PHN-8718 SIMULATIONS OF SOME ALTERNATIVE POLICIES FOR THE BRAZILIAN SOCIAL SECURITY SYSTEM by Francisco Eduardo Barrets de Oliveira Maria Helena Fernandes da Trindade Henriques Kaizo Iwakami Beltrao June 1987 Population, Health and Nutrition Department World Bank The World Bank does not accept responsibility for the views expressed herein which are those of the author(s) and should not be attributed to the World Bank or to its affiliated organizations. The findings, interpretations, and conclusions are the results of research supported by the Bank; they do not necessarily represent official policy of the Bank. The designations employed, the presentation of material, and any maps used in this document are solely for the convenience of the reader and do not imply the expression of any opinion whatsoever on the part of the World Bank or its affiliates , concerning the legal status of any country, territory, city area, or of its authorities, or concerning the deliminations of its boundaries, or national affiliation. PHN Technical Note 87-18 SIMULATIONS OF SOME ALTERNATIVE POLICIES FOR THE BRAZILIAN SOCIAL SECURITY SYSTEM ABSTRACT The paper reviews the Brazilian social security system and analyzes the reasons for its growth and financial imbalance. A key variable affecting the system's financial equilibrium is the ratio of workers to current beneficiaries, which has experienced gradual and systematic decline. The largest share of benefits, 28 percent, is paid to early retirees on the basis of completion of a minimum number of years of service. Early retirement beneficiaries receive, on average, three times the average pension for regular retirees. Building on an existing simulation model, the paper discusses the implications of reduction in early retirement benefits. With no change, the cost of the benefits package would rise from 13 percent of total benefits paid in 1980 to between 19 and 23 percent by the year 2000. The cost of early retirement will rise from 4.9 percent of the salary base in 1985 to 6-8 percent in the year 2000. The elimination of this benefit could reduce total benefits in the year 2000 from the projected level of 21 percent of the wage base, to 15 percent of the wage base. These savings would be adequate to pay for enhanced benefits for rural retirees not now covered. Prepared by: Francisco Eduardo Barrets de Oliveira Maria Helena Fernandes da Trindade Henriques Kaizo Iwakami Beltrao Consultants to the World Bank June 1987 # 1 Simulations of Some Alternative Policies for the Brazilian Social Security System Francisco Eduardo Barrets de Oliveira Maria Helena Fernandes da Trindade Henriques Kaizo Iwakami Beltrao Rio de Janeiro August, 1986 1. Retrospective Analysis of Social Security Policy in Brazil 1.1 Institutional Evolution The earliest measures in the field of social security in Brazil date from the colonial period when the first social assistance institutions, such as the Casa de Misericordia de Santos (1543) were created. Later on, during Imperial times, other predecessors of the actual system might be found in some civil and military forms of assistance. In global terms, the first measure of social protection to be known is the Law no 3.724, of January 15 in 1919, that regulated work hazards protection by making compulsory the compensation, from the employer's side, of all accidents that happened at place of work. 1) For further details see Leite, Celso Barroso and Velloso, Luis Paranhos, A Previdencia Social, Rio de Janeiro, Zahar Ed., 1962, and Leite, Celso Barroso, A Protecao Social no Brasil. 2nd edition, Sao Paulo, LTR Ed., 1978. 2 However, the effective departure for the actual social security system is the so- called Eloy Chaves Law (N*. 4.682 of January 24, 1923) that determined the creation of retirement and family payment funds to cover existing railroad firms employees. Each firm had a fund to pay for its employees when they ceased their professional activities. The Eloy Chaves Law marks the beginning of a phasis that links social security protection to certain firms, characterized by the small number of covered employees - sometimes the needed minimum to make the system work in the designed way - and by the multiplicity of institutions. The administration of the funds was made by a commitee composed of equal numbers of employers and employees, without the state interference. During the 1920's and 1930's the system was extended to firms of several professional categories reaching a number of 183 funds "(Caixas)" in 1937. From 1930 onwards the urban wage workers became more important in the country's political and economic scenario. The Ministery of Labor, Industry and Commerce was created and social security became a theme of major concern for the State. As the small number of covered employees provided insufficient resources for the maintenance of the system in a stable way, a reform was needed. A new phasis started under which the criterion for coverage became the professional affiliation. Retirement and Family Payment Institutes were created and social security coverage was extended to virtually all urban wage workers and to a large share of the self-employeed. The State, up to these times kept away from the administration of the system, took the important 3 role of organizing the institutions by selecting and nominating their presidents. The first institution to be created at the national level was the Instituto de Aposentadoria e Pensoes dos Maritimos (institute for Retirements and Family Payments of Seamen), created in June 29, 1933, covering all workers of the commercial Navy in the country. Soon after, in 1934, similar institutions were created for workers in the branches of commerce, banking, warehouses, coffee warehouses and stevedors. In 1936 the industrial workers also got their Institute which became effective only in 1938. The process of incorporation and progressive transformation of Funds into Institutes was completed in 1953 through the creation of a single fund (Caixa de Aposentadoria e Pensoes dos Ferroviarios e Empregados em Servicos Publicos). Later on it was transformed into an institute via the Lei Organica da Previdencia Social. The institutes even though providing coverage to a large share of urban workers, embodied large disparities among the several benefit plans offered, mainly due to the different financial capacity of each institution. Because the contribution depended on the employees wages, the institutes with larger resources were the ones that represented higher paid professional categories. In 1945 there was an attempt to correct this distortion through the creation of the Institute of Social Services of Brazil, which would unite and centralize the social security plan of all economically active population in the country. 4 Nevertheless the government that came into power in 1946 removed the credit for the institutes activation and consequently it was not implemented. The first concrete measure to decrease social disparities among all existing professional categories plans was the announcement of the Lei Organica da Previdencia Social (Organic Social Security Law) in August 26, 1960, that remained for 14 years under Congress exam. Its main importance lies in unifying the contributions and benefits plans of the various institutes. Following its approval, social security coverage was extended to all employers and to self-employed in general, since then included as compulsory covered by the social security policy. Six years later the institutional unification was made effective through the creation of the National Social Security Institute (Instituto Nacional de Previdencia Social - JNPS) on November 21, 1966. The new institute began to operate one year later and it united into one single structure. The six existing Institutes for Retirements and Family Payments. The extension of social security coverage was almost global in what refers to urban workers - the exception being domestic servants. Rural workers coverage was lacking. Through the 1960 decade two attempts were made in this regard. In 1963 and 1969 two institutions were created, respectively, the Fund for the Rural Workers Assistance and the Basic Plan, the latter aiming initially at agro-industry and sugar-cane workers, being later extended to other rural activites. However, none of these attempts filled their objectives to satisfaction. The effective inclusion of rural workers into social security coverage only happened through the Complementary Law that in May 25 of 5 1971 abolished the Basic Plan, replacing it by the Program for the Rural Workers Assistance (Pro-Rural). The process of expansion of social security coverage to formerly excluded categories came to a conclusion throughout the 1970 decade through the following measures: inclusion of domestic servants (1972), regulamentation of self-employed contributions making them compulsory (1973), social security assistance to persons older than 70 years and to non-covered disabled (1974) extension of social security benefits to rural employers and their dependents (1976). Therefore, the social security system encompasses the totality of people that receive cash payments in the country. This evolution is summarized in Charts 1.1 and 1.2. In 1947 through the unfolding of the former Ministery of Labor and Social Security, the Social Security Ministery was created, being made responsible for the definition and execution of all social security, medical assistance and social assistance policies. The creation of this ministery represents a significant landmark in the evolution of the Brazilian social security policy that moved further up with the creation of the National Social Assistance and Social Security System (Sistema Nacional de Previdencia e Assistencia Social - SINPAS) in 1977. With the SINPAS creation, each function of the system began to be performed by a specific organ (see chart 1.3). To accomplish that, some organs were created and others went through a re-definition of their functions. The INPS was made exclusively responsible for the maintenance and concession of benefits. Medical assistance, both to urban workers and rural employers and workers, u CHART 1.1 EVOLUTION OF THE BRAZILIAN SOCIAL SECURITY SYSTEM ELOY CHAVES LAW 1923 COVERAGE THROUGH THE FIRM D * BEGINNING OF THE FUSION OF ALL 1 E RETIREMENT AND FAMILY BENEFITS FUNDS 9 C 3 A * CREATION OF THE RETIREMENT AND FAMILY 0 D BENEFITS INSTITUTES E COVERAGE THROUGH THE PROFESSIONAL AFFILIATION D I E 0 ORGANIC SOCIAL SECURITY LAW (L.O.P.S.) 9C 6 A 0 CREATION OF THE NATIONAL SOCIAL 0 D SECURITY INSTITUTE (INPS) E INSTITUTIONAL UNIFICATION 0 CREATION OF PRO'-RURAL D 1 E * CREATION OF THE MINISTERY OF SOCIAL 9 C SECURITY AND SOCIAL ASSISTANCE (MPAS) 7A 0 D * CREATION OF THE NATIONAL SYSTEM OF E SOCIAL SECURITY AND SOCIAL ASSISTANCE (SINPAS) UNIVERSALIZATION 7 CHART 1.2 COVERED POPULATION SMALL NUMBERS FUNDS OF COVERED POPULATION VIRTUALLY THE TOTALITY OF INSTITUTES URBAN WORKERS AND A LARGE SHARE OF THE SELF-EMPLOYED INCLUSION OF EMPLOYERS AND L.O.P.S. OF SELF-EMPLOYED AS COMPULSORY COVERED TOTALITY OF THE POPULATION THAT WORK FOR A PAYMENT SINPAS (INCLUDES DOMESTIC SERVANTS AND RURAL WORKERS) 8 CHART 1.3 SINPAS OBJECTIVES AND ORGANIC STRUCTURE INPS - National Social Security Institute - provides social security SOCIAL protection to urban and rural clientele, public employees SECURITY and financial support to the aged and disabled. INAMPS - National Institute for Medical Assistance of the Social Security System - provides clinical, surgical, MEDICAL pharmaceutical and dentist services to all urban and rural ASSISTANCE clientele, public employees and their dependents. CEME - Central Medicines Unit" - distributes medicines. LBA - Brazilian Assistance Legion - provides social assistance to the needy population. SOCIAL ASSISTANCE FUNABEM- National Foundation for the Welfare of Minors - cares for deprived minors. IAPAS - Institute for the Financial Administration of the Social Security and Social Assistance System - collects and checks ADMINISTRATIVE the contributions and is responsible for the financial ASSISTANCE administration of the whole system. DATAPREV- Social Security Data Processing Enterprise - treats and processes all the information and data. ' Recently transferred to the Health Ministery. 9 was delegated to an organ specially created to accomplish this objective: the National Institute for Medical Assistance of the Social Security System (Instituto Nacional de Assistencia Medica da Previdencia Social - INAMPS). Another institution, the Institute for the Financial Administration of the Social Security and Social Assistance System (Instituto de Administracao Financeira da Previdencia e Assistencia Social - IAPAS) was created to specifically promote the administrative, finance and wealth execution of the system. Social assistance to the needy population was given to the Brazilian Assistance Legion (Legiao Brasileira de Assistencia - LBA) that went through a redefinition of its functions. Together with these, the other institutions that integrate SINPAS are: the National Foundation for the Welfare of Minors (Fundacao Nacional do Bern - Estar do Menor - FUNABEM), the Social Security Data Processing Enterprise (Empresa de Processamento de Dados da Previdencia Social - DATAPREV) and the Central Medicines Unit (Central de Medicamentos - CEME). 10 1.2 Budgetary Resources The funding of the Brazilian Social Security System is covered by revenue coming from social security contributions, the Federal Government contributions and by revenue coming from other resources. Revenue coming from contributions is basically composed of compulsory contributions coming from the firms' payroll, the employees, the employers, the self-employed, domestic servants and others. Actually, the wage share of these contributions vary across several categories as follows: Urban employees: according to their wage up to a limit of 20 minimum wages. Minimum Wage Bracket Social Security Share (%) 0 - 3 8.50 3 - 5 8.75 5 -10 9.00 10 -15 9.50 15 -20 10.00 Urban firms: it applies to the payroll up to a limit of 20 minimum wages. 11 Discrimination Social Security Share (%) Matching the employee contribution 10.0 Contribution to rural social security 2.4 Family allowance 4.0 Maternity allowance 0.3 13th monthly wage (average) 1.5 Work hazard (average) 1.8 Rural firms: the share is worth 2.5 percent of the first commercial sale of products. Self-employed and double contributors: the contribution is 19.2 percent over their work compensation, also according to a special table to estimate the share of the contribution, progressive according to the length of the contribution period. Following this general rule there are some specificities for the other contributions. For instance, government institutions do not pay the 2.4 percent for the rural social security; the non-profit organizations do not pay the 10 percent matching share of their employees, the 2.4 percent for the rural social security or the 4.0 percent for family allowance. Firms that work on a temporary basis have only to contribute the employee/employer share and the work hazard contribution. Sport associations contribute only over the employees' contribution, the 13th monthly wage and the work hazard contribution. There is still a 5 percent contribution over the net revenue of 12 sport events and specific contributions of public institutions and civil employees. The State contribution is defined by law as covering expenditures with personnel and general administration of INPS and INAMPS. Presently the State contribution comes from various shares on lottery games, horse competitions and from a tax on car fuels. All these constitute the so-called "Social Security Quota." It is also the State responsibility to pay for benefits and medical assistance of its federal employees. Other revenues such as fines, moratory interests, services rendered to a third party, real estate rentals and so on, represent a small share of the global revenue. As Chart 1.4 shows, all resources are chanrieled to a single fund - the Fund for Social Security and Social Assistance (Fundo de Previdencia e Assistencia Social - FPAS), and are distributed to the various organs according to the specific needs anchored by the several Multi-Annual Budgetary Plans of the Social Security System. In theory, these plans aim at regulating the use of the financial resources in order to assure the actuarial balance of the FPAS; in practice, however, the pre-established limits are frail, leaving to the social security administration ample room for maneuvering. 13 CHART 1.4 BUDGETARY RESOURCES FLOW SOURCES USES OTHER SOURCES I I REVENUE FROM SOCIAL CONTRIBUTIONS INAMPS STATE TRANSFERS QUOTA FLPS-NS OTHERS I FAS FPAS BENEFITS COVERAGE OTHERS LBA OTHER REVENUES SFUNABEM WEALTH L CAPITAL DATAPREV Note: CEME and DATAPREV indirectly receive resources from FPAS, via contracts for services rendered. 14 In considering the four main categories of expenditures - social security, medical assistance, social assistance and administration, budgetary plans give highest priority to the first, that is, the payment of benefits" to the contributor and its dependents. The other programs are financed, in thesis, through the residual resources net after the payment of the first obligations. In practice, what happens many times is that on the grounds of the need for an artificial equilibrium of the budget, there is an ex-post rationale for excess expenditures under the argument of the compulsory need for the payment of benefits. 1.3 Financial Flows in the financial field, IAPAS concentrates all the social security resources mobilization by receiving the contributions either directly or through the bank network and by transferring them to the other agencies. To eliminate idle resources, from 1979 onwards, IAPAS began to centralize all payments made by INPS and INAMPS. Other resources go through the IAPAS account; these are employers' contributions to other government agencies (SENAI, SESI, SESC, SENAC, INCRA, Salario Educacao, Fundo de Desenvolvimento do Ensino Profissional Maritimo, Fundo Aeroviario) that are also deductible from the employer's payroll. The latter are transferred to these other agencies according to variable criteria. 1) Benefits comprise retirements, family payments, monthly stipends, bonus and other in cash payments made to the contributors and their dependents. 15 Before September of 1979 there was an agreement between the bank network and APAS to transfer resources to the latter, 30 days after they were paid for at the bank. This way, the retention of these resources for that long period provided the banks wide compensation for their given services. However, in view of the frequent checks with insufficient funds made by the Social Security System, a new agreement with the banks was made allowing for: a) a new retention period by the banks of social security deposits: on average 8.33 days; b) payment of a compensation to the banks for their rendered services equivalent to 0.2 percent of all social security collected contributions; c) payment of 0.35 percent over all paid benefits as a compensation to the bank for the payment of benefits; d) anticipation of deposits from IAPAS 48 hours before the payment of benefits is due. If the latter is not fulfilled, the banks still pay the benefits but charge interests over the IAPAS debts equivalent to the discount rate ("taxa de redesconto") of the Central Bank. It is observed, therefore, that when IAPAS lacks the necessary coverage for the payment of benefits, a rotative credit is automatically established with the bank network. In other words, the daily cash-flow desiquilibria are covered by a succession of "credit operations" of a one day period. 16 Chart 1.5 represents in a schematic way what happens with balances and flows of resources as time elapses. It consolidates, as an example, all the accounts both of contributions and of payment of benefits of all the banks involved. It is observed that resources collected by the banks are not available to IAPAS during the length of the retention period. It is possible that during this period negative balances arise due to insufficient funds to cover the payment of benefits. In summary, this mechanism of reciprocation allows for interest rates charged to the Social Security System considerably higher than the nominal rates, even after the reduction of the length of the contributions retention by the banks. As mentioned before, the services of collection of contributions and payment of benefits are paid at a rate of 0.20 percent the first and 0.35 percent the latter, over all resources respectively collected and paid. As the sum collected grew month after month - because of contributors' wage adjustments - the payment for the service of collection went through monthly adjustments. The adjustment of the benefits' value existing every three months corrected as well the amount due by the Social Security System to the banks as a compensation forthis service. It is legitimate to assume that the rates agreed between the banks and IAPAS would fully cover the costs of these services given the voluntary agreement by the banks on the terms of the negotiation. It is worth reminding that in the majority of instances, the banks use the same structure for collecting and paying of their services as in the case of federal, municipal and state taxes, thereby diluting fixed costs. Q 17 CHART 1.5 SCHEMATIC FLOW OF RESOURCES BETWEEN THE BANK NETWORK AND IAPAS DEPOSITS BANK A BANK B BANK N CONTRIBUTIONS CONTRIBUTIONS CONTRIBUTIONS ACCOUNT ACCOUNT ACCOUNT CONTRIBUTIONS TRANSFERENCES IAPAS FLOW OTHER ACCOUNTATTHE DRAWS BANK OF BRAZIL P GBACKTHE YMENTOFBE TS BANKA BANKB BANKN PAYMENT OF PAYMENT OF PAYMENT OF BENEFITS BENEFITS BENEFITS ACCOUNT ACCOUNT ACCOUNT PAYMENTS PERFORMED BY THE BANK NETWORK 18 An important aspect to be considered is that IAPAS can only make payments without sufficient funds when paying for benefits; in fact the banks do not pay for other expenditures if these are not funded. Therefore, IAPAS needs to make a provision of funds to cover for other expenditures such as personnel, material, third party medical assistance, and so on. The net balance of these funds is daily applied in the overnight market at the Central Bank. It is also worth mentioning the double role played by the Bank of Brazil in what refers to the financial administration of the social security system. On the one hand it acts similarly to any.other bank, paying benefits and eventually anticipating funds. On the other hand, as mentioned before, it is the agent of the IAPAS single account ("conta movimento") to which the contributions collected go and from which draws are made to pay for the various SINPAS expenditures. Beginning in 1981 previewed mechanisms in the IAPAS - banks agreement went through many modifications. The interest rate paid to the banks for their anticipated payment of benefits was frozen at a level of 4.33 percent per month, regardless of the fluctuations in the discount rate (taxa de redesconto) of the Central Bank. A new mechanism was introduced via the anticipation of 19 funds by the banks to IAPAS during the first days of the month. In exchange they asked for a lengthier duration in the retention of collected resources." Amazingly enough changes of this nature were made to our best knowledge, through informal agreements between IAPAS and the various banks. It is obvious that in the current situation interest rates depend not only on the volume of collected contributions and paid benefits but also on the volume and length of anticipations and retentions additional to the agreed volume of resources. When this paper was written the social security administration was in the process of renegotiating the agreement with the bank network, aiming at a reduction in the retention period to approximately two days. Also, relating to financial aspects, the deadline for the payment of benefits was reduced from the last working day of the month to the 10th, enhancing substantial additional volume of resources, especially in 1985. 1) Up to recently a large share of contributions were paid at the banks by the employers around the last day of the month. The collected resources were sent by the banks to IAPAS approximately 8.33 days later. Because, on the other hand, the large volume of payment of benefits was due during the first 10 working days of the month, a serious imbalance existed in the SINPAS flow of resources. 20 1.4 Budgetary Regimes The multi-annual Budgetary Plans specify a regime of partial capitalization for long duration benefits (retirements and family payments) and a pay-as-you- go, with a contingency fund, for the remainder expenditures with benefits and services of all SINPAS agencies. In other words, long duration benefits would be guaranteed by real estate properties of the social security and social assistance agencies and by their rentals. Another contingency reserve should be formed to assist on eventual fluctuations during the duration of a budget for the rest of expenditures conducted under the pay-as-you-go regime. That is the regime in which all resources are used during the fiscal period. To sum up, a contingency reserve would act as mufflers for short-term fluctuations. Given the insufficient amount of resources, it was not yet possible to create the mentioned reserve. On the other hand, one may argue whether the real estate properties of the various agencies would be enough as warrant to long duration benefits. Therefore, in practical terms, the Brazilian Social Security System operates as a pay-as-you-go regime, where in-flows of one calendar year are used to cover expenditures of the same calendar year. Because the system does not allow for the creation of financial reserves, the system economic-financial equilibrium in the short run is at the mercy of cyclical fluctuations in employment and in wages, given the rigid nature of most expenditures, mainly the ones directly referring to social security benefits. 21 Given the order of magnitude of SINPAS expenditures - 6 to 7 percent of GDP - the consequences of any financial desiquilibrium are quite significant. It is worth remembering that until recently the resources given to IAPAS as cash advancements were included among the main causes for the senioriage process in the country. A second effect embodied in the pay-as-you-go system, equally important but subject to intense arguments on the measurement side, even at the international level, is the role of social security in the savings formation. Some economists argue that the existence of a social security system of this nature lowers the capital accumulation rate in the economy'). Others'), on the contrary, attempt to demonstrate that social security is neutral for the aggregrate capital accumulation. The truth is that, even in developed countries, this is a complicated question and its empirical solution has proven to be problematical 3. In the Brazilian situation the lack of specific studies on the subject provides room for only one undeniable statement. That is, the present pay-as-you-go system does not generate investment resources, in opposition to other funds (such as the FGTS and the PIS/PASEP). See, for Instance, Feldstein, M.S. "Social Security Induced Retirement and Aggregate Capital Accumulation", Journal of Political Economy, Sept/Oct 1974, 82, p. 905 - 906. - This line is presented, for instance, by Leimer, D. R. and Lesnoy, S. D. in "Social Security and Private Saving: The Time Series Evidence", Journal of Political Economy, June 1982, 90, p. 506 - 629. For recent discussions see Burkhauser, R. V. and Turner, J. A. in "Social Security, Preretirement Labour Supply and Saving: A Conformation and A Critique", Journal of Political Economy, June 1982, 90 p. 643 - 650 and Feldstein, M., "Social Security and Private Saving, Reply, "Journal of Political Economy, June 1982, 90, p. 630 - 642. 22 In the long run and assuming the maintenance of the institutional rules on the concession of benefits, the actuary equilibrium would basically depend on the level of demographic variables (fertility, mortality, migration and labor force participation) as well as on the degree of formalization in working relationships. As seen later, present trend points to an even larger deterioration in the contributors/beneficiaries ratio. This, unless substantial increases in real wages show up, would lead, coeteris paribus, to substantial increases in the wage share that go to social security. Without aiming to a full generalization, the observation of current social security systems operating elsewhere on a pay-as-you-go basis, shows that a rise in costs is practically a general rule'). 1.5 Recent Evolution of Revenue and Expenditures 1.5.1 Revenue2) Table 1.1 shows the revenue evolution in the 1971/1985 period. Itshowsthat contributions represented, on average, 88 percent of all annual revenue. On the other hand approximately 96 percent of compulsory contributions revenue come from the urban payroll, as shown on Table 1.2. 1) In fact even the actuarial calculations in pay-as-you-go social security systems forecast, in the large majority of cases, gradual increases in the wage share intended as the social security contribution. 2) The majority of this Chapter comes from Azevedo, M.E.R.M. and Oliveira, F.E.B., Previdencia Social, Documento de Trabalho No 15, CNRH/IPEA, Oct. 1984. Table 1.1 Evolution of FPAS Revenue, By Type of Revenue: 1971/1985 (Current Value) Specification Unit 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 Cr$ 10166 15300 22266 32733 49148 78793 124594 188038 304315 582 687 1 199 131 2 962 519 5 944 064 17 081 372 63 077 826 Social Million Contributions % 83,44 85,42 87,05 88,81 88,21 88.04 89,68 87,97 92,10 91,62 87,61 93,06 89,56 85,95 89,64 Cr$ 1338 1 755 2169 2562 3479 5560 9856 12743 16600 33139 130428 145 437* 545733 2198035 3092098 Federal Million Government Revenue % 10,98 9,80 8,48 6,95 6,24 6,21 7,09 5,96 4,94 5,21 9,53 4,57 8,22 11,06 4.39 CR$ 680 857 1 142 1 561 3090 5142 4487 12967 9946 20177 39116 75392 147100 594063 4195145 Other Million Revenue % 5,58 4,78 4,47 4.24 5,55 5,75 3,23 6,07 2,96 3,17 2,86 2,37 2,22 2,99 5,97 CR$ 12184 17912 25577 36856 55717 89495 138937 213748 330861 636003 1368 675 3 183 348 6 636 898 19 873 470 70 365069 TOTAL Million % 100,00 100,00 100,00 100,00 100,00 100,00 100,00 100,00 100,00 100,00 100,00 100,00 100,00 100,00 100,00 Source: Grupo de Custeio do MPAS and Balancos do FPAS. * In calculating the value of the Federal Government contributions in 1982, the balance did not include as revenue the Cr$ 180 billion in ORTN imposed to the banks as partial payment for the Social Security debts. The 1985 revenue from contributions include a resource anticipation, estimated as Cr$ 4 trillion, consequence of the reduction in the payment of the contribution date, from the last to the 10th day of the month. 24 Table 1.2 Composition of Contributions Revenue* Specification Total Share (in %) 1 - On the Urban Payroll 96 - Employee Contribution 33 - Firms contribution 63 2 - On the Rural Production 4 Total 100 Source: FPAS Balance, 1981. * In 1981. However, the percentual composition of the contributions revenue has been kept stable since 1978. 25 As to the Federal Government participation, a general trend towards a reduction of its share is observed up to 1980. Following this date there is some recovery as a consequence of the attempts to reduce the existing deficit through resources supplementation. Moreover, from 1980 onwards, the negative balance under the Federal Government responsibility became part of the SINPAS budget, so artificially enlarging the Government share. Table 1.3 shows the extraordinary growth of contributions in total revenue during the 1970 decade. In this period, contributions grew at rates larger than the GNP ones. This fact may be explained by a combination of factors. Among these the most important are: an increase in the ceiling for contributions (from 10 to 20 minimum wages), increase in the share for the self-employed contributions, creation of new shares (for instance, the 2.4 percent share over rural commercialization to cover rural social security costs) and the incorporation of groups that were not previously covered (for instance domestic servants). The jump in the revenue rate from contributions to a value of 25.2 percent in 1982 was a consequence of the various government initiatives to solve the social security deficit. Among these the most important were the attempts directly aiming at raising revenue (increase in shares, increase in ceiling of contributions, creation of a contribution for recipients of retirements and family payments). Therefore, the real rate, net of these initiatives does not run away from the declining trend of these last years. 26 TABLE 1.3 GROWTH RATES OF REVENUE FROM CONTRIBUTIONS AND OF GNP: 1971 - 1985 REVENUE FROM SOCIAL SECURITY CONTRIBUTIONS GNP YEAR In billion Cr$ of 1985 % % 1971 12.238,60 - 1972 15.463,16 26,3 11,1 1973 19.619,13 26,9 13,6 1974 23.369,40 19,1 9,7 1975 26.823,28 14,8 5,4 1976 30.866,37 15,1 9,7 1977 34.199,16 10,5 5,7 1978 36.387,67 6,7 5,0 1979 39.702,21 9,1 6,4 1980 40.428,96 1,8 7,2 1981 41.635,54 3,0 -1,6 1982 52.258,33 25,5 0,9 1983 51.038,96 -2,3 -3,2 1984 53.141,63 4,1 4,4 1985 63.077,83 18,7 7,4 Prices of 1985 - The index was based on wage adjustments. Sources: Grupo de Custeio do MPAS and Balancos Gerais do INPS. Notes: - Given that wages and benefits are adjusted by the same indices considered by the wage policy, the deflators used were: 1970f79: Ministery of Labor Standard of Living Index and 1979 onwards: INPC (National Consumer Prices Index) variations. - If one takes out from the 1985 contributions revenue, the effect of anticipating the payment of contributions, from the last to the 10th day of the month, estimated as CR$ 4 trillion, the contributions would decrease to CR$ 59,077.83 and its rate to 11.2 percent. 27 Together with these factors of an institutional nature, the accelerated processes of urbanization and economic growth experienced by the country led to a substantial increase in the number of social security contributions (see Table 1.4). This growth was due not only to the growth in employment, especially during the 1970 decade, but to the fast incorporation of workers into the social security system, which did not contribute to social security before because they lacked a normalized labor situation. Data relative to the first four years of the 1980 decade simply mirror the effects of the deep economic recession that the country went through during this period, onto the social security contributions. These effects may be also observed in Table 1.4 where there is a clear reduction in the contributors' rates of growth. In 1985 there is an extremely high (18.7 percent) real rate of growth of social contributions. At first sight, in the absence of better data on wages and employment growth, one may say that the larger part of this growth is due to the fast wage re-composition as well as to the re-initiation of the economic growth in what refers to the number of employments in the Brazilian economy. One must also take into account that this high rate of growth also reflects the effect of the anticipation of the due date at which contributions should be paid, to the 10th working day of the month. If one would take out of the total revenue, the share that came through the anticipation of the due payment date, the revenue would have declined to CR $ 59,077.83 billion and the rate of growth to approximately 11.2 percent. 28 TABLE 1.4 GROWTH IN THE NUMBER OF SOCIAL SECURITY CONTRIBUTORS: 1971 - 1985 SOCIAL SECURITY CONTRIBUTIONS YEAR NUMBER RATE OF GROWTH (IN 1000) (In %) 1971 9,690 - 1972 10,436 7,70 1973 11,963 14,63 1974 14,973 25,16 1975 16,347 9,18 1976 18,595 13,75 1977 20,957 12,67 1978 21,166 10,26 1979 22,436 6,00 1980 23,782 6,00 1981 24,448 2,80 1982 24,815 1,50 1983 25,063 0,99 1984 25,065 0,00 1985 25,302 1,01 SOURCE - IBGE - Anuario Estatistico do Brasil, several years. 29 1.5.2 Expenditures Table 1.5 shows the evolution of SINPAS expenditures as if it had existed as such since 1971, via a consolidation of balances from INPS, IPASE and FUNRURAL. In a general way, one may say that expenditures on social security, including expenditures with the properly called benefits and expenditures with the direct administration of the program, are represented by INPS expenditures. At first sight one verifies that, on average, social security is responsible for 2/3 of total expenditures. It is yet obvious that there is a trend toward the increase of INPS share in total expenditures, as it grew from 65.67 percent in 1971 to 74.08 percent in 1983. The inverse occurs in relation to INAMPS expenditures, composed of all expenditures with medical assistance and its administration, which relative share decreased from 27.91 percent in the first year to 21.29 percent in 1983. In 1984 and 1985 one observes a slight recuperation of INAMPS share that is concomitant to a reduction in the INPS share. IAPAS expenditures relative to the fiscal, financial and real estate administrations reveal a general downward trend, with the exception of 1981. The abrupt increase in the expenditures level for this year is due to the IAPAS appropriation of all interests paid to the bank network due to the high existing debts. LBA and FUNABEM are responsible for stricto sensu social assistance by rendering services to non-covered populations. One clearly observes that even though showing a sudden increase in its share, especially after 1979, these expenditures still represent a small part of the total. Table 1.5 Evolution of Expenditures of the Various SINPAS Agencies: 1971/1985 (Current Value) INPS INAMPS IAPAS LBA FUNABEM SINPAS Years Cr$ Million % Cr$ Million % Cr$ Million % Cr$ Million % Cr$ Million % Cr$ Million % 1971 7684 65,67 3265 27,91 631 5,39 78 0,67 42 0,36 11700 100,00 1972 11437 67,99 4436 26,37 808 4,80 97 0,58 44 0,26 16822 100,00 1973 15735 67,77 6230 26,84 1077 4,64 118 0,51 55 0,24 23215 100,00 1974 22990 68,16 8943 26,51 1583 4,69 147 0,44 68 0,20 33731 100,00 1975 34290 65,12 15377 29,21 2659 5,05 225 0,43 98 0,19 52649 100,00 1976 56625 62,41 28657 31,59 4959 5,47 303 0,33 179 0,20 90723 100,00 1977 89459 64,91 42115 30,56 5045 3,66 866 0,63 324 0,24 137809 100,00 1978 140149 65,87 63422 29,81 6894 3,24 1683 0,79 614 0,29 212762 100,00 1979 229088 68,84 91 791 27,58 7 305 2,19 3 639 1,09 991 0,30 332814 100,00 1980 464415 68,02 186773 27,35 20236 2,96 8738 1,28 2652 0,39 682814 100,00 1981 1015381 68,26 362112 24,34 84344 5,67 19280 1,30 6467 0,43 1487584 100,00 1982 2240086 72,19 722678 23,29 92853 2,99 35815 1,15 11550 0,37 3102982 10Q,00 1983 5003083 74,08 1479425 21,91 177671 2,63 70.019 1,04 23056 0,34 6753254 100,00 1984 14.061824 70,54 5051 147 25,33 604099 3,03 163128 0,82 55302 0,28 19935500 100,00 1985 46.839.749 69,07 18.130.820 26,74 2 122 749 3,13 542 581 0,80 174 051 0,26 67809950 100,00 Avg. 71/85 - 67,93 - 27,02 - 3,97 - 0,79 - 0,29 - 100,00 Sources: Grupo de Custeio do MPAS for data up to 1977, Balancos Gerais do FPAS and Balancos da LBA and da FUNABEM. Note: Data referring to years prior to 1978 were estimated arming at a reconstruction of SINPAS as if it had existed since 1971. 31 though showing a sudden increase in its share, especially after 1979, these expenditures still represent a small part of the total. If one carefully analyzes INPS expenditures in relation to benefits (see Table 1.6), it is clear the existence of extremely high rates of growth during the last years, superior, on average, to the 10 percent per year level (see also Table 1.7 for rates of growth of expenditures on benefits for selected countries). Among the several causes leading to this phenomenon, the main ones are: a) accelerated growth of the number of system beneficiaries, as a consequence of the retirement from activity of individuals that, in the past, led to the high rates of growth of contributors' numbers; b) gradual increase in life expectancy for the population in general and for beneficiaries in particular; c) creation of new benefits without the adequate pre-coverage period: the concession of a for life monthly income to aged and disabled not covered by any social security system, even though meritory from a social justice point of view and implying relatively small expenditures, is a typical example of that; also the creation of PRO-RURAL turned, as seen later on, into a factor of influence in the expanding expenditures with benefits; d) increase in the average value of some given benefits, vis-a-vis these under maintenance, that is, newer benefits of a same nature, tend to have a comparatively higher average value than the benefits stock; and 32 TABLE 1.6 EVOLUTION OF EXPENDITURES WITH BENEFITS - 1971/85 EXPENDITURES WITH BENEFITS (IN MILLION CR$) RATE OF GROWTH YEARS INFLATED SERIES CURRENT VALUE INFLATED AT 1985 VALUE* (IN %) 1971 7.126 8.192.510 1972 10.049 9.698.536 18,4 1973 13.257 11.155.20:2 15,0 1974 18.640 12.709.0710 14,0 1975 28.437 14.821.888 16,6 1976 46.333 17.333.950 16,9 1977 86.931 22.733.911 31,2** 1978 138.591 25.611.870 12,7 1979 213.815 26.208.554 2,3*** 1980 433.533**** 28.726.178 9,6 1981 968.547 32.115.574 11,8 1982 2.135.000 35.966.048 12,0 1983 4.840.900 39.796.188 10,6 1984 13.969.096 43.892.296 10,3 1985 46.565.521 46.565.521 6,1 Sources: Grupo de Custeio do MPAS and Balancos Gerais do INPS. * For the 1970/79 period the inflator used was the standard of Living Index estimated by the Ministery of Labor and used as a basis for wage adjustments; from 1979 onwardswe used National Consumers Price Index (INPC) variations. ** The value between 1976 and 1977 is probably distorted as a function of problems in the construction of the inflators. *** The anomalous variation in 1979 reflects balance consolidation problems during SINPAS implementation. ** It includes social security contributions from the Federal Government starting in 1980. 33 Table 1.7 Real Rates of Growth For Expenditures With Benefits For Selected Countries RATES OF GROWTH FOR EXPENDITURES WITH BENEFITS (ANNUAL AVERAGE IN %) COUNTRIES 1960-73 1974-80 West Germany 7,7 3,7 New Zealand 4,7 12,3 Sweden 6,2 12,8 Netherlands 10,8 4,8 Great Britain 6,3 3,0 Denmark 7,0 2,9 Norway 10,9 6,6 Italy 10,2 3,7 Ireland 7,5 8,1 USA 8,5 4,8 Australia 6,2 4,2 Finland 9,8 2,0 France 10,7 9,2 Canada 8,1 5,6 Japan 16,6 21,0 Source: Kaufmann, F. and Leiseing, L., "Demographic changes as a problem for social security systems", International Social Security Review, 4/84, table 2, p. 394. 34 e) the wage policy that, during the period in which Law No. 6,708 (Nov. 1979 to Jan. 1983) was effective, allowed for benefits value adjustments on average, higher than the National Consumers Price Index variations. Inversely to what was observed in regard to revenue, there was a substantial reduction in the real rate of benefits growth during the 1984/1985 period. Though official statements link this fact to a higher effectiveness in the administration, particularly in what relates to fraud control, the lack of data makes any conclusion to be premature at this stage. 1.5.3 Analysis of the Active/Inactive Ratio The equilibrium of a pay-as-you-go social security system, holding everything else constant (values of contributions and of benefits, rules for concession and maintenance of benefits, and so on), depends on the relationship between the number of contributors and of beneficiaries. The institutional evolution experienced by the social security system was characterized, in Brazil, by a progressive incorporation of new groups of affiliated persons, within a process of universalization of coverage that lasts through today, as in most occidental countries. Figure 1.1 shows the growth of contributors and beneficiaries of the Brazilian social security system during the 1923/1983 period. Given the established vesting period (5 years), the first beneficiaries - retirees and family dependents - only qualified for benefits from 1929 onward imaking, therefore, for a 1/13 inactive/active ratio, as shown on Figure 1.2. Starting in 35 Figure 1.1 Actives and Inactives (In 1000) Actives 25000' Creation of the National Social Security and Social Security and Social Assistance System (SINPAS) -1977 Creation of social security support for persons older than 70 or disabled. 20000 Creation of the Social Security and Social Assistance Ministery (MPAS) -1974 1923: EWo Chaves Law inclusion of domestic servants in the INPS social security system. Law No. 5.890 that regulated the subscription of self-employed_ contributors. -1973 1933- Creation of the instute for Retirements and 15000' Family Payments for Seaman (IAPM) Creation of the Assistance 1971- 1 to the Rural Worke Program - -1934 - Creation of the Institute for Retirement and Family Payment for Commerce Workers (IAPC) -Creation of the Institute for Retirement and Family Payment for Bank Employees (JAPB) Creation of the National Social Security institute (JNPS)- 1967 10000- 9Creation of the Institute for Retirement and Family Payment Inactives for industrial Workers (IAPI) Announcement of the Organic Social Security Law that included as compulsory contributors the self- employed in general, it also created the Insitute for Retirement and Family Payment for Railroad and Public Employees (IAPFESP) - 5000 Creation of the Funds for Retirement and Family Payment for Railroad and Public Services Employees 1923 28 33 38 43 48 53 58 63 68 73 78 1983 Sources: IBGE, Anuario Estatistico do Brasil (1946/1982), and several publications of INPS, DATAPREV and IAPAS. FIGURE 1.2 BRAZIL: NUMBER OF ACTIVES FOR EVERY INACTIVE 30 25 20 15 10\- A 5 0 - 1929 1934 1939 1944 1949 1954 1959 1964 1969 1974 1979 1984 Source: Figure 1.1 37 therefore, for a 1/13 inactive/active ratio, as shown on Figure 1.2. Starting in that year, the regular entry of beneficiaries into the system was higher than that of contributors - coming from an affiliation with newly created social security funds - leading the ratio to an increase reaching 1/8.59 in 1933. The creation of Institutes for Retirement and Family Payments led to a continuous reduction in the inactive/active ratio that reached 1/30.36 in 1983, following the IAPI creation. Following the large increase in contributors' numbers (111.6 percent in relation to the previous year) consequence of the IAPI creation, there was a progressive increase in beneficiaries' numbers. This made the ratio to continuously increase, with some breakthrough points such as the creation of INPS in 1967, the inclusion of domestic servants and the regulamentation of the self-employed contributions in 1973. Other measures, such as the creation of PRO-RURAL in 1971 and the institutionalization of the social security support to people either 70 years or older or disabled, in 1974, led to a large increase in beneficiaries numbers and contributed to make the inactive/active ratio reach a low 1/2.74 in 1981. The use of absolute numbers in dealing with contributors and beneficiaries disguises some important distributive characteristics that should be taken into account in a more detailed analysis of the problem. For instance, a beneficiary that has been a rural worker gets half one minimum wage as retirement payment due to age. An urban worker, on the other hand, for the same benefit, gets on average, 1.7 times the minimum wage. 38 The truth is that even acknowledging some limitations in this analysis, the dependency ratio in the Brazilian social security system has reached considerably high levels. As an illustration, Table 1.8 shows the inactive/active ratio for some countries. Table 1.8 Inactives/Actives Ratio for Selected Countries: 1981 Countries Inactives/Actives Ratio Brazil 1:2.74 USA 1:3.23 Japan 1:5.52 France* 1:2.67 West 1:1.77 Germany Source: International Labour Review, May/June, 1983 * 1980 ** 1979 As in every international comparison, one must be exremely careful in interpreting numbers. Given the differences in per capita income a ratio of 1 1.77 in West Germany could be better than the one found in Brazil. The contributory capacity of actives is larger in developed countries, enabling them to support a heavier mass of inactives. 39 The age composition of the population and the type of benefits also effect these relations. For instance, having a younger age distribution, Brazil should present a more favorable ratio than Japan. However, the opposite happens because Brazil offers a more generous benefit's plan. 1.5.4 Budgetary Results The analysis of the accounting results of the social security system (Table 1.9) hides the severity of its economic - financial problems. In fact, these results reflect, in general, that the Government, facing the perspective of large deficits, decided to act in a preventive fashion by raising the social security revenue. One must bear in mind that the previously mentioned "resources anticipations" made by the banks to IAPAS and the reduction in the collection of payments period tend to distort the results, as they enable the appearance of larger than normal revenue. Nevertheless, it is important to emphasize that the significant increase in contributions that existed in 1981') was spent in only one year. Already in 1983 the system was back onto a deficit situation. This fact is even more serious as it is known the Federal Government fully paid for his debts to APAS It showed its effects on the 1982 revenue, leading to a growth rate of 25 percent, as previously mentioned. 40 Table 1.9 SINPAS Budgetary Results (FPAS) - 1971/85 Million CR$ - Current Value Years Revenue Expenditures Deficit Superavit 1971 12.184 11.700 - 484 1972 17.912 16.822 - 1.090 1973 25.577 23.215 - 2.362 1974 36.856 33.731 - 3.125 1975 55.717 52.649 - 3.068 1976 89.495 90.723 1.228 - 1977 138.937 .137.809 1.128 - 1978 213.748 212.762 - 986 1979 330.861 332.814 1.953 - 1980 636.003 682.814 46.811 1981 1.368.675 1.487.584 118.909 - 1982 3.363.348 3.102.982 - 260.366 1983 6.636.898 6.753.254 116.356 - 1984 19.873.470 19.935.500 62.030 - 1985 70.365.069 67.809.950 - 2.555.119 Sources: Grupo de Custeio do MPAS and Balancos do FPAS. 41 in December of 19811), what substantially reduced interest expenditures in the following years. Finally, it is observed that were it not for the anticipation on the due payment date, the system balance in 1985 would present a small deficit. 2. Causes for the Desiquilibrium. Short-Term and Structural Factors. Obviously the social security deficit has its origin in a set of short-term and structural factors, being this separation only a matter of analytical interest. Among the main short-term factors from the revenue side one may include the effects of a reduction in the contributors growth rate and in the average contribution wage, caused by the serious economic recession of the 1980's first years. The wage policy which was implemented after 1979 aside of having contributed to a larger employment rotation, induced extremely pervasive effects over social security in what relates to ways to adjust wages and benefits2). 1) Further details are discussed later'on. 2) Legislation at that time specified that benefits were adjusted at the same date as the minimum wage and according to coefficients as defined by the wage policy. 42 As Table 2.1 shows, profiles of the contributions payroll and of the payment of benefits payroll are very different according to wage brackets. One may observe, for instance, that while 58.78 percent of the urban payroll are in theO - 3 minimum wages bracket, 80.83 percent of the benefits' payroll are in the same interval. Table 2.2 was obtained by applying the various adjustment factors established by the various laws in effect during the 1979/1983 period to the wages and benefits distributions, by estimating an average factor in each case. Besides the fact that during a considerable period, benefits had an average adjustment higher than the INPC, it is worth noting that the average benefit's adjustment factor was always higher than the average factor employed for wages. These differences tend to be higher if one considers that the wage adjustment factors are "theoretical", that is, they are not corrected for eventual increases in labor rotation which would lead to a reduction in the average effective wage adjustment factor. Given the cumulative nature of the adjustments, the impact of these differentials both on the revenue and on the expenditures side is quite significant from 1979 onwards, being a factor towards the economic - financial desiquilibrium. 43 Table 2.1 Distribution of the Contributions and of Benefits Payroll by Minimum Wage Brackets: 1981 Share of the Contributions Share of the Benefits Payroll, Minimum Payroll (In %) Adjusted by the Wage Policy Wage Bracket Number of . Number of Contributors in Payroll Value in Beneficiaries in Payroll Value in Each Bracket Each Bracket Each Bracket Each Bracket 0 - 3 29,00 58,78 55,55 80,83 3 - 7 30,64 23,35 31,83 17,39 7 -10 11,64 7,54 11,31 1,46 10 -15 11,44 6,76 0,91 0,20 15 -20 7,28 3,57 0,15 0,07 20 and 10,00 - 0,25 0,05 Over Total 100,00 100,00 100,00 100,00 Sources: Contributors payroll: RAIS, 1981; Benefits payroll adjusted by the wage policy - DATAPREV, Beneficios em Manutencao (1981) 44 Table 2.2 Average Adjustment Factors For the Contributions and Benefits Payroll Factors Law N Decreet Law Decreet Law Decreet Law 6,708/79 N- 2,012/83 I 2,024/83 No 2,065/83 Average adjustment factor 1.027 0.942 0.953 0.878 for the wages' payroll. Average adjustment factor 1.080 0.987 0.996 0.957 for the benefits payroll Source: Table 2.1 The extremely high growth of benefits may be considered as the main structural component of desiquilibrium. In other words, even taking into account short-term favorable factors and assuming that growth in numbers and in the average value of contributions to be constant, there would still be, in the long run, sizeable deficits in the system. In fact, as Tables 2.3 and 2.4 show, the number of benefits in maintenance grew at rates higher than 10 percent in the large majority of cases. As to the growth in the average value of benefits, preliminary estimates indicate a significant trend towards its elevation during last years. For instance, the average value of retirements due to disability, length of service, age, and special professional category affiliation grew, in the 1971/1979 period by respectively 52, 34, 46 and 68 percent, in considering an inflated series where the factors applied to benefits themseilves'). One may also 1) For a detailed analysis see Oliveira, F.E.B. et allii, Tendencias a Medio Prazo da Previdencia Social Brasileira: Um Modelo de Simulacao, Textos de Discussao Interna No 73, IPEA/INPES, Rio de Janeiro, Jan. 1985 45 Table 2.3 Urban Social Security: Rates of Growth of Benefits In Maintenance: 1971/1983 GEOMETRIC RATE MEDIAN OF AVERAGE OF IN THE PERIOD ANNUAL RATES ANNUAL RATES Retirements Disability 9,12 7,36 9,19 Length of-Service 10,64 11,42 10,67 Age 10,47 12,92 10,61 Special professional 14,65 13,30 14,79 affiliation Family payments 8,41 7,86 8,40 Bonus to remain in work 25% -11,70 -18,36 -9,57 20% 10,50 0,48 11,89 Life Monthly Income Disability* 22,24 4,30 34,39 Age* 13,03 -0,47 19,71 Monthly Payments Illness 4,07 5,51 4,65 Work hazard 24,09 19,58 24,93 Supplementary** 74,80 58,44- 87,34 Detention 5,59 10,75 6,57 Source: INPS * Rates refer to the 1975/1983 period. ** Rates refer to the 1977/1983 period. 46 Table 2.4 Rural Social Security: Rates of Growth of Benefits In Maintenance: 1973/1983 GEOMETRIC RATE MEDIAN OF AVERAGE OF BENEFIT IN THE PERIOD ANNUAL RATES ANNUAL RATES Rural Worker Retirement due to disability 30,63 21,40 32,89 Retirement due to age 7,77 5,31 7,99 Family payments 36,51 20,51 42,78 Social security support 14,03 13,30 19,88 Benefits due to Work Hazards Retirement due to disability 42,05 37,88 47,93 Family payments* 30,08 24,61 30,99 Support during illness** 47,44 41,67 49,21 Rural Employer Retirement due to disability* 127,68 43,49 460,99 Retirement due to age* 41,23 15,76 63,83 Family payments* 148,01 70,82 444,56 Source: INPS * Rates refer to the 1977/1983 period. ** Rates refer to the 1979/1982 period. 47 observe that during the given period, the effects of increases in benefits under maintenance due to the effect of the Law N' 6,708 (new wage policy) were not yet felt. This way, it is quite probable that the increase in the average value of the benefit would actually be due to the recent increases in value. Even though at this point in time one does not have available data for a full explanation of the observed fact, it is quite possible to especulate that events similar to the successive increases in the contributions' ceiling from the past, may be contributing, in a large measure, to the increase in the average value of benefits that are given away today. Frauds are rampant within the social security system. They can be a mere evasion of contributions or excesses in medical assistance or even more, the undue payment of benefits. It is obvious, therefore, that a serious and continuous work might help to solve, to a large extent the SINPAS economic and financial problems. Certainly because of the size of the bureaucratic machines involved, gains of that nature may only be expected in the middle to long runs. A second question very often argued about is the so-called Government debt. Because the Government has not recently reduced its relative contribution to the social security system - quite on the contrary, it increased its resources in relative terms - an argument developed over the ways to cover the deficit rather than over fighting its causes of origin. The option becomes to finance the deficit either through taxes or through social security contributions. In a country like Brazil where the social security system is not universal, it is difficult to come to an easy answer. It is worth remembering that taxes from the 48 population in general, would be financing social security protection to a segment of it, that is, the covered population. As mentioned before, the Government has been forced to adopt a series of measures aiming at a reduction of the social security deficit. Starting in 1979, one may single out: -the anticipation of resources obtained through the reduction in the length of resources retention by the banks from 30 to 8.33 days (1979); -the extension of the basis of incidence of contributions to SESI, SENAI, SESC and SENAC, through the increase in the ceiling from 10 times the reference value to the ceiling of social security contributions, and the destination of these resources to SINPAS (1981); -the raising of the shares and of the ceiling of the social security contributions (1981); -the Government payment of the IAPAS debts to the bank network in December 1981, via the emission of special ORTN'). -the institution of contributions over retirement and family payments; -the anticipation of the deadline for the payment of contributions to the 10th working day following a complete month in the payroll, (1984); and, -more recently, because of a renewal of the social security deficit, the Government has used its last financial resources to reduce the cash flow problems of the social security system, namely: the reduction in the length of resources retention by the banks to practically zero and, the anticipation of the contributions payment made by the employer also for the 10th working day subsequent to the complete month in the payroll (1985). 1) Government titles 49 One may observe that Government action has been so far in the line of adding or anticipating revenues, without much consideration to the rhythm of expenditures evolution. 3. Middle-range Trends 3.1 The simulation model Aiming at an evaluation of the future evolution of the social security system financial balance we decided to rely on a simulation model previously developed.') The model is composed of three basic modules: a demographics, a contributors and a beneficiaries one. The demographic module provides year after year (1981 to 1999), the population distribution by sex and single years, from the ages 0 to 99. The population is projected through the components method and is presented by the followings partitions; total, urban and rural population, active and non-active population. Starting from the 1980 Population Census, we made exogenous hypotheses on the behavior of fertility, mortality, urban/rural distribution and working status of the population. 1) Oliveira, Francisco E. B. de et allii Tendencias e Medio Prazo da Previdencia Social Brasileira: Um Modelo de Simulacao IPEA/INPES Textos para-Discussao Interna, No. 73, Rio de Janeiro, Jan. 1985. 50 The 1980 Population Census provided the initial age structure for the projection, by sex and single years of age. This census age structure was corrected for under-enumeration between the 0 to 9 ages and for digit preference (0 and 5). Fertility was projected assuming a Total Fertility Rate decline from 4.35 in 1980 to 2.91 children per woman in the year 2000. The projected mortality decline was less pronounced; it went from a life expectancy at birth of 60.95 for men and 66 for women in 1980 to 65.74 for men and 71.60 for women, in the year 2000. Both the fertility and mortality evolutions were assumed to be better described by a logistics; a similar function was used to project the urban/rural composition. The percentages of urban population varied from 66.97 percent for men and 69.28 percent for women, in 1980, to 79.65 percent for men and 81.58 percent for women, in the year 2000. A graphical representation of the projection hypotheses appear in Figure 3.1. Figure 3.2, on the other hand, depicts the observed and projected age structures. In what refers to economic activity, there was one hypothesis for men and one hypothesis to describe the future economic activity of women. We assumed the men economic participation to be constant at the 1980 observed levels. For women we assumed that urban participation levels would be constant at 1980 urban Rio Grande do Sul state and rural participation levels were held constant at the 1982 (National Household Survey) rural ones. This hypothesis implied an increase of 25 to 30 percent, depending on the age group, over the 1980 levels. The selected age-specific activity rates are presented on Figures 3.3 and 3.4 for the urban and rural population, respectively. .51 FIGURE 3.1 PROJECTED HYPOTHESES FOR FERTILITY, MORTALITY AND PERCENTAGE OF URBANIZATION. Urban * Popul. TFR 70 84 5 4,35 82% 68 80 4 67 66 76. 3 y 2,91 Total Fertility Rate (T.F.R.) 64 72 2 ( -%Urban Population 62 68 1 " 68 1+- Life expectancy at birth '61 58 64 0 0 * * * 1980 1985 1990 1995 2000 Years Figure 3.2: Observed and Projected Population Pyramids to asI Legend 1980 s 3- - - - 2000 I I I T I r ' I 00 I rI Men r Women I ~-- g-- so- 40I ,---I so --- -J 3 5 L is~ 30 so I * 0 Irng Percentages Age Specific Figure 3.3: Observed and Adjusted Age Specific Activity Rates Activity for the Urban Population, by Sex Rates (%). 100- 90. 80. 70 Men 60 50 Women 30 20 10- 1 5 20 25 30 35 40 45 50 55 60 65 70 75 CO CS Ages Figure 3.4: Observed and Adjusted Age Specific Activity Rates for the Rural Population, by Sex Age Specific Activity Rates 100- 90- Men 80- 70 60- 50 Women 40* 30 20 10 10 15 20 25 30 35 40 45. 50 55 60 65 70 75 60 65 Ages 55 The contributors simulated module provides, also year by year, by sex and single age groups, the urban contributors, rural employers and rural workers, based on the urban/rural economically active population projection. The age and sex economic participation rates onto the formal labor market, needed condition to be considered an urban social security contributor, are exogenously provided. The partition of the rural economically active population into employers and workers, obeys similar criterion. The National Household Surveys (PNADs) were employed to estimate the full range of contributors - state, municipal and federal - and their share of the SINPAS affiliated. The 1979 and 1981 PNADs provided the percentages of economically active contributors by sex and 5-year-age groups. Interpolation curves were adjusted to them to obtain the distributions by single years of age in 1980. SINPAS share in the contributors group was assumed to be similar to the 1983 PNAD one. For that particular year we had special tabulations that provided the information by single years of age. The distribution of contributors in the economically active population was altered throughout the projection according to two hypotheses - a pessimistic and an optimistic one. These hypotheses would be detailed later as scenarios are presented. The beneficiaries simulation module employs the results of the previous modules and using the initial stocks of the main benefits given by both INPS and PRO-RURAL in 1980, simulates their evolution until 1999. For that it uses a 56 1st order Markovian Model') where the transition probabilities were estimated from the 1980, 1981 and 1982 flows of benefits'). The maintenance of the present set of institutional rules, implicit in the various probabilities of transition from the state of contributors into the state of beneficiaries, reflects the assumptions that the social security system is not going to keep on liberalizing the conditions to enter into the inactivity state. However, if one recalls the latest initiatives of the Legislative in this field (for instance, ruling on early retirement for teachers), one may assume this to be an optimistic hypothesis. 3.2 Scenarios and Policies Aiming at exemplifying the potential use of the simulation model we organized two scenarios: an optimistic and a pessimistic one. Both are identical in what relates to the demographic module hypotheses and differ in the simulation of urban contributors and consequently of SINPAS contributors). 1) Except family payments which were projected through time series. 2) For a more detailed description of the model, the data adjustments and working hypotheses, see Oliveira, Francisco E. B. de et allii; Tendencias a Medio Prazo da Previdencia Social Brasileira: Urm Modelo de Simulacao, IPEA/INPES, Textos para Discussao Interna No. 73, Jan. 1985. 3) In the case of rural contributors only rural employers are part of the group. That is the reason for keeping constant at 1980 levels the rural contributors' distribution. 57 The optimistic scenario takes-off from the 1980 estimated urban contributors' distribution. From this point on, the curve reaches a minimum, given by the 1983 PNAD distribution, in 1985 and gradually reaches a maximum, in the year 2000, following a linear trend. According to this course the curve is made tangent in 1990 to the 1978 PNAD distribution, that implied higher levels than the ones found in 1980 and, in 1995 the curve represents the average between the former situation and the upper limit, represented by the economically active population in the year 2000. These values are also corrected by the share of SINPAS contributors in the whole contributors' group. Figures 3.5 A and B and 3.6 A and B present, respectively, the urban men and women contributors, without and with the correction for the SINPAS share in the total, and their evolution according to the previously mentioned course. The pessimistic scenario also takes-off from the estimated urban contributors' distribution. For the 1985, 1990, 1995 and 2000 we have adopted the 1983 PNAD distribution, smoothed for single years of age, for having registered the so far lowest participation rates. The transition from 1980 to 1985 is linear and gradual and the distribution is henceforth kept constant. For each of these scenarios there are three alternative policies tested, implying changes in the present institutional rules for concession of benefits. The first alternative, so-called policy 1, conditions retirement due to length of service to an age limit. In this example, the age limit was fixed at 55 years and the change would become effective in 1987, remaining into effect until the year 2000. In the so-called policy 2 alternative, one completely eliminates the length of service retirement. This change is equivalent to allow the benefit to Figure 3.5A: Urban Male Contributors Urban Contributors/Urban Economically Active Population Ratio Optimistic Scenario Contributors/ Ec. Active Pop. (in %) 100 .-- 90 ............ 80 *Legend ***********. 1980 _\---- -- 1990 709 - :--.1. - - -- - ooo 60- 50- 401 so- 20- to-i .0 i 0 0 5 10 Is 20 23 30 35. 40 45 30 3 60 63 70 15 s0 eS 9Q Ages Source: IBGE, Anuario Estatistico. Several years. Figure 3.5B: Urban Female Contributors Urban Contributors/Urban Economically Active Population Ratio Optimistic Scenario Contributors/ Ec. Active Pop. (in %) 100 S0 so . ,--------.-. Legend 70.*o-*. 1990 '995 60 40 - 30 20 . 10 I Il II 11 0- 0 5 10 15 20 25 30 35 40' 45 50 55 60 65 70 75 so as So Source: IBGE, Anucrio Estatistico. Several years. Ages Source: IBGE o0 Figure 36A: Urban Male Contributors Urban SINPAS Contributors/Urban Economically Active Popidation Ratio Pessimistic Scenario SINPAS Contributors/ Urban Ec. Active Pop. 90 so. Legend 1980 . 1985.90.95.00 60 I 50- 40 I i 30 20 to% 0 Figure 3.6B: Urban Female Contributors Urban SINPAS Contributors/Urban Economically Active Population Ratio Pessimistic Scenario SINPAS Contributors/ Urban Ec. Active Pop. (in %) 100- 90 no Legend 1980 Go- 95,90,95.0 50 40 30 20 0* 0 5 50 15 20 25 30 35 40 45 50 55 60 65 70 73 80 85 S Ages Source: IBGE, Anuario Estatistico Several years._ 62 people aged at least 60, if women, or 65, if men, age limits that qualify persons for retirement on age grounds. Finally, in policy 3 alternative, we projected a constant reduction in the value of length of service retirement, equivalent to 5 percent for every year of age inferior to the age limit of 55 years. The reduction would reach a maximum of 50 percent for people wishing to retire at age 45 or before that. Throughout the simulation of these policy alternatives the concession rules for length of serice bonuses followed the restrictions imposed on the length of service retirement. 3.3 Results 3.3.1 Present Situation a) Urban social security In looking at Figures 3.7 and 3.8 that display the evolution of population volumes, one observes that the only significant difference lies in the evolution of contributors' volumes. While in the optimistic scenario the number of contributors goes from 20 million in 1980 to approximately 45 million in 1999, in the pessimistic scenario the final number does not reach 35 million. It is also interesting to mention that the beneficiaries' volume is practically non- sensitive to hypotheses of variation in contributors' level. Nevertheless, in what relates to rates of growth, Figures 3.9 and 3.10 point to differences of considerable significance. In the optimistic scenario the contributors population grows at declining rates in the first five years (from 2.15 to 1.75 percent per year). As we established 1985 as the end of the recession year, according to our hypothesis, there is an expansion in both Figure 3.7: Urban Social Security: 1980/1999 Optimistic Scenario (Current rules) (in 1.000) 140.000 Urban Population 120.000 100.000 80.000 60.000 Urban Economically Active Population -. - .a * N 40.000 . o w - - - - g - - I wo - - - Contributors n . - * - 20.000 - * - . * -""Beneficiaries 0 8 8 a 8 8 8 8 8 8 0 9 9 9 9 9 9 9 9 s 1980 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 1999 Years Figure 3.8: Urban Social Security: 1980/1999 Pessimistic Scenario (Current rules) (in 1.000) 140.000 Urban Population 120.000 100.000 80.000 60.000 Urban Economically Active Population .. 40.000 . -w -w - - - - - - - - - - - - -. .- -Contributors 20.000 M * * *.. N a --m-...w Beneficiaries 1980 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 1999 Years Figure 3.9: Urban Social Security: Population Annual Rates of Growth: 1981/99 Optimistic Scenario (In 1.000) (Current rules) 8 Contributors 7 6 Beneficiaries 5 m I Urban Economically Active Population 'I No -11t '' I "" *o 1 1 3 Urban Population I 2 1. 1 0 - - - - - - - - - - - - - - - - - 1980 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 1999 Years Figure 3.10: Urban Social Security: Population Annual Rates of Growth: 1981/99 Pessimistic Scenario (in 1.000) (Current rules) 9 8,5 8 7 6 5 Urjpn Economically Active Population A Wa*t aU _ Cnntributors ow eas mum a men a ses a num a sena a m . 3 U * - - I Urban Population * 1 0 - - *- - - - . a - - - - * * * * * * * 1980 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 1999 Years CK 67 supply and demand of labor, leading to a jump in the contributors' growth rate to 7 percent, a higher level than the one experienced by the beneficiaries population. Starting in 1990, the rate remains constant at a level between 4.5 and 5 percent. However, in the first five years, the retirement rates of growth due to disability or due to special professional affiliation are quite high and regardless of the scenario, they remain higher than 10 percent per year.) In what relates to volumes, though, differences are minor and do not even deserve comments (See Figures 3.11 through 3.16). However, it is on the expenditures side where differences between the optimitic and the pessimistic scenarios are more intense. As an indicator of cost we have selected the expenditures as a percentage of the urban contributors' payroll (see Figures 3.17 and 3.18). As for the total value of benefits, this grows from 13 percent in 1980 to 17.5 percent in 1985. From this date onwards, the differential weight of contributors embodied in each scenario makes that in the pessimistic scenario, the share of expenditures reach a 22.5 percent while in the optimistic, the final value is only 18.2 percent. This evolution is in fact pushed by the retirements, which in the pessimistic scenario grow from 9 to 16.8 percent of the urban payroll in 1999; in the 1) Given that the differing hypotheses between the optimistic and the pessimistic scenarios only start to play after 1985, this evolution displays the 1980 observed levels, corrected as explained in the text. Figure 3.11: Urban Social Security: Benefits of Continuing Payments Optimistic Scenario (In 1.000) (Current Rules) 8.000 7.000 6.000 5.000 Retirements 4.000 0000 -0 3.000 Family Payments 2.000 -.. . Life Monthly Payments 1.000 a **" a a Epsom Bonuses 0 * * * * 9 0 0 * * 6 1980 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 1999 Years Figure 3.12: Urban Social Security: Benefits of Continuing Payments Pessimistic Scenario (Current Rules) (In 1.000) 7.000 6.000 5.000 Retirement 4.000 3.000 w,oo ***00 Family Payments goo ... 0 &' 2.000 W - w Life Monthly Payments 1.000 - --- - --- --- -- --- -- --- -- --- Bonuses . . .------- - u 8 - - 5* ** -- 8 8 9 9 1980 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 1999 Years Figure 3.13: Urban Social Security: Retirement Payments: 1980/99 Optimistic Scenario (In 1.000) (Current Rules) 8.000 7.000 6.000 5.000 4.000 Total Retirements - - - - 2.000 - -.- -*- 0 - Length of Service Retirements . . 1.000 . * * - Age Retirements - - 2 83 - - 9 97 98 19 Year 0 Special Professional Affiliation. Retirements- 00 Figure 3.14: Urban Social Security: Retirement Payments: 1980/99 Pessimistic Scenario (Current Rules) (in 1.000) 7.000 6.000 5.000 Total Retirements 4.000 3.000 . 0 0 0 - w Disability Retirements ...** 2.000 .. - -- as*,,O... owl*wo Length of Service Retirements 1.ooowo . .woo--* 1.000 so WN MM - @no *-S-*- - Age Retirements , ,. - U " 0 - eZiaTProTessi nal Affiliation Retirements 1980 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 1999 Years Figure 3.15: Urban Social Security: Annual Rates of Growth of Benefits of Continuing Payments Optimistic Scenario (In 1.000) (Current Rules) 13 12 11 10 Disability Retirements 9 Special Profession Affiliation Retirements 8 7 egth of Service Total Retirements Retireent 1 - - - i - . Age Retirements 5 4 3 2 1980 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 1999 Years FIQur6 3.16:U『ban SQCial Se【uritV:Annua!RateS of Growth of BenetitS Ot COntinUinQ PaVm6ntS PeSS廳m面St豆【S(ena『10 (Current Rules) (In%) ！〕一一 1980 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 1999 YearS閱 騙 (in 1.000) Figure 3.17: Urban Social Security: Benefits of Continuing Payment as a Percentage of Urban Contributors Payro If Optimistic Scenario 23 22 21 20 19 18 Total Benefits 17 16 15 1 14 4 1 0000000 13 4 1 Retirements own was wow mom mom 12 41 00 MW mom W. W.M.M, vw wow 11 000 10 .00 AL .00 8 7 6 5 4 Family Payments 0 MOM E 3 Eventual Support Payments 2 a m a E a a g n 0 a a 0 0 a a a a 0 m a a a w 0 a m a w w a s a w a a a Bonuses 0 Life Monthly Payments - - - - - 0 a 0 0 6 0 1980 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 1999 Years Figure 3.18: Urban Social Security: Benefits of Continuing Payment as a Percentage of Urban Contributors Payroll (in %) Pessimistic Scenario 22 21 20 19 Total Benefits 18 17 16 -.. 15 -oo -w* wo 13 Retirements . **-* 1 3 1,, ..*wo o 12 -o -0 - - - 10 9 8 7 6 5 4 Family Payments 0 .. a * a U 2 Eventual Support Payments 1 Bonuses 0 1980 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 1999 Years 76 optimistic scenario the final percentage is 13.4 percent. It deserves to be noted that even though family payments represent a minor item in total expenditures, as compared to retirements, their cost goes from 2.5 to 5 percent of the urban payroll in the pessimistic scenario while in the optimistic their cost does not reach the 4 percent. Within retirements (see figures 3.19 and 3.20) the largest cost increases are represented by the length of service and by disability retirements, in this order. While in the pessimistic scenario the cost of lenght of service retirement goes from 4.9 to 8 percent of the urban payroll and the cost of disability retirements increases from 2.9 to 6.2 percent, in the optimistic scenario the first increases by only 1 percent while the second, disability retirement, reaches only 5 percent. b) Rural Social Security In the case of the rural population it does not make any sense to talk about two scenarios, except in what relates to the cost of benefits as a percentage of the urban payroll. In fact, the obtained results are quite similar to the basic scenario results of the model's prior version.1) There is a decrease in the absolute volume of the rural population (from 38 to 35 million), a stabilization of the rural economically 1) Oliveira, Francisco E. et allii, op. cit.. p. 111 - 129. Figure 3.19: Urban Social Security: Cost of Retirements as a Percentage of the Urban Payroll: 1980/99 Optimistic Scenario (Current Rules) (in %) 14Total Retirements 13 12 11 10 9 8 7 Length of Service Retirements 5 , w Disability Retirements a , ""a a * spa suo own Wsp so"- * 4 * ** 2 Age Retirements M0N a UM a OWSUME U NEWS a "M a swnMd" UEM WNNDW a MWE m ENMNN ONd 1 Special Professional Affiliation Retirements S a g ... - - 4 8 0 1980 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 1999 Years Figure 3.20: Urban Social Security: Cost of Retirements as a Percentage of the Urban Payroll: 1980/99 Pessimistic Scenario (Current Rules) (in %) 17 16 15 Total Retirements 14 13 12 11 10 9 8 . * 1-Annth of Service Retirements ..- -* - 7 ,,- ..o Disability Retirements 6 0,0 lo -o wo sa lo Ww - I' a woo 2 Age Retirements 1 w a U 0 * . . , . Special Professional Affiliation Retirements 1 8 8 8 8 8 -8 8 1 Years 1980 81 82 83 84 85 86 87 88 89 90 91 -92 93 94 95 96 97 98 1999 Years 003 79 active population in approximately 15 million and a doubling of the number of beneficiaries (from 2.5 to close to 6 million), all evident in Figure 3.21. This evolution matches the negative or close to zero rates of growth for the rural total and economically active populations, which appear on Figure 3.22. It is also consistent with the evolution of the beneficiaries' rate of growth which, even though showing a declining trend, is kept at a high level (from 7.5 to 3.8 percent per annum). b.1) Rural Worker As we see in Figure 3.23, the number of rural workers remains more or less stable at approximately 14 million. In what relates to benefits, there is an impressive variation in retirements (from 1,750,000 to 2,750,000) and family payments (from 500,000 to 2,700,000). The growth in retirments (Figur 3.24) is exclusively owed to the increase in disability retirements which would have gone from 200,000 in 1980 to 1,200,000 in the year 2000, despite the reduction in its growth rate (Figure 3.25); in the meanwhile age retirements kept constant at a level of 1,400,000 throughout the projection period. In what relates to costs, as these are viewed as a percentage of the urban payroll, differences between the pessimistic and the optimistic scenarios showed up (Figures 3.26 and 3.27). In the first five years of the projection, 1980 - 85, in both scenarios, by definition, benefits grow in the same way. (from 2.5 to 2.8 percent of the urban payroll). From that point on, in the optimistic scenario the percentage drops to the starting level; in the meanwhile, in the pessimistic scenario, there is an uphill evolution, reaching 3.4 percent at the end of the period. This occurs mainly due to retirements and Figure 3.21: Rural Social Security: Populations - 1980/99 (Current Rules) (in 1.000) 40.000 Total Population 35.000 30.000 25.000 20.000 Rural Economically Active Population 15.000 - 10.000 Beneficiaries w w.. . - * wn 5.000 O n- * ' *w 1980 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 1999 Years Figure 3.22: Rural Social Security: Annual Rates of Growth (Current Rules) (in %) 7,0 6,0 5,0 Beneficiaries 4,0 3,0 2,0 110 Rural Economically Active Population 0 'RON - (1,0) namea B Rural Population (2,0) 1980 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 1999 Years 0" Figure 3.23: Rural Social Security: Rural Worker Benefits (In 1000) (Current Rules) 15.000 Rural Worker 14.000 13.000 12.000 11.000 10.000 9.000 8.000 7.000 D.UUU 5.000 4.000 3.000 Retirements 2.000 _ 2.000 o.... .- - Family Payments It . - * - * l" gn- a ~~~ m a a a a a a a a a a amaaaaaaaa aaaaaaUa 00 % *FmanaI Support 1980 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 1999 Years Figure 3.24: Rural Social Security: Rural Worker Retirements: 1980/99 (In %) (Current Rules) 2.250 Total Retirements 2.000 1.750 Age Retirements 1.500 1.250 1.000 750 Disability Retirements, * 500/ 250 *0 125 1980 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 1999 Years Figure 3.25: Rural Social Security: Annual Rates of Growth of Rural Workers' Retirements: 1980/99 (Current Rules) (in %) 30 25 15 Disability Retirements 10 Age Retirements 0 r I I I -1 1- -" - 7I 1980 81 82 83 84 85. 86 87 88 89 90 1 92 93 94 95 96 97 98 1999 Years Figure 3.26: Rural Social Security: Cost of Rural Workers Benefits as a Percentage of Urban Contributions Payroll: 1980/99 Optimistic Scenario (Current Rules) (In %) 3,5 3,0 Total Rural Workers' Benefit 2,5 2,0 ... - - om* *t. Retirements 1,0 Family Payments , * . . - - ..* woo Life Monthly Payments 1980 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 1999 Years Figure 3.27: Rural Social Security: Cost of Rural Workers Benefits as a Percentage of Urban Contributions Payroll: 1980/99 Pessimistic Scenario (Current Rules) (in %) 3,5 3,5 Total Rural Workers Benefits 3,0 2,5 4,U Retirements 1,5 w E = m M m m ...- *- 1,0 Family Payments . w* o Life Monthly Payments 198O 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 1999 Years 87 family payments. As to the retirements, in the optimistic scenario their cost declined from 1.5 to 1.1 percent, while in the pessimistic scenario the cost remained around 1.4 percent. In what relates to family payments in both scenarios there occurs an increase in their costs; in the optimistic scenario the cost rises from 0.45 to 1.05 percent but in the pessimistic, it reaches 1.4%. The cost of life monthly payments is kept at a close to constant level, inferior to 0.5 percent, despite the growth in numbers (Figure 3.28) from 300,000 to 900,000, mainly due to the growth in disability payments. b.2 Rural Employer In the case of rural employers there was no variation between the optimistc and the pessimistic scenarios, given the working hypotheses. As Figure 3.29 shows, the volume of rural employers suffers minor variations, from 300,000 to 320,000, throughout the projection period. Retirements, on the other hand, show a more significant growth, changing from 90,000 in 1980 to 120,000 in 1999. As to the family payments, these double in the period, going from a volume of almost 14,000 to 100,000. In Figure 3.30 one sees that as in the case of rural workers, the increase in the volume of retirements is accounted for the increase in disability retirements which grow from an initial 5,000 to almost 40,000 in 1999. Figure 3.28: Rural Social Security: Financial Support - 1980/99 (Current Rules) (In %) 800 Total of Financial Supports 700 600 500 4UU 00 Financial Support for Age Reasons 0 0 .. -M 300 ,.. g 200 ...o * 00 Financial Support for Disability Reasons a0 100 1980 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 1999 Years Figure 3.29: Rural Social Security: Rural Employer Benefits - 1980/99 (Current Rules) (in 1.000) Rural Employers 320 300 280 200 180 150 140 120 Rural Employer Retirements 10 80 ..** 60 a . * -- - 40 f - . * o w Family Payments 20 * 0 1980 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 1999 Years 0 Figure 3.30: Rural Social Security: Rural Employer Retirements - 1980/99 (Current Rules) (in 1.000) 130 Rural Employer Retirements 120 110 100 90 SNo - W M Age Retirements 80 - ., 70 60 50 40 Disability Retirements 30 ,,,oo - * v M' 20 ,. o 10 0 1980 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 1999 Years 91 In Figure 3.31 it is clear that this evolution in the number of benefits doubles their cost as a percentage of the urban payroll, moving it from 15 to 28.5 percent'). 3.3.3 Alternative Policies As mentioned before, to allow for changes in the concession of length of service retirement, we made working hypotheses simulating some changes. In what is called policy 1 we condition length of service retirement to a minimum age limit of 55 years; in policy 2 we eliminate length of service retirement and in policy 3 we introduce a fine to people who retire before age 55, inversely related to the age of retirement. A result is clear from the examination of Figures 3.32 and 3.33. In both optimistic and pessimistic scenarios, there is a constant difference between the total number of retirements that occur in each scenario, at present rules of concession of benefits and by altering them according to the simulated policies. According to the optimistic scenario we would have approximately 7,500,000 retirements at the end of the period, at the present rules of concession of benefits or if adopting policy 1, 6,700,000 retirements, if adopting policy 2 and again 7,500,000 if policy 3 was the chosen one, being 12 percent of these affected by the partial reduction in the value of the benefit. 1) In the absence of better data, the average contribution wage of rural employer was made equal to the average urban contribution wage. Therefore, the employers' payroll was obtained, year after year, as a product of the number of rural employers times a constant, which was the average contributions wage. Figure 3.31: Rural Social Security: Cost of Rural Employers .Benefits as a Percentage of Urban Contributions Payroll: 1980/99 (Current Rules) (in %) 30 Total Benefits 20 Retirements 10 woo Family Payments . - - -a moo a o 0 - 8 *4* 95 9 1980 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 1999 Years Figure 3.32: Urban Social Security: Currently Retired Persons: 1980/99 (in %) Optimistic Scenario 8000 Policy 1 O ol 7000 Policy 2 Present Policy 00*00*0 6000 5000 Total Retirements 4000 3000 2000 Length of Service Retirements 1000 4 M E u a a 1881 0 0 0 .. 8 8 1980 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 1999 Years Figure 3.33: Urban Social Security: Currently Retired Persons: 1980/99 (in %) Pessimistic Scenario 8000 7000 Policy 2do 6000 Present Policydog 5000 Total Retirements 4000 3000 2000 Length of Service Retirements 1000 .._. . -- - 1980 81 82 83 84 85 86 87 88 89 90 9T 92 9 4 95 96 97 98 1999 Years 95 In the case of length of service retirements, directly affected by the policy modifications and embodying a higher average value, reductions would be proportionately larger. At the present rules of concession one would have in 1999 a volume of 1,900,000 length of service retirements as opposed to 1,600,000 if policy 1 was adopted, 1,250,000 if policy 2 and finally 1,800,000 with policy 3. In the latter, 50 percent of the retirees would suffer a permanent reduction in the value of their benefits. In the pessimistic scenario, because we assume a smaller number of contributors, we would also have a smaller number of retirements. According to the present rules we would have 7,200,000 retirements at the end of the period; if policy 3 would be implemented on the other hand, this amount would be only 6,300,000, that is 88. percent of the volume obtained according to the present rules. In this scenario, length of service retirements would also be affected by a permanent reduction in the value of benefits; this would affect approximately 50 percent of the retirees. In the pessimistic scenario, because of the growth momentum of contributors in the past, that would generate less ties in the future, one would have a relative decrease in the contributors/inactives ratio. However, the shares in costs produced in this scenario would be considerably higher than the ones occuring in the optimistic scenario. In the first, for the total number of benefits at the present rules of concession one would obtain a share of 21.4 percent out of the urban payroll. This value would consistently decrease If any of the alternative policies were adopted and would reach it lowest value, 19.3 percent if policy 2 was implemented. In the optimistic 96 scenario the take-off, under present rules, is 17 percent and it would be reduced to 14.8 percent if policy 2 were adopted. Therefore, judging from the selected cost indicator, policy 2, where length of service retirement is eliminated, would provide the highest economy in costs. According to Table 3.1 where the various simulated shares for length of service retirements are presented, one observes that if, on the one hand, the assumed variations in the contributors' volume make the length of retirements shares grow, the adoption of any of the alternative policies would lead to substantial reductions, reaching even 50 percent in the case of policy 2 (optimistic scenario). Looking at Figures 3.34 and 3.35 one sees that there is a significative.reduction in the values of the annual rates of growth, regardless of the scenario. In 1990, for instance, the growth rate at present rules of concession would be 6 percent but if policy 2 would be implemented, this value would decline to 3.7 percent; in the pessimistic scenario these values would be 5.8 and 3.4 percent, respectively. It is also clear that there is a trend towards rates' conversion, as time elapses, in all policy alternatives, towards values around 4.8 to 5.4 percent, maybe asa consequence of the aging of the population. Finally, the inertia that the whole social security system embodies to respond in a fast and clear fashion to changes, of the nature of the ones herein simulated (level and profile of contributors and radical changes in the rules of concession for length of service retirements), slows down in the pretended evolution of the active/beneficiaries ratio, according to the various simulated policies (See Table 3.2). In analyzing the column relative to the current rules, 97 Table 3.1 Shares of Length of Service Retirements Optimistic Scenario Pessimistic Scenario Years Current 1 2 3 Current 1 2 3 1980 0,0489 0,0489 0,0489 0,0489 0,0489 0,0489 0,0489 0,0489 1981 0,0513 0,0513 0,0513 0,0513 0,0513 0,0513 0,0513 0,0513 1982 0,0539 0,0539 0,0539 0,0539 0,0539 0,0539 0,0539 0,0539 1983 0,0565 0,0565 0,0565 0,0565 0,0565 0,0565 0,0565 0,0565 1984 0,0591 0,0591 0,0591 0,0591 0,0591 0,0591 0,0591 0,0591 1985 0,0617 0,0617 0,0617 0,0617 0,0618 0,0618 0,0618 0,0618 1986 0,0611 0,0611 0,0611 0,0611 0,0629 0,0629 0,0629 0,0629 1987 0,0605 0,0584 0,0559 0,0598 .0,0642 0,0620 0,0594 0,0636 1988 0,0600 0,0562 0,0514 0,0589 0,0654 0,0613 0,0561 0,0643 1989 0,0597 0,0545 0,0474 0,0581 0,0666 0,0610 0,0532 0,0651 1990 0,0594 0,0531 0,0440 0,0574 0,0679 0,0610 0,0507 0,0658 1991 0,0598 0,0527 0,0415 0,0575 0,0692 0,0613 0,0485 0,0666 1992 0,0601 0,0524 0,0394 0,0576 0,0705 0,0617 0,0468 0,0674 1993 0,0606 0,0523 0,0376 0,0576 0,0717 0,0623 0,0453 0,0681 1994 0,0610 0,0522 0,0362 0,0577 0,0730 0,0630 0,0441 0,0688 1995 0,0614 0,0523 0,0351 0,0578 0,0742 0,0639 0,0435 0,0694 1996 0,0619 0,0525 0,0343 0,0580 0,0755 0,0648 0,0430 0,0701 1997 0,0623 0,0526 0,0336 0,0581 0,0768 0,0656 0,0427 0,0707 1998 0,0628 0,0525 0,0329 0,0583 0,0780 0,0662 0,0423 0,0712 1999 0,0632 0,0525 0,0322 0,0584 0,0792 0,0668 0,0420 0,0717 Figure 3.34: Urban Social Security: Rates of Growth of Urban Benefits: 1980/99 Optimistic Scenario (in %) 10 9 8 7 6 Present Policy 5 Policy 1 -= = =-_ F- _ _ MO M 4 3 2 1 1980 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 1999 Years Figure 3.35: Urban Social Security: Rates of Growth of Urban Benefits: 1980/99 Pessimistic Scenario (in %) 9 8 7 6 Present Policy 5 4 Policy 1 3 2 1 1980 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 1999 Years 100 one sees that there occurs, in either scenario, a deterioration in the active/beneficiaries ratio as it falls; in the pessimistic scenario, it moves from 2.79 in 1980 to 1.77 in 1999. The adoption of any of the alternative policies would improve this desequilibrium, especially in the optimistic scenario, leading to a recuperation of the ratio almost to initial levels, more so in the case of policy 2. It is obvious, anyway, that this equilibrium is a temporary one as the rate of growth of the economically active population, even in the optimistic scenario, does not reach the estimated rate for the beneficiaries growth. The solution would then come via a differential in values, there is, the wages of the economically active population, in real terms, would have to grow at a rate that would compensate the difference between the growth of contributors and of the beneficiaries. 4. Conclusions and Final Comments In light of the limitations imposed by the quality of the available data, the present report points to a trend towards a desequilibrium, in the middle range, of the Brazilian Social Security System under the so-called pessimistic scenario and to a relative stability under the"optimistic scenario. The periods 1984/85 and 1985/86 seem, otherwise, frontally contradict the model simulations. Nevertheless, we believe these periods to be atypical, especially in what relates to the growth in contributions. According to the data from the Ministery of Labor, there was a real growth in wages of 18.96 percent in the first period, subdivided into 4.89 percent due to an increase in the level of employment and 13.41 percent due to real wage gains (above the INPC). Taking into account that benefits did not profit from similar gains, it is 101 Table 3.2 Active/Beneficiaries Ratio Optimistic Scenario Pessimistic Scenario Years Years Current 1 2 3 Current 1 2 3 1980 2,75 2,75 2,75 2,75 2,79 2,79 2,79 2,79 1981 2,62 2,62 2,61 2,62 2,64 2,64 2,64 2,64 1982 2,49 2,49 2,49 2,49 2,50 2,50 2,50 2,50 1983 2,38 2,38 2,37 2,38 2,38 2,38 2,38 2,38 1984 2,27 2,27 2,27 2,27 2,26 2,26 2,26 2,26 1985 2,18 2,18 2,18 2,18 2,16 2,16 2,16 2,16 1986 2,20 2,20 2,20 2,20 2,12 2,12 2,12 2,12 1987 2,23 2,24 2,25 2,24 2,09 2,10 2,11 2,09 1988 2,25 2,28 2,30 2,25 2,05 2,07 2,09 2,05 1989 2,27 2,31 2,34 2,27 2,02 2,05 2,07 2,02 1990 2,30 2,34 2,38 2,30 1,99 2,03 2,06 1,99 1991 2,29 2,35 2,39 2,29 1,96 2,00 2,04 1,96 1992 2,29 2,36 2,40 2,29 1,93 1,98 2,02 1,93 1993 2,29 2,36 2,41 2,29 1,91 1,96 2,00 1,91 1994 2,29 2,36 2,42 2,29 1,88 1,94 1,99 1,88 1995 2,28 2,36 2,42 2,28 1,86 1,92 1,97 1,86 1996 2,28 2,36 2,43 2,28 1,84 1,90 1,95 1,84 1997 2,27 2,36 2,43 2,27 1,81 1,88 1,93 1,81 1998 2,27 2,36 2,43 2,27 1,79 1,85 1,91 1,79 1999 2,26 2,35 2,43 2,26 1,77 1,83 1,89 1,77 102 quite easy to understand the existence of this "short term superavit". On the other hand, the available information show that both the number of contributors and the revenue from contributions grew at substantially lower rates. Even the wage's payroll indicated a probable deterioration in the social security apparatusto collect contributions. In the 1985/86 period, data from the Monthly Employment Survey (Pesquisa Mensal de Emprego) conducted by IBGE showed that, in the metropolitan areas of Rio de Janeiro and Sao Paulo, the average real wage gains of employees linked to the formal labor market were 13.21 and 21.92 percent, respectively. These numbers were obtained for the period from March to February of each year and took into account the Expanded Consumer Price Index (Indice de Precos ao Consumidor Ampliado - IPCA) as a deflator. For the same period, employment would have grown by 8.2 percent in the Metropolitan Area of Sao Paulo. For the 1986 year, these increases are accounted for by the initial success of the "Cruzado Plan", which by conceding a real wage gain of 8 percent to all wage workers and of 15 percent to the ones on one minimum wage, allowed for a retaking of the economic growth and, consequently, of employment. It is important to understand, however, that the sudden reduction in inflation levels, even though desirable in general terms, might bring quite negative impacts over the social security system equilibrium, as it tendsto increase the average valLie of benefits.') This fact 1) For an in-depth discussion of the subject see Oliveira, Francisco E.B., Beltrao, Kaizo 1. and Ribeiro, Marco A. S., Algumas Consideracoes sobre os Efeitos da Reforma Monetaria no Campo Social: Seguro Desemprego e Previdencia Social,Textos para Discussao Interna, IPEA/INPES, Rio de Janeiro, May, 1986. 103 occurs because the value of benefits is estimated as a weighted average of the last 12 or 36 contribution wages of the contributor, being the last 12 months not subjected to any correction for inflation. Therefore, the higher the inflation, the higher the permanent reduction factor applicable to the value of the benefit. So, if inflation levels are kept low, expenditures with benefits tend to grow at a fast speed. The rate of growth would also be contingent to the renewal in stocks of each type of benefit. The truth is that, in the political sphere, there is a clear trend towards greater liberalization of the benefit's plan, partially motivated by the official speech that "the social security deficit was reduced to zero". This statement, even though true in the short-run, supports the demand for higher value benefits and services in the social security system. The latter even though justifiable on social grounds may not be compatible with the real economic and financial capabilities of the system. It is worth it to emphasize some recent initiatives either taken or under study by the Government: a) Extinction of the social security contribution for retirees and family dependents; b) Creation, in August 1986, by the President, of the Social Security and Social Assistance Superior Council, comprising 21 representative members from the employees, employers and Government; c) Creation, also in August 1986, by the Health, Education and Social Security Ministers, of the Sanitation Reform Committee, also encompassing representatives from the employees, employers and Government; and, 104 d) Implantation, in 1986, of the Working Group for the Social Security System Reform, at inter-ministerial level and comprising, as the former ones, representatives from the employees, employers and beneficiaries. Many proposals have been discussed by these groups. Among the most important of them we single out: a) Diversification of the financing sources for the social security system, by replacing part of the revenue coming from the urban payroll for a share of the operational profit of firms; b) Increase in the progressiveness of the social contributions share by raising the shares on the payroll and by removing the ceiling on the share over wages; c) Improve the plan of benefits of rural workers and of urban workers linked to the informal labor market leading in the long-run to a single plan (Basic Plan of Benefits); d) Extension to rural workers and employers of the medical assistance provided to urban contributors; and, e) Implantation according to family income levels of selectivity criteria in the concession of some benefits of an assistant nature: family allowance, burial support, and so on. Finally, it is also worth mentioning that the Brazilian Social Security System comprises serious problems in what relates to equity that deserve to be seriously studies in the core of future changes in the institutional rules.