WEST BANK AND GAZA PUBLIC EXPENDITURE REVIEW The Palestinian Pension System: A Roadmap for an Unfinished Reform Final Report 12 September 2023 Acknowledgment This report was prepared by Oleksiy Sluchynsky (Senior Economist, WB), Lidiia Tkachenko (Demography and Social Insurance Specialist, WB Consultant), and Eimar Coleman (Senior Institutional and Social Security Reform Adviser, WB Consultant). The PROST simulations were conducted by Lidiia Tkachenko. Nur Nasser Eddin (Senior Economist, WB) and Luan Zhao (Senior Economist, WB) provided valuable contribution to the report. The team appreciates feedback on this report received from the Ministry of Finance and from the Palestinian Pension Authority. The team would like to thank Ayman Aldoqi (General Director of Administration and Planning, PPA) and his colleagues for their invaluable guidance, feedback, and coordination. The management and staff of the PPA generously contributed their time and provided operational data as well as answers to SIAD questionnaires at the request of the World Bank team. The views expressed in this report are the responsibility of the authors. They do not necessarily reflect the position of the institutions they represent. 2 Contents Acknowledgment ...................................................................................................................... 2 Executive Summary.................................................................................................................. 4 1. Background ...................................................................................................................... 5 2. Institutional and Operational Setup .............................................................................. 5 3. Socio-economic Impact and Policy Challenges ............................................................. 8 4. System Finances ............................................................................................................. 12 5. Fiscal Simulations: Status-quo and Parametric Reforms .......................................... 17 6. Conclusions and Outline of the Proposed Reforms .................................................... 24 Annex I. History of the PA’s Policy Initiatives and Reforms in Pensions ......................... 30 Annex II. Modeling Methodology Overview ........................................................................ 33 Part A1. Assumptions........................................................................................................... 33 Part A2. PROST Model Overview ....................................................................................... 38 Annex III. Administrative Diagnostics of the PPE .............................................................. 41 A1. Internal Governance ...................................................................................................... 41 A2. Investment and Management of Funds ......................................................................... 46 A3. ICT Management and Data Policies .............................................................................. 46 A4. Identification and Registration ...................................................................................... 49 A5. Contribution Collection................................................................................................. 50 A6. Benefits Management ................................................................................................... 51 A7. Grievances - Appeals and Complaints .......................................................................... 53 A8. Communications ........................................................................................................... 53 Summary of Observations and Recommendations .............................................................. 54 3 Executive Summary • This work builds on the Public Expenditure Review (PER) pensions analysis conducted in 2016. It takes stock of the current issues, collecting new data, updating financial analysis and fiscal projections, and offering new policy recommendations. • The pension system for the public sector employees and security personnel of the Palestinian Authority (PA) is a system in transition from fragmented arrangements to a unified pension scheme, open to participation of the non-government civil society institutions and their employees. While the legacy schemes have practically been phased out (with only the group of the beneficiaries of those schemes remaining), the new scheme still suffers from challenges, some of which have been imposed by the external environment while others reflect intrinsic deficiencies of its design. • On the demographic side, the system dependency is projected to worsen dramatically in the longer term, while changes in the next decade will be moderate, which could be used as an opportunity for further reforms. • On the benefit side, the Defined Benefit (DB) component remains financially unbalanced: benefit promises are not in line with contribution rates and retirement ages, including early retirement privileges. Balancing these three parameters is essential to guarantee the financial sustainability of any DB pension scheme financed on a pay-as-you-go basis. • The Defined Contribution (DC) component has never been effectively delivered as a tool for strengthening the income security of the retiring public sector employees. It operates on accrual basis, offers zero nominal interest, and pays only face value of the total contributions in a lump-sum. • On the financing side, the premise of the 2006 reform that the government would pick up most of the transition cost of the reform by paying for the liabilities of the legacy schemes, while at the same time contributing to the new scheme, turned out to be unrealistic. While pensions have been consistently paid out, the mounting stock of contribution arrears, currently at over 15% of GDP (despite the fact that some of those contributions have been deducted from employee wages) serves as evidence of a reality that is far from the reform’s original intent of fiscal self- sufficiency. • The proposed reforms should proceed on three parallel tracks: (i) financial restructuring, (ii) parametric adjustments, and (iii) systemic changes (including fiscally neutral change in the benefit package and revisions to the DC benefit). As result of the proposed changes, the public debt towards the Palestinian Pension Agency (PPA) could be instantly removed from the MOF treasury books in exchange for the government taking full responsibility for funding future DB pension payments (at least in the proposed new non-contributory element). The DC component could be transformed into a severance benefit, as part of rationalizing the early retirement mechanism. Additional measures should touch on strengthening the institutional and operational capacity of the PPA. It is important to note that any public sector pension reform – its objectives and results – should be evaluated primarily in the context of an overall compensation package in the public sector. 4 1. Background This work builds on the Public Expenditure Review (PER) analysis of the pension system conducted in 2016 but also brings important innovations. It updates the financial analysis conducted then and offers new policy recommendations, with the key objective of improving the affordability of the pension system, subject to benefit adequacy. The principal difference of this work is in the adopted approach, which looks at the pension system holistically and offers advice from the angle of rationalizing the overall public expenditure policy in pensions, rather than working with each pension scheme individually. In doing so, it takes a forward-looking approach and minimizes the dependency path of the convoluted legacy system. The work also incorporates a component that looks at the Palestinian Pension Agency’s (PPA) administrative systems using a new World Bank tool – Social Insurance Administrative Diagnostic (SIAD). The findings reveal a system with unfinished reforms and call for a comprehensive restructuring of the pension financing arrangements, along with a change of the construct of the system. This would lead to a system that can be finally and truly consolidated across different territories, sectors, and employee cohorts and would be closer to its financially sustainable path. This work is largely diagnostic in nature, offering high level policy recommendations and suggesting a comprehensive follow up analysis to elaborate the parameters of the reformed scheme. While presenting a status quo analysis and illustrating effects of several parametric adjustments, the main set of recommendations is centered around the structural adjustments and financial restructuring in the public sector pension scheme, which would require a round of follow up work, elaborating and modeling various reform cases, including careful fine-tuning of the transitional provisions. While we focus only on the public sector pensions, any future developments there should be well coordinated with the plans and policies of introducing pension provisions for the private sector, ensuring pension benefit portability and promoting labor mobility across sectors. 2. Institutional and Operational Setup The pension system for the public sector employees in the Palestinian territories is an intrinsic part of the broader system of public finances of the Palestinian Authority (PA), with benefit expenditures1 constituting 3.5% of GDP or 13% of the PA’s budget in 2022. As a share of GDP, this constitutes a 17% increase in the overall expenditures compared to 2013, the year of the data used in the last PER analysis 2 . The system has gone through rounds of reforms and adjustments – both parametric and systemic3. However, reforms remain unfinished as there continue to be 1 Total benefit and administrative expenditures, including all schemes and benefits. Estimated at the total of NIS 2.3 billion, on the basis of data from PPA. 2 The pension expenditures in 2013 were reported as NIS 1,089 million, or around 3% of GDP. 3 The pension system is governed by the Palestinian Pension Law 7 of 2005 – incorporating previous retirement regulations or laws (Civil Retirement Law No 34 of 1959 and its’ amendments, Insurance and Pensions Law No. 8 5 concerns regarding the soundness of the system’s finances and its sustainability – financially and politically – in its current form. It appears that some critical solutions are largely beyond the control of the PPA and should be assessed in the broader context of the PA’s public sector finances and operations. The year 2021 marked the end of a 15-year transition from a highly fragmented setup to a unified pension scheme introduced in 2006, as the last contributors of the legacy schemes have retired. Prior to the 2006 reform, the PPA operated three different schemes with different parameters. Two of those schemes covered civil servants: one was for the West Bank and was inherited from Jordan (hereafter referred to as Scheme IV), and the other was for Gaza, inherited from Egypt (hereafter referred to as Scheme I) 4 . In 2004, a new Security Services Pension Law (SSPL) was enacted to provide pensions for all security personnel aged 46 in 2006 and above through a scheme, hereafter referred to as Scheme III. However, since September 2006, and as a result of the unified pension law (UPL) that was enacted in 2005, all government employees including security staff under the age of 46 became members of a new scheme (typically referred to as Scheme II). Older employees were “grandfathered� and remain beneficiaries of their original pension schemes (Schemes I, III, and IV). Hence, beneficiary coverage in different schemes is age dependent. While the legacy schemes I, III, and IV were a defined-benefit (DB) type, Scheme II, in addition to a reduced DB component introduced a new defined-contribution (DC) component (which was never fully implemented5). Today, all participating employees are covered under scheme II only (referred to in this report as the “new scheme�), while schemes I, III and IV (“legacy schemes�) only pay benefits to the current retirees and will continue doing so until the last member retired under those arrangements will pass away and all surviving family members will cease to be eligible. of 1964 and its’ amendments, and Insurance and Pensions Law of the Palestinian Security Forces No. 16 of 2004 and its’ amendments), and by the subsequent Presidential Decree 29/2018. 4 The PA has been conducting efforts to integrate the two civil service schemes since 1998, which is why in 2000 the West Bank civil servants’ scheme (Scheme IV) was closed off and new entrants were enrolled into the Gaza scheme (Scheme I). 5 By design, DC schemes are fully funded and provide for self-financing of future pensions, reducing reliance on the public finances. However, the UPL did not provide explicit requirements regarding investment of pension assets or any particular form of income accounting in individual accounts, leaving significant ambiguity on how those contributions and accounts could be used. 6 Table 1: Contributors and beneficiaries of the schemes managed by PPA Civil Service Security Sector Non-Gov Sector New Scheme Legacy Schemes New Scheme Legacy Scheme New Scheme Total II I IV II III II 2021 Insured 82,069 84 - 52,479 24 13,049 147,705 Pensioners Old Age 3,536 10,821 18,452 18,057 5,271 267 56,404 Disability 1,277 286 - 25 - 31 1,619 Survivors 1,637 4,924 incl. 2,088 2,817 125 11,591 2013 Insured 74,438 6,455 4,670 63,724 1,202 3,054 153,543 Pensioners Old Age 6 4,954 771 5,561 - 21,433 10,141 Disability 159 347 8 - - 514 Survivors 1,777 6,951 4,863 2,976 3,264 4 19,835 Source: Authors’ compilation based on data from PPA. Note: Survivors are individual beneficiaries. The system covers risks resulting in long-term pension benefits (old age, disability, and survivorship). It does not protect against various short-term risks that may result in a temporary loss of income, such as sickness, maternity, accident, injury (some of which may be covered by the public sector employee compensation policies but would remain outside of the PPA’s operational domain). While, in contrast to most countries, the private sector in the Palestinian territories does not have its own state mandated pension program, the UPL provided for the voluntary participation of the employers and employees of the non-governmental organizations in the new pension scheme under the same parametric provisions. There are close to 100 organizations, 14,000 employees, and already around 500 retirees from the sector of non-governmental organizations (NGOs) who participate in the new scheme. While analysis of these arrangements falls outside of the scope of this work, the implications of the assessment of the civil service arrangements (see below) will likely be similar. The system is administered by the Palestinian Pension Agency, an independent public legal entity reporting directly to the President of the Palestinian Authority. The PPA, as a separate public agency, was established in 2005 under the Palestinian Pension Law 7 of 2005 and it took over from the General Pensions and Insurance Corporation (GPIC). A rapid administrative diagnostic6 of the PPA (see Annex III) indicates that it is a well-structured organization with the operations solidly rooted in the ICT7 culture. At the same time, the assessment points to certain shortcomings. For example, from the perspective of fiscal analysis, there is limited availability and poor quality of the records of the legacy schemes III and IV (e.g., in age/gender dimension), which is due in part to the weak data sharing provisions between the PA and the PPA. Apart from implications for this analysis, this is a serious operational shortcoming, as it limits transparency of the budgeting process and weakens financial controls. Further observations and recommendations are provided in Annex III. 6 The Bank team deployed the Social Insurance Administrative Diagnostic (SIAD) to assess various aspects of the PPA’s governance, institutional, and operational setup. 7 Information and Communications Technology. 7 3. Socio-economic Impact and Policy Challenges The system provides a decent rate of income replacement for those who retire under one of the above-mentioned schemes. The civil servants and employees of the non-government sector retire today with a pension that represents, on average, around 60% of the average wage for the sector, which is well above the minimum of 40% recommended by the ILO for a full career worker. The security sector employees are awarded pensions at a rate exceeding the average wage in the security sector (which likely reflects significant pre-retirement wage hikes, perhaps associated with rank promotions around that time). Notably, the main benefit is supplemented, universally, by a monthly “family and personal allowance�8. (It is referred to further in text as a personal allowance, as we only assess impact of the personal NIS 300 benefit). This supplement is reflected in all the reported and projected pension benefit numbers and is procedurally funded by the MOF. Table 2: Income standards of employees and Old Age pensioners in 2021, NIS Average Average of new Average of all Pension-to-wage contributory wage old age pensions old age pensions ratios (all pensions) Civil Service (II+I)* 3,850 2,413 2,453 64% * Security Sector (II+III) 3,615 4,329 3,241 90% Non-Gov Sector (II) 3,857 2,302 2,310 60% Source: Authors’ compilation based on data from PPA Notes: * - average for all schemes in the sector The new scheme is still transitioning to its demographic steady state, with system dependency projected to worsen over the next few decades. It is important to look at the demographic transition of the pension system overall, rather than each of its four schemes in silo. The old age dependency of the system overall (measured as the number of old-age beneficiaries over the number of active contributors) – in both the civil and security sectors – will increase from around 40% in 2021 to around 70% by 2080. The major driving factor of this change is the projected significant improvements in the life expectancy9. Another notable aspect is a rapid worsening of dependency of the security sector scheme – from around 40% to 60% over a 15-year period beginning in 2030. Despite the large increases in the old age dependency in the longer run, the changes over the next 10 years will be moderate, which could be seized as an opportunity for further reforms. 8 This benefit was inherited from the legacy Scheme I. The amount of the payment is set by the President, but its value has been fixed at the level of NIS 300 since 1995. The family top-up includes a NIS 60 per spouse and NIS 20 per child. This payment is assigned to all pensioners, regardless of the type or rate of the pension, and also regardless of other sources of income. For the survivors, that amount is split in accordance with the rules of inheriting the shares of the main pension. 9 The UN population projections, which we used in our analysis, suggest the expansion of life expectancy at the age of 60 from around 16 to 25 years for men and from around 20 to 27 for women. This will be approaching the average span of years of service in the public sector scheme. In other words, unless things change, the beneficiary on average would be receiving pension for the same period of time that she or he had worked prior to separation from service. With all other parameters, this is typically a clear indication that system will not be fiscally sustainable (with such a demographic transition, a contribution of around 20% of wage can no longer support a pension of over 60% of wage for the same periods of time on both ends). 8 Figure 1: Projected old age dependency in the pension system 80% 70% 60% 50% 40% 30% 20% 10% 0% Civil Service Security Sector Source: Authors’ calculations. Notes: Civil Service - Schemes I, II (civil), and IV; Security Sector - Schemes II (security) and III. Despite significant public resources dedicated to the pension system, its coverage is narrow, by design. The system covers just over 5% of the Palestinian working age population, around 10% of all those employed in the economy, and around 20% of the population over the age of 60. Since this analysis is focused on a professional pension system for employees of the public sector, we do not perform projections for the total labor force. Rather, the projections assume that as the population goes through demographic changes, the public sector’s employment share in the total labor force will remain fixed. The modeling does not assume any expansion in the contributory coverage of the private sector. Such a policy is, anyway, not encouraged, unless the system is made parametrically sustainable first (See below). The legacy schemes offered a relatively generous benefit, hence the 2006 reform meant to address that concern 10 in addition to providing greater flexibility regarding financing the public pension system. The key parametric change was the reduction of the “pension accrual� factor11 in the benefit calculation formula from 2.5 to 2% in the new scheme. The reduction in the benefit was compensated by launching a new complementary DC component (with part of the old contributions to be channeled to finance that new component). Notably, the reduction in DB contribution was larger than the DB benefit reduction12 -- which negated the potential fiscal gain within the DB program. The ideology of the reform was, presumably, to reduce dependency on the budget subsidies in the long run, by shrinking the DB component in the new scheme (on both the contribution and benefit sides), but the intrinsic sustainability of the DB scheme was addressed only to a limited extent. 10 See Annex I for history of the reforms as well as historic parameters of the legacy schemes. 11 This parameter links pensionable wage to the calculated pension. It shows how each year of service adds to pension benefit. In the new scheme, each additional year worked adds 2% of the final pensionable wage to individual’s monthly pension level. 12 Under schemes I and III before the reform, someone retiring with 25 years of service would be eligible for a pension with a replacement rate of 62.5%, while after reform the same service provided for a 50% replacement rate, which is a 20% reduction in the benefit due to the reform. At the same time, the contribution rate to the DB component was reduced from 22.5% to 16%, which is equivalent to a 29% reduction. 9 Table 3: Parameters of the New Scheme (Scheme II), 2023 DB: 9% Employer and 7% Employee Contribution rate DC: 3% Employer and 3% Employee Retirement age 60 years Required minimum service (vesting) 15 years Basic replacement with minimum service 30% Incremental replacement (accrual rate) 2% Maximum replacement 80% Number of Years for Wage base 3 Wage valorization Not provided Benefit indexation Inflation* Source: Authors’ compilation. Notes: * The law requires pensions to be indexed to inflation, subject to government approval. For our long-term assumptions about indexation, see Annex II. While the legacy schemes will no longer produce new pensioners, their impact will continue to be present for a long time in the benefit calculations of those that retire under the new arrangements. Indeed, the newly retired pensioners of the new scheme have pensions composed of two parts – one calculated under the new rules and based on service rendered under the UPL, after the year 2006, as in the table above, and another part resulting from the individual’s service prior to 2006. The latter part still follows the calculations under the rules of one of the legacy schemes (see Annex I). Furthermore, the interim agreement with the MOF is that as long as someone participated even a year in the pre-reform system, while retiring under the new system, such a retiree will still be a financial responsibility of the MOF, pending reconciliation of contributions and arrears. The UPL requires that pension benefits be indexed, at least once every 3 years, in accordance with CPI, subject to a government decree. However, in reality, while wages have largely been growing in line with consumer prices, pensions have remained unadjusted since 2014. While the public sector wage has remained fixed for a long period of time (with the last CPI allowance approved in 2014), there has been a personal service-related adjustment in wages of 1.25% annually. The combination of such adjustments and various promotion related wage increases resulted in a continuous growth in average public sector wage. Pension adjustments, however, did not follow the intent of the law. In the future, in periods of significant real wage growth, decision will need to be made on the benchmark for pension indexation. Some countries, to better sustain the pension benefit value, use a hybrid approach of an index that combines growth of wages and price inflation13. Retirement under the statutory retirement age remains very common: the age of almost a quarter of the current civil service regular retirees and of 80% of the retirees in the security sector is under the statutory retirement age. The mandatory retirement age is 60. A civil servant may be eligible for early retirement after completing 15 years of contributory service and reaching the age of 55. A full pension is granted if the separation is a decision of the employer (government), or if the 13Given the operational uncertainty with the current setup and assuming that the wage growth will slightly accelerate in the future, in our analysis, we assumed a more conventional mechanism of pension indexation of 50% inflation and 50% wages, to balance the social and fiscal considerations. 10 employee earned 25 years of service at the age of 50 or 20 years of service at the age of 55 (in the civil sector). Otherwise, the benefit is reduced14. Almost all individuals who separate from service early, do so under an official decree, which exempts them from paying such penalty. Security service employees, as well as contributors in the risky professions, may receive a full pension after 20 years of contributory service and reaching the age of 50. Any separation from service triggers costs for the pension system, which is suboptimal from the policy perspective. For example, an individual who has not accumulated the statutory minimum of 15 years of service for pension, at the point of separation, will typically claim contributions from that service in a lump sum. If such an individual decides to rejoin the service later, she or he will have to start accumulating pension service from anew. Such benefit fragmentation does not respond to a core social principle of the pension benefit provision. Ideally, as most pension programs globally provide, all contributions along with the pension rights should be preserved until the retirement age, and only then the application for benefit and decision on eligibility should be made. In fact, it may not be unrealistic to imagine someone separating from the civil service after a short period and then at some point joining a non-governmental agency that participates in the same system (see below). While this is something that could be encouraged from the labor policy perspective, such cases may end up being treated less favorably from the perspective of pension policy, producing two different lump sum payouts instead of a regular stream of pension at retirement15. Although it affects a small number of employees 16, a separation with insufficient service for retirement penalizes the employee, but an assistance benefit is made available. Individuals who were in service for less than 15 years do not accrue rights for a regular pension. Instead, they have their nominal contributions paid back without any interest or inflationary adjustment in value 17 . At the same time, such former employees of the public sector may be eligible for a “basic pension� of NIS 700, funded by the MOF and, in theory, means-tested. This may be quite unfair: someone who worked for 5 years only and in low-grade position or for 14 years and has progressed in ranks ends up being eligible for the same benefit. We provide corresponding recommendations below related to both the “basic pension� benefit and the vesting period of the main scheme. The surviving beneficiaries of the deceased members constitute a sizable group of the pension system beneficiaries: around one third of all PPA’s beneficiaries are either disabled or receive survivorship benefits. While this is not uncommon for the region, these arrangements result in high costs to the budget 18. The majority of the current 14 The UPL provided for a 4% reduction for each year under the statutory retirement age, but that was subsequently changed, by a President’s decree in 2007, to a flat 5% reduction until the retirement age. 15 Although, there are provisions allowing such individuals to pay all such contributions back to PPA and to re- instate the record of service that could produce regular benefits after re-joining the system. 16 Under 100 individuals at the age of retirement, but also some number of those who separated early. 17 At the same time, those with insufficient service when reaching the retirement age are allowed to purchase the missing years of service (up to 6 years) to earn the right for a regular pension. 18 A wide range of cases are entitled to receive benefits from the survivorship program including: (i) the widow or widows of a participant; (ii) children and brothers younger than 21 who were dependent on the participant at the time of death; (iii) any children and brothers between the ages of 22 and 25 years old who were dependent on the participant at the time of death and who are continuing their university or higher education; (iv) any children and brothers who were dependent on the participant at the time of death and are incapable of earning a living due to health issues; (v) any daughters and sisters who are unmarried, widowed, or divorced; (vi) parents of the participant; and (vii) husband of the participant if, at the time of her death, he was medically unfit to support himself. 11 survivor beneficiaries are from the legacy schemes I, III, and IV, but the rules in the new Scheme II were not tightened enough, so this trend will likely continue. Since rules cover a broad range of surviving dependents, and there are strong social and cultural reasons that prevent their adjustment, these pensions will continue posing a financial and administrative burden on the system. They also disincentivize participation in productive employment, particularly among youth and women. This area remains a focus for future policy deliberations. 4. System Finances Financing of all the schemes and benefit is not very transparent. The PPA runs a very complex financing mechanism with multiple schemes, in different territories, financed from multiple sources, in different currencies, with some components operating only on accrual basis. Table 4 provides a glimpse into that challenging operational context. The premise of the 2006 reform that the government would pick up most of the transition cost of that reform turned out to be an unfeasible plan. The transition costs of a pension reform arise when an old scheme closes and all the contributions are channeled to a new scheme, typically of a different design, leaving the bill of the old scheme without a stable source of financing. For example, when the government introduced a new DC element of the system, a 6% contribution was redirected to its financing. That immediately created a transition cost, as pensions of the old DB system (which in part were supported by those 6% in the past) still needed to be paid. At the same time, if a reform introduces changes in the DB scheme, there is no need to establish a new scheme each time such parametric changes within the DB scheme take place, partitioning such schemes into separate financing mechanisms, and instituting the transition costs that fall on the government. Yet, the reform of 2006 produced exactly such kind of a transition, where the new scheme was largely seen as self- financed and pre-funded (both DB and DC), while the legacy DB schemes (except scheme I) fell under financial responsibility of the government. 12 Table 4: Pension schemes finances, 2021, NIS millions NIS Civil Service Security Sector Non-Gov Sector New Scheme Legacy Schemes New Scheme Legacy Scheme New Scheme II I IV II III II MOFl Acrued Contributions DB 633.0 358.0 - - - DC 237.3 134.2 - PPA Actual Revenues ER and EE contributions - 65.9 Transfers from MOF 40.0 712.5 412.2 - Other revenues 13.5 - Investment income 33.3 Expenditures Regular Pensions incl 507.6 412.2 All Supplements 620.8 712.5 - DC payouts - - - Admin expenditures 13.6 Source: Authors’ compilation on the basis of data from PPA. Notes: DB contributions – authors’ estimate In 2021, PPA claims that financing of the civil service pensions was largely conducted at the expense of PPA’s reserves. Indeed, in the context of very harsh fiscal constraints, the PA is made to pay for two different types of expenditures. De jure it is required to cover costs of (i) the current pensions that were not adequately funded in the past (Schemes I, III, and IV), and (ii) pre-funding of the future pensions (Scheme II). Table 4 below provides more details on how the cost accounting of the pensions in the Palestinian territories works. A rough assessment suggests that close to 70% of all the pension payouts today is directly or indirectly a legal responsibility of the MOF – this is on top of the MOF’s statutory contributory obligations to Scheme II that are of the same magnitude! If there was any logic in setting up the legal framework enabling such a mechanism, it certainly failed, as the de facto situation of covering only the current pension payouts and a reality of very significant and growing contribution arrears is calling for an urgent and complex reform of the public expenditure policies towards pensions. It is unlikely that the government will be able to pay that stock of arrears back. Hence, some restructuring may need to be considered. Table 5: Funding responsibilities towards different pension benefits Pension Benefit Funding Responsibility Scheme I PPA Scheme II civil service and NGOs PPA Scheme II security sector MOF (*) Scheme II (those with any pre-2006 service in Schemes III or IV) MOF (*) Scheme III MOF Scheme IV MOF Family and Personal Allowance, MOF/PPA(**) Source: Authors’ compilation on the basis of information from PPA. 13 Notes: (*) – pending reconciliation of contribution arrears between PPA and MOF (*) – personal allowance of NIS 300 is funded by the MOF, while the family top-ups (NIS60 per spouse and NIS20 per child) are funded by the PPA Indeed, the financing of the pension system remains a major challenge for the government, as the accumulated contribution arrears stand over NIS 11 billion. Around NIS 3.7 billon of the total stock constitute the accrued contributions to the DC component of the new scheme (with no interest accrued by the PPA on those accounts)19. While the PPA applies financial sanctions for the non-payment of contributions, this does not seem to be the best instrument to deal with this systemic issue. A quarter of the total stock of arrears constitutes interest and penalties. Since the PA is ultimately responsible for paying all pensions, such penalties do not serve any practical purpose. The PA’s arrears towards the DB scheme could have been largely avoided if a different transition mechanism was chosen at the time of reform. The collected contributions could have been by law directed to finance current pensions across all schemes, hence gradually clearing all past liabilities. In a way, this could well be the best investment of the surplus of the young Scheme II. At the same time, financially, this would also secure full compliance of the government with its contributory mandate. This approach would deliver a fiscal reality with no dichotomy between de facto and de jure. That new reality would require some additional policy adjustments to address (i) the lack of parametric balance in the DB scheme (see below), and (ii) inefficiencies of the operational setup of the DC scheme. The new DC component was meant to compensate the DB benefit reduction for the members of the new scheme when it was introduced in 2006, but it was never properly implemented. Today, the PPA maintains close to 137,000 individual DC accounts across different sectors, with total value of close to NIS 3.7 bln, on the accrual basis. This amount includes only contributions and no interest (which would normally have accrued over the past 15 years on those contributions). However, these records and amounts are only on paper (in electronic form, to be precise). In reality, while the contributions get deducted from the employees’ pay, they are not transferred by the MOF Treasury to the PPA to fund the individual DC accounts and consequently no interest accrues on those accounts. On cash basis, contributions largely finance general budget. On accrual basis, the contributions registered by the PPA in DC accounts deteriorate in value over time. Upon retirement, the PPA simply pays to the workers the total accrued amount of the contributions as a lump sum and/or in installments20, in turn billing that amount to the MOF. From the semantical perspective, this scheme operates as a non-contributory and zero-interest Notional Defined Contribution (NDC) scheme, which does not have any precedents globally. It also does not follow any meaningful logic and remains suboptimal from a social perspective. The change that the UPL meant to introduce in 2006 resulted in lower (DB) benefits for pensioners and higher risks for public finances. Simply put, the lump sum payments take away from the regular pension benefit. First, the employees would likely have been better off by putting all those DC contributions 19According to the PPA. 20In principle, the Law provides for a choice of a form of payment – regular or one-off – but only one-off lump sum payments are most common. 14 in bank deposits. Second, such a reduced DB benefit without a compensating regular payout of the DC scheme produces political pressure to undo reforms in the DB program and increase DB benefits with an increased dependency on the budget – again. Third, when the lump sum DC money is used up and an individual is left with a lower DB benefit, there is an increased risk of poverty, which would become an additional challenge for the government. Finally, the focus on a lump sum may lead to the perverse incentives pushing some to leave the service early, with adverse impact on the HR policy of the PA government. Therefore, this issue requires remediation. While the new scheme practically does not accumulate any new assets in either the DB or DC components from the public sector, the PPA still administers some financial reserves (around NIS 1bln). A quarter of this amount – a relatively small amount on its own – is associated with pension liabilities of one of the legacy schemes (see Table 6), and the rest is likely associated to a significant extent with participation of the non-Government sector (as both DC accounts and DB reserves). Given the highly volatile situation with financing the pension program, this stock of assets helps smooth any transfer shortcomings from the government, so pension benefits can be paid consistently. Funds are managed professionally (see Annex III) 21 . However, these reserves will not be near enough to cover the long term financing gap of the pension program. Table 6: Portfolio structure of the PPA’s reserves, USD Overseas Investments Local Investments Investment Type Scheme I Scheme II Scheme I Scheme II Deposits 0 - 6,568,350 16,567,500 Current Accounts 2,170,372 - 2,989,793 16,179,824 Bonds 19,437,423 - 0 0 Shares 19,197,793 - 21,076,278 161,330,605 Alternative Investments 892,682 - 0 0 Real Estate 0 - 17,165,071 42,781,112 Loans to Beneficiaries 0 - 0 31,742,317 Total 41,698,271 - 47,799,491 268,601,358 Source: PPA There are several issues pertaining to the policy and operations of the PPA’s financial reserves management: • The PPA will be close to its financing break-even point when on the accrual basis the expenditures of its own financial responsibility (costs of Scheme I and part of Scheme II pensions) would approach the accrued DB contributions of Scheme II. All this time since 2006, the PPA has been running accrual surpluses and accrued assets in the new scheme – on paper. In reality, its revenues (mostly MOF transfers) are made only to match the annual pension expenditure needs. So, the current stock of the PPA’s financial assets remains vulnerable to the discretionary funding policy of the PA’s government. Should the MOF in certain years transfer less funds than the payments of the legacy schemes would require, this may have an immediate and dramatic impact on the reserves. Indeed, the value of reserves is meager, accounting for only around a half of the annual expenditure needs of the PPA. In fact, the year 21 A detailed assessment of investment management practices is outside of scope of this review. 15 2021, allegedly, did establish such a precedent when the MOF did not transfer a significant portion of its obligations to finance pensions, so the PPA reserves in that year were used to cover the shortfall. This questions the viability and importance of the PPA’s asset management function. Some clear policy in respect to those assets needs to be developed regarding when and how they can be used in financing the current benefits, should the MOF continue experiencing financial hardships related to the pension scheme. This policy should specifically safeguard assets in individual DC accounts (at the moment, the only cash balances are in DC accounts of the non- government sector). • It is not totally clear what kind of operational separation exists between assets of different schemes. For example, would it be possible indeed for the PPA to use the assets accumulated as part of the participation of the non-government sector to cover the deficit of the civil service or security sector schemes? Ideally, the records and finances of the non-government sector should not be mixed with the finances of the public sector. 16 5. Fiscal Simulations: Status-quo and Parametric Reforms The fiscal analysis was conducted using PROST, as a tool of long-term pension modeling (see Annex II for brief description of the PROST model and assumptions). We conducted three separate sets of projections: (a) for the Civil Service, Scheme II (civil) and Scheme I, combined, given a common source of financing by the PPA, (b) for Scheme IV, financed by the MOF, and (c) for the Security Sector personnel, Scheme II (security) and Scheme III, combined. The revenues were projected on the accrual basis. The projections indicate that in the long run, the DB pension expenditures will expand, by almost 1% of GDP in the civil service, counting all the schemes collectively (See Figure 2.A and Table A4 in Annex II). This reflects, among other things, largely the same reality that was depicted in Figure 1, which is the projected increases in life expectancy, which makes financing the current pension scheme increasingly more costly22. Note that except for the personal allowance, which is funded by the MOF, the source of funding of the schemes was not differentiated (as noted in Table 5, part of the Scheme II financing attributable to service rendered under schemes III and IV is a responsibility of the MOF). The indicated revenues are accrued and remain stable as a share of GDP (as per our assumptions). Projections show that the deficit of the DB system in the long term may run to over 1.5% of GDP for the civil service scheme and over 1% of GDP for the security sector scheme (Figure 3). Under an alternative scenario of the civil service growth, where it remains stable in size over the next three years and then grows with the rate of 80% of the population growth, the pension expenditure is lower by almost 0.5% of GDP in the long run and remain relatively stable over the projected period (Figure 2.B). However, revenues are smaller too, hence change in the deficit in the long run is insignificant. On the accrual basis, if all contributions were duly paid, the total stock of pension financial reserves attributable to the Civil Service part of the program would sustain a non-deficit financing for around three years, after which those assets would get fully depleted. Furthermore, if, highly hypothetically, the accrued value of the government overdue DB contributions and penalties (assessed at NIS 11 bln, minus accrued value of the DC accounts of NIS 3.7 bln) was transferred to the PPA, adding to the stock of reserves, these reserves would get exhausted by around 2040. This indicates an intrinsic lack of fiscal sustainability of the system. 22Another factor is that we indexed pensions (except personal allowance) in the medium to long run by a composite index of prices and wages (see Annex II for assumptions), to sustain their value relative to wages. This adds around 0.4% of GDP in costs in the long run, compared to just the price indexation. 17 Figure 2: Projected consolidated DB expenditures, accrued revenues: CS schemes, % GDP A. Baseline Assumptions 3.0% 2.5% 2.0% 1.5% 1.0% 0.5% 0.0% Estimated personal allowance expenditures (I+II+IV) CS Pensions IV CS Pensions (I+II) CS Scheme II DB+Other Revenues (accrued) B. Alternative Assumptions for CS Growth 3.0% 2.5% 2.0% 1.5% 1.0% 0.5% 0.0% Estimated personal allowance expenditures (I+II+IV) CS Pensions IV CS Pensions (I+II) CS Scheme II DB+Other Revenues (accrued) Source: Authors’ calculations Notes: Revenues of DB component only (16%) 18 Figure 3: Projected consolidated DB expenditures, accrued revenues: SEC schemes, % GDP 2.5% 2.0% 1.5% 1.0% 0.5% 0.0% Estimated personal allowance expenditures (II+III) Sec Pensions (II+III) Sec Scheme II DB+Other Revenues (accrued) Source: Authors’ calculations Notes: Revenues of DB component only (16%) The social dimension should also be considered, specifically “replacement rates� (or rate of pension relative to the wage rate). We are focusing specifically on how the replacement rates of the newly retired individuals would change over time. The simulations show that the pensions in the civil service may gradually come down in relative terms by over 5% of wage in medium-to-long run. We observe a similar effect in the security sector. The gradual decline is largely due to the fact that service earned prior to 2006 and rewarded with the accrual rate of 2.5% for each year of service under the old rules is gradually phasing out of the system, being replaced with the service earned in the new system with the accrual rate in the benefit formula of 2.0%23. Figure 4: Replacement rates of the new CS Old Age pensioners, % of average CS wage 75.0% 70.0% 65.0% 60.0% 55.0% 50.0% CS DB Male CS DB Female Source: Authors’ calculations 23In fact, the base year of 2021 marked around a half-way of that transition, so the average baseline accrual rate was estimated at around 2.25. 19 The simulations of the parametric reforms indicate that they may help reduce the long-term deficit and sustain the standards of living of the pensioners. The following set of (moderate) parametric changes to the Civil Service pension plan were considered, as an illustration of their impact (applying those changes from the year 2025): • The statutory retirement age is gradually raised from 60 to 63 for both women and men, by three months annually, over a 12-year period. This is to keep pace with the increasing life expectancy in the medium term. • Early retirement is restricted to 5 years under the retirement age, starting from 2025, without exceptions and without any penalty. • The qualifying minimum service is reduced from 15 to 10 years (the minimum base benefit correspondingly reduced to 20%), starting from 2025. • The pensionable wage definition was expanded to cover 5 years of the last individual’s wages, adjusted with the rate of inflation (to limit manipulations with pensionable wages pre-retirement). These moderate parametric changes produce some limited impact on the costs of the scheme, while helping stabilize the relative level of pension benefits in the long run. As Figure 5.A indicates, while a long run reduction of about 0.3% of GDP relative to status quo is achieved, expenditures stay on their upward trend (see also Table A5 in Annex II). At the same time, as individuals retire later in life and with longer service, the average pensions remain stable in relative terms (i.e., the benefit is higher than in the baseline case). We also note that the required contribution rate to keep the system balanced in the long run in case of a moderate reform is still very high, around 28.5% of the payroll, compared to 30% in the status quo (Tables A4 and A5 Annex II). A deeper reform produced a more profound fiscal impact but did not get anywhere close to covering the financing gap. Specifically, we have also modeled a set of more significant parametric changes for the civil service sector pensions. In addition to the above changes, they included: (i) further increase in the statutory retirement age to 65 for women and men24, and (ii) gradual reduction in the accrual rate of the pension formula from the current 2% to 1.75%. Results (Figure 5.A) suggest that such measures will certainly make a stronger contribution towards maintaining the pension expenditures around 2% of GDP in the long run. However, that would come at cost of lower pensions; in fact, the replacement rate of an average pension would decline following a very similar trajectory as in the baseline (Figure 5.B). 24Significant increases in the statutory retirement age need to be carefully calibrated against the risks of expansion of the disability and survivorship benefits. 20 Figure 5: Consolidated CS DB pension expenditures and replacement rates, under status quo and reforms (Moderate and Significant Reform Scenarios) A. Pension expenditures as %GDP 3.0% 2.5% 2.0% 1.5% 1.0% Expenditure Status quo Expenditure M Reform Expenditure S Reform B. Replacement rates as %Wages 80.0% 70.0% 60.0% 50.0% 40.0% 30.0% Replacement Rate Status quo Replacement Rate M Reform Replacement Rate S Reform Source: Authors’ calculations Notes: Expenditures include schemes I+II(CS)+IV (DB component only) as well as personal allowance Identifying a practical solution among the options of the parametric changes to close the financing gap completely may be difficult, therefore other solutions may need to be explored. Indeed, closing the combined deficit of all schemes at over 2 % of GDP may be beyond the reach of policy makers, as it would require measures such as drastic increases in contribution rates25 and/or dramatic reductions in the benefit. At the same time, financial self-sufficiency of the scheme does not need to be an ultimate objective of the policy makers in such a reform, as long as any non-contributory financing of the scheme is seen as part of the contract and compensation policy within 25For example, the contribution rate may need to be gradually raised to up to 30% for the civil service (Table A4 and A5 in Annex II) and 40% for security sectors correspondingly, to eliminate the deficit of the DB scheme completely. 21 the public sector. Indeed, the overall compensation package of the public sector employees, including pensions and other benefits, is funded largely through the general taxes. At the same time, certain policy gains could be made through improvements in transparency and flexibility of the scheme, on which we provide further guidance in the section with recommendations. The DC scheme was meant to provide an additional layer of income security, yet it has not lived up to the expectations of the 2005 reform. It operates on accrual basis, offers zero nominal interest, and pays out only face value of the total contributions in a lump-sum. We first looked at the financial balances of the scheme. We modeled it on an accrual basis, using information from the PPA on the average accrued account balances, by age and gender, accumulated so far and with projected due contributions. Notably, the way it operates now, on an accrual basis, it is an attractive proposition for the MOF, as the MOF transfers to the PPA to cover those DC payouts only half of the total amount of the due employer and employee contributions would be. This means that the DC component would have remained largely balanced even if funded only by the employee 3%-contribution. However, it does so at the expense of the members, as from the original design perspective it is certainly a money- losing proposition for them: simply put, their funds sit in no-interest virtual accounts for years or decades. Figure 6: Projected DC contributions and benefit expenditures: SC scheme, % GDP 0.45% 0.40% 0.35% 0.30% 0.25% 0.20% 0.15% 0.10% 0.05% 0.00% Due DC contributions Lump-sum DC benefits Source: Authors’ calculations We also looked at the option of DC balances being annuitized, producing payments for life. Zero nominal interest is assumed, as the current PPA’s practice suggests, and an alternative of interest equaling the inflation (or, zero real interest) and accruing after the year 2022 under the reform scenario. Under such scenario, the DC scheme today would be able to provide a supplementary 4-5% of the replacement rate. After a short expansion in the medium term26, it would go on the declining trend, given suboptimal investment performance and developments in the life expectancy (see Figure 7). 26The scheme is still young (covering only half of the average career), and so individual accounts will continue expanding in size 22 Figure 7: Indicative supplementary replacement rates of DC annuity in CS scheme, % of average wage 7.0% 6.0% 5.0% 4.0% 3.0% 2.0% 1.0% DC Annuity, zero real return DC Annuity, zero nominal return Source: Authors’ calculations Notes: Assumed life-time annuitization In summary, the policy effectiveness of the DC component remains questionable. Its’ practical implementation is not in line with the original design or contributory requirements of the law. The benefit does not provide for an increment to the regular pension and beneficiaries do not take a full advantage of the funds the law mandates to be put aside for retirement. 23 6. Conclusions and Outline of the Proposed Reforms The public sector pension reform in the Palestinian territories has not fully succeeded, and further policy adjustments are needed to make the system sustainable. In what follows, we propose a general roadmap of the reform guided by the above analysis. Further simulations will be required to converge on a specific set of parameters. The reforms should proceed on three parallel tracks: (i) financial restructuring, (ii) parametric adjustments, and (iii) systemic changes (including revisions to the DC benefit). Additional measures to strengthen the institutional and operational capacity of the PPA are outlined in Annex III. It is important to note that any public sector pension reform – its objectives and results – should be evaluated primarily in the context of an overall compensation package in the public sector. I. Financial restructuring calls for rationalizing the institutional and financial responsibilities of the PPA and the MOF towards the pension system to improve its transparency. Bank projections show that pension expenditures will always exceed pension contributions, even with the proposed parametric reforms. If the PA implicitly guarantees all such pension payments, no matter the circumstances in the PPA, it would be logical to propose that the MOF should legally assume the full financial responsibility for all current and future pension obligations, by promptly transferring all due contributions and by financing any deficit of the consolidated system, without differentiating between schemes. In exchange, all the DB contribution arrears that the MOF has accumulated so far, along with interest and penalties, would get annulled and voided (pending decision on the DC component – see below). Among other things, this would result in a reduction of the current explicit public debt27. The MOF would also avoid paying for both the legacy pensions and the new pension rights, as explained above, easing the fiscal burden of the overall pension bill. The PPA would abandon keeping different books for different schemes, bringing all the current and future beneficiaries into a consolidated pool of clients of a unified scheme. The legal responsibility to self-finance schemes I and II should also be lifted from the PPA. A separate policy decision and the guidelines would be required on how best to use the remaining assets associated with the civil service pension rights in schemes I and II. One option would be to maintain those assets as a buffer fund in case any future revenue shortfalls. II. Parametric adjustments will provide for certain improvements in the financial sustainability of the system, alleviating the costs of future financing of pension liabilities. Building on the illustration of the policy simulations above, the following measures of the parametric reform are recommended (we also note which specific measures have been modeled in this analysis): ▪ To keep pace with the increasing life expectancy, the statutory retirement age could be gradually raised to 63, for women and men. This could be done very gradually, by three months annually, over a 12-year period (which we have modeled). ▪ To limit the costs of early retirement as well as to reinforce the objective of the system in providing old age retirement income (rather than a loss of income insurance), early retirement should be restricted in all cases to 5 years under the 27Even if the contribution arrears that the MoF has accumulated so far would be annulled and voided, the fiscal burden of the future pension payments would not disappear. Pension liabilities would remain implicit, and their stock would have to be diligently monitored through mechanisms of regular actuarial valuations of the scheme. 24 retirement age (initially 55), no matter what total service has been accumulated. Retirement below that age would be allowed only in cases of disability. So, a concept of a “minimum retirement age� would need to be introduced. This recommendation has been modeled above under the reform scenario. Additionally, it would be highly advisable to reduce the benefit, at least temporarily, until reaching the statutory retirement age. The reduction could be actuarily fair (age specific) or have a flat reduction rate. Alternatively, in case the DB benefit is split into two components, contributory and non-contributory (see below), only the contributory component could be made available to individuals under the statutory retirement age (without any reduction), while the other component (non- contributory and financed by the MOF Treasury) would only be triggered by reaching the statutory retirement age (see Table 7 as an illustration of the proposed approach). Importantly, this rule should apply no matter whether the individual separated voluntary or due to the employer’s resolution. The key principle should be upheld – the retirement benefit is for old age, not to cover temporary loss of income of a working age individual (for which a separate insurance benefit could be introduced, funded from additional employee contributions, as we suggest below). Finally, individuals that get separated from service at an age below the allowable early retirement would have to wait until reaching that age, possibly seeking opportunities of re-employment in the public sector. Table 7: Proposed new age eligibility rules for old age retirement in the Civil Service Age of separation or Eligibility to benefit retirement Under 55 DC payout (optional)(*) 55 and under 60 DB (contributory), DC payout DB (contributory), DB (non- 60+ contributory), DC payout Source: Authors’ design Notes: (*) For example, as a severance benefit and an early separation insurance ▪ Correspondingly, no DB contributions should be paid back to employees who have not accumulated sufficient minimum service at the point of separation. If that happens at an age below the minimum retirement age, such individual should wait until the regular retirement age (or a minimum retirement age) before she or he could apply for a DB benefit, exploring options of re-employment in the interim with the public or non-government sector. Possibly, an option of an early separation insurance (a transformed DC component) should be available to such individuals. Again, pension rights, significant or not, should not be accessed just at any age (except in the cases of disability or death). ▪ To allow for a greater number of the civil and security personnel to access their regular DB pensions and to eliminate one off contribution pay-backs, the qualifying minimum service (vesting) could be reduced from 15 to 10 years (with a corresponding reduction in the minimum base benefit to 20% – or lower), which we have modeled above. This measure will have implications of reduced costs but also strengthened social outcomes. This will eliminate the need for paying out lump sums to those with service of 10 to under 15 years. Instead, it will produce a pension benefit of around NIS 700 or higher for life, payable by the PPA. In turn, it will 25 eliminate the need for the life-long “basic pension� assistance of NIS 700 financed by MOF. So, instead of receiving effectively a double-compensation of a lump sum and a regular basic pension for life, such individuals will be receiving only a regular pension, with the financing responsibility shifting from the MOF to the PPA. ▪ We also recommend making the “basic pension� (NIS 700) benefit, for those who did not earn sufficient rights for regular pension in the public sector, vary in amount depending on the duration of service. We did not model this program in our analysis. ▪ Benefit indexation policy should consider and carefully balance fiscal and social implications, by instituting an index that combines wage growth and inflation (we modeled such an approach under both status quo and reform scenarios)28. The add- hoc benefit increases should be replaced by a rule-based approach that defines both the rate of periodic adjustments and their timing. ▪ It would be advisable to explore options of further (gradual) reduction in the accrual rate of pension benefits (e.g., from the current 2% down to around 1.75%), as a way to narrow the financing gap further; for example, such reduction could apply to the minimum basic amount of pension. This could be explored in greater details in a follow up analysis, subject to agreeing on a pragmatic set of assumptions with the PPA. ▪ The definition of the pensionable wage at retirement could be expanded to cover at least 5 years of wages, valorized at a minimum with the rate of inflation (which we have modeled in our reform case). ▪ The reform program should include adjustments to survivorship and disability benefits (which has been largely kept out of scope of this analysis). ▪ Unless the systemic changes that are proposed below are acceptable, eligibility to the family and personal allowance of NIS 300 could be linked to the main pension amount, so that this allowance is not added to the total amount above a certain level of pension. Alternatively, the age eligibility to that allowance could be increased by 5 years or more, so that it becomes a seniority top-up to the pension. Both options will result in cost savings. III. Systemic changes could focus on rationalizing the benefit package. It is very likely that most pragmatic options of the parametric reforms will not make the system financially self-sustainable. So, if the system remains contributory, as is, it will always operate with a deficit and will be subsidized by the MOF Treasury. Alternatively, the DB scheme could be made partially non-contributory. This would be a better option in our view. Under such a scenario, the current DB benefit would be split into two components: one component would be earnings related and explicitly linked to the contribution in a way that is financially balanced, while another component would be a non-contributory base pension carefully calibrated and periodically adjusted, building on the concept and precedent of the personal allowance (NIS 300) payable to all retirees. Such an approach would allow for a more transparent budgeting process, would more explicitly link to the wage bill policy, and would give more flexibility to the government in responding to various policy challenges. For example, while the contributory component would be linked to earnings and its adjustments ideally would 28In the proposed context of the two components (contributory and non-contributory) of the DB benefit, an option could be to have the contributory component wage indexed, while the non-contributory component could be inflation indexed. This gives a lot of flexibility in achieving various objectives of the benefit indexation policy. 26 not be subject to discretion of the government, the value of the base part of the benefit could be set periodically as part of the annual budgeting and wage bill approval process. 29 Similarly, eligibility to such contributory and non-contributory elements could be defined in different terms (e.g., in different ages). For instance, an individual may gain access to the contributory part at the point of early retirement, while access to the non-contributory part would be legally restricted to the statutory retirement age. Calibrating the exact split into the contributory and non-contributory parts would require additional rounds of simulations to ensure that this measure is fiscally neutral. The preliminary analysis of the expenditure gap in the long term (Figures 2 and 3) suggests that share of the flat non-contributory component could be up to a half of the current benefit (or around NIS 1200 for the civil service). In case such a non- contributory base pension is introduced, it would be restricted only to the retiring public sector employees, and, if desired, its value could be differentiated according to age and/or according to length of service – both such design elements have international precedents and serve certain policy objectives. IV. The DC benefit is an important component of the overall benefit package, so the government should carefully consider available options. The status quo option effectively converted the DC component into an end-of-service indemnity scheme with a lump sum benefit. While such construct resulted in a reduced overall pension benefit, it could be totally acceptable as a particular form of a social contract and compensation package, and as such it goes outside of the pension policy discussion. The only recommendation in such case would be to consider lifting the responsibility of the government to make the matching 3% contribution to that scheme – and possibly making the employee contributions to such scheme voluntary. Overall, this would perhaps be the least costly solution to the government. In line with that concept, and in reference to the above discussion about the early retirement, the DC component could be de-jure transformed into end-of-service indemnity scheme/ early separation insurance. So, with a 3% contribution, an employee would pre-fund the costs associated with the risks of his/her unexpected separation from service before reaching the retirement age (which is a different risk from the inability to work due to old age or disability). This could be a DC or a DB benefit. If such individual risk did not materialize, depending on the funding mechanism, the benefit may be payable at the point of retirement (see Table 7). If, however, there is an interest to revive the full potential of this component to contribute to the retirement income, it could be made possible in several ways, as follows: ▪ If there is an intent to retain the DC component and make actual financial contributions to the DC scheme in the future, all past accrued liabilities could be converted to formal recognition bonds30. So, the PPA would bill the government each time someone retires, similar to the current process, but in a more structurally and financially organized way. Such bonds would earn some notional interest, agreed by the MOF. This would, however, cement the current financial obligation (arrears) of the MOF towards that scheme (confirming corresponding pension part in the formal public debt of the Palestinian government) at an amount of close to 29 Indeed, different indexation policies could apply to contributory (insurance) and non-contributory (income guarantee) components of the DB benefit package. 30 A special financial instrument, typically non-tradable, to recognize the accrued pension rights – both historic and incremental future rights – and keep explicit record of the future pension liabilities. 27 NIS 4 billion (value of the accrued DC balances to date). Under this scenario, as a cost saver, it is recommended to not pay a lump sum from the DC component to those who apply for early retirement. Their DC savings (even if notional) would be considered as financing part of the costs of early retirement. ▪ If DC contributions, as a matter of policy, ever materialize in the future, the associated financial assets should be segregated from the DB assets and possibly follow a different investment policy. ▪ If there is no intent or fiscal space to make financial contributions to the DC accounts, or if there is an interest to move away from the financial accounts with explicit financial balances and correspondingly explicit public debt resulting from the accumulated arrears, the DC component could be converted to a Notional DC (NDC) scheme. This changes the financing and accounting nature of the scheme. The explicit public debt pertaining to the pension scheme would be erased and all the liabilities would become notional. Yet, they would be explicitly recorded in the individual accounts. No prefunding would be required. This component could then also be merged with the contributory DB component – both converted to a NDC plan. ▪ Under either of the above arrangements, there is no good reason why payouts from such accounts (DC or NDC), if reconfirmed as pensions from the policy intent perspective, could not be easily organized in a form of regular payments, promoting consumption smoothing throughout the retirement. Certainly, for the PPA with its advanced systems of electronic recordkeeping, it should not be a significant additional administrative burden. In fact, one simple option of such a conversion could be implemented through the existing mechanisms offered by the PPA, whereby the lump-sum DC payout could be used to purchase additional service in the DB component, leading to the higher DB benefit for life. To strengthen capacity of the budget planning and actuarial projections, the PPA may want to establish and further build a permanent capacity of such an analysis in-house, perhaps as part of a team of data analytics experts. In fact, considering the advanced state of recordkeeping practices at the PPA, it is recommended to adopt the tools of microsimulations to have the analysis and all the calculations very nuanced. Any reform of that type and scope should be accompanied by a strong communication campaign. As the administrative assessment of PPA indicates (see Annex III), the public communication capacity of the PPA needs to be strengthened. Among other activities, the agency may need to run opinion poll of the members of the system about their perception of the PPA, the system it operates, and the reform plans. The PPA should use that information as a basis to develop a comprehensive communication plan and campaign. It should also establish a hot line to address reform related inquiries of the civil service and the security personnel. Finally, expanding the coverage of the existing system to the new categories of the population should not be a strategic priority for the PPA at the moment, unless the fiscal sustainability of the scheme is addressed. For sure, it would be a mistake to pursue widening of the system’s coverage, aiming to expand the PPA’s revenue base, as it would only expand the fiscal burden of the unsustainable pension liabilities and perpetuate the financial challenges of the government. The policy deliberations around the coverage expansion should be taken in a broader context of the work on the social security law. 28 29 Annex I. History of the PA’s Policy Initiatives and Reforms in Pensions31 The Unified Pension Law (UPL) was approved in 2005 and all government employees, including civil and security services who were less than 46 years of age in 2006, were enrolled into a new scheme (Scheme II). It was designed to have two components: a defined-benefit (DB) and a defined contribution (DC). The old schemes and their beneficiaries were grandfathered. In 2008, the PA adopted the Palestinian Reform and Development Plan (PRDP) 2008- 2011 in which the PA committed, among other things, to the preparation of a pension reform action plan covering administrative and parametric changes to drive the pension system on the path of fiscal sustainability. In 2009, the MoF issued a decree requiring all civil servants to remain in each grade for a minimum of five years before being eligible for promotion and in the case of security services personnel, a minimum of four years. This would limit manipulations that could affect the pension rights. All pensions as of January 1, 2009, were calculated using the average of three years of wages to determine the pensionable salary. As such artificial end-of-career salary increases, which had a negative effect on pension liabilities, were brought under control. In July 2010, the Council of Ministers approved an action plan aimed at reducing the heavy burden that the pension system imposes on the budget. The plan, however, only consisted of small measures that were likely to produce desirable effects in the short- term but were not enough to ensure long-term sustainability. The plan included a set of administrative reforms to strengthen the institutional capacity of the PPA in addition to a number of parametric changes aimed at improving the financial performance and fairness of the existing pension schemes. Proposed parametric reforms covered by the plan included: i) increasing the mandatory retirement age from 60 to 62; ii) early retirement would only be allowed if pensions are reduced according to an actuarially fair coefficient; iii) lump sums for those who contributed to Scheme I for more than 28 years would be calculated using the same average salary as the one used to calculate regular pensions; iv) buying additional years of contributions would only apply with a limit of up to 3 years at the time of retirement for Schemes I, III, IV; v) new entrants to Scheme II would be allowed to purchase voluntary contributions of up to 5 years in advance within the first year of joining the system; vi) pensions would be indexed to the CPI published by the Palestinian Central Bureau of Statistics; and vii) the DC component of Scheme II would be modified to introduce notional accounts instead of funded accounts as currently stipulated by the law. Progress made in the implementation of the above reform action plan was uneven. Most of the administrative reforms were implemented, leading to an overall improvement in the organizational aspects of the PPA. However, none of the parametric reforms were carried out. In recent years, the most important changes to the pension rules were introduced by the Presidential Decree No. 29 of 2018 «Amending the Public Retirement Law No. 7 of 31 The first part of this summary is based on World Bank, PER, 2016. 30 2005 and its Amendments». These changes were probably intended to harmonize the conditions of the old and new pension schemes, as employees began to retire, whose careers were partly under one of the old pension schemes, partly new scheme II. As the body of the retirees of the new scheme was growing, it also became clear that benefits under the Scheme II were lower than under the legacy schemes, and this caused discontent. The changes instituted by the Decree, therefore, were as follows: • Service in the old schemes was equated to the participation in the new scheme regarding the entitlement to a pension; • Investment revenues of the PРA, base salary pension and all types of contributions were exempted from income tax; • Frequency of conducting actuarial analytical studies to redefine contribution and benefit rates increased from three to five years; • For years of service earned in the old schemes, it is allowed to use the accrual rate that was in these old schemes (i.e. 2.5%, not 2% as in the new scheme II); • Maximum replacement rate for calculating the pension amount in the DB formula increased from 70% to 80%; • Payment of the full amount accumulated on the personal account of the DC pillar was allowed when the employee retires for any reason, including due to early resignation (previously it was possible to receive such payment only upon reaching the mandatory retirement age); • Choice of the payments type from the DC pillar was limited to the lump sum payment of the entire amount or a monthly payment an agreed schedule (previously there were also options for a monthly lifetime payment and a combination of a lump sum payment with a monthly payment); • It was forbidden to combine two salaries to calculate a pension. For an employee who has two pensionable salaries, only the higher one is taken into the formula to calculate the pension; • It was clarified that all pensions are calculated on the basis of the average salary for the last three years; • The MOF Treasury was made to finance the financial obligations of pension under DB and DC schemes in the event that funds were unavailable to pay beneficiaries and their heirs for any reason. Thus, the latest changes in the pension legislation rather increased the financial obligations of the public budget. 31 Table A1: Consolidated parameters of pension schemes Source: Updated on the basis of World Bank, PER, 2016 32 Annex II. Modeling Methodology Overview Part A1. Assumptions 1.1 General framework The financial projections were produced with the World Bank’s PROST (Pension Reform Options Simulation Toolkit) model. The analysis covered both the Scheme II (New Scheme) and Schemes I, III, and IV (Legacy (Old) Schemes). The civil and security sectors were modeled separately, with the consolidation of new and old schemes in each sector as follows: Civil Security Financial responsibility PPA MoF MoF Pension schemes II (Civil) + I IV II (Security) + III All contributors and new pensioners were modeled under the New Scheme, while participants in the old schemes were gradually being phased out of the total stock of pensioners. We did a detailed modeling of both the DB and DC components, including the family and personal allowance (NIS 300) as part of the total DB pension. The simulation period runs from 2021 (the base year) to 2080 (the ending year) – approximately a lifetime of one generation (working age and retirement age), which is viewed as an adequate time span for pension system modeling. Information for the base year was kindly provided by the PPA. We present the key data and assumptions used in the modeling exercise as well as some projection results. 1.2 Macroeconomic indicators Table A1 below presents key economic indicators. Real GDP growth and inflation rate were provided by the World Bank macroeconomic team, while other indicators were simulated by the authors using PROST. In the medium to long run, the real GDP growth is assumed at 3% (is slightly higher than the population growth rate), and the inflation rate was set at 2%. Wage growth rates were calibrated based on our core assumption that public sector payroll as share in GDP remains stable (about 10.5% of the civil and security sectors combined). Table A1. Projected macroeconomic context 2021 2022 2023 2024 2025 2030 2040 2050 2060 2070 2080 Macroeconomic Indicators Nominal GDP, NIS bln 58.5 63.1 66.1 68.9 71.8 91.9 150.5 246.6 404.0 661.9 1,084.3 Real GDP Growth 7.1% 4.0% 3.0% 3.0% 3.0% 3.0% 3.0% 3.0% 3.0% 3.0% 3.0% Inflation Rate 1.2% 3.7% 1.6% 1.2% 1.2% 2.0% 2.0% 2.0% 2.0% 2.0% 2.0% Nominal Wage Bill, NIS bln Civil Sector 3.8 4.1 4.3 4.5 4.7 6.0 9.8 16.1 26.3 43.1 70.5 Security Sector 2.3 2.5 2.7 2.8 3.0 3.9 6.0 9.9 16.1 26.4 43.2 Total 6.1 6.6 7 7.3 7.7 9.9 15.8 26 42.4 69.5 113.7 33 Nominal Wage Bill as % of Nominal GDP Civil Sector 6.5% 6.5% 6.5% 6.5% 6.5% 6.5% 6.5% 6.5% 6.5% 6.5% 6.5% Security Sector 3.9% 4.0% 4.0% 4.1% 4.1% 4.3% 4.0% 4.0% 4.0% 4.0% 4.0% Total 10.4% 10.5% 10.5% 10.6% 10.6% 10.8% 10.5% 10.5% 10.5% 10.5% 10.5% Source: WB macro indicators and Authors’ PROST Simulations 1.3 Demography The UN World Population Prospects 2022 is the source of our demographic data and assumptions. Population projections for the Palestinian Territories are in line with the long-term international trends and demographic transition theory. The total fertility rate is assumed to decrease, while the mortality rates are also assumed to decline gradually over the whole period (Table A2). Table A2. Projected Life Expectancy and Total Fertility Rate (for the General Population) 2021 2022 2023 2024 2025 2030 2040 2050 2060 2070 2080 Total Fertility Rate 3.4 3.4 3.3 3.3 3.3 3.0 2.6 2.4 2.2 2.0 2.0 Life Expectancy: Male At Birth 70.8 71.5 72.2 73.0 73.9 74.7 76.4 78.2 80.0 81.8 83.3 At Age 60 16.1 16.6 17.2 17.8 18.4 18.9 20.0 21.2 22.5 23.9 25.1 At Age 65 12.7 13.1 13.6 14.1 14.7 15.1 16.0 17.1 18.3 19.6 20.7 Life Expectancy: Female At Birth 75.7 76.3 76.9 77.5 78.1 79.0 80.5 82.0 83.3 84.5 85.7 At Age 60 19.7 20.2 20.6 21.1 21.7 22.2 23.2 24.3 25.3 26.3 27.2 At Age 65 15.7 16.1 16.5 17.0 17.5 17.9 18.9 19.9 20.8 21.7 22.6 Source: UN World Population Prospects 2022 (Online Edition) and Authors’ PROST Simulations The decreasing fertility and increasing longevity are the two driving forces for population aging. Given the fact that both working age population (15-59) and the old age population (60+) are both increasing, the old-age dependency ratio is relatively constant for a few years, however, from around 2040 the old-age dependency ratio starts considerably increasing. By 2050, the old-age dependency ratio is projected to be more than 18% and will double again by 2080 (Table A3). Table A3. Summary of Population Projections (for the General Population) 2021 2022 2023 2024 2025 2030 2040 2050 2060 2070 2080 Total Population, thousands persons Total 5,130 5,253 5,378 5,505 5,634 6,301 7,684 9,051 10,307 11,420 12,332 Age 0 - 14 2,009 2,033 2,056 2,078 2,097 2,172 2,347 2,474 2,482 2,463 2,422 Age 15 - 59 2,844 2,930 3,019 3,109 3,205 3,709 4,691 5,563 6,295 6,839 7,112 Age 60 and over 277 289 302 317 332 420 646 1,014 1,530 2,118 2,799 Age Structure of Population Total 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% Age 0 - 14 39.2% 38.7% 38.2% 37.8% 37.2% 34.5% 30.5% 27.3% 24.1% 21.6% 19.6% Age 15 – 59 55.4% 55.8% 56.1% 56.5% 56.9% 58.9% 61.0% 61.5% 61.1% 59.9% 57.7% Age 60 and over 5.4% 5.5% 5.6% 5.8% 5.9% 6.7% 8.4% 11.2% 14.8% 18.5% 22.7% Dependency Rates Age 0-14 / Age 15-60 70.6 69.4 68.1 66.8 65.4 58.5 50.0 44.5 39.4 36.0 34.1 60 + / Age 15-59 9.7 9.9 10.0 10.2 10.4 11.3 13.8 18.2 24.3 31.0 39.4 Support Ratio: 10.3 10.1 10.0 9.8 9.6 8.8 7.3 5.5 4.1 3.2 2.5 Age 15-59 / 60 + 34 Source: UN World Population Prospects 2022 (Online Edition) and Authors’ PROST Simulations Population pyramids (Figure A1) illustrate the projected aging trend. Based on the above assumptions, the population is projected to grow from 5.1 million in 2020 to about 12.3 million by 2080. Figure A1. Projected Population Pyramids 2021 2080 99 99 92 92 85 85 78 78 71 71 64 64 57 57 50 50 43 43 36 36 29 29 22 22 15 15 8 8 1 1 94 74 54 34 14 6 26 46 66 86 94 74 54 34 14 6 26 46 66 86 Source: UN World Population Prospects 2022 (Online Edition) and Authors’ PROST Simulations 1.4 Policy, parametric, and procedural assumptions • The number of contributors in the legacy schemes (I, III, IV) was in insignificant in 2021 and was ignored. • The baseline scenario assumes that the existing parameters and rules of the pension scheme II remain unchanged throughout the forecast period. • The number of employees in the public sector (they are also contributors to the II scheme) is growing at the same rate as the total population, so it remains stable in relative terms. Subsequent rounds of simulations will need to explore option where the size of the public sector remains stable in absolute terms or is expanding less rapidly. • The structure of new contributors by gender and age remains unchanged throughout the projection period. • Contribution revenues were projected on the accrual basis. • Given the long run horizon of the projections, all DB pensions are indexed to 100% of inflation until 2030 and 50%/50% inflation/wages correspondingly, thereafter. For the same reason, we assumed that value of the personal allowance (currently, NIS 300 monthly) would be automatically indexed with inflation and be added uniformly to all the new pensions. 35 • We did not include calculations of the NIS 700 “basic pension� allowance as a compensation to those who did not earn rights for pension. • As a prevailing practice, all employees separate from service by decree of the employer, hence no reduction in benefit due to early retirement simulated. • Considering average service of 27 years at retirement and 15 years into the reform process from 2006, we assumed that in 2023, the average accrual factor (which combines careers from before and after reform) is 2.25, and it would gradually transition to 2.0 by the year 2036. • Since the detailed age/gender data was not available for Scheme IV, we utilized and extrapolated the distributions of Scheme I to approximate the composition of Scheme IV. • We estimated a NIS 300 million among the stock of the PPA pension reserves attributable to the Civil Service DB schemes. We did not attribute any reserves to the security DB scheme. • For simplicity, all PPA administrative expenditures were attributed to and included under the CS scheme for calculation purposes. • DC scheme calculations: o The distribution of accrued values in individual DC account balances by age and gender (Separately for civil service and for security) was provided by the PPA o All new contributions are subject to the full statutory rate of 3%+3% o Accounts earn zero interest (as per current policy) o Payout takes various forms. For the purpose of assessment of the income replacement capacity, we assumed full annuitization (i.e., payments for life). 36 1.4 Summary of financial projections Table A4. Summary of Projections: Civil Service Schemes I and II, DB only, Baseline I D 2021 2022 2023 2025 2030 2035 2040 2050 2060 2070 2080 PAYG Total Revenue 635 697 723 780 973 1,247 1,595 2,632 4,298 7,005 11,605 Employer and Employee Contributions 607 654 685 744 952 1,220 1,561 2,575 4,205 6,853 11,354 Pensioner Contributions Government Contributions & Transfers Other Revenue 13 15 15 17 21 27 35 57 93 152 252 Investment Returns 15 28 24 20 PAYG Total Expenditure 566 622 712 877 1,403 2,141 3,090 5,179 8,390 15,995 29,294 Pension Payments 551 606 695 859 1,378 2,107 3,044 5,104 8,268 15,784 28,931 Other Payments 3 3 3 4 7 11 15 26 41 79 145 Administrative Costs 12 13 13 14 18 24 30 50 81 132 219 PAYG Current Balance 69 75 12 (97) (430) (894) (1,495) (2,547) (4,092) (8,990) (17,689) PAYG Reserve (Explicit PAYG Debt) 369 444 456 316 PAYG Implicit Debt 33,342.7 36,440.9 38,996.8 44,094.2 61,347.7 83,719.1 112,226.1 196,840.1 352,719.3 611,881.2 1,000,892.8 Total PAYG Debt 32,973.4 35,996.6 38,540.7 43,778.6 61,347.7 83,719.1 112,226.1 196,840.1 352,719.3 611,881.2 1,000,892.8 As % of GDP PAYG Total Revenue 1.1% 1.1% 1.1% 1.1% 1.1% 1.1% 1.1% 1.1% 1.1% 1.1% 1.1% PAYG Total Expenditure 1.0% 1.0% 1.1% 1.2% 1.5% 1.8% 2.1% 2.1% 2.1% 2.4% 2.7% PAYG Current Balance .1% .1% .0% (.1%) (.5%) (.8%) (1.0%) (1.0%) (1.0%) (1.4%) (1.6%) PAYG Reserve (Explicit PAYG Debt) .6% .7% .7% .4% Total PAYG Debt (Implicit & Explicit) 56.3% 57.0% 58.3% 61.0% 66.8% 71.2% 74.5% 79.8% 87.3% 92.4% 92.3% Contribution Rate from Employees and Employers Average Collected 16.0% 16.0% 16.0% 16.0% 16.0% 16.0% 16.0% 16.0% 16.0% 16.0% 16.0% Required to Balance Fund from Base to Current Year 14.2% 14.6% 15.1% 16.3% 18.8% 21.1% 23.2% 26.1% 27.2% 28.5% 30.0% Table A5. Summary of Projections: Civil Service Schemes I and II, DB only, Moderate Reform I D 2021 2022 2023 2025 2030 2035 2040 2050 2060 2070 2080 PAYG Total Revenue 635 697 723 780 965 1,202 1,498 2,513 4,061 6,609 10,850 Employer and Employee Contributions 607 654 685 744 944 1,176 1,465 2,458 3,973 6,466 10,615 Pensioner Contributions Government Contributions & Transfers Other Revenue 13 15 15 17 21 26 33 55 88 143 236 Investment Returns 15 28 24 20 PAYG Total Expenditure 566 622 712 881 1,353 1,873 2,601 4,941 7,505 14,157 26,655 Pension Payments 551 606 695 862 1,328 1,841 2,560 4,869 7,392 13,963 26,318 Other Payments 3 3 3 4 7 9 13 24 37 70 132 Administrative Costs 12 13 13 14 18 23 28 47 77 125 205 PAYG Current Balance 69 75 12 (100) (388) (671) (1,103) (2,428) (3,444) (7,548) (15,804) PAYG Reserve (Explicit PAYG Debt) 369 444 456 311 PAYG Implicit Debt 31,610.4 34,488.3 36,819.2 41,408.1 56,611.8 73,445.8 95,100.1 176,196.0 305,447.1 551,617.0 914,765.0 Total PAYG Debt 31,241.1 34,044.1 36,363.5 41,096.9 56,611.8 73,445.8 95,100.1 176,196.0 305,447.1 551,617.0 914,765.0 As % of GDP PAYG Total Revenue 1.1% 1.1% 1.1% 1.1% 1.1% 1.0% 1.0% 1.0% 1.0% 1.0% 1.0% PAYG Total Expenditure 1.0% 1.0% 1.1% 1.2% 1.5% 1.6% 1.7% 2.0% 1.9% 2.1% 2.5% PAYG Current Balance .1% .1% .0% (.1%) (.4%) (.6%) (.7%) (1.0%) (.9%) (1.1%) (1.5%) PAYG Reserve (Explicit PAYG Debt) .6% .7% .7% .4% Total PAYG Debt (Implicit & Explicit) 53.4% 53.9% 55.0% 57.2% 61.6% 62.4% 63.2% 71.4% 75.6% 83.3% 84.4% Contribution Rate from Employees and Employers Average Collected 16.0% 16.0% 16.0% 16.0% 16.0% 16.0% 16.0% 16.0% 16.0% 16.0% 16.0% Required to Balance Fund from Base to Current Year 14.2% 14.6% 15.1% 16.3% 18.7% 20.5% 21.9% 24.7% 25.9% 27.1% 28.5% Source: Authors’ calculations Notes: “PAYG� refers to the DB scheme. Schemes I and II (CS) only. 37 Part A2. PROST Model Overview PROST is a computer-based pension model developed to simulate the behavior of pension systems and assess their financial sustainability under various economic and demographic hypotheses during a lengthy period of time. This model may be adapted to a wide range of various pension systems in different countries; it also allows modeling various options of pension system reform. The model consists of a Work Book to enter input data and five Output Modules. For input data, a user would enter a number of demographic and economic parameters and parameters of the pension system of a given country, as well as hypotheses on future changes. This information is recorded in the input data file that includes six Worksheets: General Economic variables (GDP and wage growth, inflation, interest rate), parameters of the pensions system that do not depend on age (pension fund balance and pension expenditures in the baseline year, retirement age, contribution rate, pension indexation rules, etc.) and some demographic variables. Population Population by age and sex in the baseline year, fertility and mortality rates and migration, disaggregated by age and sex. Labor force Labor force participation and unemployment rates disaggregated by age and sex as well as distribution of wages and pension by age and sex. Pensions Information on the number of contributors, pensioners, coverage ratios and share of retirees, average contributory period at retirement and income replacement rates for new pensioners, disaggregated by age and sex. Profile Information on typical individuals such as sex, contributory period, individual wage, life expectancy, etc. Reform Parameters required to simulate systemic reforms (any combination of legacy schemes of PAYG type, FF and NDC), including a strategy for switching to a new system, form of paying pension rights accrued in the old system, contribution rate, rules of annuitization and payments from FF and NDC, income replacement rates/formula for calculating new pensions in DB, indexation, etc. The general calculation scheme can be presented in a most simplified form as shown below. PROST tracks down sex and age cohorts over time and estimates the population projections that, together with labor market hypotheses, are used to predict the number of contributors and pensioners. In their turn, these predictions help calculate the flows of revenues and expenditures. Then the model calculates the balance of cash flows and implicit pension debt. Also, the model projects the equilibrium rate of contributions and admissible income replacement rates that ensure the balance of the pension fund in each year of the projection period. Finally, PROST calculates parameters with regard to individuals, i.e., how much contributions a given individual would pay to the system and what he/she would receive from 38 PAYG or from a multi-pillar system. This allows analyzing both intra- and intergenerational redistributional processes. Based on the features of the pension system and available data, a user may select a method to calculate some variables. In particular, the number of contributors and pensioners may be calculated through using either a Margin Method or a Flows Method. The Margin Method first calculates the total number of contributors/pensioners (Margin) for each year, and then the inflow (new contributors and pensioners) is calculated as changes in the margin: Inflow(a,t,g) = Margin(a,t,g) – Margin(a-1,t-1,g) + Outflow(a,t,g), The Flows Method first calculates the inflow, and then the margin is calculated as the margin of the previous year in each age/sex group adjusted for net inflow (inflow – outflow): Margin(a,t,g) = Margin(a-1,t-1,g) - Outflow(a,t,g) + Inflow(a,t,g), where: a = age, t = year, and g = gender. Since РROST tracks down the contributory period accrued by each sex and age cohort, the calculated number of new pensioners, regardless of what method was used, is then adjusted in a way so that the contributory period accrued by a cohort is equal to the total contributory period declared by the cohort for calculating pensions at retirement. After the number of new pensioners is adjusted, the margin of pensioners is recalculated using the Flows Method. A user may also select the way of simulating pensions of new pensioners, i.e., either using the pension calculation formula or using the income replacement rates for each age and sex. General Calculation Scheme As mentioned earlier, the PROST outputs are grouped into five Output Modules. Each module contains a number of Excel Worksheets and a summary graph for key indicators: Demographic projections Projections of the population and demographic pyramid, longevity tables, changes in life expectancy, population dependency ratios, etc. Pension system demographics Projections of the labor force number and employment rate, the number of contributors and pensioners, 39 demographic structure of the pension system, and system dependency rates. Finances of one-pillar PAYG Macroeconomic projections, projections of wage, pensions for all or new pensioners, pension system’s revenues and expenditures, contribution rate and average income replacement rate needed for a balanced system, and implicit pension debt. Finances of a Multi-Pillar System Projections of pensions for all or new pensioners in each of the three pillars (PAYG, NDC and FF), PAYG and FF revenues and expenditures, implicit pension debt in PAYG, reform and reform outcomes (compare projections of the level of pensions and finances in the one-pillar PAYG scenarios and multi-pillar system scenarios). Individual Accounts Paid contributions and received pensions during the entire life and summary analytical indicators of participating in the system for six types of individuals specified in the input data file table, Section Profile, for PAYG (in the form set forth by law as well as in options with an adjusted contribution rate and adjusted level of pensions) and for a multi-pillar system (for those who switched to multi-pillar system and for those who stayed in PAYG DB). 40 Annex III. Administrative Diagnostics of the PPE This assessment was produced using the Social Insurance Administrative Diagnostic (SIAD), a tool developed by the World Bank. There are currently eight modules within the tool, covering the following areas of administration, that were used in the assessment: ▪ Internal Governance ▪ Investment and Management of Funds ▪ ICT Management and Data Policies ▪ Identification and Registration ▪ Contribution Collection ▪ Benefits Management ▪ Grievances – Appeals and Complaints ▪ Communications It should be noted that in some instances it would only be possible to more fully flesh out the various topics and findings by undertaking deeper and more extensive analysis. This level of effort was outside the scope of this assessment. A1. Internal Governance The Public Sector Pension Scheme in the Palestinian Territories (West Bank and Gaza) is administered by the Palestinian Pension Authority (PPA). The PPA is an independent public legal entity reporting directly to the President of Palestinian Authority. The PPA was established in 2005 under the Palestinian Pension Law 7 of 2005 (hereinafter ‘the Law’) and it took over responsibility for pensions system management from the General Pensions and Insurance Corporation (GPIC). 1.1 Board of Directors The PPA is governed by a Board of Directors (BoD) and a Head of Agency/General Director (CEO). The BoD comprises of nine persons, three of whom are nominated by the Council of Ministers and the remaining six are nominated (one per institution) by PPA, General Personnel Office, Administrative Head of Police and Security Forces, Local Government/Municipalities, Association of Retirees, and Ministry of Finance. The nominations are submitted to the President who has the final authority to approve the Board. The President also appoints the Chairman and Deputy Chairman of the Board and decides on remuneration level for the Board. The Law prescribes the attributes that BoD members must have, and these act as guidance in their selection. Members of the BoD serve for four years, with one period of renewal permitted. The BoD nominates the General Director and submits the nomination to the President for final approval. The General Director is ex-officio a voting member of the BoD. The BoD has a wide range of responsibilities, as prescribed in the Law (Articles 40 through 55, and Articles 61 through 81). These are typical of the strategic governance 41 role of a Board. It is noted that a major responsibility of the BoD relates to the investment function – policy, asset classes, risk, audit etc. The BoD has also two permanent sub-Committees (Investment, Audit) and establishes other sub-Committees from time to time as required. The BoD also contracts in expertise for specialist tasks, such as developing the PPA Strategic Plan, undertaking Actuarial Studies, and undertaking Investment Management. The PPA has a formal code of conduct that must be adhered to by BoD members, and it also requires disclosure of any conflict of interest by BoD members. 1.2 Organisational Structure The General Director is supported by four Assistant General-Directors (AGD), and each AGD manages a group of Departments. The Organisation Chart below shows a very detailed and comprehensive organisational architecture. 42 Chairman of the Council Director of the Chairman's Office (Board of Directors) Audit Committee Investment Committee Preserver External Specialized Committees Auditor Investment Manager Head of Agency Financial Control Department (Minister's Follow-up and Coordination degree) Department Administrative Control and Internal Control and Total Total Quality Department Office of the Chairman Quality Department of the Commission Public Relations and Department of Legal Studies Media Department and Consultations Legal Affairs Unit Technical Judicial Follow-up Committees Department Advisers Assistant to the President Assistant Chairman for Assistant Chairman for Assistant Chairman of the of the Authority for Subscriptions and Public Technical Affairs Authority for Financial Support Services Affairs Services Affairs and Investment Department of Settlement and Settlement and IT Management Administrative Affairs, Public Services Manage Pensions Department Pensions Department Investment Finance Studies, and Planning for Southern for the Northern Management Department Subscriptions Governorates Governorates Department Department of Studies and Operators Services Registration and Settlement Investment Collections Planning Department Matches Settlement Department Systems and Department (Department Department Department Study and Department Government Sector Government Sector Programming Manager) (Civil and Security Follow-up (Civil and Security Risk and Actuarial Offices & Branches Department of Forces Non-Governmental) Forces Non-Governmental) Department Payments Studies Department Purchase and Sector Settlement Sector Settlement Department Computerized Coordinator Addition of Service Service (Department Contribution Specific Systems Department Manager) Accounts Periods Pension Department Pension Department Shareholding Department of Human Resources Government Sector Government Sector Investments Department and Personnel Specific Contribution Department of (Civil and Security (Civil and Security Database and Services Service Counting and Forces) Forces) Department Financial Audit Reporting (Department Processing of Loans and Department Maintenance and Non-governmental Non-governmental Department Manager) Contribution sector pensions sector pensions Replacement General Services Periods Department department department Department Department of Reports and Infrastructure and Archiving and Technical Support Supplies and Documents Director of Registration and Registration and Department Maintenance Warehouses Contributions and Audit Department Audit Department Department Department (Department Reports Follow-up Manager) Department 1.3 Procedures Management The PPA has a broad range of formal policies and procedures that guide its work. These cover budget formulation and management, investment management, responsibility delegation policy from Board to senior management, human resource policy, procurement and outsourcing, and operational risk management. The PPA has a set of written Standard Operating Procedures (SOPs) that govern each key work function and business processes. These SOPs are routinely reviewed and updated, with the last update being undertaken in 2020. There is a formal process to encourage and review PPA staff suggestions to amend and improve the SOPs and business workflow procedures. 1.4 Budget and Finances The PPA administration budget is approved annually by the BoD. The Law (Article 5) permits the PPA to be funded from a variety of sources: deduction from contributions collected, payments to PPA from other bodies contracted to pay PPA benefits, a (unspecified) proportion of investment returns, and other (unspecified) sources related to PPA (undefined) activities. No more than 2% of contributions payable can be deducted to fund the PPA administration budget. Any amount exceeding 2% requires Parliamentary approval. The amount allocated to PPA in 2021 was 0.778% of the contributions. 1.5 Audit – Internal and External The PPA’s BoD has a full-time Audit sub-Committee, whose responsibility is to oversee PPA’s financial affairs, investment activities, external audit, and internal audit. This work also includes reviewing the internal and external controls of the Custodian (financial institution that holds PPA’s funds in safe-keeping) and Investment Managers contracted by the PPA. The PPA also has a permanent Internal Audit Unit, with 2 staff, and which reports directly to the General Director. This Unit produces an annual Audit Plan setting out the areas and time-frames for audit activities. However, such audits focus only on financial matters such as investments assets and operational finances. The Internal Audit Unit does not undertake auditing in relation to PPA’s business processes, data management, or record-keeping activities. The Unit produces an annual Audit Report for PPA senior management. The PPA is subject to an annual External Audit by a designated government agency. The External Audit covers most aspects of the PPA’s business including - finances, investments, operations, business processes, and data and records-keeping. The PPA does not have a systematic follow-up process to address audit findings and recommendations or brief the BoD on actions taken post-audit. The results of the External Audit are not published or made available to the public. 1.6 Strategic Planning The PPA undertakes Strategic Planning and last produced a Strategic Plan in 2022. This Plan identifies strategic priorities and goals. It encompasses human resources, financial management, workload volume predictions, and ICT alignment activities. An action plan to achieve those priority outcomes is also included. The Head of ICT is an active participant in the Strategic Planning process. 1.7 Quality Assurance The PPA does not have a formal Quality Assurance (QA) system in place. Implementing a QA system would include establishing acceptable performance and service thresholds for each key function. Then, PPA would undertake systematic sampling of actual levels of performance/service, identify areas of under-performance or deficits in levels of service, identify reasons for such under-performance/ service level deficits, identify changes required to achieve QA standards, develop and implementing those changes, and monitor results. 1.8 Risk Management The PPA does have a Risk Policy (overseen by the BoD), but it has a very narrow focus and concentrates on financial risk only, is not integrated within the PPA’s Strategic Plan, and is updated only once every 5 years. However, comprehensive risk management encompasses all critical areas of the institution, including organisation, technology, service, human resources, finance, and physical assets. The PPA does not have a Risk sub-Committee (Board level), or a formal Risk Officer position, or a Risk Unit. Furthermore, the PPA does not keep a Risk Register – a prioritised list of major risks and the likelihood of them materialising. 1.9 Actuarial Review The PPA periodically undertakes an Actuarial Review of the Pension System (Defined Benefit component). The Law originally prescribed that such a review be undertaken every two years, and 5 years. This was later amended by Presidential Decree 29/2018 (Article 6), which stated “The Authority shall conduct every five years an actuarial advisory study to redetermine the percentage of contributions and benefits, taking into account the indicators of other regulations, in order to reach the financial balance." The projected time-frame within an Actuarial Review is 50-100 years. The last Actuarial Review was undertaken (by the ILO) in 2017. The Actuarial Report is submitted to the President, Government, Parliament, and External Financial Auditor. The Report is not made public or published on-line. 1.10 Annual Reporting The Law requires the PPA BoD to produce an Annual Report and to present this Report to the Prime Minister. It also prescribes the topics to be covered by that Report. The Law also requires that the Report sets out (ex-post) information about PPA’s business activities, investments, service delivery, audits, financial performance, and human resources. This Report is made available to the BoD, PPA management and other 45 institutional stakeholders. However, the Report is not published online, or made available to the public. 1.11 Human Resources Management The PPA has a formal HR unit (Department of Human Resources) within the Directory of Administrative Affairs, Studies, and Planning. The Director of that Directory is a member of the PPA’s senior management team and participates in the development of the PPA Strategic Plan. The PPA’s staff are all public servants, and the Department of Human Resources participates in the staff recruitment, selection, and promotion processes. This Department also undertakes staff requirements forecasting. PPA staff are recruited via multiple steps – open advertisement, interview, and examinations. Selected staff are then placed on probation – to confirm their suitability. There is a staff evaluation process, though it is not linked to a financial reward programme. Technical and professional staff positions have formal written job descriptions. There is a PPA Human Resources committee, and whilst it is not a sub-Committee of the BoD, it comprises of the three Assistant General-Directors, the Administration and Planning Director, and the Head of the Counsel. Its’ duties include review and analysis of HR and administrative issues in PPA and deliver conclusions and recommendations to the General-Director. A2. Investment and Management of Funds The PPA is legally tasked with managing the investment of contribution funds. The PPA’s investment function is guided by the PPA’s Investment Policy Statement and is under the management control of the Assistant General-Director for Financial Affairs and Investment. There is a specific Department comprised of three units (Investment Study and Follow-up, Specific Shareholding, Loans and Replacements) that administers the investment function. Investment decisions are under the final approval authority of the BoD sub-Committee for Investments. PPA has a written code of ethics that govern all staff involved in the investment function. Investment execution and management (active/passive) of PPA’s assets are contracted out to the Bank of Palestine. The PPA’s investment assets are safeguarded by a Custodian (the National Bank) contracted by the PPA. The PPA publishes quarterly reports on the financial situation of the investment portfolio. The last report was in final quarter of 2022. The information is provided in Excel spreadsheets only, as PPA does not use special software to help manage its’ investment function. A3. ICT Management and Data Policies 46 The PPA has a strong culture of Information and Communication Technologies (ICT), a relatively mature ICT system, and a well-organised ICT administration system. There is a formal Head of ICT Management Division, and this person is a member of the PPA’s senior management team. The ICT Department has ten staff who are professionally qualified ICT specialists. The PPA has a comprehensive set of written policies governing the work of the ICT Department and system. These policies cover the areas of ICT governance, ICT risk management, data governance and data management. There is an ICT Help Desk to support business users, and a mechanism for users to provide feedback to (ICT Department) on their use of the software applications. The ICT system has a client-server architecture and is deployed, via PPA’s private VPN, to all (three) PPA branch offices. The PPA does not use cloud services. It does not have any preferred hardware brand or supplier and does not have a maintenance contract in place. The PPA has plans to significantly upgrade its’ ICT system within the next three (3) years. 3.1 Software Software development is undertaken both in-house and by outsourced contractors. At present there are three (3) such contracts being managed. The main software applications were developed in-house. The last comprehensive business requirements/gap analysis was undertaken in 2013. The PPA has a separate unit (Department of Systems Programming) responsible for software development and maintenance. This unit applies formal development analysis, standards, and methodologies to all software development tasks – ensuring reliability and consistency across all modules. However, the skill levels in this unit have been diminished (e.g. with retirements of skilled staff, and lack of new major applications developments). There is a lack of documentation underpinning the current software applications. However, there is the original source code for its main software applications. Most software is covered by up-to-date licenses, except for the Oracle applications. The core business functions of client registration, contribution collection and accounting, individual insurance accounts management, benefits processing, benefits payments, and client case management are covered by the main software applications. In addition. there are a small number of support applications (e.g., OpenText - document management/archiving, Oracle Financials – accounting, workflow management). The last significant upgrade to the core software applications was undertaken in 2017. The last significant upgrade to the operating system was undertaken in 2016. The last upgrade to the system hardware was undertaken in 2015. The PPA reports that a main challenge for it is the lack of experienced in-house technical staff, and the lack of training for ICT staff. 47 3.2 Data Centre The PPA has its own on-premises Data Centre where the main hardware and software applications are housed. This Data Centre manages all the PPA’s ICT services. This Data Centre is not benchmarked against international standards published by organisations such as the Uptime Institute, or Telecommunications Industry Association. The PPA does not have a physical back-up Data Centre, or any virtual back-up facility (e.g., using cloud services). 3.3 Risk Management and Disaster Recovery The PPA’S ICT Department has its own stand-alone written Risk Management Plan. This plan covers most critical ICT-centric issues including data destruction, technology, hardware, networks, software applications, systems security, human resources, internal accident (e.g., fire, accidental loss of data), and natural disaster event (fire, flood, lightening, earthquake). However, the PPA does not have a written Business Continuity (BC) plan or an Emergency Response Team – that would cover how the PPA’s core business functions could continue operating in the event of a catastrophic event impacting on the ICT system. The PPA does have a Disaster Recovery (DR) plan, and a DR plan is a key sub-set of a full BC plan. The PPA undertakes systematic tests to ensure the DR plan is functioning. However, it is not understood how a DR Plan can be considered complete without a backup Data Centre functionality (either in the cloud or in a physical premises). The systems administrator uses a formal process to restrict access to areas of the system (including use of applications and access to data), and to remove permissions from persons who leave the organisation or who change roles within the PPA. The PPA has a formal policy (including security) to manage the use of mobile devices for PPA business. The PPA’s ICT system access points (pcs/tablets/terminals/phones) are all covered by secure authentication procedures that involve the use of strong password and lock-out functionality. In addition, the PPA uses anti-malware and cyber- threat detection tools, and keeps these protections updated. 3.4 Data Governance and Management The PPA does not have data privacy-specific guiding legislation at national level32 that governs the matter of data privacy and data protection. However, the PPA does have a Data Working Group/Committee that is focused on Data Governance and Management policies. These policies include data organisation standards, data formats and data entry standards, data storage/access/retrieval standards 32 For example: the EU General Data Protection Regulation 48 and procedures, data integrity, database management, data exchange (internal and external), and data reporting. The PPA does not have a Chief Data Officer that is responsible for developing and managing data-specific policies – including data safeguarding, data manipulation, data storage, data sharing, data amendments. The PPA’s ICT systems development/upgrading/amendment plans are reviewed formally to ensure their conformance with PPA’s Data Governance policy. All PPA staff must sign a data non-disclosure/confidentiality agreement. However, there is no formal sanction system applicable to those who breach such agreements. The PPA does not have a well-established system of automatic data exchange with other agencies (e.g., to verify civil status, etc.). It is understood that this is mainly due to the lack of digitalisation of such records in such institutions. The PPA undertakes regular sampling of its’ beneficiary data quality and integrity. 3.5 Reports, Business Intelligence, and Data Analytics The PPA produces a range of reports that provide information on service, delivery, operational performance, and technical issues. These reports are generated in accordance with prescribed formats and content. Reports are made available to managers in Directories and in branch offices. The reports are produced on a daily and monthly basis, or by special request. There is no central dashboard from which managers can obtain critical information about the business and technical systems performance or detect service delivery issues. The PPA uses its own software to generate reports from data analysis. These reports are generated directly from the live production databases only. The PPA does not have a Data Warehouse – from which it could generate reports without potentially affecting the live database. PPA does not use Business Intelligence software to, e.g., undertake deeper data analytics, or to detect trends. A4. Identification and Registration All insured members must be registered in the PPA by the employer. Registration with the PPA is separate (and not linked) to registration with any other institution (e.g., health insurance, tax authority) and there is no linkage (or registration data exchange) between registration with PPA and registration with other institutions (health insurance, tax authority etc.). Registration is undertaken electronically, although the original contract of employment must still be submitted to PPA in hard copy (paper) format. 49 All insured members are registered using their National ID33 as the primary registration key. However, for persons who work in the non-government sector, there can be more than one linked employer number. The PPA checks existing registrations in its database before it registers a new member. The primary data collected and recorded by PPA. The PPA uses the National ID to identify and register contributors and beneficiaries in all Schemes. In cases where the subject has no National ID (e.g. foreigners) the PPA uses an alternative identification method (usually passport number etc.). Notwithstanding the fact that most pensioners would be in possession of a National ID card, the PPA issues its’ own card to pensioners. This card is used for identification purposes. It is a plastic card that contains a photo and basic information of the beneficiary, including National ID. It also contains a card number and Beneficiary Number that is generated by PPA. It has no magnetic band or any electronic chip onboard. It is issued from PPA through the IT Department after the approval of the Pensions Department. It is not issued for survivors, but only for the primary beneficiaries. A5. Contribution Collection The process of calculating and paying contributions due is as follows: 1. Employer submits a declaration form (extract from payroll) to PPA for pre- approval – with details of each active employee, including their wages and the time worked. 2. The PPA uses the payroll information to calculate the contributions due to be paid in respect of each employee. 3. The PPA notifies the employer about the calculated contributions due to be paid. 4. The employer then remits the contributions to PPA – the Government sector uses a wire transfer, whilst non-Government employers use a bank cheque. The contributions are paid into a single account that covers all Schemes, though the payment declaration form shows the amount due for each Scheme. 33 The PA’s Ministry of Interior (national civil affairs registry) issues National ID numbers to persons at birth in the Palestinian Territories, and to other qualified persons if they were born abroad or were exiled by Israel. This ID is used by the PPA, the Tax Administration and by most other public institutions and service providers (e.g. medical insurance, social assistance). This ID is also supposed to be recognized by Israel in accordance with the Oslo Accords. The National ID card does not contain any electronic bio-metric data but does have a photo included. It also contains personal information about the cardholder, as well as date of issuance. The data on the card is written in both Arabic and Hebrew languages. The card also contains an ID number. The ID is just a simple sequential number, but it is a unique number and is adopted in almost all institutions as identification for residents. The number does not contain perso nal identification data within its’ structure. 50 When the payment has been made the PPA checks that the paid amount balances with the amount notified to the employer by the PPA at the contribution calculation and approval stage. The PPA does not have a control and compliance unit that is responsible for ensuring employers are complying with their contribution and reporting obligations. The PPA does not have field inspectors that visit employers’ premises or inspect employers’ books to ensure compliance with their social insurance obligations. The PPA has the authority to impose sanctions on employers in cases off non-payment of contributions, late payment of contributions, or irregularities in registration of employees. A6. Benefits Management The PPA manages the following categories of benefits: • Old-Age (Retirement) – defined benefit • Old-Age (Retirement) – defined contribution • Old-Age (Retirement) – basic pension (means-tested) • Disability Benefit • Survivors Benefit • Family Allowance/Child Allowance • Funeral Grant • Marriage Grant • Loans to active contributors 6.1 Old Age – Defined Benefit Employers notify the contributor that they should apply for their Old-Age/Retirement benefit. There is no facility to apply online (web service) for the benefit. Applicants must supply the following evidence to the PPA: ▪ National ID ▪ Employer workbook showing details of applicant’s employment history ▪ IBAN number – for bank account that benefit will be paid into ▪ Marriage certificate – for applicants claiming additional Wife Allowance component ▪ Birth certificate(s) – for applicants claiming additional Child Allowance component The PPA uses its internal electronic and paper-based records to assist in determining the contributor/member’s application for benefit. Applicants who were born after 1 September 1961, and who were, up to September 2006, enrolled in a Scheme that has been closed have their employment/insurance 51 periods transferred to the new Unified Pension Scheme (Scheme II). These periods of insurance are added together to determine their total number of years insurance. However, the amount of pension is calculated separately for the periods they were in each Scheme in accordance with the rules of those Schemes – and then combined to make up their pension amount payable. 6.2 Disability Benefit This benefit, whilst it requires registration as a Contributor/Member, is not contingent on having a minimum number of insurance periods. Applicants are medically examined by a committee of medical professionals appointed by the Ministry of Health. Following this initial examination persons who are deemed medically eligible are then medically re-assessed every 24 months. The results of medical assessments are used solely by the PPA and are not shared with any other institution. 6.3 Survivors Benefit Articles 32 through 36 of the Law prescribes a detailed and complex system of qualification for a survivors’ benefit, and apportionment of such benefit amongst eligible family members. Continued entitlement conditions are also set out in these Articles. Further analysis of this benefit is outside of scope of this note. 6.4 Basic Pension In addition to benefits payable under the social insurance Schemes, the PPA pays a basic pension to persons who were formerly members of a Scheme but who do not qualify under the rules of the Scheme (don’t have enough contributions/insurance periods). This benefit is means-tested, and the PPA collects proof of means (from inspections undertaken by the Ministry of Social Security) to verify eligibility. This payment is funded by the Ministry of Finance and payment is administered through the PPA system. 6.5 Benefit Payments and Control Beneficiaries are required to have a bank account, and to submit their IBAN number when making application for benefit. Payments are made monthly and are deposited directly into the beneficiary’s bank account. There is no other payment method or payment channel used by PPA. The PPA periodically (annually and sometimes more frequently) undertakes a ‘proof of life’ review of beneficiaries. This requires the beneficiary to personally attend a PPA office and show proof of ID. In addition, the Civil Registry periodically supplies – offline and in batch mode - the PPA data on registered deaths, in digital format. These data are used as a control measure – to identify persons who have died but where PPA have not been notified. The PPA has a policy of prosecuting persons who fraudulently obtain benefits. However, PPA does not have a special control unit (internal or external) to ensure beneficiaries continued eligibility and to detect potential fraud. 52 From the fraud control perspective, it is unclear why the PPA Beneficiary Card is a necessary addition to the National ID Card, as it seems to duplicate all the data on the National ID card. Furthermore, the PPA Beneficiary Card has not been produced in the same secure environment or according to the same data management rules as the National ID card. A7. Grievances - Appeals and Complaints The Law (Article 39) provides for the right of participants to appeal against decisions taken by the PPA. This right also includes the right to apply to the courts to seek a judge’s decision. The PPA has a formal regulated system of grievance redress (appeals and complaints resolution). This system differentiates between appeals and complaints. The PPA has a senior manager (akin to a Chief Appeals Officer) who oversees the system. Both appeals and complaints are handled at PPA’s regional and central office levels. Whilst there is no specific form that must be used to submit an appeal or a grievance, appeals can be lodged through personal attendance at PPA office, by letter, or by telephone call. There is no online facility to submit an appeal, and appeals are not accepted if they are submitted by email. Whilst the Law requires the PPA to act ‘promptly’ to resolve appeals, there is no specific time-limit set for PPA to resolve appeals or complaints. In addition to using the appeals channel, persons are also entitled to directly ask the courts to investigate their appeal, though it is unclear (in the Law) if they must first engage with the PPA’s appeal process. Though appeals data are examined to identify potential improvements/adjustments to operations or service delivery, complaints data are not similarly analysed. A8. Communications The PPA has a special unit ‘Public Relations and Media Department’ whose responsibility is to develop and implement PPA communications plans. This Department reports directly to the PPA’s General Director. The Head of the Department is a member of the PPA’s senior management team. The PPA co-operates with other institutions (e.g., Tax Administration/ health insurance fund, trade unions, employers) to provide Scheme information to employees, and other relevant persons. The PPA does not have a written communications strategy or plan. It does not undertake targeted public awareness campaigns – to explain pension-related issues to the employees and servicemen of the public sector, to inform them of changes to the pensions system, to remind employers and employees of their rights and obligations to 53 the pension system. It does not have a formal internal communications plan either – to brief staff, keep staff updated on specific topics etc. The PPA does have a written document that specifies customers rights, responsibilities, PPA’s service standards, and information about how to make a complaint. It does not undertake client opinion surveys – to measure attitudes towards the PPA and the pensions system, to measure the public’s understanding of the pensions system, to gain suggestions about improvements to the PPA’s service etc. Furthermore, the PPA doesn’t utilise outreach methods (focus groups, customer panels, customer liaison committees etc) to seek advice or gain input to its business strategy or seek policy implementation feedback. The PPA does not operate a call-centre. While its website contains information brochures for employers and employees, it does not provide a Frequently Asked Questions section and is not linked to other governmental websites. The PPA uses social media and its website to disseminate information. However, the PPA does not collect data on how or through what medium/channel the public accesses information about the pensions system or about PPA’s services. The website is lacking in interactivity with the clients, as it does not enable members to view their contributions history, to submit benefit applications, or to submit appeals. Summary of Observations and Recommendations Overall, the PPA appears well structured. It has administration units with responsibilities for most functions expected of a public pension institution. Its’ formal governance system reflects international best practice – in terms of having a representative Board, Board sub-Committees, General-Director (Chief Executive), and Assistant General-Directors. The roles and responsibilities of these formal positions are clear and well-articulated in the Law (2005). In addition, there are written job descriptions for each senior managerial position. The organisational architecture below these (most) senior management levels is quite logical and structured, with groupings at Directory, Department, and Unit levels. The PPA is organisationally structured and positioned to implement the Law as it is written. Nonetheless, the reality is that some components of the Law (Defined Contribution, basic pension for all-population) are not fully operable in practice. Furthermore, the contribution collection function is seriously impaired – due to the lack of finances available to the Government, with consequential knock-on effects for availability of funds to the investment function. The ICT function is well organised, and the system contains most of the functionality expected of a pensions ICT system. However, there are weaknesses with regard to web/e-services, data governance, and disaster recovery. Whilst the PPA does have a basic online presence, it lacks sufficient functionality. There is an opportunity to pivot the technology towards web-services and reduce reliance on physical attendance by clients at PPA offices and using legacy channels to obtain data. 54 There are also weaknesses in critical areas such as: ▪ Risk management (focus is too narrow) ▪ Audit (focus is too narrow) ▪ Data analysis and reporting tools (e.g., underdeveloped Business Intelligence) ▪ Quality assurance-management (function is too weak) ▪ Appeals (function is underdefined) ▪ Communications (focus is too narrow) In respect to the above findings, the following summarizes our key recommendations: e-Services • Adopt an e-services digitalisation strategy. • Re-develop the PPA website to encompass more (automated) functionality – including benefit claims submission, employer reporting, appeals submission, email acceptance, FAQs, etc. • Re-design function and process workflows to facilitate automatic (and digitised) data reception and transmission. • Develop interoperability functionality – to facilitate automatic data exchange with other public sector institutions and data sources. ICT and Data Governance • Appoint a Chief Data Officer • Develop and implement a data governance plan – which will include (a) an analysis of how each type of data is managed and safeguarded (against error, fraud, accident etc.), and (b) a definition of what the PPA’s master data are. Business Intelligence and Data Analytics • Deploy an offline data warehouse to enable business analytics and reports to be generated without using the live production database. • Deploy Business Intelligence tools – to enable better and more accurate insight to: business systems efficiency, policy outcome/effectiveness, identify trends, highlight performance issues, better respond to requests for data etc. Disaster Recovery and Back-up Capacity • Develop a full Business Continuity Plan • Develop a physical back-up Data Centre and/or buy-in cloud back-up and storage facilities - as part of Disaster Recovery Planning. Audit and Risk • Broaden the Audit Plan to include audit of business processes, data management, record-keeping, service delivery, disaster recovery etc. 55 • Mandate the Internal Audit Unit to undertake such expanded scope of auditing. • Link Audit Plan to the Risk Plans. • Develop a PPA Risk Strategy – with detailed Risk Plans for every PPA function and responsibility. • Create a Risk Unit • Appoint a Chief Risk Officer Quality Management • Develop and implement a Quality Assurance (QA) system to enable performance measurement across all key areas of PPA administration. Such a QA system would include specifying acceptable performance and service thresholds for each key administration function and using those standards to benchmark actual performance and identify needed modifications to systems or practice. • Integrate a QA Plan into the PPA Strategic Plan. Appeals • Analyse complaints data. Examine ways in which Scheme information can be improved – to reduce possibility of appeals that are due to misunderstanding. • Impose time-limits for PPA resolution of appeals and complaints, as part of service standards • Develop case management system (as an MIS module) to register all appeals, track progress, and record results. Publish annual report on appeals system, using anonymised cases to highlight important cases and decisions. Include ratio of accepted/rejected cases – per Scheme. • Benchmark the existing appeals system against best practice principles. Communications • Review existing communications function – internal and external. Identify gaps and opportunities to strengthen and improve communications. • Assess options of establishing a call-centre for clients. • Develop comprehensive communications strategy – covering all areas of communications. • Develop of system and process client opinion surveys – to measure client and public attitudes towards the PPA • Develop communications training plan for PPA management and staff. 56