Social Security in Brazil: Importance, Problems and Reform Agenda

Transcription

1 Social Security in Brazil: Importance, Problems and Reform Agenda Pedro Jucá Maciel Brazilian National Treasury Ministry of Finance Visiting Scholar at Stanford Center for Latin American Studies August, 12 th 2010 The opinions expressed and arguments employed herein do not necessarily reflect the official views of the Brazilian National Treasury Ministry of Finance

2 Agenda 2 1. Social Security Systems and Public Policy 2. Brazilian Demographic Profile and Future Trends 3. Social Security and Social Development. 4. Social Security and Fiscal Policy in Brazil 5. Brazilian General Social Security Regime (RGPS) 6. Public Sector Social Security Regime (RPPS) 7. Policy Recommendations

3 Social Security Systems and Public Policy 3

4 Why do governments have to be involved in this matter*? High Transaction Costs: Private firms run in a higher administrative costs (including substantial commissions for the salespeople) to provide retirement annuities than public sector. US Department of Labor (1998) estimates in 1% administrative cost of public sector in providing social security against 6% of private insurances companies. Inability of Private Markets to Insure Social Risks: Private firms are unable to provide guaranties against social risks in comparison to the Government. Gov. is in a position to meet its obligations by raising taxes or engaging in risk sharing across generations. Adverse Selection: Life insurance companies are interested in people who will live longer. Retirement annuities companies look for people who will die sooner. In practice, it is costly to identify those risks (due to adverse selection of people). Gov., however, can force all individuals to purchase the insurance. In doing so, it is engaging in some distribution: good risks are paying more and bad risks are paying less. Moral Hazard: This failure is faced by both public and private sector. It happens when individuals have no incentives to reveal their real abilities for a specific situation. If you guarantee individuals 100% retirement annuities after certain age, even people who are productive and healthier to work will have no incentives to continue working. Social Security is a merit good!!!: It is a paternalistic idea that government should require individuals to save money for their benefits in the future (for their own good). In fact, social security systems reduces poverty among elders and the beneficiaries are not only the recipients, but also their children and sons, who otherwise would have to support them or felt guilty in not doing so. * Stiglitz (1999) 4

5 Why we should always being engaged in reforming the system? 5 In Summary: In a social security system Pay-as-you-go (such as the Brazilian one), the dynamic financial sustainability depends on the ratio of the numbers of workers to the number of retirees. Sources of Fiscal Imbalance: Decrease in Population Growth: Brazilian Population Growth is declining very fast! Decrease in Retirement Age: Some groups has special retirement age such as teachers and women. Increase in Longevity: Brazilian population s life expectancy is increasing 0,36 yrs per year. Reasons for Reforms (not only fiscal imbalance justifies reforms in the System): Fiscal Imbalance: It is the essence. It means the financial sustainability of the system. Adverse Effect on Savings: Provision of social security reduces individual s need to save, thus investment and productivity growth (Long Run Growth). Policy makers must design mechanisms (in reforms) to correct incentives and avoid this problem. Adverse Effect on Labor Supply: It induces individuals to retire earlier. Reform should reward workers who keep their labor activities after retirement age. Brazilian Fator Previdenciario works in this direction. Low (perceived) Rate of Return: Pay-as-you-go does not provide a competitive rate of return as regular investments. Complementary private system can be encouraged in order to achieve this goal. Inequities: Need of transparency in system s: (i) subsidies, (ii) beneficiaries, (iii) impact over poverty and income distribution (as a regular social program).

6 Brazilian Demographic Profile and Future Trends 6

7 Brazilian Demographic Profile: Yes, We Have a Problem! Age Age 7 Age Age Source: IBGE Projecao da Populacao (2008 Review)

8 Brazilian Demographic Profile: Age Groups Brazilian population is getting old fast. In 2010, the 65 yrs old or more group represented 6,8% of the population. It is projected to reach 22,7% in 2050 (3 times more). Between (in 10 years!), the share of working age group will start to decline. After 2035, there will be more elders than children in Brazil. Brazilian Population Distribution by Age Group ( ) 8 Source: IBGE Projecao da Populacao (2008 Review)

9 Brazilian Demographic Profile: Growth Rate The 65 yrs old or more group growth rate will increase up to 4.2% a.a. in 2025 and will stabilize in 2.5% a.a. in Working age group growth rate has a decline trend reaching the a negative rate value in Brazilian Total Population will decline after 2040!!!! Geometric Growth Rate of Brazilian Population by Age Group ( ) 9 Source: IBGE Projecao da Populacao (2008 Review)

10 Brazilian Demographic Profile: Dependence Ratio Inverse of the dependence ratio represents the numbers of working age people per elder or youth. This is a very important indicator for the financial sustainability of a payas-you-go retirement systems such as the Brazilian one. Nowadays, there are 10 working age people per Elder. In 2050 there will be only 3! This trends means that the Brazilian Social Security will face real increase in the expenditures (quantity) and smaller revenue collection base => Higher Deficit!. Brazilian Demography: Inverse of the Dependence Ratio Youth and Elders ( ) 10 Source: IBGE Projecao da Populacao (2008 Review)

11 Demographic Profile: International Comparison UN Data show the Brazilian Median Age is getting older very fast, from 29 yrs in 2010 to 46 yrs in Brazilian Population will be as old as the European and older than the Northern America in This is a huge change! World's Population: Historical Data and Projections for Median Age by Region (Review 2008) Year BRAZIL South North America America Europe Asia Africa Oceania Source: Population Division of the Department of Economic and Social Affairs of the UN, World Population Prospects: The 2008 Revision 11

12 Brazilian Social Security and Social Development 12

13 Brazilian Social Security System Coverage: Contribution Most Brazilians workers (48.4%) belong to the Regime Geral de Previdencia Social (General Social Security Regime). Regime Proprio de Previdencia Social (for civil public servants and the military regime) represents 7.3% of contributors. Special Beneficiaries (specific categories): 8.8%. 34.1% of the workers are unprotected by the System (informal economy) where 15,7% (most women) receives less than a minimum wage (MW) and 17,8% more than MW. Social Security Coverage: 16 to 59 years Old Population (2008) 13 ** Rural Workers; *** Social security beneficiaries who is still working, but not paying for the System. Source: PNAD/IBGE 2008 in SPS/MPS (2009)

14 Brazilian Social Security System Coverage: Evolution Social Security Coverage: Evolution of Working Age Population enrolled in the System 14 Social Security Coverage: Evolution of 60 years Old or More Elders who receive social security benefits Source: PNAD/IBGE 2008 in SPS/MPS (2009)

15 Brazilian Social Policy: Programs and Instruments. Brazilian Main Social Programs (Excluding Education and Health Care Programs) 15 Programs Description Cost in 2009 Beneficiaries General Social Security Regime (RGPS) Public Service Regime (RPPS) Social Assistence Programs (LOAS and RMV) It includes retirement benefits (age, length of contribution, and invalidity), survivor pensions, temporary aids, etc. It includes public servants retirement benefits and survivor pensions from Federal, State and Municipal Governments It gives one minimum wage benefit to +65 yrs old elders whose income is less than 1/2 minimum wage (formal or informal sector) or to families with a physically disabled member whose income is less than 1/2 MW. 7.1% GDP 23.2 millions 4.3% GDP 3.1 millions 0.60% GDP 1.6 millions Unemployment Security Benefit given to unemployed formal sector workers for a period of 0.62% GDP 7.3 millions 3 to 5 months long Abono Salarial Annual benefit (one monthly minimum wage) given to all formal workers whose income is less than 2 MWs 0.25% GDP 15.5 millions Bolsa Familia Source: STN (2010) and SPS/MPS (2010) Monthly conditional cash transference given to poverty-stricken families found in a situation of poverty (formal and informal sectors) in a condition of school attendance (5-17 yrs old kids/teens) and vaccination (0-5 yrs old kids). 0.42% GDP 11.1 millions Object of this research Instruments (Price Effect): Minimum Wage Adjustment: It is the minimum benefit for the social security system (RGPS, RPPS, LOAS and RMV) representing 2/3 of the total benefits and influence also private sector formal workers salaries (informal sector salaries indirectly). Adjustment Rate Rule = Inflation t-1 + GDP t-2. Above 1 Minimum Wage Retirement Adjustment Index: Government has increased slightly less than the minimum wage (most times just by the inflation rate).

16 Brazilian Social Security System: Impacts on Inequality Brazilian Social Security System (General and Public Regimes) impacts positively income distribution. It reduces Gini Income Concentration Index by 0.04 p.p.!!!however!!! Public Service Regime (alone) creates a very negative impact on income distribution. According to World Bank (2005), 50% of the total benefits are given to the richest 20% of the population and only 10% to the poorest 20%. Income Inequality (Gini Index) and the Social Security System in Brazil Source: Castro (2009)

17 Brazilian Social Security System: Impacts on Poverty 17 Violet line represents the observed poverty line and the light violet line represents a estimated poverty line considering the nonexistence of social security benefits. Social Security system has an important role in poverty reduction in Brazil mainly among the oldest segments of the population. Young age groups poverty reaches almost 50%!!! Observation: this information does not evaluate the efficiency of poverty reduction comparing Social Security system to other social programs. Percentage of Population Earning less than ½ Minimum Wage (Poverty) considering Social Security Income (2008) Source: PNAD/IBGE 2008 in SPS/MPS (2009)

18 Is raising Social Security Benefits Value (Minimum Wage) an Efficient Way to Combat Poverty? Minimum Wage 2008: R$ 415. Definition poverty line 2008: per capita income = R$ 200. Extreme poverty = R$ 100. Poverty Level: Total Poverty = 22% of the population, Extreme Poverty = 8% of population. Most beneficiaries are non-poor families with no children.!!!97.8% of extreme poor families are excluded of the program and 80% of them have at least 1 child!!! Social Security Benficiaries: Distribution of Families According to Poverty Level and Numbers of Children (2008) Families Non-Poor Poor Total Extreme Poor Total Numbers mi mi mi Families Beneficiaries of Social Security 35.5% 9.0% 2.2% No Children 80.0% 25.8% 11.2% 1 Child 14.1% 28.1% 13.2% 2 Children 4.5% 23.1% 23.8% 3 Children 1.1% 14.1% 24.5% 4 or More Children 0.3% 8.9% 27.2% 18 Families Non-Beneficiaries of Social Security 64.5% 91.0% 97.8% No Children 47.9% 17.2% 19.7% 1 Child 32.2% 30.2% 29.2% 2 Children 15.8% 28.0% 22.9% 3 Children 3.4% 14.9% 14.8% 4 or More Children 0.7% 9.7% 13.4% Source: Giambiagi and Tafner (2010) based on PNAD 2008

19 Social Security Benefits Value Adjustment (Minimum Wage Increase) 19 vs. Bolsa Familia Expansion Simulation proposed by Giambiagi and Tafner (2010): In 2009, the government increased the minimum wage (lowest benefits) in 5.8% + inflation and benefits above the minimum wage was adjusted by inflation (5.9%) Base Scenario. If government adjusted the value of all benefits just by inflation and re-distributed theses savings to the Bolsa Familia Program, what would be the impact over extreme poverty? Answer: We could have reduced the extreme poverty 13% more than the current minimum wage policy! Simulation: Increase in the Social Security Benficiaries VS. Expansion of Bolsa Familia Program Benefit Type Cost in 2008 (R$ mi) (i) Real Increase Adjustment for the Minimum Wage and (ii) Inflation to the others Cost in 2009 (R$ mi) Adjustment equal to Inflation for all beneficiaries Adjustment equal to Inflation (all beneficiaries) and redistribution of the remaining balance to the Bolsa Familia Retirement Pensions Total Social Security Expansion Bolsa Familia Total Spending % Poverty 26,25% 25,81% 26,98% 25,98% % Extreme Poverty 9,89% 9,78% 9,80% 8,50% % Change Extreme Poverty -1,11% -0,91% -14,05% Source: Giambiagi and Tafner (2010) based on PNAD 2008

20 Social Security Systems and Fiscal Policy 20

21 Brazilian Spending on Social Security: International Comparison In 2005, Brazilian Public Sector spent 10.2% of GDP in retirement benefits (old age benefit and pensions) in both General (RGPS) and Public Sector (RPPS) Regimes. This number is 32% higher than the average of OECD countries who have a much older demographic profile! Moreover, IBGE Population Review 2008 shows the Brazilian Demography Profile is getting older very fast and will be similar to European countries in 2050! It seems we have a big problem to solve! 14% 13% 12% 11% 10% 9% 8% 7% 6% 5% 4% 3% 2% 1% 0% Public Expenditure on Retirement Benefits 2005 (% GDP) Korea Ireland Iceland Canada New Zealand Australia Netherlands United States Slovak Republic United Kingdom Norway Switzerland Luxembourg Czech Republic Denmark OECD Turkey Spain Belgium Hungary Finland Japan Brazil* Portugal Sweden Germany Source: OECD Pensions at a Glance 2009 and STN (2009) * Brazilian Data include RGPS and RPPS (Federal, State, Municipal Governments) and exclude LOAS. ** OECD countries data include in kind (non-cash) benefits (eg. housing) and mandatory private. 21 Poland Greece France Austria Italy

22 Brazilian Spending on Social Security: International Comparison Dependence Ratio (in the Graph below) = proportion of 65 yrs old population in relation to the yrs old population. Brazil is an outlier! Its spending is relatively much higher to the population profile compared to other countries. International Comparison: Expenditure on Social Security System vs. the Dependence Ratio 22 Source: Rocha and Caetano (2008) based on OECD, UN, World Bank and other Domestic Sources.

23 Public Spending: Importance of Social Security in 2008 Social Security System is the most important sector of public spending in Brazil. It took 31% of the public sector budget in It represents more than 2 times the spending on Education. Looking on the bright side Reforms in the Social Security System can raise considerable amount of resources to be re-allocated in other important development areas such as infrastructure and human capital. Brazilian Public Sector Expenditure in 2008 (By Sectors) Defens e, 2.2% Others, 8.1% 23 Labor/Work, 2.4% Urbanism, 3.4% Social Assistance, 3.7% Social Security, 30.9% Transportation, 3.9% Public Security, 4.0% Judiciary, 4.9% Administrative, 6.4% Education, 14.3% Health, 14.0% Source: Secretariat of the National Treasury (2009)

24 Social Security Taxation: International Comparison Brazil has one of the highest tax rates to finance its Social Security System. Giambiagi (2000) recognizes the Brazilian Government have increased payroll. contributions rates due to population aging associated with generous rules. Social Security Revenues come from payroll taxes. High Payroll taxes is an argument used by some Brazilian companies (sectors) to justify the Brazilian international lack of competitiveness and also one of the reasons (incentive) for the high rate of economy s informality and low rate of employment. International Comparison: Taxation for the Social Security System Funding 24 Source: Rocha and Caetano (2008) based on OECD, World Bank and other Domestic Sources.

25 Brazilian Social Security Spending: Recent Trends 25 Expenditure on the General Social Security Regime (RGPS) 7,50%. % G D P 7,00% 6,50% 6,00% 5,50% Expenditure: General Social Security Regime spending increased from 5% to 7.2% in the last 12 yrs. 5,00% 4,50% 4,00% Source: STN (2010) Public Service Retirement Regime (RPPS) spending has presented a more stable trend in relation to the GDP, around 3.7%, but it is still very high compared to other countries. % G D P 4,0% 3,5% 3,0% 2,5% 2,0% 1,5% 1,0% 0,5% 0,0% Expenditure on Public Servants Retirement Benefits by Government Level Source: STN (2009) Federal State Municipal Total

26 Brazilian Social Security Spending: Recent Trends General Social Security Regime is not only the most expensive component of the Federal Budget, but also it is also the component that most increased from 1991 to If the Brazilian Society doesn t wish to increase tax burden in the future (the current consensus in Brazil), for a given fiscally responsible scenario, the social security reform is a necessary and inexorable agenda! % GDP 8% 7% 6% 5% 4% 3% 2% 1% 0% Source: STN (2010) G ro w th R ate (a.a.) 8,0% 7,0% 6,0% 5,0% 4,0% 3,0% 2,0% 1,0% 0,0% Federal Spending Growth Rate Between (% a.a.) 26 Transferences Payroll General Social Other Current Total Primary GDP to States and Security and Capital Spending Municipalities Source: OECD Regime Economic Expenditures Survey: Brazil (2009) Source: Giambiagi and Tafler (2010) Federal Government Primary Spending ( ) Payroll General Social Security Regime Other Current and Capital Expenditures Others

27 Social Security Deficit: General Social Security Regime (RGPS) General Social Security Regime: Quantity and Cost of Benefits by Category (April 2010) 27 QUANTITY COST (R$) AVERAGE COST (R$) CATEGORY Total % do total Beneficiaries Beneficiaries Beneficiaries % do Total TOTAL total Urban Rural Urban Rural Urban Rural TOTAL , , ,62 778,55 465,49 GENERAL REGIME BENEFITS , , ,46 836,76 464,73 Social Security Contributory , , ,13 846,78 464,83 Retirements , , ,13 908,70 464,51 by Age , , ,24 641,02 463,69 by Invalidity , , ,83 687,33 467,53 by Length of Contribution , , , ,82 675,12 Suvivor Pension , , ,44 716,73 463,92 Temporary Benefits , , ,75 859,20 492,14 Sickness Benefits , , ,93 873,23 509,04 Partial Invalidity , , ,21 478,19 255,91 Imprisionment Benefit , , ,39 597,59 497,80 Maternity Benefit , , ,56 532,46 473,93 Others 617 0, , ,03 477,03 Labor Accident Insurance , , ,31 640,98 438,25 Retirement by Invalidity , , ,22 869,64 469,39 Suvivor Pension , , ,33 749,54 484,51 Sickness Benefits , , ,49 926,92 507,16 Partial Invalidity , , ,79 445,10 255,64 Supplementary Benefit , , ,07 136,07 SOCIAL ASSISTANCE BENEFITS , , ,04 511,11 509,37 Social Assistance Pension (LOAS) , , ,19 509,19 for the Age , , ,48 509,48 for the Impaired , , ,93 508,93 Lifelong Indemnization Pensions , , ,71 965,71 Old Social Assistance Benefit (RMV) , , ,02 508,73 509,37 for the Age , , ,73 509,60 509,86 for the Impaired , , ,77 508,46 509,17 OTHER TREASURY OWED PENSIONS (EPU) SOURCE: DATAPREV, SUB, SINTESE , , , ,07

28 Social Security Deficit: General vs. Federal Public Service Regime Fiscal deficit comparison between the General Social Security Regime and the Federal Public Regime reveals an very unfair situation where 1 mi beneficiaries (Fed. Public Regime) cost (in terms of deficit % GDP) more than 23 mi beneficiaries in the General Regime. Reason? Generous rules that Public Service Regime used to enjoy (changed only by 93, 98 and 2003 reforms): (i) no retirement minimum age, (ii) low rate of contribution (4%), (iii) 1988 Constitution public regimes unification, and so on... As public servants contribute for social security based on the full salary (no limit) and they have the retirement benefit based on the last salary (or even active workers salaries!), their average benefits are much higher than the General Regime. Social Security Balance Federal Government (% GDP) Public Servants Regime (RPPS): 1.0 mi Beneficiaries* 28 Revenues 0.3% Expenditure 2.0% Balance -1.7% General Social Security Regime (RGPS): 23.2 mi Beneficiaries Revenues 5.8% Expenditure 7.2% Balance -1.4% Total Federal Social Security Regime Revenues 6.1% Expenditure 9.2% Balance -3.1% Source: Giambiagi and Tafner (2010) * SIAPE Dez 2009 Source: SRH/MPOG Boletim de Estatistica de Pessoal Jan 2010

29 Brazilian General Social Security System (RGPS) 29

30 Social Security: a Historical Perspective 1920 to 1930 (Eloy Chaves Law): - Elloy Chaves Law (1923) was the first regulatory law for the Caixas de Aposentatorias e Pensoes (CAP) - system organized (mostly) by state owned firms. The system was characterized by few numbers of members, many institutions and low amount of financial resources to 1960 (Vargas s Model): - In order to increase the number of members of the pension system, the Government created the Institutos de Aposentadorias e Pensoes IPA for the unionized professionals. The system expanded in terms of number of members and funding to 1980 (Military Regime): - Social Security Law (1960) standardized IPA s pension regulatory characteristics. Moreover, the law expanded the system for employers and autonomous workers. - Military regime expanded the system for rural workers (1969), maids (1972) and temporary workers (1974). It launched new benefits such as sickness/accident insurance and maternity benefits (salary) and changed rules that used to guarantee the financial equilibrium of the system such as contagem do tempo de trabalho ficticio. - Ministerio da Previdencia (2009) recognizes theses reforms as the main source of the social security financial imbalance after the 80 s. Military Government used to give social security benefits to compensate decrease and extinction of political rights ( moeda de troca ) 1980 to 1990 (Constitution): Constitution brought new reforms (benefits): (i) minimum benefit (piso) equal to one minimum wage (before Constitution there were different values for rural and urban workers); (ii) 5 yrs reduction for rural workers retirement age; (iii) establishment of a unified regime for public servants (including servants who were not in the this system); (iv) expansion of proportional retirement benefit for women. 30

31 Basic Concepts and Current Rules for the RGPS Basic Concepts of the General Social Security Regime: Minimum Benefit: One Minimum Wage (R$ 415 in Apr/2010) Maximum Benefit: Social Security Cap (R$ in Apr/2010) Benefit Salary: It is the basic value used to calculate the retirement benefits. It represents the average of the worker s 80% highest salaries since July Retirement Benefit = Benefit Salary x Social Security Factor. Social Security Factor: It is calculated taking into consideration the age, life expectancy and length of contribution (Women receive 5 yrs contribution bonus): LenghtContr 0.31 Age + ( LenghtContr 0.31) F = 1 LifeExpec SOCIAL SECURITY FACTOR (2008) RETIREMENT AGE C 30 0,518 0,536 0,556 0,576 0,598 0,619 0,643 0,670 0,695 0,721 0,753 O 31 0,536 0,555 0,575 0,596 0,619 0,640 0,666 0,693 0,719 0,747 0,780 N 32 0,555 0,574 0,595 0,617 0,640 0,662 0,689 0,717 0,744 0,772 0,807 T 33 0,573 0,593 0,615 0,637 0,661 0,684 0,712 0,741 0,768 0,798 0,833 R 34 0,592 0,612 0,635 0,658 0,683 0,706 0,734 0,764 0,793 0,823 0,860 I 35 0,610 0,632 0,654 0,679 0,704 0,729 0,757 0,788 0,818 0,849 0,887 B. 36 0,629 0,651 0,674 0,699 0,726 0,751 0,781 0,812 0,843 0,875 0, ,648 0,670 0,694 0,720 0,747 0,773 0,804 0,837 0,868 0,901 0,941 T 38 0,666 0,690 0,715 0,741 0,769 0,795 0,827 0,861 0,893 0,927 0,968 I 39 0,685 0,709 0,735 0,762 0,791 0,818 0,850 0,885 0,918 0,953 0,995 M 40 0,704 0,729 0,755 0,783 0,812 0,840 0,874 0,909 0,943 0,979 1,023 E 41 0,723 0,748 0,775 0,804 0,834 0,863 0,897 0,934 0,969 1,006 1, ,742 0,768 0,796 0,825 0,856 0,886 0,921 0,958 0,994 1,032 1,078 Elaboração: SPS/MPS. Proposed Simulation: 31 On average, Brazilians start to work at 20 yrs old. When they reach 60 yrs old, the factor will be They will receive 2% more of the Benefit Salary when they retire on the age of 60 yrs. This age is lower than most developed and developing countries apply.

32 Types of Benefits (RGPS) and LOAS (Social Assistance) Retirement due to Age: It is given to the insured who has a minimum age of 60 yrs for women and 65 yrs for men (Rural workers can retire 5 years earlier). Minimum length of contribution is 15 years. The benefit correspond to 70% of the benefit salary + 1% per year of contribution. Retirement due to Length of Contribution: It is given to the insured who has a minimum length of contribution of 35 yrs for men and 30 yrs for women (Rural workers and teachers can retire 5 years earlier). Minimum age = 53 yrs (men) and 48 yrs (women). The Retirement Benefit correspond to 100% of the Benefit Salary times the Social Security Factor. Transition rules apply for workers who were enrolled in the system before Special Retirement: Benefit is awarded to the working insured who exercise activities subject to special conditions that are harmful to their health or physical integrity. According to the risk, length of contribution varies from 15 to 25 years. Disability Retirement: Benefit is awarded to the insured who is totally or definitely incapacitated for the job. Minimum of 12 months of contribution. Retirement Benefit is equal to 100% of the Benefit Salary. Survivors Pension: This benefit is given to the dependants in the case of the death of the insured. The rights for this benefit go as follows: husband/wife/companion; non-emancipated child under 21 yrs and handicapped child of any age. The value correspond to 100% of the retirement benefit that the insured used to receive. LOAS / RMV (Social Assistance Program): This program gives one minimum wage benefit to +65 yrs old elders whose income is less than 1/2 minimum wage (formal or informal sector) or to families with a physically disabled member whose income is less than 1/2 MW. 32

33 Retirement Minimum Age: Life Expectancy 33 Life Expectancy is a usual argument (from people who are against reforms) to justify differences in the minimum retirement age in Brazil compared to other countries. However, life expectancy at birth is not a good indicator for the social security system (especially in developing countries) because it is biased by the high infant mortality rate. The most appropriate indicator is the life expectancy at age 60 or 65. Brazilian life expectancy at birth increased 18 yrs from 80 s to 2005 (men) and 15.8 yrs for women. Life expectancy at 60 yrs increased 3.2 yrs for men and 5.3 for women. Brazil has a considerable difference in the life expectancy at birth indicator compared to European countries, however, at age 60, the differences are just 2 yrs (m) and 3 yrs (w). Thus, life expectancy can not justify significant differences in the retirement age between Brazilians and Europeans countries. They have close numbers and Brazilian society will reach the European demographic profile soon (IBGE projections). Life Expectancy at Birth and at Age 60 According to Gende for Men and Women Countries Brazil Men Women 1970/ / Life Expectancy at Birth Life Expectancy at Age Average European Countries Life Expectancy at Birth Life Expectancy at Age Source: Giambiagi and Tafner (2010)

34 Retirement Minimum Age: International Comparison The Brazilian average retirement age is lower than most countries. The main reason is the still existence of the retirement due to length of contribution regime combined with a very low minimum age (53 yrs men and 48 yrs women). Length of Contribution Regime represents almost 50% of the retirement expenditure and 30% of the numbers of beneficiaries. The average retirement age in this regime is 54.4 yrs (men) and 51.3 yrs (women). It is almost 10 years younger than the average of OECD countries. Expected retirement lengths in OECD countries are 16 yrs (men) and 21 yrs (women) and in Brazil 23 yrs men and 29 yrs women. In the case of women, it is almost the same length that they have to work (contribute)! Teachers Special Treatment: Teachers retirement age is 5 years less than the regular categories. There is no reasonable justification for this privilege (e.g. health risk or harmful activities). If the Brazilian society believes this is a special (honorable) profession, raise their salaries now! Don t give earlier retirement compensation. Retirement Age: International Comparison 34 (Average) Source: Rocha e Caetano (2008)

35 Women Retirement Age 35 Equal rights, but equal responsibilities too! In a modern society, it is clear that women have a key role in the labor market. They have better education indicators (on average) and able to share market s jobs in a equal position to the men. Recent data of labor market show that the participation rate of women in the labor force increased from 35% in the 1990 to 50% in 2008 (men reduced from 75% to 70% in the same period). The Brazilian Social Security System should not pay the price of any kind of discrimination against women. 66% of the countries in the World apply same statutory age for men and women. Within the minority group, which Brazil belongs (34% countries who apply different ages), we usually find a difference in less than 5 yrs (the Brazilian rule). Social Security Systems - International Comparison of Gender Staturory Pensionable Age Rules Region Number of Countries According to Statutory Pensionable Age by Gender Different Ages Same Ages Percentage of Countries According to Statutory Pensionable Age by Gender Different Ages Same Ages The Americas Europe Asia and the Pacific Africa 7 36 TOTAL % 44% 42% 19% 34% 68% 56% 58% 81% 66% Brazil Applies Diferent Ages Source: Social Security Systems Throughtout the World 2008 and yrs Men / 60 yrs Women

36 Survivor Pensions: Does it need to be reformed? 36 Survivor pensions represent 25% of the social security spending in Brazil. Women represents 88% of the program beneficiaries. The Brazilian survivor pension spending is approaching to 3,5% GDP. Again, Brazil is an outlier compared to other countries [OECD (2003)]. Brazilian Pension System still have generous rules compared to the other countries. Criteria Brazil International Experience Minimum Length of Contribution Survivor Pensions Rules: International Comparion None Elegibility Criteria Requires a minimum period of contribution to request the benefit Marriage Status Beneficiary doesn t need to be married Minimum (Period) Length of Marriage or Civil Union Age No minimum age New Marriage Status Pension doesn't change in a new marriage situation Restrictions apply specially for suvivors younger than 45 years old Usually, pensions are cancelled after a new marriage Item Brazil International Experience Number of Dependents Influence Reduction of the benefit in case the beneficiary receives other source of income None. Pensions are always 100% independently of the numbers of dependents. None Age Influence None Benefit Calculation Formula Source: Soares and Caetano (2008) based on Missoc (2005) and Social Security Administration (2005) Usually, pensions represents 70% of the previous retirement benefit. Additionally, there are 10% more per dependent until it reaches the maximum value of 100% Reduction of the value applies. In some case, it is prohibited acumulate pension and other source of income. Usualy, younger survivors receive lower benefit value than olders

37 Social Security Assistance: LOAS and RMV In April 2010, Social Assistance Benefits (LOAS) represent 10% of the total expenditure of the Social Security Ministry and 13% of the beneficiaries of the overall Social Security System. LOAS spending is increasing fast and the coverage of the benefits among the >= 65 yrs old population raised from 6.2% in 1996 to 12.7% in Poverty Reduction: PNAD data reveal LOAS is an efficient program for poverty reduction (example: 55% of the beneficiaries are illiterate and most beneficiaries would be below the poverty line without the benefit). Problem 1 (Incentives): The problem associated to this program is the creation of bad incentives in the labor market. As the LOAS beneficiaries receives the same benefit as most formal worker retiree (minimum wage), there is no incentive (premium) for workers (who receive a minimum wage salary) to be in the formal sector. Problem 2 (Income Benefit Calculation): Giambiagi and Tafner (2010) show that LOAS can create a unfair situation due to its calculation formula. Reason: LOAS benefit is not considered individual s income for the S. Security System. Problem 2 (an example): Expenditure on Social Assistance Benefits (LOAS) 0.65% 1) If a married couple has a member retiree by the RGPS who receives a minimum wage (MW), the 2 nd member 0.60% will not be qualified to receive the LOAS benefit 0.55% because their per capita income will be exactly ½ MW. 0.50% 2) However, if a married couple has a member (who was in the informal sector all life) and already receives 0.45% LOAS benefit (a minimum wage), the 2nd member will 0.40% be eligible to receive the LOAS benefit because their 0.35% per capita income will be zero as LOAS benefit is not 0.30% considered income for this purpose. 3) So the married couple who used to work in the 0.25% informal sector will receive twice more benefits than couple who was in the formal sector. % G D P Source: STN (2009) 37

38 Social Security Assistance: Rural Retirement Rule 38 Importance: Rural Workers retirement systems are an important progressive program that does not only creates incentives for workers stay in the rural sector (avoiding migration to the urban centers) but it is also important for small municipal s public accounts and economies. Given its non-contributive bias (Figure below), this program works as a social assistance program. Rule Part 1: Rural Retirement Benefit is given just for subsistence agriculture (agricultura familiar) workers. Their contributions are based on a fraction of commercialized product from worker s excess production. Rule Part 2: However, during the program s implementation, it became clear that most workers rarely had excess production. Thus, The Constitution guaranteed the right of retirement for rural workers independently of their contribution. They just need to prove their required working time (length) [5 yrs less than the urban workers]. Urban vs. Rural Revenues and Spending (2009)!!!Important!!! Although it looks like the social security system is running on surplus if you don t consider rural retirement system, this doesn t mean we don t have a problem to face. Source: SPS/MPS (2010) According to Caetano and Soares (2008), LOAS and Rural Workers Programs just explain ¼ of the distortion of the social security (high) spending in Brazil compared to other countries.

39 Social Security Assistance: Rural Retirement Rule Rural Workers Retirement Regime is under discussion on the Brazilian Congress. According to Campos (2006), the main topics for its reform are: 1) Improve the parameters that define the concept of subsistence agriculture (agricultura familiar). 2) Expansion of the benefits for other categories (such as rural workers who live in urban areas and temporary agriculture workers). 3) Discuss the working length in the subsistence agriculture eligibility criteria as a parameter for the benefit request. 4) Maintenance of the contribution criteria based on the excess of commercialized agriculture products. 5) Development of an identification systems for rural workers beneficiaries % Expenditure on Rural Workers Retirement Benefits 1.40% 1.35% %GDP 1.30% 1.25% 1.20% 1.15% 1.10% 1.05% 1.00% Source: STN (2010)

40 Public Sector Pension Regimes 40

41 The size of the Public Pension System The size of the System: The Brazilian Public Service Pension System is large and complex. It includes employees of Executive, Legislative and Judiciary Branches and also for Army Force (Federal) and Military Police (State Level). There are 27 states and more than municipal systems. According to OECD (2010), their rules differ slightly among each other, however they are usually more generous than the private system (RGPS). 41 Source: MPS Overview of the Brazilian Social Wefare (2009). Inverse of Dependence Ratio = # Working Employees / # (Retirees + Pensioners): In a Pay-as-you-Go System it is important you observe the number of people who contribute for the financial sustainability of the system. - Federal Government = 1:1 - It very low compared to international/sub-national levels and, considering the average age of federal employees (next slides) this is a real problem! - State Governments = 2:1. - Municipal Governments = 4:1.

42 Rules and Recent Reforms of Public Sector Regime: Length of Contribution: It requires 35 years for men and 30 years for woman. Minimum Retirement Age (only after 1998 Reform!!!): Men 60 years and women 55 years Incentives Structure: Creates incentives (premiums) for servants who will postpone retirement request after eligibility criteria [2003 Reform]. Contribution (% Salary): 11% of salaries (Before 1993, public servants just needed to contribute 4%!!!). Contribution Base and Cap: Nowadays, the mandatory contribution is based on full salary and the benefit is equal to the final salary (or Active workers salaries). Only after 2003 Reform regulation (Compl. Law) public servants will have a contribution and benefit caps- same limit of private sector workers. Complementary Contributive Regimes: 2003 Reform establishes the possibility to implement complementary regime, however it is still dependent on a Complementary Constitutional Law. Supportive Retired Servants Contribution: 2003 reform required retired public servants to continue contributing to the system (but less than the working employees). Limit Earlier Retirement for invalidity or illness: Annuities based on the contributive period (proportional). Forced Retirement: is given to men and women after 70 years old. Annuities Cap (2003 Reform): Application of general wage cap (Federal, State and Municipal) avoiding abusive high benefits. 42

43 Evolution of Benefit of the Public Service Regime Disability retirement shows a wave behavior throughout the last 13 years. It reached 30% of retirement requests in 2004 and 2005! However, It is observed a decline trend over the past 5 years. It seems the Federal Government has improved its controls over these benefits. Forced Retirement (compulsoria) reached 10% in 2005, however nowadays is less than 5%. Thus, voluntary retirement represents 90% of the requests. 43 Source: SRH.MPOG (2010)

44 Retirement Age of Federal Public Servants It seems the 2003 Reform was effective in increasing the retirement age from 58 (men) in 2003 to 61 in 2008 and 54 years for women in 2003 to 58 (2008). However, we still have 75% of retirement applications requested by people younger than 65 yrs old (minimum retirement age in most countries). 44 Dec/2009

45 Where do Retirees/Pensioners of the Fed. Gov. Come From? 45

46 International Comparison of Public Service Regimes Palacios and Whitehouse (2006), Bontori (2002) and OECD (2005 and 2007) recognize the Brazilian Public Sector Pension System as one of the most benevolent in the world. In fact, Brazil spends MUCH more in the civil pension system than most countries in the world (available data). 46 Source: Palacios and Whitehouse (2006) Source: Palacios and Whitehouse (2006)

47 Government Workers Aging in Brazil vs. OECD Brazilian Public Sector must be ready to deliver new services for a society that is getting older faster. Brazilian public sector servants are older not only compared to other OECD countries, but also to the Brazilian overall labor force. Almost 40% of servants are more than 50 yrs old! Old Labor Force brings challenges and opportunities to the Brazilian Public Sector: - Sooner there will be a fast growth in the government retirement system spending. - Replacement of old workers can provide a very important opportunity to: (i) reallocate labor in most pressured areas (demand for each public services changes over the time), (ii) bring new people with new competencies and qualifications and (iii) reduce payroll cost (new workers receive lower salaries than older). Percentage of 50 Years Old or Older Public Sector Workers (Federal Government) in Brazil and OECD 47 Source: OECD (2010)

48 International Experience in the Public System: OECD General Idea: More than 50% of OECD countries have special regimes for public employees. OECD (2010) recognizes that public special regimes are usually more generous than private sector workers regimes because: (i) Public annuities are usually based on the final salary (rather than on the contribution history); (ii) Permit provisions for earlier retirement; (iii) Recognition of non-contributive periods such as for education and children care. Most OECD countries are engaged in reforms for the Public Service System. The general trend is to extinct special rules for public employees in favor of equality between private and public workers (Unify these systems). This would encourage mobility between private and public sectors workers (with productivity gains for both sectors) OECD Countries Reforms Trends [OECD (2010)]: Annuities Calculation Factor: Most OECD countries increased their contribution period to 40 years and changed their calculation formula to harmonize with private sector. Increase the retirement age: Most countries require 65 years and they also harmonized the retirement ages for men and women. Reduction in the compensatory benefits: Pensions are reduced (or banned) if verify that the survivor (dependent) has already sufficient income for living [Austria Case]. Limit Earlier Retirement: Discourage workers to request earlier retirement (e.g. for invalidity or illness) by calculating the benefit proportional to the contributive period. Create Incentives for Deferred Retirement Option: Create financial incentives for employees to continue working after the retirement age. Implement/Development Complementary Contributive Regimes: Reduce the mandatory public contributive value and incentive contributive complementary regimes. 48

49 Policies Recommendations 49

50 Policy Recommendations: General Principles for Reforms 1. Gradualism: avoiding political difficulties for reform implementations [Giambiagi and Tafner (2010)] Retirees and pensioners should not suffer any change in their rules as the Brazilian Constitution states (Direitos Adquiridos). Among active workers, reforms should be applied based on the previous working experience. Older workers should receive less changes proportionally to younger workers. The young population who have not entered the labor market yet will receive the full reform changes. First Best vs. Feasible Reforms: Sure, the gradualism reform principle will only provide Brazilian convergence to international benchmarks in the long run (more than 30 years), however this is possibly the only way to approve reforms due to political difficulties. 2. International Comparison Brazil is not a isolated island in the world s economy. International experience comparisons should be considered as a very important reference for future reforms. Fact is: Brazil is spending much more in the social security system, given its demographic profile, than other countries. It is also observed that Brazil still have generous rules that permit workers to retire earlier or benefits that last longer. Given the huge change in the population dynamics (that we will face throughout the next 40 years), the time for adjustments is now! 3. Isonomy: Long Term Convergence Among Retirement Regimes There is no reasonable justification to keep different rules for public servants (all levels), militaries, teachers and other general workers. Special Regimes: professions whose exercise activities are subject to special conditions harmful to health or physical integrity should continue to be treated differently. 50

51 Policy Recommendations: the New Common Regime 1. Establishment of a 65 years old Minimum Retirement Age International Evidence: Most developed countries are applying 65 years old minimum age and many are reforming their system to reach this level. As the social security reform will (probably) be fully implemented just for new workers (over the next 30 years), the Brazilian demography that time will be the same as Euro Countries. If any worker found himself in a situation that he can not work until 65 yrs old, he should obviously apply for the invalidity retirement, following this criteria rule. 2. Length of Contribution should not be used as a Criteria for Retirement Requests, but as a base to a New Social Security Factor which gives a premium (higher retirement value) for workers who have longer contributive period. Few countries in the world (most in the Middle East) adopt length of contribution as a criteria for Retirement Requests. As the minimum age for this kind of retirement criteria is just 53 years old for men (48 yrs women), the cost due to this criteria is still very high. Correcting Incentives: length of contribution should be considered as a premium in the value of the benefit but workers must complete 65 yrs old to request the benefit. 3. Apply Same Age for Statutory Retirement Age between Men and Woman International Evidence: Brazil belongs to the minority group (1/3 of countries) who still discriminates retirement ages for men and women. This proposal can save considerable amount of resources in the long-run as the fraction of women in the labor force is increasing. 4. Apply for Teachers the same Statutory Retirement Age as other workers. If the Brazilian society wants to give a benefit, rise their salaries now, not in the future! Low salaries and early retirement policy creates no incentives to attract the best professionals to the teachers career which is an essential policy to improve education quality in Brazil. 51

52 Policy Recommendations: the New Common Regime 5. Benefit Cap (Teto Previdenciario) Value Adjustment just by Inflation According to Giambiagi and Tafler (2010), 95% of salaries in Brazil are below the benefit cap. Average retirement benefit is equal to 130% of the average salaries of working age population. These indicators are higher than most countries experience.?why Brazil (Brazilians) doesn t have a higher saving rate? One of the reason is simple: consume as much as possible today and depend on the government tomorrow. Uncertainty creates incentives to people increase savings (Keynes). Government should encourage people to develop their own strategy of savings. Complementary System: International experience shows that if people desire to receive your last working salary in the retirement, they have to engage in a complementary plan or invest their money in other opportunities (e.g. savings, real estate, stocks, Tesouro Direto, etc). 6. Reforming the Survivor Pensions System Convergence of the Brazilian generous rules to the international experience (Slide 36). Changes needed to be addressed to eligibility criteria (length of contribution, age, marriage status) and benefit value calculation formula (number of dependents, age and other source of income). 7. Public Service Regime (RPPS) Regulation of the 2003 Reform must be designed considering the long term convergence of both system. Retirement Benefit should be based on the average contributive history, not in the active workers salaries. Higher Spending: Assuming a retirement substitution rate of 1 (1 retirement will imply in 1 hiring), Brazilian Federal Government will have its despesas de pessoal under pressure in medium term (40% of total employees are 50 yrs old or older). Public service personnel renew provides opportunities to the government for better human resources management (Slide 47), however planning is needed [OECD (2010)]. 52

53 Policy Recommendations: Social Assistance Programs 1. Reforming the Social Assistance Programs (LOAS and RMV) Avoiding Distortions: LOAS benefit must be considered as income for the social security system. Correcting Incentives (politically difficult): Fix the value of the benefit as a share of the minimum wage (such as 90%). This reform aims to create incentives for workers to stay and contribute to the social security system (formal sector economy). Nowadays, many formal workers receive a minimum wage as a retirement benefit which is equal to the LOAS benefit (informal sector workers). 2. Rural Workers Retirement System Rural workers retirement system spent R$ 45 bi in 2009 (3 times more than Bolsa Familia). As a social program, we have to monitor its effectiveness for poverty reduction and its incentives for people to stay in the rural sector (avoiding urban problems). Possibly, the most difficult problem in the Rural Retirement System design is related to the information asymmetry given the particularities of the rural activities. How can we select and subsidize who really needs? and How can we improve our tax collection from people who can contribute more (reducing the deficit)? It is important to improve government s information of the rural sector (not only for the social security system, but also for the general revenues collection). 3. Minimum Wage Rule If we apply today s Minimum Wage Rule in the long run (real increase), all Brazilian workers will receive a minimum wage someday. It is clear this rule is unsustainable. Brazilian society needs to decide what is its minimum wage long term value and define strategy to gradually reach that level. The most important binding constraint to increase the value of the minimum wage is the social security spending. 53

54 Policy Recommendations: Administrative Reforms* 1. Benefits Administrative Reforms: Periodical Beneficiaries Re-registration and Better Death Registration Systems. Sickness Temporary Benefit Requirement: It became an important topic/program due to its increase from 0.5 million requests in 2000 to 1.7 million in The idea is to create incentives for (i) a better diagnosis of the beneficiary disease/problem (avoiding frauds), (ii) workers to reveal their real health situation (if they are able to return to work) and (iii) better integration of the Social Security and the health care network. Rural Regime: Today s proofs of rural sector employment required by the Social Security are very flexible (e.g. registers in commercial firms or churches). It permits people who never worked in rural activities to request the benefit. 2. Revenues Administrative Reforms: Information System Improvement for Data Analyses: Information exchange provide clues of tax evasion through data analysis (as the Income Tax System works). Exchange can be implemented between (i) General Revenue Tax System (SRF) and Social Security System or (ii) GPS and GFIP. Information System Integration between Federal and Social Security Revenues : Integration of the systems can result in gains through economies of scale and synergies. 3. Social Security Debt s Recovery Reforms: Compensation between Federal and Social Security Revenues Social Security Debt Payment through Judiciary Credits (Precatorios Previdenciarios). Limit Debt Schedule Payments Length Social Security Debt Securitization. Classification of Debts to be Received by Recovery Risk Degree. 54 * (Caetano [2006])

55 55 Thank you! Contacts:

56 Backup Slides 56

57 Brazilian Demography Profile: Growth Rate 57 Source: IBGE Projecao da Populacao (2008 Review)

58 International Experience in Social Security: Europe p Source: Social Security Programs Throughout the World (2008)

59 International Experience in Social Security: Europe p Source: Social Security Programs Throughout the World (2008)