墨西哥1997年的养老金改革

POLIcy RESEARCH WORKING PAPER 1933 The 1997 Pension Reform Under Mexico's reformed pension system, private in Mexico pension funds could become the single largest financial Gloria Grandolini industry in a decade. Their Luis Cerda efficiency and investment Luis Cerda returns will profoundly affect the welfare of retirees, the finances of government, the development of capital markets, and the rate of savings. The World Bank Latin America and the Caribbean Regional Office Finance, Private Sector, and Infrastructure Unit June 1998U  POLICY RESEARCH WORKING PAPER 1933 Summary findings In 1995-96, Mexico shifted to a multipillar approach to component, and the moral hazard implications of the old-age security. The objective of the publicly managed lifetime-switch option. But Mexico achieved radical first pillar is redistribution; a fully-funded second pillar reform with its pension system within a difficult political provides for mandatory individual savings accounts and and economic environment. competitive but exclusive and specialized pension fund And the timing of reform was appropriate. The age management; the third pillar is voluntary savings. structure in the existing system is very young, so This package could provide effective income security coverage could increase. Also, reform took place after and protection against old-age poverty, in a manner the inflationary 1980s and the recent financial crisis, compatible with goals of savings and economic growth. which eroded the real value of old pensions, the acquired It offers Mexico's first real opportunity to shift to a pension rights of the transition generation, and the defined-contribution model and to expand and deepen minimum pension for minimum-wage retirees. domestic capital markets by creating a new class of If returns on invested contributions are high enough, institutional investors - although in the short term its much of the transition generation will choose the impact on capital markets will be limited by the need to defined-contribution alternative over the old pay-as-you- focus on the security of pension fund investments. go system. This will release the government from The reformed system provides for a probably pension liabilities, except for the minimum pension irreversible shift toward private intermediation of most guarantee for new affiliates. domestic investment funds, Further efforts to improve Ensuring the system's long-term success will require the pension system should encourage efficiency, improved financial performance from INFONAVIT, the confidence, and economies of scale. authorities' political will and technical ability to enforce There are weaknesses in Mexico's pension design - pension laws and regulations, and the system's flexibility especially the limited scope for workers in the private in the face of changing circumstances. sector, the continued role of the housing-fund This paper - a product of the Finance, Private Sector, and Infrastructure Unit, Latin America and the Caribbean Regional Office - is part of a larger effort to study contractual savings development in Latin America. Copies of the paper are available free from the World Bank, 1818 H Street NW, Washington, DC 20433. Please contact Cara Zappala, room IS- 074, telephone 202-458-7945, fax 202-522-2106, Internet address czappala@worldbank.org. Gloria Grandolini may be contacted at ggrandolini@worldbank.org. June 1998. (43 pages) The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the view of the World Bank, its Executive Directors, or the countries they represent. Produced by the Policy Research Dissemination Center  THE 1997 PENSION REFORM IN MEXICO by Gloria Grandolini (World Bank) and Luis Cerda (Secretaria de Hacienda y Credito Publico) The views presented in this paper are those of the authors and do not necessarily represent the opinions of the institutions with which they are affiliated. This paper includes contributions and suggestions by Carlos Noriega, Fernando Solis, Jose Maria Rivera, Rafael Trava, Marcela Abraham, Gabriel Martinez, Eduardo Gonzalez, Carlos Soto, Enrique Davila, John Stein, Mike Lubrano, Tom Glaessner, Hemant Shah, Dimitri Vittas, Estelle James, Ulpiano Ayala, Steve Weisbrod, and Truman Packard.  Mexico: The 1997 Pension Reform 2 TABLE OF CONTENTS INTRODUCTION 3 I. MEXICO'S TRADITIONAL PAYG SYSTEM: FEATURES AND NEED FOR REFORM_ 3 The Pension System Before 1997 3 The Rationale for Pension Reform 5 II. THE 1992 SAR EXPERIENCE: HOW PENSION REFORM SHOULD NOT BE DONE 8 The Second Pillar: Individual Retirement Accounts in the SAR 8 Lessons from the SAR Experience 9 -II. THE 1997 PENSION REFORM 1 11 Context for Reform 11 Private Management of Retirement, Old Age and Severance Funds 12 The Rules of Pension Fund Administration_ 14 Disability and Life: IMSS and Insurance Companies Living Together _19 Looking South 19 IV. ESTIMATION OF THE FISCAL COSTS OF THE REFORM 20 What is the Bill? A Summary of Results 20 Fiscal Costs: Assumptions and Scenario Analysis 24 V. KEY CHALLENGES TO THE REFORMED SYSTEM 32 Correcting Design Issues 32 Confronting Implementation Challenges_ 36 CONCLUSIONS 40 BIBLIOGRAPHY 41  Mexico: The 1997 Pension Reform 3 INTRODUCTION This paper presents an analysis of the Mexican reform as a case study of the pension reforms taking place in Latin America. The goal is to provide insights into the rationale of the reform and attempt to identify problems tha.: might arise during the implementatic. of the new pension scheme. Demographic changes, health costs, as well as an increased level of benefits not properly backed in sufficient actuarial reserves indicated that the public pay-as-you-go (PAYG) pension system operating in Mexico, would become financially unsustainable by the end of the decade. A limited and failed attempt at reform was made in 1992 by establishing a mandatory individual retirement savings account and by raising the contribution quota to the Mexican Social Security Institute (Instituto Mexicano de Seguridad Social - IMSS) to cover increasing financial requirements. This was done with the levying of a higher pay-roll tax that encouraged evasion among workers and employers. The IMSS worsening financial situation, labor market distortions, unfairness linked to the benefits' scheme, and the need to generate higher domestic savings led to a growing consensus for pension reform in Mexico. In December 1995, the Mexican Congress approved legislation to reform the existing social security law and shift the pension system for workers in the formal private sector to a privately managed defined-contribution system. Congress approved a second legislative package on the implementation of the reforms in April 1996. The Mexican design is based on a multipillar approach to old age security. Fundamentally, the reformed scheme consists of a publicly managed first pillar with a redistributive objective, a fully- funded second pillar with mandatory individual savings accounts and competitive but exclusive and specialized pension fund management, and a third pillar consisting of voluntary savings. The reform went into effect on September 1, 1997, it covers approximately 11.5 million private sector workers and it is estimated that approximately US$4.0 billion will flow into the system annually. The paper is organized into five sections that discuss: (i) the initial conditions of the public pension system and the forces driving the Mexican reform, including the need, timing, and context for reform; (ii) lessons from the 1992 Retirement Savings System (Sistema de Ahorro para el Retiro - SAR) experience; (iii) the characteristics of the 1997 reform, including comparisons with pension reforms in other Latin American countries and the legal, regulatory, and supervisory framework for fund administration; (iv) the results of the estimation of the fiscal costs of transition; and (v) the key challenges to the reformed system. I. MEXICO'S TRADITIONAL PAYG SYSTEM: FEATURES AND NEED FOR REFORM The Pension System before 1997 Since 1944 the IMSS has managed, among others, the old age, severance, disability and life insurance (Invalidez, Vejez, Cesantia en Edad Avanzada, y Muerte - IVCM). The IMSS is the largest institution in Mexico providing social security to the population. The total number of affiliates to the IMSS is around 11 million and all affiliates are private sector workers (Table I). The Mexican social security system for formal private sector workers and the self-employed as it evolved for almost fifty years through 1992 consisted essentially of two elements of support: retirement, pensioners' health benefits, and disability pensions (IVCM) based on payroll contributions of 8.5 % of wages for formal private sector workers administered by IMSS; and  Mexico: The 1997 Pension Reform 4 additional retirement income from contributions to a specialized housing fund based on an employer-paid contribution of 5 % of wages for private sector employees administered by the National Workers' Housing Fund Institute (Instituto del Fondo Nacional de la Vivienda de los Trabajadores - INFONAVIT). Originally, the IMSS's pension program was designed to be a collective fund financed from contributions from workers, employers and the state. The original IMSS-IVCM can be characterized as a partially-funded defined benefit scheme. However, since the very beginning, it operated as a PAYG scheme as the fund's actuarial reserves were used to finance other social insurance activities, particularly health. Until the 1970s there were no actuarial reserves to face pension liabilities. This was possible because a stable macroeconomic situation and a young population generated enough resources to finance current pension liabilities and at the same time finance health requirements. Formnal 19.8 59.1 Informnal 13.7 40.9 TOTAL 33.5 100 ~NINUNi 111>KYw:i111i 1 , Public Sector Schemes 1.4 11.4 ISSSTE, States, and Armed Forces (a) 1.3 10.6 PEMEX (b) 0.1 0.8 Total Covered 12.3 100 Not Covered 21.2 Self-employed or underemployed 19.5 Unemployed 1.7 Sources: IMSS, PEMEX, and Encuesta Nacional de Empleo, 1995. (a) Institute for Security and Social Services for Government Workers (Instituto de Seguridad y Servicios Sociales de los Trabajadores del Estado) (b) The State-owned petroleum company (Petr6leos Mexicanos) To receive disability benefits, a worker had to contribute to the IVCM system for at least 150 weeks. Despite the existence of a formal process for reviewing disability requests, a striking fact is that 30% of current pensioners are receiving disability pensions. To be entitled to receive an old age pension, the insured was required to be 65 years old, be currently working and have contributed a minimum of 500 weeks. The same conditions applied for severance but the minimum age was 60 years. In case of death, the survivor's pension is paid to the widow and/or dependents, or, in case they do not exist, to the contributor's parents. The insured had to contribute at least 150 weeks. The system contained a guaranteed minimum pension equal to the minimum wage. To address the erosion of the real value of pensions, in 1989 pensions were indexed to changes in the minimum wage. At least 500 weeks of contributions were necessary to qualify for the minimum pension. Pensions are paid according to a formula contained in the old Social Security Law which favors workers at the bottom of the wage scale. The pension base is calculated as a percentage of the average wage in nominal terms for the last 5 years, plus a fraction for each year of contribution in excess of 10.  Mexico: The 1997 Pension Reform 5 As wages are higher, the percentage of the pension base is reduced. Pensions may never be higher than the average wage a worker received during the previous 5 contributing years. The old PAYG system appeared to be quite effective in redistributing income to low-income affiliates. However, as in many other theoretically redistributive PAYG systems, there is a considerable difference between promised and actual redistribution (Friedman, 1996; Schmidt-Hebbel, 1994; and World Bank, 1994). In the Mexican case, as a result of the eligibility requirements (10 years vesting period) and the benefit formula, the system was in fact favoring particular groups of affiliates, which were not necessarily in the lower income brackets. This perverse redistributive effect is related to: * The eligibility requirements. To receive a retirement pension, workers had to accumulate at least 500 weeks of contributions (equivalent to 10 years). Women and other affiliates with high rotation could loose all rights to a pension. The redistribution then would go from those workers working less than 10 years to those working 10 years or more. Also, high-income workers who in economic terms contributed less than low-income workers could still receive a much higher pension. This could happen if the rich worker contributed only for the minimum of 10 years, while a low-income worker contributed for more than 20 years. * The structure of the benefits and contributions. The benefit formula in the old PAYG system allowed those people who worked for more than 10 years to receive relatively higher pensions. However, there were incentives for some workers to either work for only 10 years (since the largest part of a pension would be gained by only working this minimum time), or to belong to the formal sector for only the minimum of 10 years. In these cases, workers with longer periods of contribution subsidized workers contributing for only the minimum period required, regardless of their income. The benefit formula also created incentives to under-report income, allowing high-income workers to contribute the same as some low-income workers, and upon retirement, to receive the same pension. The Rationale for Pension Reform Beginning in the early 1 990s, a consensus began to develop that the social security system, and in particular the IMSS system, needed to be reformed. A long debate ensued on the options for reform and the depth of the measures to be implemented. The three critical reasons for pension reform were that despite the relatively young population (nearly half is under the age of 25) the Mexican pension system was plagued by financial disequilibrium, inadequate pensions, and high evasion. These factors are the most common instigators of pension reform throughout the developing world where populations are still young, public PAYG system finances are on a downward trend, and informality and poverty are widespread (World Bank, 1994; James, 1996; and Demirguc-Kunt and Schwarz, 1996). Demographic Trends: Avoiding Future Shocks and Taking Advantage of Present Opportunities In 1960, Mexico's population structure was relatively young. Of 35 million inhabitants, only 5.6% were older than 60 years, while 55% of the total population was below 20 years of age (Figure 1 a). Over the next several decades Mexico's population structure began to shift as the rate of population growth decreased and life expectancy began to rise (Figure lb). By 1990, Mexico's population had more than doubled at more than 81 million inhabitants.  Mexico: The 1997 Pension Reform 6 Age Age Rateper10 ~~~~~ I ~~~~~~~~~~72 i.ij;,~ ~ ~ ~ ~ ~~~~~~~~~~~~~~~~~~~~~~~N 6 S,&, 71-75 i1990 1 5 61 65 51 55 -~~~~~~~~ D~13960 66 25XC__-5g g u 0 ~g64 3 4145 l- 62 31 35 60 21-25 5 56 11-15 5 1970 1975 1980 1985 1990 1993 2j 0-5