Report No. 19669-NI Nicaragua Pension Reform Proposal April 3, 2000 Finance, Private Sector and Infrastructure Central America Country Management Unit Latin America and the Caribbean Regional Office ocurment of the World Bank CURRENCY EQUIVALENTS (Exchange Rate Effective 3/20/00) Currency Unit = C6rdobas I C6rdoba= US$ 0.0827 US$ I = 12.44 C6rdobas FISCAL YEAR January 1 to December 31 ABBREVIATIONS AND ACRONYMS CREPEN Comisi6n para la Reforma de Pensiones de Nicaragua (Pension Reform Technical Commission) DB Defined Benefit DC Defined Contribution EAP Economically Active Population ESAF Enhanced Structural Adjustment Facility FPSI Finance, Private Sector and Infrastructure Sector GDP Gross Domestic Product GON Government of Nicaragua HD Human Development Sector IMF International Monetary Fund INEC Instituto Nacional de Estadisticas y Censos (National Statistics and Census Institute) INSS Instituto Nicaragiuense de Seguridad Social (Nicaraguan Social Security Institute) IPD Implicit Pension Debt IVM Invalidez, Vejez y Muerte (Old Age, Disability and Survivor's Insurance) PAYG Pay-as-you-go PROST Pension Reform Options Simulation Toolkit SBIF Superintendencia de Bancos y Otras Instituciones Financieras (Superintendency of Banks and Other Financial Institutions) Vice President: David de Ferranti Country Manager/Director: Donna Dowsett-Coirolo Sector Manager/Director: Danny M. Leipziger Task Team Leader/Task Manager: Miguel Navarro NICARAGUA: PENSION REFORM PROPOSAL Table of Contents GLOSSARY iii EXECUTIVE SUMMARY vii I. INTRODUCTION 1 II. THE CURRENT SOCIAL SECURITY SYSTEM: THE OLD AGE, DISABILITY AND SURVIVORS INSURANCE (IVM) ......................... 2 System Characteristics .............................................................. 2 Description ..............................................................2 Revenues ..............................................................2 Benefits ..............................................................3 HI. SHORTCOMINGS OF THE IVM SYSTEM ........................................4 Systemic Problems ..............................................................4 Low Payroll Taxes ............................................................. 4 Generous Level of Benefits ...................................................5 Financial Unsustainability ................................................... 5 Evasion ..............................................................6 The Reference Salary, the Averaging Period, and the Effect of Inflation ..............................................................6 Other Problems ........ ......................................................7 Low Contribution Base (incomplete coverage) ............................ 7 Institutional and Administrative Weaknesses .............................8 Labor Market Distortions ..................................................... 8 Investment Returns ............................................................ 8 Non-Contributive Pensions ...................................................8 IV. LONG TERM FINANCES AND THE RATIONALE FOR REFORM ........ 10 V. THE REFORM PLAN ............................................................. 12 Reforming the PA YG System ................................................ 12 Evaluating Parametric Reform Options for Nicaragua ................. 12 Creating a Mandatory, Fully-funded, Defined-contribution System Based on Individual Capitalization Accounts ............................. 13 Design Issues ............................................................. 15 Should the Current PA YG System be Fully or Partially Closed?...... 15 How to RecognizeAcquired Rights? ....................................... 15 How to Finance the Transition? ............................................ 17 ii April 2000 Implementation Issues .................................................. 18 How to Administer the New System? ....................................... 18 How to Regulate and Supervise the New System? ....................... 18 How is the fee structure established? How Can Administrative Costs be Contained? .................................................. 19 How are the Funds Going to be Invested? ................................. 20 Other Issues .................................................. 20 Disability and Survivors Insurance ......................................... 20 Minimum Pension Guarantee or Social Assistance Pension ........... 20 Collection and Transfer of Contributions ................................. 21 Communication and Publicity Campaign .................................. 22 Regulations Required to Implement the Reformed System ............. 22 VI. THE NEW SYSTEM AND THE NECESSARY FINANCIAL SECTOR ARRANGEMENTS ................................................................... 22 BIBLIOGRAPHY ......................................................... 24 ANNEX I SIMULATIONS: TECHNICAL NOTES ANNEX II FINANCIAL REQUISITES FOR PENSION REFORM Miguel Navarro (LCSFP) prepared this report under the supervision of Hermann von Gersdorff (LCSFP). Anniruddha Bonerjee (HDNSP) carried out the simulations and contributed to the analysis of the reform strategy. Gloria Grandolini (CFO) and Dimitri Vittas (DECRG) acted as peer reviewers. Truman Packard (LCSHS); Lisa Taber (LCSER); Fernando de Mergelina, Cara Zappala, Antonio Paniagua (LSCFP); and Donald Mclsaac (BFI) provided additional contributions. The team wishes to thank the staff of the Instituto Nicaraguense de Seguridad Social, the Comisi6n para la Reforma de Pensiones de Nicaragua and the Secretaria Thcnica de la Presidencia for their input in the preparation of this report. Nicaragua: Pension Reform Proposal iii Glossary Actuarial fairness (Equidad actuarial) a method of setting insurance premiums according to the true risks involved. Adverse selection (Seleccion adversa) a problem stemming from an insurer's inability to distinguish between high- and low-risk individuals. The price for insurance then reflects the average risk level, which leads low-risk individuals to opt out and drives the price of insurance still higher until insurance markets break down. Average effective retirement age (Edad dejubilacion efectiva media) the actual average retirement age, taking into account early retirement and special regimes. Benefit rate (Tasa del beneficio) the ratio of the average pension to the average economy-wide wage or covered wage. Contracting out (Contrataci6n) the right of employers or employees to use private pension fund managers instead of participating in the publicly managed scheme. Defined benefit (DB) (Beneficio definido) a guarantee by the insurer or pension agency that a benefit based on a prescribed formula will be paid. Defined contribution (DC) (Aporte definido) a pension plan in which the periodic contribution is prescribed and the benefit depends on the contribution plus the investment return. Demographic transition (Transici6n demografica) the historical process of changing demographic structure that takes place as fertility and mortality rates decline, resulting in an increasing ratio of older to younger persons. Full funding (Financiamiento total) the accumulation of pension reserves that total 100 percent of the present value of all pension liabilities owed to current members. Implicit public pension debt (Deuda implicita del sistema piblico de pensiones (neta)) (net) the value of outstanding pension claims on the public sector minus accumulated pension reserves. Intergenerational distribution (Distribuci6n entre generaciones) income transfers between different age cohorts of persons. Intragenerational distribution (Distribucion dentro de una misma generacion) income transfers within a certain age cohort of persons. iv April 2000 Legal retirement age (Edad dejubilaci4n legal) the normal retirement age written into pension statutes. Means-tested benefit (Beneficio supeditado a la necesidad) a benefit that is paid only if the recipient's income falls below a certain level. Minimum pension guarantee (Pensi6n minima garantizada, o garantia de una pensi6n minima) a guarantee provided by the government to bring pensions to some minimum level, possibly by "topping up" the capital accumulation needed to fund the pensions. Moral hazard (Peligro moral) a situation in which insured people do not protect themselves from risk as much as they would have if they were not insured. Old age dependency ratio (Coeficiente de dependencia de los ancianos) the ratio of older persons to working age individuals. The old age dependency ratio used in the text refers to the number of persons over 60 divided by the number of persons aged 20 to 59. Pay-as-you-go (PAYG) (Pagos con ingresos corrientes, o sistema de reparto) in its strictest sense, a method of financing whereby current outlays on pension benefits are paid out of current revenues from an earmarked tax, often a payroll tax. Pension coverage rate (Tasa de cobertura de las pensiones) in this report, the number of workers actively contributing to a publicly mandated contributory or retirement scheme, divided by the estimated labor force. Pension spending (Gasto en pensiones) in this report, pension spending is defined as old age, retirement, survivors', death, and invalidity-disability payments based on past contribution records plus noncontributory, flat universal, or means- tested programs specifically targeting the old. Portability (Transferibilidad) the ability to transfer accrued pension rights between plans. Provident fund (Fondo deprevision) a fully funded, defined contribution scheme in which funds are managed by the public sector. Replacement rate (Porcentaje de reemplazo) the value of a pension as a proportion of a worker's wage during some base period, such as the last year or two before retirement or the entire lifetime average wage. It also denotes the average pension of a group of pensioners as a proportion of the average wage of the group. Nicaragua: Pension Reform Proposal v System dependency ratio (Coeficiente de dependencia del sistema) the ratio of persons receiving pensions from a certain pension scheme divided by the number of workers contributing to the same scheme in the same period. System maturation (Maduracion del sistema) the process in which young people who are eligible for pensions, in a new system, gradually grow old and retire, thereby raising the system dependency ratio to the demographic dependency ratio. In a fully mature system all old people in the covered group are eligible for full pensions. Universal flat benefit (Beneficio universal uniforme) refers to pensions paid solely on the basis of age and citizenship, without regard to prior work or contribution records. Vesting period (Tiempo de servicio) the minimum amount of time required to qualify for full ownership of pension benefits. vi April 2000 Nicaragua: Pension Reform Proposal vii Executive Summary Nicaragua is in the midst of reforming its social security system. This system is based on a defined benefit pay-as-you-go (PAYG) scheme that is not longer viable, given its low payroll taxes, overly generous benefits, short averaging period and high rates of evasion. When combined with the institutional and administrative weaknesses of INSS, the result is, according to the simulations, a cash deficit in the system since 1997 and the depletion of its reserves in the immediate future. Therefore, unless the structure of the system becomes sustainable and more equitable in the long run, benefits will have to be drastically cut or suspended and/or, contribution rates increased considerably. With Nicaragua's population still relatively young, its current low coverage and low age dependency ratio (retirees to workers, reduced coverage and manageable implicit pension debt (84.9% of GDP in 1996 compared to a likely increase to over 400% of GDP by 2030), the time is propitious for the Government to reform the system at a reasonable cost. This can be achieved by undertaking an in-depth systemic reform with the following characteristics: * The parameters of the system need to be re-defined and a mandatory, defined contribution system based on individual capitalization accounts introduced simultaneously or shortly thereafter. * The current PAYG system should be closed to new entrants, those younger than 45 years of age should move to the new mandatory system and those between 45 and 50 should be given the choice to remain in the old system or switch to the new system. * Acquired rights to those moving to the new system should be recognized through a recognition bond. * A minimum pension guarantee should be established. * The transition should be financed out of general revenue, payroll taxes, pension fund reserves, treasury bonds or other possible sources like INSS's health surpluses, or, as a last option (not considered by the Government at this point) borrowing from international organizations. * The system should be administered by a limited number of pension fund managers who would bid for a license through an international competitive bidding process. * Funds should be invested both domestically and abroad to obtain the highest possible rates of return and allow for proper diversification of the portfolio and protection from local economic and political manipulation. * The system should be regulated and supervised by an independent Superintendency of Pensions. At the same time, the Government must also consider undertaking measures to strengthen its financial and insurance markets; develop new instruments; foster international competition and investment; and improve its supervisory capacity in all areas of the financial sector. To support this reform, a communication campaign must be implemented to create consensus and explain the reform to all participants, decision makers and the general public. viii April 2000 The successful implementation of the reform will guarantee the long term viability and sustainability of the system by making it more equitable. Workers will have ownership of their resources, confidence will increase and so should coverage. This should help alleviate old age poverty and increase the overall savings rate. This boost in the savings rate will, in turn, trigger the demand for new and longer term financial instruments, stimulate local financial markets, and increase the availability of resources for productive investment and development in Nicaragua. Nicaragua: Pension Reform Proposal I. INTRODUCTION icaragua is in the midst of reforming its social security system. State provided N social security in Nicaragua can be traced back to the 1930s when a pension system was created for teachers older than 60 years of age' and formalized in the 1950s with the establishment of the Social Security Institute (Instituto Nicaraguense de Seguridad Social - INSS) as an autonomous institution in charge of administering the "Seguro Social Obligatorio" (mandatory social insurance). The system was set up as a defined benefit pay-as-you-go scheme (PAYG), funded through compulsory contributions to cover employees' Old Age, Disability, and Survivors Insurance (Invalidez, Vejez y Muerte - IVM); employees' Health Care (Seguro de Enfermedad y Maternidad - SEM); and Workers Compensation Insurance (Riesgos Profesionales - RP, added in 1959). 2. In 1979, with the rise of the Sandinistas to power, a law creating the Sistema Unico de Salud was enacted and the Ministry of Health became responsible for the provision of all health related benefits. INSS hospitals and clinics were transferred to the Ministry of Health and lNSS' role was reduced to collecting and transferring health contributions and providing old age, disability, survivors and workers compensation insurance. INSS' reduced responsibilities were short lived. The Government made INSS responsible for the provision of social assistance to non-insured sectors of the population. New unfunded programs (see non-contributory pensions below) were created, increasing to almost 20 the number of programs administered by INSS2. INSS' expanded mandate contributed to the rise in its operational costs and financial distress that, today, is threatening the sustainability of Nicaragua's social security system. 3. Although most of the programs currently under INSS administration are characterized by high administrative costs, fraud and evasion, and increased inequalities and inefficiencies that have negatively impacted INSS finances, this is not a problem unique to Nicaragua but common to PAYG schemes throughout the world. To avoid any further deterioration of INSS finances that could spread into other areas of the economy', i Teachers had to be at least 60 years of age and worked at least 20 years to be eligible to receive a Government assisted pension. 2 Health Care: Medical attentions, subsidies for Health Care, Maternity Leave, Prinzapolca Sector. IVM: Partially Disabled, Total Disability, Old Age, Survivors, Orphans, Ascendants. Workers Compensation (RP): Medical Attention, Subsidies, Partial Incapacity, and Pensions for Partial Disability, Total Disability, Survivor, Orphans, and Ascendants. Funeral Services 3 Since the Govermment is the guarantor of the social security system, the deficits incurred by INSS would have to be covered with fiscal resources that could otherwise be used for other productive activities or social areas. Page 2 April 2000 the Nicaraguan Government has taken the first steps toward reforming the system, including submission of a draft reform law for the consideration of the Assembly. 4. The report, based on the on-going policy discussions with the Government, is intended to guide the reform process by identifying issues that need to be addressed and specific recommendations for the design of the different components of the new system. While the report focuses on the reform to the largest component of the social security system, the IVM Insurance and, in particular, the Old Age Insurance (Pensions), references will also be made to Disability and Workers Compensation Insurance. The recommendations made in this report are based on the simulations carried out using the World Bank's Pension Reform Options Simulation Toolkit (PROST), discussed in Annex I. II. THE CURRENT SOCLIL SECURITY SYSTEM: THE OLD AGE, DISABILITY AND SURVIVORS INSURANCE (IVM). System Characteristics 5. Description. Nicaragua's national pension system consists of a publicly managed pay-as-you-go (PAYG) defined benefit (DB) plan for Old Age, Disability, and Survivors Insurance (IVM). Contributions to the IVM system are mandatory for salaried workers and voluntary for independent workers. An average worker has to contribute for a minimum of 750 weeks (15 years) and reach the age of 60 to be eligible for a pension calculated on the average weekly salary of the last 250 weeks (5 years) of work4. The pension replacement rate is 40% of base salary plus 1.365% accrual for every year after the first three years. 6. Revenues. Nicaragua's social security system is supposed to cover all urban and rural sectors of the economy, including private and public sector employees, as well as the self-employed and members of small family businesses5. The main source of revenue comes from mandatory contributions equivalent to 17% of payroll, of which 5.5% are for IVM. Table 1 presents the distribution of tripartite contributions6 for the main benefits offered to covered workers. 4 For those who have contributed for more than 1,000 but less than 1,250 weeks, their pensions are calculated on their average weekly salary of the last 200 weeks (4 years); and, for those who have contributed for more than 1,250 weeks, their pensions are calculated on their average weekly salary of the last 150 weeks (3 years) 5The military is exempted from the national PAYG system administered by INSS. 6 Although contributions are supposed to be tripartite (employer, employee and government), the government contribution has never been paid. Nicaragua: Pension Reform Proposal Page 3 Table 1. Payroll Tax Rates for Social Security (% of Total Payroll) Contributor - Sickness & IVM Workers War Victims Total _________ I . |Maternity | Compens. l Employer 6.0 3.5 1.5 1.5 12.5 Employee 2.25 1.75 0.0 0.0 4.0 State 0.25 0.25 0.0 0.0 0.5 Total 8.5 5.5 1.5 1.5 17.0 7. Benefits. INSS pays out benefits for old age, disability, survivors/orphans and workmen's compensation according to the rules established in Table 2. In addition, INSS pays sickness and maternity benefits, family allowances and pensions for war victims. The focus of this report is limited to the first group of benefits. Table 2. Summary of Benefit(s) Rule Benefit Type | Eligibility Condition Replacement Rate7 OldAge Urban/Rural Age 60 (males & 40% of base + 1.365% females) for every year after the 15 years of contributions first three years Miners, teachers Age 55 (males & and physically or females) mentally impaired 15 years of contributions Disability Disability At least 3 years of Same as old age. certification for contributions in last 6 Disability pension total or partial years converted to old age disability pension at age 60. Survivors/ Survivors of At least 3 years of 50% of pension for Orphans contributor, contributions in last 6 survivor. 25% of retiree or disabled years pension for orphan Workmen's Work related No minimum qualifying 60% of earnings in last Compensation disability period 2 months of I § 0 _ contributions Minimum Pension Guarantee 100% of minimum wage (currently 600 C6rdobas or approximately US$50) after 15 years of contributions. Source: Social Security Programs Through the World, International Social Security Association - 1997 7In cases of old age pension, the minimum benefit has to be greater than I minimum wage (500 C6rdobas) and less than 24 minimum wages (12,000 C6rdobas). Page 4 April 2000 III. SHORTCOMINGS OF THE IVM SYSTEM Systemic Problems Table 3. Payroll Tax Rates for Pensions (%) 8. Low Payroll Taxes. Mandatory contributions to Latin America Before RAfter the IVM system are only lrReform 26 5.5% of the average weekly Blgentina 26 26 salary category', which puts Bolivia 9 14.5 Nicaragua in the lowest tier Brazil 30 1 of payroll taxes for pensions Chile 19 13 in Latin America (Table 3). Colombia 7 14 This low rate is not consistent Costa Rica 7 with the level of benefits Dom. Rep. 10 offered and will have to be Ecuador 20 - significantly increased, to El Salvador 3 13 7.3% by 2000, 17.6% by Guatemala 5 - 2030 and 36.4% by 2060 if Honduras 3 - the current system is retained Mexico 15.5 21 under its present benefit Nicaragua 5.5 - fonnula (Figure 1). Peru 9 13 Uruguay 27.5 27.5 Source: World Bank (1994) and staff research Figure 1. Contribution Rates 40% 35%1/ 30% 2i% v | Actual 15% 20% -- - . - -- -. - -. - ~~~~-- - -- Required 10% --- 5%~ 0% 8 To estimate the amount of the contributions to INSS, salaries are stratified by a progressive income category and a weekly salary level is determined for each bracket. The actual contribution to INSS is then based on a contribution table that sets the contribution amount for each bracket (average weekly salary category). Nicaragua: Pension Reform Proposal Page 5 9. Generous Level of Benefits. The current level of benefits surpasses what is actuarially sustainable and is not consistent with the level of contributions. Figure 2 shows that the on-going average replacement rate in Nicaragua should be reduced significantly over the next 50 years if the level of contributions remains at its current level. This financial deterioration is symptomatic of public PAYG systems in the region that offer defined benefits that are not actuarially related to contributions. Figure 2. Affordable Replacement Rates 60% 50% 40%= Ongoing 20% - -. .. 0% .... 10. Financial Unsustainability. As a consequence of the mismatch between the level of contributions and benefits, the PAYG system has been in a cash deficit since 1997. Preliminary calculations show an estimated year-end (1999) negative current balance (revenues minus expenditures) of 80 million C6rdobas (US$7 million) or 0.6 % of GDP (see Figure 3)9. To maintain the current system the Government will have to make budget transfers that will continue to rise and could reach 10% to 15% of GDP in the medium term"0. To avoid the system's bankruptcy and a default on its obligations, current payroll taxes would have to be increased and/or benefits reduced immediately. 9 The World Bank's Pension Reform Options Simulation Toolkit (PROST) projects GDP exogenously. It is important to keep this in mind when viewing all graphs and tables expressed as a % of GDP. To avoid the need to make budget transfers, the Government can decide to use other sources of financing like the health surpluses INSS currently produces. Page 6 April 2000 Figure 3. Revenues and Expenditures as % of GDP 20% 18% . . . 16%- 14% . 12% 10% 1_R 2% >__ 11. Evasion. Although the evasion rate has been decreasing in Nicaragua in the last year, it is still high at an estimated 55% of the system's potential income. It is generally believed that high rates of evasion are due to certain design features of the current social security system, such as a mismatch between contributions and benefits; an inadequate collection and inspection system; an outdated information system; and a lack of clear and enforceable penalties for non-compliance. Workers in Nicaragua have an incentive to contribute for only 15 years, for which they receive a very high rate of return under the present benefit formula and evade thereafter regardless of their relative youth. Higher payroll taxes could further increase the incentives to evade if the new system does not fully address these problems. 12. The reference salary, the averaging period, and the effect of inflation. The reference salary for the calculation of a pension is equal to average earnings during the last 5 years of a worker's career (based on contributions of 15 years). Basing the pension on salary toward the end of a worker's career is inequitable because it redistributes income to workers with rising age-earning profiles. An end-valued benefit formula encourages fraud in the form of under-reporting earnings in early years and over- reporting earnings during the last few years of work, thereby adding to the financial unsustainability of the system. These problems could be mitigated if the pension depended on average salary over some longer period of time (e.g. 10 years). For Nicaragua, however, the possibility of extending the averaging period should be carefully reviewed because contributions and pensions will have to be revalued for inflation. Otherwise, the system will penalize those who retire when inflation is high. This is further complicated in Nicaragua because contributions are determined based on the average weekly salary categories and not on individual salaries. Nicaragua: Pension Reform Proposal Page 7 Other Problems 13. Low contribution base (incomplete coverage). Of an estimated economically active population (EAP) of 1,695,400 (34.4% of the total population) in 1999, only 864,654 (or 51% of the EAP) hold a job in the formal sector of the economy and only 288,743 of these (or 17.0% of the EAP) contribute to the social security system on a regular basis (Table 4). These numbers indicate a high degree of informal sector activity and/or evasion. While the economic stagnation and hyperinflation of the 1980's, along with the lack of incentives (workers and employers still perceive contributions as a tax instead of the pre-payment for future benefits) and poor enforcement capacity may be to blame, coverage remains low by Latin American standards (Table 5). For this reason, INSS revenue flows, and particularly the IVM system, will continue to be negatively affected unless remedial measures are taken to curb evasion and increase coverage. Table 4. Contributors 1990-97 Year Total % Urban % Male EAP Contributors Contributors Contributors Contributors as % of EAP 1992 210,458 95 59 1,346,400 15.63 1993 204,033 95 58 1,398,100 14.59 1994 204,509 97 57 1,442,800 14.17 1995 210,610 98 57 1,487,700 14.16 1996 226,914 98 55 1,534,100 14.79 1997 242,642 98 55 1,580,300 15.35 1998 267,080 98 55 1,630,200 16.38 1999 288,743 98 55 1,695,400 17.03 Source: INSS, Department of Planning Table 5. Coverage of Pension Systems in Latin America in 1997 (million persons) Private System Private System Public Systems Total coverage Affiliates Contributors Affiliates (% of EAP) Argentina 6.26 3.07 2.33 80 Chile 5.74 3.22 1.26 98 Colombia 2.35 1.57 2.60 30 Bolivia 0.36 n.a. 0 12 Nicaragua n.a. n.a. n.a. 15 Peru 1.73 0.76 1.3 32 Uruguay 0.46 n.a. 0.50 70 Sources: Queisser, M (1998); von Gersdorff, H. (1997); INSS Department of Planning Page 8 April 2000 14. Institutional and administrative weaknesses. In addition to all the system weaknesses described above, it is well known that INSS is suffering from several years of neglect (10 years of a centrally-planned regime under the Sandinistas and 5 years of economic and political reconstruction under the Chamorro administration), reflected in today's weak administrative systems and procedures. Information systems are not fully automated or integrated for collections, payments and records. This leads to excessive administrative costs of 13% of pension expenses (based on transaction expenses estimated by INSS), deficient client services, poor collection mechanisms, and inadequate data for determining pension amounts and evasion due to the slowness of INSS in finalizing the separation of health and pension accounts. 15. Labor market distortions. As long as contributions are perceived as a tax, there will be an incentive to evade or under-report, as well as for preferring informality and/or opting for early retirement. These labor market distortions have a direct effect on the supply and demand for labor and on productivity. A well structured defined contribution system could revert this tendency by linking benefits to contributions and by penalizing early retirement. At the same time it could help increase participation from rural contributors that currently represent only 2% of total contributors (Table 4). 16. Investment returns. Given the lack of financial instruments in the poorly developed financial sector in Nicaragua (Box 1), INSS pension reserves have been mainly invested in government bonds and (only recently), in certificates of deposit from commercial banks. Such a strategy has proven ineffective in the past in maintaining or increasing the value of the reserves, particularly during the hyperinflation period of the 1980s when real interest rates were mainly negative on most instruments. This period of negative interest rates ended during the 1990s when the economy started to stabilize, but the damage to INSS financial stability is still visible. Retums have not being enough to cover the losses suffered during the 1 980s. 17. Non-Contributive Pensions. Of the 95,749 pensions paid by INSS in 1999, 37,264 (or 38.9%) were non-contributive pensions. These pensions, given to individuals who have never contributed to the system but have performed services for the country"1, have been partially financed with a mandatory employee contribution for War Victims equal to 1.5% of the total payroll. INSS has covered all deficits and administrative expenses, which has contributed to the weakening of INSS' financial position. The two main categories are: (i) war victims benefits (including those paid to the Sandinista Popular Army, defense forces, the Patriotic Military Service, local militia, military service reserves and the resistance forces), and (ii) special pensions (including those paid to miners, charitable pensions, public servants, special state pensioners, National Sovereignty Defense Forces and circus performers). Nicaragua: Pension Reform Proposal Page 9 Box 1. Nicaragua's Financial Sector REGULATION AND SUPERVISION. The Financial Sector is regulated and supervised by the Superintendency of Banks and Other Financial Institutions (SBIF). SBIF focuses mainly on banking and, to a lesser extent, on insurance and capital markets. BANKING SECTOR. Nicaragua's financial system is dominated by the banking sector. The sector has experienced rapid growth since 1991, when private banks were re-established, banking supervision and regulation introduced, interest rate and credit controls eliminated and an inter-bank foreign exchange market established. Nevertheless the banking sector remains small and fragile, poorly capitalized, with high intermediation margins and administrative expenses. Total assets are low and still diffused among 11 small commercial banks. Total equity in the banking sector has increased from US$70 million in 1997 to over US$90 million in 1998, still capitalization levels remain low. During 1998 and 1999, the Superintendency of Banks and other Financial Institutions (SBIF) has introduced several measures to bring legislation and prudential regulation up to international standards but the sector needs to be strengthened and additional legislation introduced (e.g. deposit insurance and bank resolution procedures, registry of appraisers and external auditors, etc.). CAPITAL MARKETS. Money market funds, leasing, commercial paper, private placement and securitization by non-bank issuers are practically non-existent in Nicaragua. The market, composed of the Managua Stock Exchange and 13 brokerage houses is highly concentrated (four brokerage houses account for 66% of traded volume in 1998). The Exchange largely trades CBN and Government paper (mainly CENIS and Compensation Bonds), with only a small share of fixed income private issues. Market activity increased significantly between 1995 and 1998 with trading growing by 46% per year on average and a total volume traded of 6.1 billion C6rdobas (approximately US$500 million or 30% of GDP) during 1998 alone. Trading and brokerage activities are self-regulated by the stock exchange, exposing it to insider- dealing and fraudulent practices. If SBIF is to retain the mandate to regulate and supervise the capital markets, its capacity should be strengthened and the legal and regulatory framework improved. INSURANCE INDUSTRY. The insurance industry in Nicaragua is underdeveloped but is in a state of transition. The insurance state monopoly (INISER) was abolished in 1995 and, since then four new insurance companies have been established with considerable success. Total insurance premiums have increased but Nicaragua still remains one of the smallest insurance markets in Latin America, with approximately 0.5% of the market in the region. The bulk of the insurance business is in the various property/casualty lines (particularly auto insurance) and very little is life insurance. In general, the regulatory framework is adequate, with minimum capital requirements and solvency standards at reasonable levels and only minor improvements needed (e.g. financial analysis and monitoring, examinations, etc.). Page 10 April 2000 IV. LONG TERM FINANCES AND THE RATIONALE FOR REFORM 18. While the short term viability of the system is already compromised, its long term viability is even more in doubt. The system's reserves would be depleted in the immediate future and its dependency ratio (the number of benefit recipients supported by each active worker) would increase from 11.2% in 1996 to 28.6% in 2030 and to 69.3% by 2060. At the same time, the Implicit Pension Debt (IPD - the present value of the lifetime pension promises the Government has made to current retirees, plus accrued rights of current workers) of Nicaragua's PAYG system would rapidly rise from about 84.9% of GDP in 1996 to over 400% by the year 2030 (Figure 4). While Nicaragua is still among nations with a low to moderate IPD (see Table 6), it would rapidly surpass most of them if the system is not reformed in the short term. Nicaragua has a fairly young pension system and, in absolute terms, its IPD is comparable to countries such as China, Colombia and Mexico, where the IPD is roughly of the same magnitude. However, given the imbalances between contributions and benefits under the current system, it is inevitable that the IPD could double and rise to levels that could exceed 150% of GDP as evidenced by countries like France, Germany and other OECD Countries. Figure 4. IPD as % of GDP 600% 500%/a 400% 300% 200% 100% 0% el99 epg wob eoOw@o o ow° ttN i 19. Nicaragua's pension system is not sustainable due to the imbalances between benefits and contributions, high administrative costs and high rates of evasion. At the same time, the system is not attractive to workers because: (i) despite the disassociation between contributions and benefits, the current pensions they are eligible to receive are still very low"2; (ii) the system perpetuates inter- and intra-generational inequalities"3; (iii) the system does not allow workers to engender a sense of ownership of their "savings", instead, they perceive their contributions as a tax; and (iv) the institutional weaknesses of INSS lead to poor performance with respect to record keeping, investrnents, and the pension calculation and payment of pension benefits. 12 No indexation of average weekly salaries is used for the calculation of the pension. 3 Inter-generational inequalities allow current generations to receive higher rates of return while future generations receive lower rates of return. Intra-generational inequalities allow higher income groups within the same generation (and women, in most cases) to receive higher rates of return. Nicaragua: Pension Reform Proposal Page 11 Table 6. IPD as % of GDP in Selected Countries Latin America IPD as % of GDP Year Bolivia 40 1994 Brazil (RGPS) 175 1997 Chile 126 1981 Colombia 87 1994 Mexico 73 1994 Nicaragua 84.9 1996 Peru 37 Early 90's Uruguay 214 Early 90's Venezuela 30 Early 90's Africa Cameroon 44 Early 90's Senegal 27 Early 90's Asia _ China 63 Early 90's Philippines 43 1995 Europe Croatia 350 Early 90's France 216 1990 Germany 157 1990 Hungary 144 1997 Ukraine 141 Early 90's Turkey 72 Early 90's Sources: Queisser, M (1998); World Bank (1994); INSS Department of Planning and staff research. Figure 5. Elderly Dependency Ratio 20. Given the system's weaknesses and Nicaragua's 14,000 35.0% relatively young population 12,000 30.0% structure (in 1996 only 5% of A 10,000 25.0% the population was over 60), 0 .0% current low old age 8,000 0 dependency ratio (Figure 5, i 600- -150%X rtoof retirees to workers), ,0 00 and low coverage, the time is 2,000 - 5.0% propitious for the Government to reform the -~~~~~~~~~~~~~.0% Government to reform the 1996 2000 2010 2020 2030 2040 2050 2060 system at a reasonable cost and make it more equitable ITOtsI Popuia inPopulation aVer 60 OIdiWOkIn1 and sustainable in the long run. Source: PROST Page 12 April 2000 V. THE REFORM PLAN 21. Reforming the PAYG systems The initial steps toward a sustainable pension system in Nicaragua should include the re-definition of its parameters to avoid any further deterioration of INSS finances and the progressive downsizing of the current PAYG system. A parametric reform can be designed to increase contribution rates; raise the retirement age; standardize eligibility requirements for all workers; reduce replacement rates; increase collection efficiency; tighten eligibility requirements for disability benefits; change the reference wage; and penalize early retirement. At the same time and/or shortly after the introduction of the parametric changes, a deeper structural reform of the system toward a defined contribution system, should be undertaken. Such a measure will consolidate the system and ensure its long term viability and sustainability. 22. Evaluating parametric reform options for Nicaragua. Various proposals have been discussed with the Govemment over the last two years. These include increasing contributions, increasing the retirement age, increasing the vesting period and reducing the replacement rate. Several simulations have been carried out, assuming that the reform is undertaken in the year 2000. Box 2 shows the parametric reform measures used for the simulations in Annex I, while Table 7 and Figure 6 show a comparison of the results against the no reform scenario (Base Case). Box 2. Main Parametric Reform Measures Used for Simulations in Annex I * Raise contribution rate from 5.5% to 10%14 immediately starting in 2000 * Raise retirement age from 60 to 65 years gradually between 2000 and 2006 * Increase vesting period from 15 to 25 years, gradually between 2000 and 2020 * Increase wage base for the calculation of benefits from 5 to 10 years Table 7. Comparison of Results between Base Case and Parametric Reform Year of Fund Required PA YG IPD as % of GDP Deficit Balance = 0 Contribution Rate (°/) 2030 2060 1996 2030 2060 Base Case 1997 2000 17.6 36.4 84.9 408.4 432.8 [ P. Reform 2040 2048 6.7 20.5 63.2 250.4 255.6 Source: PROST 4 The real increase in contributions is only 4% since it is contemplated that 0.5% will come from the workers compensation insurance premium currently paid to INSS. Nicaragua: Pension Reform Proposal Page 13 Figure 6. Comparison of Results between Base Case and Parametric Reforms (Current Balance Comparison as % of Nominal GDP) 5% 0% = ,,E... -10%- Base Case -10% Reform -15% - Scenario ---.-.-. -20% 23. As observed in Figure 6, a parametric reform can only alleviate the short- to medium-term finances of the current system and moderate the cost of transition to the new system. In order to make substantial gains while retaining the PAYG structure, INSS would have to double its collection efficiency (which is unlikely given the current institutional capacity). Nevertheless, if the system is to be sustainable in the long run and the problems and distortions mentioned earlier are to be avoided, it is indispensable to introduce a structural reform. For this reason, our proposal builds on a parametric reform to create a defined contribution system based on individual capitalization accounts that should be introduced simultaneously with the parametric changes (Box 3). 24. Creating a mandatory, fully-funded, defined-contribution system based on individual capitalization accounts. The publicly managed PAYG pension system is replaced by a mandatory, fully-funded defined contribution system based on individual capitalization accounts administered by privately owned pension fund management companies. In such a system, the role of the government is reduced from directly providing old age security to regulating, supervising and (depending on national preferences), guaranteeing a minimum provision of old age security via minimum pension guarantees or a social assistance pension (Annex I) through direct budgetary transfers. 25. The system should be mandatory because people do not save enough on their own for retirement and may become a burden on the government and society at large. By forcing people to save on a regular basis the risk of facing the "old age crisis" diminishes"5. 5 From James, E (1996) "New System for Old Age Security: Why, How and So What?" paper presented at the "Pension Systems: From Crisis to Reform: conference organized by EDI, November 21-22, 1996. Page 14 April 2000 Box 3. Main Structural Reform Measures used for Simulations in Annex I * The reform starts in the year 2000 and includes all the parametric changes described in Box 2 and Annex I. * The PAYG system is closed to new entrants. Current pensioners and those older than 45 years of age remain in the PAYG system. Switch to the Defined Contribution System (DC) is mandatory for new entrants and those younger than 45 years of age. Switch is optional for those 45 to 50 years of agel6. * Accrued rights to those moving to the DC system are paid at retirement through recognition bonds. * From total contributions to the DC system of 10% (6% from the employer, 3.5% from the employee and 0.5% from the workers compensation insurance currently paid to 1NSS), 7 percentage points will go to the individual capitalization accounts and 3 percentage points will be paid as fees to cover administrative costs and other insurances. After the third year of operation of the new system, this fee will not exceed 2.5%. The difference (minimum 0.5%) will be deposited to the individual accounts. 26. The system should be fully funded to avoid future tax increases, benefit cuts or retirement age increases. It also prevents inadvertently large inter-generational transfers from young people to older workers. Finally, full funding helps build a stock of long-term national savings and helps finance future pensions out of the returns to savings. 27. The system should be based on defined contributions because linkages with benefits are much clearer. This will help reduce or discourage evasion, informality, early retirement and other labor market distortions, since contributions are no longer perceived as a tax. 28. The system should be privately managed to maximize the likelihood that economic rather than political objectives will determine the investment strategy, thereby producing a better allocation of capital and the highest risk adjusted return on savings. Pension funds managed by governments are likely to be poorly invested due to multiple objectives. In contrast, pension funds managed privately are more likely to be invested more broadly and generate better returns (see para. 49, pg. 20). 16 Establishing a specific age to switch between systems gives more certainty when calculating the cost of the transition. Allowing people of certain ages to choose between systems avoids disputes with those just below or above (within months or days) the specified age, but makes it more difficult (uncertain) to calculate the transition cost since it is not known how many will choose each of the two systems. Nicaragua. Pension Reform Proposal Page 15 29. The design and implementation of such a system requires the following considerations to be taken into account: Design Issues 30. Should the current PAYG system be fully or partially closed? If the PAYG system is partially closed, no new entrants are allowed, while some workers (e.g. certain age cohorts) are forced to remain in the old system or are given the option to switch to the new system. In either case, the PAYG system is modified and allowed to coexist with the new system or phased out over time. Partial closure is recommended for countries such as Nicaragua, that do not have a pre-existing surplus (in the general treasury or the social security system) or any other source of funds that can be used to finance part of the pension debt (see how to finance the transition below). 31. If sources of fimds can be identified, Nicaragua would be better off using those resources to fully close the PAYG system. Under full closure, all contributors are forced to switch to the new system and a mechanism is established to recognize acquired rights (Annex I). 32. Given the current precarious financial situation of the PAYG system in Nicaragua, it is advisable to partially close the current system. Closing the PAYG system for new entrants and forcing those younger than 45 to switch to a new system, will allow the Government to gradually close the PAYG system while the new system based on Defined Contributions will take root and develop as the provider of old age income. 33. How to recognize acquired rights? Regardless of the way the PAYG system is closed, the INSS has to recognize acquired rights to those already participating in the old regime that will be forced to move to the new system. Acquired rights can be recognized through a recognition bond (as in Chile), through a compensatory pension (as in Argentina and Bolivia) or with a lifetime switch option (as in Mexico). The recognition of acquired rights could be a difficult task for Nicaragua, particularly in the absence of sufficient pension fund reserves, and a lack of reliable data to reconstruct employee records (working/contribution history). 34. Issuing recognition bonds, places an explicit value on the debt owed to each worker but it is not met until workers retire. The value of the recognition bond equals the present value of a pension replacing a certain percentage of earnings (replacement rate) for a full contribution period. The value of the bond can be adjusted to inflation; earn interest; have a ceiling; and/or require a minimum period of contributions (Box 4). Payment, in the absence of pension reserves, should be made directly by the Government through reserves, general revenue, and/or other forms of revenue or taxation (see how to finance the transition below). Bonds are disbursed at retirement and can be paid through a lump sum or a monthly supplement to other pensions. By imposing restrictions on the payment of recognition bonds governments can reduce the fiscal cost of the transition. Page 16 April 2000 Box 4. Recognition Bonds: The Chilean Case The formula used in Chile for the calculation of the recognition bond equals the present value of a pension with an 80% replacement rate for a full contribution period. The bonds are adjusted for inflation and carry a 4% annual real rate of return. 35. A compensatory pension is a promise to gradually pay the debt owed to each worker for his/her previous contributions through monthly payments at retirement. The compensation is calculated based on the number of years contributed to the old system and a pre-defined replacement rate and a salary of reference contributions (Box 5). Payment should be made directly by the government through current contributions (if a public pillar is retained) and/or budgetary transfers. Box 5. Compensatory Pension: The Bolivian Case17 The formnula used in Bolivia for the calculation of the compensatory pension is equal to the actual years of contributions x 70% replacement rate! required years of contribution x last salary under old pension system. Both the reference salary and the compensatory pension are subject to a ceiling of 20 times the minimum wage. The compensation amount is indexed in US dollars. 36. Both of these methods (recognition bond and compensatory pension) acknowledge the debt owed to each worker, but postpone the day when cash will be needed. The difference is that while the recognition bond creates economic certainty and brings transparency to the process by making the pension debt explicit and legally binding, the compensatory pension leaves the pension debt implicit and vulnerable to political manipulation (it remains as a promise that future governments can choose not to fulfill or could easily modify). 37. A third alternative is the lifetime switch option implemented in Mexico where transition workers compare their benefits under the old and the new system and choose the most advantageous option upon retirement. If they opt for the old system, the accumulated balance in their individual accounts is turned over to the public system and the government assumes the responsibility to pay the benefits according to the formula used in the old system. Although this option seems to have the lowest cost in the short run (since no real recognition of acquired rights takes place), it is relatively uncertain in the long run. The cost will be known when workers reach the retirement age 17 von Gersdorff, H. (1997). Nicaragua: Pension Reform Proposal Page 17 (comprehensive analytical work is required to minimize the long term uncertainty and determine the cost of this option). Also, it has been pointed out that the broad guarantee represented by Mexico's life-time switch option introduces various sources of moral hazard for all actors in the new system. 38. Evidence suggests that the cost of the first two options can be further minimized by severely discounting the recognition of acquired rights (as mentioned above). This is possible (particularly in the recognition bond option where uncertainty is reduced through a legally binding document), because the willingness of younger generations to switch to the new system is inelastic (regardless of acquired rights recognized). Given these alternatives and the need to make the new system as transparent and free from political manipulation as possible, the best alternative for Nicaragua is to issue recognition bonds. 39. How to finance the transition? If all or part of the payroll tax is diverted to funded individual accounts, it could become difficult for the Government to pay current pensioners and old workers under the old system given the current level of reserves at INSS. Before making the transition, every effort should be made to reduce the IPD (through the parametric reform and the restrictions imposed on the recognition of acquired rights mentioned above), build up reserves and/or establish alternative funding mechanisms for the payment of current pensioners and those workers remaining under the old system. 40. While parametric changes are the first step, the following alternatives should also be evaluated, within the context of Nicaragua's overall stabilization program and financial constraints: * Penalize early retirement, and enforce tight eligibility for disability benefits. * Increase the system's revenues by reducing evasion and increasing collections. * Improve investment strategy for pension fund reserves at INSS and use them to pay off part of the pension debt. * Use privatization proceeds. * Reduce Government spending or increase other taxes and make direct allocations to cover part of the pension debt. * Issue general treasury debt and/or borrow from international organizations to cover the remaining financing gap"8. 18 Given that Nicaragua is a heavily indebted country, and is following a very strict fiscal policy to be eligible to receive debt relief (HIPC), the Government does not consider this alternative an option. Page 18 April 2000 Implementation Issues 41. How to administer the new system? Given the state of development of Nicaragua's financial sector and its weak regulation and supervisory capacity, privately owned pension fund management companies, devoted exclusively to this business, should administer the pension funds. The options to consider are whether to limit the number of fund managers (as in Bolivia) or leave it to the market to determine the number (as in Chile, Mexico and El Salvador). Since there is a positive relationship between the size of the market (number of insured workers) and the number of fund managers, Nicaragua (with approximately 200,000 participants expected to join the new pension system) should consider limiting the number of fund managers to 2 or 3 through an international competitive bidding process that could lower the cost of the system (see below) but allows competition and free choice by workers. 42. In such a system, licenses to fund managers are auctioned and awarded based on the lowest fees by bidders that have been previously certified on the basis of their past experience and performance as well as compliance with minimum capital requirements. Managers might forn joint ventures with foreign financial institutions that could bring foreign expertise but, in no case should the still weak local banking sector be allowed to manage pension funds on its own (Box 1). 43. After an initial period of 3 to 5 years and once the number of contributors increases to approximately 250,000, the Government could evaluate the possibility of granting an additional license (following the samne international competitive bidding process). Such a decision could help bring administrative costs even lower and increase the choices available to participants. However, for such a mechanism to function properly and to avoid high marketing costs, switching should be limited and allowed only under very strict conditions. 44. How to regulate and supervise the new system? The need to strengthen and modernize the current Superintendency of Banks and Other Financial Intermediaries- SBIF (Box 1) along with the absence of reputable international pension fund administrators could be a barrier to the healthy start up of the new privately managed pension system. For these reasons, Nicaragua should create an independent Superintendency of Pensions in charge of regulating and supervising the new pension system. The compulsory participation of all workers, the specialization of products involved in the new system and the expectation that private pension funds will perform better than the PAYG systems, are additional arguments in favor of an independent Superintendency of Pensions"9. The new Superintendency could be auto-financed through regulatory fees (charged to pension fund administrators) and most have the latest technological capabilities to ensure that pension fund managers are complying with all the regulations. 19 Demarco, G. and Rofinan, R. (1998). Nicaragua: Pension Reform Proposal Page 19 45. How is the fee structure established? How can administrative costs be contained? The fee structure will depend on the administrative structure chosen and the regulations established by the Superintendency of Pensions. If an international competitive bidding takes place, fees are determined by the lowest bids, while in a free entrance structure fees are determined by each fund manager and used as a competitive tool to gain market share. In any case, a ceiling could be established by the Superintendency of Pensions to maintain administrative costs within a certain range. 46. In a competitive market, fees can be based on new contributions (one-time, front- loaded), on assets (monthly or annually) on returns (monthly or annually) or on a combination of the three. Fees can also be charged as a flat rate (fixed amount). 47. Administrative costs20 are the major factor in determining which commission structure should be used. This is especially true in small countries such as Nicaragua, given the economies of scale involved in pension finance. For this reason, Nicaragua should carefully define (and/or regulate) start up and administrative costs (particularly marketing and sales commissions) to avoid having an overly costly system (high fees) for the majority of workers. However, regulation should neither restrict nor discourage competition in the long run. 48. While limiting the number of licenses in the short run (as described above) could help Nicaragua achieve economies of scale, there are other measures that might be considered and may lead to lower administrative costs, including: * Centralizing collections of contributions. * Separating record keeping from investment activities into two different entities. An international bidding process, such as the one mentioned above, should take place to select one company to do the administration of accounts (record keeping) and allow for several pension find managers (with tight eligibility requirements or through a bidding process) to participate as pension fund managers (investing). * Permitting pension funds to have lower fees for those who remain with the funds for long periods of time (to reduce switching after the market is open for competition). * Allowing sales commissions to be earned only on new affiliates to the system. 20 Costs typically associated with the mandatory, defined contribution system based on individual capitalization accounts are: start-up costs (fixed costs associated with setting up the new system in the first years of operation); collection and record keeping costs (to ensure that contributions are made and workers are kept informed); investment expenses (costs associated with investment research, brokerage fees and profits for money managers); and marketing expenses and agent commissions. Page 20 April 2000 49. How are the funds going to be invested? (The Investment Regime) Privately managed pension funds should not be subject to the same restrictions imposed on publicly managed pension funds. While regulation is necessary, it should be flexible enough to allow investments in a mix of domestic and international currencies and capital market instruments. This will facilitate obtaining the highest possible rates of return and allow for a proper diversification of the portfolio, while protecting funds from local economic and political manipulation. 50. In Nicaragua, the money and capital markets are underdeveloped. While pension reform can stimulate their development, this should not be seen as a reason for restricting investment to the local markets. Nicaragua could initially invest locally in Government bonds, certificates of deposits (CDs) from commercial banks and private sector bonds2' that comply with adequate risk ratings, to allow pension fund managers to take advantage of existing high real interest rates22. Nevertheless, international investment should be allowed to further increase diversification and protect funds from economic, political and country risks. 51. The investment regime should include, not only the classes of instruments and domestic/international diversification allowance, but also clear rules for classification of risks, rating of instruments, mechanisms for valuing securities, custodial arrangements, resolutions for conflict of interests and limitations and/or prohibitions for related party transactions. Other Issues 52. Disability and Survivor Insurance. With a defined contribution system, the two available options to administer disability and survivor insurances are either leaving them under INSS control or acquiring a policy with an insurance company. Since INSS role in the reform will most likely be limited to the collection and transfer of contributions, the best and most economic alternative in Nicaragua is to acquire a policy with an insurance company. However, it should be noted that international competition in this market should be allowed given the infancy state of the insurance sector and the lack of satisfactory supervision over this market. 53. Minimum Pension Guarantee or Social Assistance Pension. Given the current Constitutional obligation by the Government to provide social security to the population as a whole, it is necessary for the Government to participate in the new pension system as a guarantor. This safety net, which aims at reducing old age poverty, can be provided 21 Goverrnent debt is almost all indexed to the US dollar. Approximately 70% of it is long term, including an important percentage of securities with a maturity of 15 years. Private sector bonds are limited but also offer long term indexed instruments (367 to 720 days). 22 Real interest rates in Nicaragua fluctuate between 8 to 12%. Nicaragua: Pension Reform Proposal Page 21 either through a minimum pension guarantee (as in Chile) or a flat benefit per year (social assistance pension). Financing should come from general tax revenues or through a redistribution mechanism included in the reform package (a redistributive pillar). 54. Given the alternatives and costs associated with each, we consider that Nicaragua will be better off adopting a minimum pension guarantee targeted mainly to low income workers. Benefits and eligibility requirements should be established up-front to avoid being overly generous and discouraging workers from participating in the pension system. Otherwise, the guarantee could become too onerous for the Government in terms of tax revenues or contributions required to fund it. 55. A minimum pension guarantee, could be given to those workers who have contributed to the system for a period of at least 25 years and are eligible for retirement but whose accumulated funds are not adequate to provide them with an income equal to a pre-defined minimum pension. In this case, the Government would only fund the difference between the accumulated balance in the individual account and the minimum pension. 56. Collection and Transfer of Contributions. The collection of contributions can be centralized or decentralized. In the first case, a single agency (INSS) could be responsible for collecting contributions and distributing them to the different agencies, funds and/or individual accounts. In the second case, each agency or fund administrator could make its own collection and credit the correspondent individual account. In either case, the process is divided into two independent flows. One that includes all financial transfers and another that includes the information transfers. Of course, both flows should be identical in order to guarantee an efficient overall process23. 57. In choosing the more suitable alternative, the following aspects should be taken into consideration: * Efficiency: in a defined contribution system based on privately managed individual accounts, the rapid and safe collection and transfer of funds is the primary responsibility of the collection agency. * Security: while funds should be rapidly and safely collected and transferred, control mechanisms are necessary to ensure the precision of such transfers. Clear regulation, cross controls and enforcement power should be carefully analyzed. * Cost: economies of scale versus market competition should be considered when defining the most adequate collection system, especially if contributors would bear the cost. 23 Rofinan, R. and Demarco, G. (1998). Page 22 April 2000 58. Based on the above, and given the limited coverage that most financial institutions have in Nicaragua, it is recommended that, at least initially, the collection of contributions continues to be centralized by INSS, while allowing pension fund administrators to set up their own systems if they decide to do so in the future. 59. Communication and publicity campaign. A well structured communication and publicity campaign should be planned to create consensus about the need for reform, explain the changes and help people make informed decisions during the implementation stage. It should provide clear and truthful information in the most dynamic way and cover every single group that will be affected by the changes (current and future employers and employees, independent workers, etc.). 60. The general campaign should use mass media (television, newspapers, radio, magazines, etc,) but seminars, booklets, etc. should be designed specifically for certain groups or regions. The objective being to create confidence and trust in the new system. 61. Regulations required to implement the reformed systemn The Independent Superintendency of Pensions should draft and issue, among other regulations, the following: * entry/exit regulations * disclosure requirements * capitalization requirements * marketing * commission and fee structure * investment restrictions, including foreign investment VI. THE NEW SYSTEM AND THE NECESSARY FINANCIAL SECTOR ARRANGEMENTS 62. A major caveat in implementing a mandatory, fully funded, defined contribution system based on individual capitalization accounts is that financial markets must be developed sufficiently to manage the funds safely and with diversified investment instruments (James, 1996). In Nicaragua, financial markets are not well developed and few financial instruments are available. This, however, does not mean Nicaragua should postpone the implementation of its reform; on the contrary, pension reform and financial market reform should go hand in hand and the discussion of the financial requisites for pension reform (Annex II) incorporated into the broader analysis of overall financial sector development. Nicaragua: Pension Reform Proposal Page 23 63. Given Nicaragua's financial sector conditions, if private management of pension funds is to be introduced, the following caveats and guidelines are especially important as deeper financial sector reforms are simultaneously undertaken: * The funds and securities should be placed in a safe custodial arrangement (e.g. Central Nicaraguiense de Valores, S.A). * Banks should be required to meet stringent solvency requirements before they are eligible to serve as custodians or allowed to form partnerships with international firms to manage pension funds. * International institutions should be encouraged to participate as they bring in capital and expertise for the administration and investment of pension resources. * International diversification of investment should be allowed to diversify financial risk and guard funds from local economic and political manipulation. * The new independent Superintendency of Pensions should be in charge of auctioning pension fund management licenses, defining the bidding procedures and reviewing applications. It should also be in charge of enforcing minimum capital requirements for pension funds; ensuring compliance with investment limits and international diversification guidelines; making and/or reviewing audit reports; ensuring information disclosure; preventing fraud; and solving misrepresentations and conflicts of interest. * Competition (both domestic and foreign) in the insurance industry should be encouraged. The insurance industry should be developed further to meet the growing demand for annuities, as well as managing survivors' and disability insurances. Page 24 April 2000 BIBLIOGRAPHY Arrau, P., Valdes-Prieto, S., and Schmidt Hebbel, K. (1993), "Privately Managed Pension Systems Design Issues and the Chilean Experience", manuscript, April. Bustamante, J. (1999), "Disefno Global del Sistema de Pensiones para Nicaragua", Proyecto de Reforma al Sistema de Pensiones de Nicaragua, Managua, Nicaragua: March. Cifuentes, R., and Larrain, F. (1997), "Pension Systems in Central America: An Assessment of the Current Status", HIID Development Discussion Paper No. 2, Central American Project Series, Harvard, MA.: September. Demarco, G., and Rofman, R. (1998), "Supervising Mandatory Funded-Pension Systems: Issues and Challenges", second draft, paper prepared for the World Bank's Pension Primer Series, Washington D.C.: August. Demirguc-Kunt, A. and Schwarz, A. (1996), "Taking Stock of Pension Reforms Around the World", paper presented at the Conference Pension Systems: From Crisis to Reform organized by EDI, November 21-22, 1996. Grandolini, G. and Cerda, L. (1998), "The 1997 Pension Reform in Mexico", Policy Research Working Paper No. 1933, The World Bank, Washington D.C.: March. Holzmann, R. (1997), "On the Economic Benefits and Fiscal Requirements of Moving from Unfunded to Funded Pensions", AICGS Research Report No. 4, The Johns Hopkins University. James, E. (1996), "New System for Old Age Security: Why, How, and So What?", paper presented at the Conference Pension Systems: From Crisis to Reform organized by EDI, November 21-22, 1996. Kane, C. and Palacios, R. (1996), "The Implicit Pension Debt", Finance and Development, June. Melinsky, E.. (1999), "Instituto Nicaragiuense de Seguridad Social, Regimen de Invalidez, Vejez y Muerte, Valuacion Actuarial al 31/12/99", Melinsky, Pellegrinelli y Asoc., Managua, Nicaragua: June. Queisser, M. (1997), "Pension Refonn and Private Pension Funds in Peru and Colombia", Policy Research Working Paper No. 1853, The World Bank, Washington, D.C.: November. Queisser, M. (1998), "The Second-Generation Pension Reforms in Latin America", Development Centre Studies, OECD, Paris: April. Nicaragua: Pension Reform Proposal Page 25 Rizo, S. (1996), "Reforma de la Seguridad Social en Nicaragua", INSS, Managua, Nicaragua. Rizo, S. (1996), "Seguridad Social para el Siglo XXI", INSS, Managua, Nicaragua. Rofman, R., and Demarco, G., (1998), "Collecting and Transferring Contributions in Multipillar Pension Schemes", preliminary draft, paper prepared for the World Bank's Pension Primer Series, Washington D.C.: June. Schultess, W. and Demarco, G. (1996), "El Sistema de Jubilaciones y Pensiones de Argentina a Dos Afios de la Reforma", paper presented at the Seminar on Social Security Reform organized by the CEPAL, October 7-8, 1996. Shah, H. (1997), "Toward Better Regulation of Private Pension Funds", Policy Research Working Paper No. 1791, The World Bank, Washington, D.C.: June. Siles, C. (1999), "El Modelo del Cambio", Comisi6n para la Reforma de Pensiones de Nicaragua (CREPEN), INSS, Managua, Nicaragua: May. Singh, A. (1996), "Pension Reform, the Stock Market, Capital Formation and Economic Growth: A Critical Commentary on the World Bank's Proposals", International Social Security Review, Vol. 49, 3/96 (21). Vittas, D. (1992), "Contractual Savings and Emerging Securities Markets" Policy Research Working Paper No. 858, The World Bank, Washington D.C.: February Vittas, D. (1996a), "Sequencing Social Security, Pension, and Insurance Reform", Policy Research Working Paper No. 1551, The World Bank, Washington D.C.: December. Vittas, D. (1996b), "Pension Funds and Capital Markets", FSD Note No. 71, The World Bank, Washington D.C.: February. Vittas, D. (1998b), "Regulatory Controversies of Private Pension Funds", Policy Research Working Paper No. 1893, The World Bank, Washington, D.C.: March. Vogel, A. (1998), "Reforma al Sistema de Pensiones en Nicaragua", INSS, Managua, Nicaragua.: December Von Gersdorff, H. (1997), "Pension Reform in Bolivia: Innovative Solutions to Common Problems", Policy Research Working Paper No. 1832, The World Bank, Washington, D.C.: September. World Bank (1994), "Averting the Old Age Crisis", World Bank Policy Research Report, Oxford University Press, Inc., New York. Page 26 April 2000 Nicaragua. Pension Reform Proposal Paze A] Annex I - Simulations: Technical Notes INTRODUCTION The simulations for this report were based on PROST, version 8, the toolkit developed by the World Bank, HDNSP, for analyzing pension reform. This model has been developed in VBA and can be run with Windows 95/98/NT using Office 97 or later versions. Like any simulation model, the outcome from PROST depends largely on the nature and quality of data as well as on the set of assumptions being used for the simulations. Ideally, quantitative analyses should be robust to a wide range of economic scenarios and developments. This annex summarizes some of the key assumptions for the Nicaragua Pension Reform simulations. The individual input files are available upon request. It should be noted, that the nature and quality of data available currently imply that the output from PROST should be viewed as a diagnostic guide rather than as a numeric prescriptive tool. However, despite data limitations, two critical messages emerge from the simulations: * JUnder present parameters, the INSS pension arrangements are not financially sustainable. Current expenditures are already in excess of revenues. The shortfall is expected to exceed 10% of GDP in the next 30 years. Thus the young sons and daughters of current workers in Nicaragua will have to shoulder the burden of an unsustainable pension system unless reform measures are implemented quickly within a fleeting window of opportunity. * Parametric reform provides short term financial relief but does nothing to improve the long run financial viability of the system. Even the aggressive reform exercises being simulated do not improve the long-term sustainability of the system. The Government should seriously consider systemic reform, keeping in mind that coverage expansion is likely to be slow and there would probably be need for social assistance programs to individuals who are unable to come under the scope of any kind of pension regime in Nicaragua. BASE CASE ASSUMPTIONS The start of the simulation horizon was 1996, the year for which the most complete data and documentation exist' for PROST. More recently, newer data have been made available but the quality of these data have not been evaluated by Bank staff. The end year for the simulation horizon was 2060. Although it is possible to select longer simulation horizons, the horizon chosen is long enough to demonstrate the character of the pension system in Nicaragua and how it would behave under different reforms and scenarios. ' These data were compiled by Juergen Boehmer (CREPEN) and Aniruddha Bonnerjee (HDNS) Page A2 April 2000 * Population and Mortality Demographic information are from INEC. Our assumptions for mortality improvements in the future are easily translated into improvements in life expectancies according to Table 1. Fertility is assumed to decline from about 5.05 children per woman of child bearing age to about 2.25 by 2025 and then to 2.00 by the year 2030. This set of assumptions generates population projections that are comparable to independent published population projections2. Towards the end of the simulation horizon, an adult male retiree would expect to live for about 23 years more while a female would expect to survive another 26 years. Thus future contributors would have to support the elderly for a longer duration of time on the average. Table 1. Mortality Assumptions 1996 2000 2005 2010 2015 2020 2030 2040 2050 2060 Male Life Expectancy: At Birth 64.6 65.0 65.4 65.9 66.4 66.9 67.9 69.4 70.9 72.6 At Age 20 52.2 52.5 52.8 53.1 53.5 53.9 54.6 55.7 56.9 58.3 At Age 60 18.3 18.5 18.7 19.0 19.2 19.5 20.1 21.0 22.0 23.2 At Age 65 14.8 14.9 15.2 15.4 15.7 16.0 16.5 17.4 18.4 19.5 Female Life Expectancy: At Birth 68.6 69.0 69.4 69.9 70.4 70.9 72.0 73.5 75.2 77.0 At Age 20 55.4 55.7 56.1 56.5 56.8 57.2 58.1 59.3 60.6 62.2 At Age 60 20.8 21.0 21.3 21.6 21.9 22.2 22.9 23.9 25.1 26.4 At Age 65 17.1 17.3 17.5 17.8 18.1 18.5 19.1 20.1 21.3 22.6 Figure 1. Population Pyramid 1996: Eg: World Population Projections, 1995. Bos, Vu, Massiah and Bulatao. Nicaragua: Pension Reform Proposal Page A3 Estimated population projections and the implied dependency rates are shown in Table 2. As can be seen, demographic developments are going to increase the population dependency rates almost three times by the end of the simulation horizon. Although Nicaragua does not have a very old population structure at present, our simulations indicate an aging process that would gather momentum after a few decades. Table 2. Population Projections Population 1996 2000 2010 2020 2030 2040 2050 2060 Total Population 4,491 5,102 6,791 8,467 9,934 11,082 12,112 12,836 Male 2,210 2,518 3,371 4,217 4,955 5,533 6,051 6,409 Female 2,281 2,584 3,420 4,251 4,979 5,549 6,061 6,428 Pop. Dependency Rate 100.7% 92.4% 84.2% 71.0% 59.0% 52.8% 57.7% 62.8% Age 0-14 / Age 15-Ret. Age 90.4% 83.0% 75.0% 60.3% 46.2% 36.1% 34.1% 31.8% Ret. Age + / Age 15-Ret. Age 10.2% 9.4% 9.2% 10.7% 12.9% 16.7% 23.6% 31.0% Support Ratio: Age 15-Ret. Age/ 9.8 10.6 10.8 9.3 7.8 6.0 4.2 3.2 Ret. Age + Total Fertility Rate 5.04 4.67 3.84 2.98 2.00 2.00 2.00 2.00 * Contributors and Retirement PROST simulations assume that coverage rates in the future would rise from roughly 15-20% to about 50% for males and 40% for females. It is assumed that increasing labor force participation (especially of women), expansion of coverage to other sectors and better administration (hence less evasion) would increase coverage. Our assumption for coverage expansion is based on the hypothesis that by the year 2010, the younger cohorts, say in the age group 30-39, would hit a coverage rate of 50% for males and 40% for females. This assumption implies that by 2030 or so, 50% of males and 40% of females in the age group 30-59 would be covered. Compared to the base year data, our coverage expansion path implies that coverage will increase roughly 1.5 times for males and about 2 times for females by 2030. Given the pace of development in Nicaragua, this assumption could become a likely scenario. This coverage expansion will have a direct impact on retirement rates so that by the year 2031, retirement rates will also approach 50% and 40%. Due to the coverage expansion, for each age cohort, more individuals will be exposed to the probability of becoming disabled. We assume that disability rates will also rise proportionately with coverage rates since more individuals are now exposed to the probability of becoming disabled. We have not assumed any changes to survivor rates because of the lack of data on the processes generating such pensioners and the inherent difficulty in modeling such behavior. Page A4 April 2000 Table 3. Demographics of The Pension System under Base Case 1996 2000 2005 2010 2015 2020 2030 2040 2050 2060 Total 412.6 537.3 711.6 933.5 1,156.7 1,394.8 1,899.8 2,347.7 2,630.4 2,741.5 Contributors Total 46.2 69.2 98.6 138.7 195.7 282.0 542.9 904.6 1,394.7 1,901.0 Beneficiaries Old Age 21.7 38.3 60.8 92.0 137.5 211.7 442.6 753.8 1,197.3 1,647.7 Invalidity 9.1 13.1 17.3 22.8 29.9 37.2 56.6 94.7 125.8 163.7 Widows 7.8 9.4 11.2 13.5 16.6 20.3 30.4 43.7 60.3 78.1 Orphans 7.7 8.5 9.3 10.3 11.7 12.7 13.2 12.4 11.4 11.6 The system dependency ratio is expected to rise from about 10 to 11 per 100 contributors to about 70 per 100 workers by the end of the simulation horizon, an increase of roughly 5 to7 times. This figure is significantly higher than the population dependency rates because of low coverage, evasion and early retirement. To support this growing dependency rate, future payroll contribution rates would have to be raised or other parametric measures would need to be implemented. * Economic and other assumptions The macroeconomic assumptions are fairly standard and in accordance with Central Bank forecasts in Nicaragua. In the base year, GDP was estimated to have been roughly 16,623 million C6rdobas. Other assumptions are summarized below: Table 4. Macroeconomic Trends (%) Macroeconomic Trends 1996 2000 2005 2060 Real GDP Growth 4.7 6.5 5.8 5.0 Real Wage Growth of 20-year Old Males 2.0 2.0 2.0 2.0 Inflation Rate 12.1 8.0 3.0 3.0 Real Interest Rate 16.0 10.0 8.0 5.0 Minimum Wage as % of Average Wage 25.0 25.0 25.0 25.0 In Nicaragua, evasion and administrative difficulties (lack of accurate employment histories, wage data and management information systems) in collecting revenues implies a loss of operating cash. We assume that in the base year the revenue loss is about 30% from both these sources. In later years, this leakage decreases as the administrative machinery of the INSS becomes more efficient at collecting contributions. Similar to international trends, we assume that the rates of return earned on INSS investments will be slightly below average market rates of return earned by the private sector. Nicaragua: Pension Reform Proposal Page A5 Table 5. Collection and Rates of Return Assumptions (%) 1996 2010 2060 Collection Rate from Employees and Employers 70.0 75.0 95.0 Real Rate of Return on Investments 15.0 6.0 4.0 According to the base year data, the average length of contributions for new retirees is approximately 15-17 years. By the end of the simulation horizon it is assumed that due to the system's maturity and decreased evasion, a newly retired person would contribute for about 25 years on average. The wage profile used for the simulations were derived from earnings data supplied by the INSS for 1996. The average wage for all contributors was about 17,670 C6rdobas per year in 1996. Figure 2. Imputed Earnings Profile of Contributors by Age (Earnings Profile Indexed to 20 Year Old Male Earnings) 400% 350% Males 250%/ 200% 150% - Females 100% s0% -_fl I 3~ 6 9) N' V6 19 w 1 30 33 6 39 NV NS NS SN r5A ( 60 6 66 69 1'5 115 Page A 6 April 2000 * FinancialAccounts of the INSS 1996: Selected categories The following expenditures were allocated for pensions at the end 1995. Note that expenditures on pension category "Ascendencia" has been ignored3. Our simulations assume inflation indexation4. Table 6. Cash Flows - Expenditure and Revenue: Base Year, in million C6rdobas Total Receipts in Revenues for Pensions 340.0 Total Pension Payments to Female Old Age 44.00 Total Pension Payments to Male Invalidity 33.00 Total Pension Payments to Female Invalidity 34.00 Total Pension Payments to Survivors 25.0 Total Pension Payments to Male Old Age 137.43 The pension fund had a positive balance of roughly 282 million C6rdobas. However this figure may be only an approximation because (a) the IVM did not separate pension funds from other funds like health insurance (in 1995) and (b) recent changes in accounting and asset valuation standards used within the INSS have re-calibrated the 1995 pension fund balance to be roughly 360 million C6rdobas. Since the value of the pension fund at the start of the simulation horizon clearly influences future fiscal sustainability, two base cases were benchmarked using two pension fund values. However, the difference in the initial values is not large enough to generate fundamentally different conclusions other than a short term marginal improvement in the financial balances when the higher fund value is used. SIMULATION CASES The simulation exercises involved assessing the quantitative magnitude of both parametric as well as systemic reforms being contemplated in Nicaragua. Some of the major parametric reforms being contemplated are: retirement age change, increase in payroll contributions, and benefit formula change. In addition to these parametric reforms, simulations showing systemic reform outcomes have also been prepared. These simulations correspond to a shift from the reformed PAYG system to a DC based system based on a set of transition rules. Under the scenario considered most likely the following switching rules will be applied: The reform occurs in the year 2000. All new contributors and individuals less than 45 years of age are covered by the DC system. Those between 45 and 50 have the option to contribute to the system of their choice, while those above the age of 50 are expected to keep contributing to the INSS system (Figure 3). Transition costs are financed by 3These expenditures are indexed on an ad-hoc basis 4 Indexing for inflation is a more conservative approach and, in this case should be considered a worse case scenario. Nicaragua: Pension Reform Proposal Page A 7 recognition bonds whose face value is the implied pension obligation of the system to the individual. These are assumed to be redeemable upon retirement. Figure 3. Switching Rules New Entrants 15