Abstract
This Selected Issues Paper discusses the macroeconomic implications of pension reforms in Brazil. It assesses empirically the relationship between fiscal policy and the real effective exchange rate in emerging markets and draws policy implications. It reviews the current status of local capital markets in the country, the key challenges, and policy options for further development. The paper also provides a detailed description of consumer credit developments and analyzes recent indicators of household financial distress associated with the credit expansion that has taken place in the last couple of years.
I. Macroeconomic Implications of Pension Reform in Brazil1
The long-term pension challenges facing Brazil are well documented. Recognizing these, the authorities have over the years sought to advance reforms of the systems. An important signal of this commitment has been sent by the recent reform of the public system. Much of the debate over the years has focused on the fiscal implications of the outlook for and possible reforms of the pension system. However, different reform options can have very different macroeconomic implications, including for savings, growth, and external balances. To illustrate these differential impacts and inform the debate on the issue, this paper simulates the general equilibrium effects for Brazil of various pension reform options that have been used in countries around the world. All options examined help address the system’s long term funding gap and are conducive to raising real private savings and growth in the long run. However, we find that reforms that involve lower mandatory contributions or higher retirement ages have larger effects on output though a boost in labor supply. Meanwhile, reforms focused on reducing benefits would promote growth mostly through a larger impact on private savings.
A. The Brazilian Pension System
Current Structure and Fiscal Position
1. The Brazilian public pension system currently comprises three defined-benefit schemes:2 a mandatory private sector regime (Regime Geral de Previdência Social, RGPS), currently covering some 23 million beneficiaries and disbursing around 6½ percent of GDP, a mandatory public sector regime (Regimes Próprios de Previdência Social, RPPS), with about 1 million beneficiaries3 and a disbursement of some 2 percent of GDP, and a noncontributory means-tested branch for rural workers, disabled people and other low income families, which disburses less than ½ percent of GDP. Benefits are financed out of current proceeds from an 8 to 11 percent payroll tax paid by employees, a 20 percent contribution tax by employers (which also finances other social insurance benefits such as for sickness and maternity) and two other specific taxes.4 Both contributions and benefits are capped in the RGPS, but in the RPPS only future participants will be subject to such rules (see below).
2. Pension spending in Brazil is very high by international standards, considering the relatively young Brazilian workforce. Indeed, both the RGPS and RPPS are currently running deficits—1 percent and 1.4 percent, respectively, in 2010—as a consequence of relatively generous replacement rates, a low average retirement age5 and current indexation rules. The indexation of minimum pensions to the minimum wage is a particularly large driver of overall pension costs. About 40 percent of total spending pertains to beneficiaries receiving the minimum pension (2/3 of RGPS beneficiaries), which has more than doubled in real terms over the past 15 years. Nonetheless, it has contributed for the important reductions in old-age poverty seen in Brazil.
Pension Spending in International Context, 2011
Citation: IMF Staff Country Reports 2012, 192; 10.5089/9781475506716.002.A001
Source: IMF Fiscal Board Paper “The Challenge of Public Pension Reform in Advanced and Emerging Economies” and staff calculations.Pension Spending in International Context, 2011
Citation: IMF Staff Country Reports 2012, 192; 10.5089/9781475506716.002.A001
Source: IMF Fiscal Board Paper “The Challenge of Public Pension Reform in Advanced and Emerging Economies” and staff calculations.Pension Spending in International Context, 2011
Citation: IMF Staff Country Reports 2012, 192; 10.5089/9781475506716.002.A001
Source: IMF Fiscal Board Paper “The Challenge of Public Pension Reform in Advanced and Emerging Economies” and staff calculations.3. Staff estimates that the pension system faces an NPV funding gap of close to 25 percent of GDP over the next 20 years, rising to 100 percent through 2050.6 Under current rules, the financing needs of the social security system should undergo a modest rise in the coming 20 years, when the population is still relatively young. After that, the funding gap will increase sharply as the old age dependency ratio is expected to continue rising steeply (to over 60 percent in 50 years, from today’s 10 percent).
4. Comparatively, the RPPS has traditionally offered especially advantageous conditions, including very high replacement rates (still equal to 100 percent for participants who started service before 2003, compared to an average of the best 80 percent monthly salaries during the working life in RGPS),7 a short entitlement period (only 10 years of civil service to qualify for an RPPS pension) and the indexation of pension benefits to the salaries of active civil servants instead of inflation. This explains why deficits in the two subsystems are of the same order of magnitude, even though the RGPS has a much wider coverage.
Recent Reforms
5. Important changes to the RPPS were first enacted in 2003, including steps for the ongoing establishment of a dual pillar system (2012 reform). Faced with mounting pension costs—a 1998 pension reform had a relatively limited impact on curbing deficits—and in a context of a rising external risk premium, a reform of the pension sub-system for civil servants was introduced in 2003 to enhance long term fiscal prospects. The reform introduced a number of parametric changes: an 11 percent payroll tax on pension benefits,8 lower replacement rates (harmonizing the rules with RGPS for new civil servants), and the penalization of early retirement of 5 percent of benefits per year (before age 60 for men and 55 for women).9 Importantly, it also set the stage for the creation of a fully funded pillar for public servants, which institution was finally approved by the Senate in March 2012.
6. The 2012 reform introduces a defined contribution pillar to the RPPS. Benefits and contributions for new civil servants will be subject to the same ceilings as those in the RGPS, while participants have the option to enroll in a complementary defined-contribution scheme (Previdência Complementar) if they wish to receive a pension beyond the ceiling. Current active civil servants may choose to stay in the old system or switch to the new one with two pillars. Participants can choose how much to save into a retirement account, knowing that the employer will match their investment by up to 8.5 percent of their salary. At the time of retirement, they accrue the returns from this investment.10
7. During its first stage of implementation, the 2012 reform will generate a net cost driven essentially by the loss of contributions to the pay-as-you-go (PAYG) branch. The state will also be making transfers to the individual pension accounts on behalf of employees. However, since the reform affects only the RPPS subsystem and the contribution ceiling is relatively high—only 1/3 of servants earn beyond the correspondent salary base—, the transition cost is expected to be contained (about 0.1 percent of GDP).
8. As the new generations of civil servants retire and disbursements will be lower, the government will reap the benefits of the 2012 reform. On net terms, the authorities expect an improvement in the balance of the RPPS from 2033 onwards, with gains rising to 0.4 percent of GDP per year in the long run. Staff estimates point to an overall impact of around 10 percent of GDP in NPV terms up in the long run.
9. The introduction of a funded pillar into the RPPS is welcome. By reducing replacement rates for higher earners, it is expected to encourage long-term private savings and thereby support the development of financial markets. Progressiveness within RPPS system is also enhanced, as well as equity vis-à-vis private sector workers. Finally, the relatively small transition cost is an important consideration for sustainability of the reform—especially given that the fiscal framework in Brazil is anchored by a primary surplus target—in light of international experience where costly pension transitions have at times lead to some unwinding of the pension reform. The reform may thus be a stepping stone for further improvements to the system down the road.
B. Macroeconomic Implications of the 2012 reform
10. We now assess the broader macroeconomic implications of the recent reform. The analysis uses the IMF’s Global Integrated Fiscal and Monetary (GIMF) model parameterized on data for the Brazilian economy.11 The GIMF is a non-Ricardian, dynamic stochastic general equilibrium model which features—overlapping generations, finite horizons (myopia), and endogenous labor and capital markets—allowing for a meaningful discussion of the short and medium run impact of pension reforms.
11. Our baseline is an economic environment reflecting pre-Lehman fiscal trends. In particular, data as of 2007 was used to parameterize initial levels of government spending, revenue decomposition and transfers (including pensions), thereby abstracting from cyclical impact of the recent crisis on these variables. Net public debt is assumed at 40 percent of GDP in the initial steady state, close to 2011 levels.
12. The 2012 reform is introduced as a shock, first to contribution rates and later to pension benefits. By capping mandatory contributions to the PAYG pillar, the government will effectively be lowering average (mandatory) contribution rates for public servants. Based on the estimated transition cost shown in Figure 2, we proxy that change by the shift in labor taxes that, in the model, would produce such a cost (up to its peak in 2035). In other words, we assume that contributions to defined benefit schemes are generally perceived by participants as a tax, unlike what would happen in an optional defined contribution plan.12 After 2035, the average contribution rate is kept fixed and the fiscal trajectory thereafter is dictated by the reduction in pension benefits for new entrants.13 As will be shown, the quantitative impact of the reform is small in broad macroeconomic terms; but this is only a consequence of the circumscribed scope of the reform in terms of affected beneficiaries. The results do suggest a high elasticity of private savings and growth rates to the implied fiscal savings in the context of this particular reform.
Estimated Fiscal Impact of the 2012 Reform
(Percent of GDP)
Citation: IMF Staff Country Reports 2012, 192; 10.5089/9781475506716.002.A001
Source: Ministry of Finance.Estimated Fiscal Impact of the 2012 Reform
(Percent of GDP)
Citation: IMF Staff Country Reports 2012, 192; 10.5089/9781475506716.002.A001
Source: Ministry of Finance.Estimated Fiscal Impact of the 2012 Reform
(Percent of GDP)
Citation: IMF Staff Country Reports 2012, 192; 10.5089/9781475506716.002.A001
Source: Ministry of Finance.Macroeconomic impact when the reform is financed by public debt
13. For illustrative purposes, we analyze first the effects of the 2012 reform under the assumption that it is financed by public debt. The resulting path of the relevant fiscal variables, in deviations from the baseline scenario, is shown in Figure 3a.14 Primary balances worsen in the first 20 years and improve subsequently, like in the estimated net cost/benefit shown in Figure 2 above. Implicitly, the initial rise derives from the reduction in the average contribution rate—as weighted by the share of agents in the economy subject to the new pension rules—of 0.2 percentage points by 2027. Pension benefits start falling on that date, and the system matures with disbursements 0.4 percent of GDP lower than the baseline. Since public investment does not change and the impact on GDP is small (see below), net public savings mirror the dynamics of the overall deficit.
The 2012 Reform with Debt Finance: Dynamics of Fiscal Variables
(Deviations from steady state. Periods correspond to years.)
Citation: IMF Staff Country Reports 2012, 192; 10.5089/9781475506716.002.A001
Source: IMF staff calculationsThe 2012 Reform with Debt Finance: Dynamics of Fiscal Variables
(Deviations from steady state. Periods correspond to years.)
Citation: IMF Staff Country Reports 2012, 192; 10.5089/9781475506716.002.A001
Source: IMF staff calculationsThe 2012 Reform with Debt Finance: Dynamics of Fiscal Variables
(Deviations from steady state. Periods correspond to years.)
Citation: IMF Staff Country Reports 2012, 192; 10.5089/9781475506716.002.A001
Source: IMF staff calculations14. Labor supply increases, pushing up real GDP growth. The drop in compulsory contribution rates reduces a labor market distortion, raising marginal incentives to work and thereby increasing the return on capital.15 Real investment is thus higher, although with some recoil in the medium run because real interest rates rise during that period. With both higher labor supply and capital accumulation, real GDP increases above its long term trend during the first years, which ends up putting pressure on prices. Monetary policy adjusts by hiking real interest rates temporarily.
15. As expected, the reform builds incentives to raise private savings. Faced with higher net income during their working life and foreseeing lower pension transfers in the future, individuals accumulate savings during the next 30 years, incidentally in the form of optional contributions to the second pillar of the pension system. The subsequent drop in government transfers reduces disposable household income and thus the private savings ratio to GDP, but in NPV terms private savings increase. Since agents are myopic—one of the non-Ricardian features of the model—and a share of the population is liquidity constrained, consumption is not perfectly smoothed. Furthermore, private savings undershoot in the medium run, as the long run decline in pensions is not fully internalized by the currently active population.
16. National savings, on the other hand, stay roughly constant during the transition period. Initially, higher households’ savings are simply traded off by government debt accumulation. However, as pension benefits drop permanently, so do private savings, the public debt ratio and interest payments. There is therefore a small rise in national savings rates over the very long run, owing almost exclusively to lower public debt service. With the investment ratio mostly constant over the entire period, the current account starts improves permanently once the transition period is over.
Although a low impact of multi-pillar reforms on total savings has been observed in a number of emerging economies, the existing evidence is far from conclusive.16 In countries like Chile, Peru, and Latvia national savings rose in the aftermath of the reform, but in other cases—including Colombia, Mexico, and Uruguay—it either remained unchanged or dropped slightly. The relationship between these types of reforms and savings is nonetheless hard to pin down in the long term because the latter depends on a myriad of factors. In practice, reform packages often include parametrical changes with adverse effects on private savings, as is the case of increases in the retirement age (see below). Furthermore, an important determinant of the impact on savings is the financing strategy for the transition cost, as we will argue in the next paragraphs. Finally, the relatively short time period since most of multi-pillar reforms were introduced—particularly in Eastern Europe—makes it hard to fully assess the impact of those in household savings and labor incentives. Indeed, reforms of pension systems in the 1990’s advanced economies such as Sweden have been associated with increases in household savings, but these behavioral changes have been observed gradually over long periods.
The 2012 Reform with Debt Finance: Impact on Macroeconomic Variables
(Deviations from steady state. Periods correspond to years.)
Citation: IMF Staff Country Reports 2012, 192; 10.5089/9781475506716.002.A001
The 2012 Reform with Debt Finance: Impact on Macroeconomic Variables
(Deviations from steady state. Periods correspond to years.)
Citation: IMF Staff Country Reports 2012, 192; 10.5089/9781475506716.002.A001
The 2012 Reform with Debt Finance: Impact on Macroeconomic Variables
(Deviations from steady state. Periods correspond to years.)
Citation: IMF Staff Country Reports 2012, 192; 10.5089/9781475506716.002.A001
Trends in Gross Domestic Savings
(Percent of GDP)
Citation: IMF Staff Country Reports 2012, 192; 10.5089/9781475506716.002.A001
Sources: World Bank, Pension Reform and the Development of Pension Systems, An Evaluation of World Bank Assistance, IEG and WDI2011*Last available data pointTrends in Gross Domestic Savings
(Percent of GDP)
Citation: IMF Staff Country Reports 2012, 192; 10.5089/9781475506716.002.A001
Sources: World Bank, Pension Reform and the Development of Pension Systems, An Evaluation of World Bank Assistance, IEG and WDI2011*Last available data pointTrends in Gross Domestic Savings
(Percent of GDP)
Citation: IMF Staff Country Reports 2012, 192; 10.5089/9781475506716.002.A001
Sources: World Bank, Pension Reform and the Development of Pension Systems, An Evaluation of World Bank Assistance, IEG and WDI2011*Last available data pointMacroeconomic impact when the reform is financed by government savings
17. Given the primary surplus fiscal target in Brazil, the next scenario assumes that the transition cost is financed by a reduction in government consumption.17 Labor taxation and pension benefits still follow the same path as before, but government consumption now mirrors their combined budget impact, such that the primary balance doesn’t change. Thus, changes to the government’s overall balance debt reflect solely the small variation in interest rates.
Government Consumption/GDP
(Difference)
Citation: IMF Staff Country Reports 2012, 192; 10.5089/9781475506716.002.A001
Government Consumption/GDP
(Difference)
Citation: IMF Staff Country Reports 2012, 192; 10.5089/9781475506716.002.A001
Government Consumption/GDP
(Difference)
Citation: IMF Staff Country Reports 2012, 192; 10.5089/9781475506716.002.A001
18. As before, the fall in contribution rates promotes labor supply, investment and real growth—however, the impulse to national savings turns positive at all dates. Since the medium-to long term decline in total transfers—pension and others—is higher in this case,18 the impact on aggregate labor supply is stronger, which in turn brings up real GDP by more than double the amount found when the reform is financed through debt. At the same time, the decline in government consumption keeps government savings close to the baseline level. National savings will then rise already in the short to medium term, arguably in the form of increased savings in household retirement accounts.
The 2012 Reform with Debt Finance: Impact on Macroeconomic Variables
(Deviations from steady state. Periods correspond to years.)
Citation: IMF Staff Country Reports 2012, 192; 10.5089/9781475506716.002.A001
The 2012 Reform with Debt Finance: Impact on Macroeconomic Variables
(Deviations from steady state. Periods correspond to years.)
Citation: IMF Staff Country Reports 2012, 192; 10.5089/9781475506716.002.A001
The 2012 Reform with Debt Finance: Impact on Macroeconomic Variables
(Deviations from steady state. Periods correspond to years.)
Citation: IMF Staff Country Reports 2012, 192; 10.5089/9781475506716.002.A001
C. Macroeconomic Implications of Alternative Pension Reform Options19
19. Reflecting existing high costs, which will be exacerbated by demographic transition, further adjustments to the social security system will be needed in the future. Efforts will likely need to focus not only on the RPPS but also the RGPS. Although less generous, the private sector subsystem covers a much wider range of the population and is therefore bound to be most affected by ageing. Furthermore, as discussed above, the NPV of the 2012 reform is modest when compared with the actual pension gap.
20. We now present some illustrative simulations of the macroeconomic effects of possible alternative approaches, as future reforms are considered in Brazil. For the purposes of the analysis here, the focus will be on parametric changes to the pay-as-you-go (PAYG) systems that reduce their financing needs other than those associated with an expansion of the defined-contribution pillar. However, our simulations from the previous section suggest that the latter could be beneficial (by reducing the threshold further and increasing the importance of pension savings accounts), should the government identify fiscal space to finance the transition cost.
21. With a medium to long term horizon in mind, the analysis uses average pension spending in G20 countries as an indicative benchmark for Brazil. Convergence to such an average would imply a decline in social security disbursements in Brazil of about 2 percent of GDP, practically eliminating the projected social security deficit, barring ageing pressures. In these simulations we will assume that such reforms could be phased-in over the next 20 years. A gradual implementation of this nature would be associated with a higher sustainability of the associated reforms over time, which Brazil can afford as it still enjoys the demographic dividend of a young labor force with a low overall dependency ratio.
22. Two types of general instruments are considered for convergence to such a benchmark—lowering benefits and increases in the retirement age,20 in both cases assuming that the government would keep its primary surplus target unchanged.21 The former can be achieved through a number of specific policies, such as the reduction of net replacement rates (either directly or by revisiting the formula of the factor previdenciário22) or a change in the minimum pension indexation rules. For simplicity in the simulation, we will assume that pension reductions are evenly spread across liquidity constrained and unconstrained agents. Average retirement ages could increase directly through a hike of the minimum retirement age or by penalizing early retirement (for instance, extending the minimum time of contributions required to qualify for a full pension).
23. The decline in benefits raises private savings, investment, and labor supply, although with a modest real GDP gain of 0.8 percent over 20 years. When the reform is announced, current workers and beneficiaries internalize the permanent decline in future pension benefits. Thus, consumption immediately drops and savings rise, putting downward pressure on interest rates. Labor supply increases—as consumption and leisure are complementary goods—which, together with the lower interest rate, encourages higher private investment. As transfers decline, so does disposable income and consumption will continue converging to a lower level. For the same reason, the private savings ratio eases in the long run, but it is still permanently higher than in the baseline. In all, the national savings ratio increases on the back of higher private savings and a permanent (albeit small) reduction in public debt service.
24. By contrast, an increase in retirement ages depresses savings in the short- to medium term, but has a large positive impact on investment and output growth (6 percent increase). Agents foresee a shorter retirement period at the time the reform is announced, and immediately decrease savings. Interest rates go up, but the substantial rise in labor supply improves returns on capital so much that private investment rises in equilibrium. This furthers demand pressures, contributing to higher interest rates. In the medium- to long run, output rises significantly, and so does household disposable income. Thus, private savings rebounds in real terms, although its ratio to GDP is permanently lower than in the baseline. With a slightly higher deficit, the national savings ratio falls permanently in this case, although again mostly due to a denominator effect.
Decrease in Pension Benefits: Impact on Macroeconomic Variables
(Deviations from steady state. Periods correspond to years.)
Citation: IMF Staff Country Reports 2012, 192; 10.5089/9781475506716.002.A001
Decrease in Pension Benefits: Impact on Macroeconomic Variables
(Deviations from steady state. Periods correspond to years.)
Citation: IMF Staff Country Reports 2012, 192; 10.5089/9781475506716.002.A001
Decrease in Pension Benefits: Impact on Macroeconomic Variables
(Deviations from steady state. Periods correspond to years.)
Citation: IMF Staff Country Reports 2012, 192; 10.5089/9781475506716.002.A001
Increase in Retirement Age: Impact on Macroeconomic Variables
(Deviations from steady state. Periods correspond to years.)
Citation: IMF Staff Country Reports 2012, 192; 10.5089/9781475506716.002.A001
Increase in Retirement Age: Impact on Macroeconomic Variables
(Deviations from steady state. Periods correspond to years.)
Citation: IMF Staff Country Reports 2012, 192; 10.5089/9781475506716.002.A001
Increase in Retirement Age: Impact on Macroeconomic Variables
(Deviations from steady state. Periods correspond to years.)
Citation: IMF Staff Country Reports 2012, 192; 10.5089/9781475506716.002.A001
D. Conclusion
25. Current levels of pension spending in Brazil are high by international standards, particularly as the country is now enjoying the peak of its demographic dividend. The generosity of the system is believed to hold back private savings, investment and aggregate labor supply. If unadjusted, over time, spending pressures here will compromise the achievement of primary surplus’ targets without substantially squeezing discretionary spending or further increases in already high—by emerging market standards—income and consumption taxes.
26. Our simulations suggest important effects on macro variables such as savings and growth of different parametric adjustments that have been used in other countries and might possibly be considered in Brazil. In this note we discussed the macroeconomic impact of different parametric reform options, starting with the recently approved introduction of a defined contribution scheme for the public sector subsystem. We found that pension reforms increase real private savings and growth, although the elasticities to the implicit fiscal savings are quite different across the different options. Reforms that involve an increase in retirement ages or a decline in average contribution rates are supportive of higher growth through their positive income in labor supply and investment, even if the impact on savings is not necessarily higher than in options mostly focused on a reduction of benefits.
References
Arnold, J., 2011, “Raising Investment in Brazil”, OECD Economics Department Working Papers, No. 900.
Attanasio, O., Rohwedder, S., 2003, “Pension Wealth and Household Saving: Evidence from Pension Reforms in the United Kingdom,” The American Economic Review 93(5), pp. 1499–1521.
Caetano, Marcelo A., 2008, “Previdência complementar para o serviço público no Brasil”. Sinais Sociais, Vol. 3. No. 8, September/December, Rio de Janeiro: SESC.
Disney, R., 2005, “Household Saving Rates And The Design of Social Security Programmes: Evidence From A Country Panel”, CESifo Working Paper No. 1541.
International Monetary Fund, 2011, The Challenge of Public Pension Reform in Advanced and Emerging Economies, IMF Fiscal Board Paper.
Kumhof, M., D. Laxton, D. Muir, and S. Mursula, 2010, “The Global Integrated Monetary and Fiscal Model (GIMF)—Theoretical Structure”, IMF Working Paper 10/34 (Washington: International Monetary Fund)
Mascarenhas, R., Oliveira, A. and Caetano, M., 2004, “Análise Atuarial da Reforma da Previdência do Funcionalismo Público da União,” Ministério da Previdência Social do Brasil, Coleção Previdência Social, Série Estudos, Vol. 21.
Medici, A., 2004, “The Political Economy of Reform in Brazil’s Civil Servant Pension Scheme,” Inter American Development Bank, Technical Note on Pension No. 002.
OECD, 2005, OECD Economic Surveys – Brazil, Paris: OECD.
OECD, 2011a, 2011 Economic Review—Brazil, Paris: OECD.
Projeto de Lei No. 1.992-A, de 2007, Camara dos Deputados, República Federativa do Brasil.
U.S. Social Security Administration, 2009, Social Security Programs Throughout the World: The Americas, Washington: SSA Publication No. 13-11804.
Prepared by Joana Pereira.
In addition, there is also a growing network of private pension funds (mostly defined contribution). Participation in these schemes is voluntary and the government only plays a regulatory/monitoring role therein.
Applies to federal workers only.
Contribution to Social Security Financing (Contribuição para o Financiamento da Seguridade Social, COFINS) and Social Contribution on Net Profits (Contribuição Social sobre o Lucro Liquido, CSLL).
The average retirement age is 53 for RGPS beneficiaries who contributed a minimum of 35 years (for men) or 30 years (for women). Among retirees who contributed for a shorter period and among RPPS retirees the average age is close to 60.
The estimates are based on the authorities’ actuarial projections of RGPS and RPPS financing needs as of April 2012, as well as of the fiscal impact of the latest reform (see paragraphs 5 to 9) recent FAD projections of public pension increases - IMF Fiscal Board Paper “The Challenge of Public Pension Reform in Advanced and Emerging Economies.”
Furthermore, benefits in the RGPS are subject to an adjustment factor (factor previdenciário)—based on age and length of contributions—which was introduced in 1999 to account for changes in average life expectancy.
Levied on the portion exceeding 60 percent of the RGPS’s benefits ceiling, for all participants.
The minimum retirement age is 53 for men and 48 for women, provided that participants qualify for a full pension by time of contribution.
The defined contribution scheme is to be administered by a newly created Fundação de Previdência Complementar do Servidor Público Federal (Funpresp), divided into three branches for servants in the executive, judiciary and legislative power, respectively. Members of the Funpresp’s Executive Board and Financial Committee are appointed by the government, but the institutions enjoy administrative independence and are subject to a private legal regime (like public enterprises).
A detailed outline of the GIMF model can be found in Kumhof et al (2010). For calculations in this paper, the model features three regions: Brazil, Emerging Asia, and Rest of the World.
We are also assuming that the co-payments by public employers to the optional pension savings accounts are perceived as part of the tax-rate reduction, and participants would take it into account when targeting a desired pension savings amount.
The GIMF features two types of agents: a group of liquidity constrained households (LIQ agents), who do not have access to capital markets, and intertemporal optimizers (OLG agents), who can borrow and save. In this section, we assume that reductions in PAYG benefits affect OLG agents only because in reality only the highest earners will be affected.
To guarantee the dynamic stability of the model, the deficit and debt to GDP ratios need to stabilize in the long run, which in our simulations requires that the fall in primary surplus cannot made permanent. Thus, we assume general transfers to rebound after 2075.
Labor and capital are complementary factors of production in the GIMF.
The World Bank’s Independent Evaluation Group notes a generally small impact on national savings in the short to medium run in its report on Pension Reform and the Development of Pension Systems, An Evaluation, 2006. A number of papers also note a very high (low) substitutability between pension savings with a high (low) actuarial component and other kinds of financial wealth (e.g. Attanasio and Rohwedder (2003) and Disney (2005)), hinting at a low overall impact of these reforms. However, Arnold (2011) points to the generosity of the PAYG pension system as one of the main causes of low savings in Brazil.
Likewise, when the reform produces a net benefit (after 2035) we assume that government consumption rises accordingly. This assumption does not impact significantly the short-to medium term macro-impulses, but it leads to lower government saving rate and higher private savings in the very long run because transfers are permanently lower in this case (see footnote 11).
See footnote 14.
Reform scenarios explored in this section are illustrative.
Staff estimates—based on the total current level of pension and life expectancy at average retirement age—that an increase of 3 years in average retirement ages is needed to produce savings of 2 percent of GDP. Considering a 20 percent lower participation rate for workers older than the current average retirement age, such a rise corresponds approximately to a 7 percent increase of the workforce.
We adjust government consumption so that the target is met. Public investment or taxes could be used instead, although the effects on real GDP will be harder to identify in that case.
See footnote 4.
