At the start of 2016, President Dilma Rousseff announced a two-pronged strategy to deal with Brazil’s deteriorating public sector accounts: in the long term, curb government spending through social security reform; meanwhile, balance the public sector accounts by raising taxes.
The announcement had a target audience: the majority of the Brazilian population, in Congress and outside, that opposes further tax increases and wants the government to balance its books by cutting spending. To them, Rousseff offered her bargain: raise taxes until the end of her term -- in particular, by approving the highly unpopular financial transaction tax (CPMF) – and she will in turn fix the social security system, even if this means alienating the political forces that still support the Workers’ Party (to which she and former president Lula belong). Should society accept her bargain?
It is crucial that social security reform is again under consideration. Generous retirement rules and a demographic transition that is outpacing that in developed economies have led to a rapid rise in social security expenditures, which in turn has contributed to push public sector spending up and jeopardize the fiscal balance. It will be virtually impossible to consolidate Brazil’s public sector accounts without making the rules of Brazil’s social security system less generous.
Brazil has two main public social security systems: one for civil servants and another for workers in the private sector. The latter is managed by the INSS (National Institute of Social Security). For the last two decades, most of the pressure on the fiscal accounts has come from the INSS: between 1997 and 2015, INSS outlays expanded at a whopping 6.4% per year in real terms, on average, climbing from 4.9% to 7.4% of GDP (Figure 1).
Two leading factors explain the steep rise in INSS outlays: the rapid expansion in the number of beneficiaries and the large increase in the mean value of the benefits paid to them. The Ministry of Social Security does not report the number of beneficiaries. We can, though, approximate its dynamics by that of the number of benefits INSS has paid: the same beneficiary can received more than one benefit (e.g., his or her own pension, plus the retirement benefits of a deceased spouse). The numbers are revealing. Considering only the month of December of each year, in 2000-15 the number of INSS benefits went up by an average 3.5% per year, while the mean benefit value rose 2.1% per year (Figure 2).
Figure 1: Expenditures with private sector social security system (INSS),
in constant R$ of 2005 and % GDP
Source: Ministry of Finance.
Figure 2: Number and mean value of benefits paid by the INSS in December (billion benefits and constant R$ of Dec 2015)
Source: Social Security Statistical Bulletin (Ministry of Social Security).
The growing number of benefits is consistent with the rapid expansion of the population aged 60 or more, the main clientele of the INSS. In 2000-15, the number of Brazilians in that age bracket grew, on average, 3.5% per year. This compares to average annual expansions of 1.1% for the overall population and 1.5% for the working age population (15-59-year-olds).
In turn, the mean benefit value went up due to the rise in the real value of the minimum wage: as of December 2015, two out of every three benefits paid by the INSS amounted to exactly one minimum wage. And, in 2000-15, the government raised the minimum wage by an annual average of 4.6%, on top of the increase in consumer prices. This compares to an expansion of 1.5% per annum in per capita GDP.
Absent reforms, social security spending will continue to climb ahead of GDP:
1. The government projects that over the next ten years the number of Brazilians aged 60 or more will expand 4.0% per year (0.7% for the overall population).
2. Current legislation determines that the minimum wage will rise each year according to the inflation rate in the previous year, plus the rise in GDP two years before, if this was positive, or zero, otherwise.
3. Assuming that GDP will fall 3% in 2016, stagnate in 2017 and grow 2% per year in the following eight years, we should expect INSS outlays to reach 10% of GDP by 2025!
Thus, to be meaningful, Rousseff’s reform will have to significantly slow down the rise in the number and mean value of benefits. And the new rules need to kick in relatively fast, before the INSS becomes overwhelmed by unmanageable liabilities, which will then take decades to go away.
Rousseff and her cabinet have been rather vague about the content and timeframe of the reform she has in mind (see, in particular, the interview by Miguel Rosseto, minister of Social Security, to Valor Econômico). So far, the two most positive ideas raised by government officials include:
(i) Establish a minimum retirement age, which could eventually become the same for men and women. Currently, the mean retirement age in Brazil is 58 years, so presumably the legal minimum age would be set at or above 60.
(ii) Unify all the existing social security systems, setting equal rules for civil servants and private sector workers and for urban and rural workers.
On the other hand, the government has also made clear that:
(iii) Eliminating the link between the minimum wage and the value of social security benefits is off the table.
(iv) There will be a lengthy transition period both to start implementing the new rules and then before they become fully effective. Tentative dates would be 2027, to start changing the rules, and 2040, to complete the transition process.
(v) There is no consensus within the government about what should be in the reform proposal. This will only emerge after extensive negotiations within the working group created with this purpose, which consists of government officials and representatives of labor unions, business organizations and civil society.
$1(vi) The government hopes to reach a consensus with all key players on a reform proposal by mid-2016. Yet, if it does not, it will go on negotiating until a consensus emerges.
This is a clear recipe for failure. The ideas raised so far fall seriously short of what is needed to prevent an explosion in social security expenses. The level of government commitment is extremely low. The suggested timeframe means changes will only come when it is too late.
Such working procedures and the lack of clarity and commitment make clear that the government likely intends little beyond offering bait to incentivize the approval of higher taxes. More likely than not—and the government’s track record in this area is very telling—it will abandon any intention to meaningfully change the social security system as soon as it gets its side of the bargain. Congress and society in general should not accept these nonsensical negotiation terms.