings are paid. Aon Hewitt could be asked to complete the following table to summarise their modelling. This could be done separately for PECRS and the CARE scheme.

These calculations would highlight how high the risk that additional contributions are required is and hence demonstrate whether the proposed benefits are sustainable.

3. Governance arrangements

Both the chairman of the JNG and the chairman of the COM alluded to the breakdown in communications with the employer around the time of the 2007 valuation as being one of the reasons why the deficit at the 2007 valuation was not addressed. This ultimately led to the reduction in pension increases to eliminate the deficit.

Going back over time, there appears to have been a tendency towards procrastination in dealing with funding shortfalls, evidenced by the fact that the approach to dealing with the 1987 Debt issue was not resolved until 2002, and only then with a repayment plan of 82 years. There were further delays in implementing the 10 point plan (see section 3.3) to repay the debt, with the regulations dated 2005. The reasons for this apparent historical reluctance of the States as employer to address the pension funding on a timely basis are beyond the scope of this report.

With a greater focus on the governance arrangements around the new arrangements, it would appear that in future there will be more collaboration between the employer and the COM, with oversight by the Treasury.

3. Comments on sustainability

It is critical that the CARE benefits are funded on prudent assumptions, to maximise the chances of the actual investment performance exceeding that expected. If this can be achieved then it is more likely that the CARE benefits can be fully funded, without the need to resort to limiting members' pension increases and revaluation of CARE benefits.

Once Aon Hewitt have provided the probabilities noted in section 7.2.2 above, the relevant parties will be able to assess whether the risks are considered small enough that the proposals are deemed to be sustainable.

8. Fairness

1. What does fairness mean?

Intuitively the concept of fairness is about treating people equally and avoiding any discriminatory treatment. However, it is recognised that there may be some situations where there are objective reasons for treating people differently. In some circumstances the perception of what is fair and equitable can be subjective.

2. Are the current arrangements fair?

Concerns have been expressed around the fairness of the existing PECRS benefits structure in a number of areas. These stem from:

• The nature of a final salary scheme
• The cross subsidy between uniformed and non-uniformed staff
• Different accrual rates
• The approach to dealing with surpluses and deficits

1. Final salary benefits

The retirement benefits provided by PECRS are dependent on members' salary progression in the three years prior to retirement or leaving service. Therefore an employee who receives a substantial promotion three or so years before retirement will benefit from a substantial increase in their accrued pension, despite only paying contributions on their higher salary for a short period. Therefore they benefit substantially more from a final salary scheme than a member with similar length service but without a late promotion.

2. Cross Subsidy

At present some uniformed members have a normal retirement age of 55 while many non- uniformed members have a higher normal retirement age of 65. The age at which a person can retire with an unreduced pension is an important factor in determining the cost of providing that pension. It has long been recognised that the value of the pension benefits for uniformed staff is significantly greater than the pension benefits of non-uniformed staff, yet the contribution rates they pay are currently the same. Therefore the non-uniformed members are effectively cross- subsidising the uniformed members' benefits.

3. Different accrual rates

As noted in section 3.2.4, there are four different accrual rates at the current time within PECRS, ranging from 1/80ths to 1/45ths. Therefore there is a significant difference between members which appears to be unfair.

4. Method of dealing with surpluses and deficits

Although PECRS has the appearance of a defined benefit scheme to its members, in practice the States have operated it as if it were a defined contribution scheme, by keeping the rate at which they contributed to the scheme broadly fixed. As a result the States have accounted for PECRS as a defined contribution scheme within its accounts. If it were treated as a defined benefit

scheme  then  the  States  would  need  to  account  for  the  funding  position  of  PECRS  on  the prescribed basis under accounting standards applicable to defined benefits schemes within its accounts. This would be expected to result in a substantial deficit having to be disclosed. The notional PECRS deficit under accounting standard IAS19 disclosed in the States accounts for 2012 was £766.8M (2011:£698.0M).

In an effort to maintain the "fixed" contribution rate, both the actuarial valuation assumptions and the funding method have been adjusted over the years to manage the funding position. However, this approach became unsustainable as a result of the size of funding shortfall revealed at the 2007 valuation. The regulations governing the operation of PECRS permit a reduction in the pension increases, or even a cut in benefits to restore the funding position. Consequently the COM reduced the level of pension increases to RPI  - 0.3%pa following the 2007 valuation. Therefore,  despite  the  appearance  of  a  defined  benefit  scheme,  the  underfunding  risk  is effectively borne by the members.

3. Are the proposals fair?

In  considering  whether  the  CARE  proposals  are  fair,  we  have  focused  on  any  differential treatment within the proposed CARE scheme.

1. Protection

There are two groups of members who are eligible to continue accruing final salary benefits. These are:

• those within 7 years of normal retirement age on 1 January 2015
• those with a 45th accrual rate

We have  been advised  that the starting point from the TWG was that there should  be  no protection, other than that accrued benefits would not be affected. However the protection was introduced as part of the negotiation process with the JNG.

"7 Year" Protection

The members within 7 years of normal retirement age will be the oldest members of PECRS and therefore it is likely that they will already have substantial service and benefits built up within PECRS. As such it is likely to be those who already have the best benefits who are being afforded protection. This appears to be unfair to the younger members.

In addition, whilst Jersey does not yet have anti-age discrimination legislation, we understand that it is expected to be introduced at some point in the next few years and there is a risk that the protection  could  be  challenged  on  those  grounds. Even  if  it  is  acceptable  from  a  legal perspective, it does appear to be intrinsically unfair to other members. It also appears to be unfair to taxpayers. We have been advised by the treasury team that the cost of this protection is estimated to be 0.2%pa of pensionable earnings if spread over a 10 year period.

On  the  issue  of  the  fairness  of  the  protection  arrangements,  the  Chairman  of  the  COM commented to the Panel during the hearing on 23 April: