Public Disclosure Authorized

The Argentine pension system, its success and

challenges
Contributions to an informed policy debate
Public Disclosure Authorized

Ignacio Apella1

1. Introduction
The pension system in Argentina is the main recipient of public resources. This explains
spending values of around 12% of GDP, equivalent to 25% of consolidated public spending; that is,
taking into account public spending by the national government, subnational jurisdictions, and
municipalities.2 The magnitude of such spending is a source of concern to different sectors of the
political and academic circles, due to both its current level and its impact on fiscal accounts within a
context of tough restrictions, as well as in terms of medium and long-term perspectives. This high
Public Disclosure Authorized

level of spending relates to a much-extended coverage (it practically reaches out to cover the totality
of adults over the age of 65 who receive pension benefits), while the average value of the benefits they
receive is also significant.
Discussions on pension policy involve three dimensions: coverage (that is, the number of
individuals receiving benefits), adequacy of benefits, and fiscal and economic sustainability.
Given that it is impossible to optimize these three dimensions simultaneously, the policy challenge is
to seek agreements in balancing these three dimensions, and once this balance is attained to ensure
their sustainability over time.
Concerns about and interventions on some of these three dimensions were evident at different
points in history. In the first place, during the late 80s the difficulty related to financial/actuarial
sustainability. After implementing a structural reform in the early 90s, which implied the incorporation
of a capitalization pillar and stricter conditions of access3, difficulties in financing --associated to
Public Disclosure Authorized

transition costs-- became the main concern in the short to medium term. To cope with such transition
costs, new taxes and a 15% of the co-participation revenues were incorporated on provincial
jurisdictions. Secondly, stricter eligibility terms had an impact on system coverage. During the first
decade of the 21st century, once the macroeconomy was more stable after the economic crisis that
resulted in the failure of the convertibility plan in 2002 and following a growth period fostered by a
boom in commodity prices, the focus of attention shifted to the system’s gap in coverage. To tackle

1 Senior Economist for Social Protection and Labor Global Practice, World Bank. The author wishes to thank Juan Martín
Moreno, Julian Folgar and Montserrat Pallares-Miralles for their valuable comments and contribution in the preparation
of this document. Special thanks to Jordan Schwartz, Pablo Gottret and María Eugenia Bonilla-Chacin for their support
to this project. Contact iapella@worldbank.org
2 Ministry of Finance

(https://www.argentina.gob.ar/hacienda/politicaeconomica/macroeconomica/gastopublicoconsolidado)
3 The retirement age for women was increased by 5 years, going from 55 to 60, and for men from 60 to 65. Total years of

contribution also increased by 5 years, going from 20 to 30 years.

-1-
this issue, two semi-contributory benefits programs were introduced, the moratorias previsionales (social
security moratorium): the first one was adopted in 2005 (Act Nº 24476 and Executive Decree
N°1454/05) and the second in 2014 (Act N°26970), further extended in 2019; later on, in 2016 the
Pensión Universal para el Adulto Mayor – (PUAM [Universal Pension for the Elderly])4 was created and,
finally, in 2021 a program recognizing contributions for care services was implemented.
After reaching out to cover almost one hundred percent of the passive population, financial
sustainability becomes once again the eye of the storm and triggers reform-centered debates
and discussions. As the system shifted to a scheme that combined contributory and non-
contributory sources of funding and benefits, this implied a higher level of spending, exacting a heavy
toll on the financial sustainability during the second decade of the 21st century.
The Sistema Integrado Previsional Argentino (SIPA [Argentine Integrated Social Security
System]) is the main component in the Argentine pension system; it is a scheme resulting
from a complex combination of rules and regulations that have accumulated over the years.
Subsequent reforms throughout its history have led to a situation that sees the coexistence of
beneficiaries retired from different special pension regimes (cajas previsionales) that existed up until the
80s; the Sistema Integrado de Jubilaciones y Pensiones (SIJP [Integrated Retirement and Pensions System]),
a multi-pillar system with individual capitalization created in 1994; the special social security
moratorium regime adopted in 2005 and extended in 2014 and 2019; the SIPA, created in 2008 and
the Universal Pension for the Elderly (PUAM), created in 2016. The complexity stemming from this
multiplicity of reforms, together with non-compliance (or partial compliance) with the constitutional
provisions that established the mobility of pensions (that is, they must be adjusted periodically), has
resulted in a system that is highly involved in judicial proceedings, as most beneficiaries have, at some
point, filed legal actions on the grounds that their pension benefits had been altered. Accordingly, a
considerable degree of uncertainty was generated, which affected both beneficiaries and system
managers.
Given the current configuration and future demographic trends, the Argentine pension
system faces two major obstacles relative to equity and sustainability in the medium to long-
terms. On the one hand, there are difficulties in terms of allocation efficiency and vertical equity that
result in a pay-as-you-go system that is very generous and subsequently puts pressure on actuarial and
fiscal balance. On the other hand, population aging calls for a declining trend of the pension support
ratio—which measures the ratio between active contributors to the system and beneficiaries; this,
inevitably, leads to reconsidering the Bismarckian spirit and the urgent need to design new Beveridge-
like transfer schemes with complementary non-contributory pillars.5
The purpose of this study is to contribute to an informed debate on public policy relative to
income protection schemes for the older adults in Argentina. In particular, this report analyzes
and discusses some ideas that help reflect on short-term public policy tools to improve efficiency and
equity, and also on the concept of a pension system in the context of population aging in the long
term, by taking into consideration some reform options that typically come up in public debate.

2. Characterization and diagnosis of the pension system

4 For further details see Rofman and Apella (2014) and Bertín (2019)
5 A Bismarckian system refers to a contributory social security scheme that replaces the worker’s labor income upon
retirement, whereas a Beveridge system is connected to the provision of a flat transfer in old age that may (but not
necessarily) incorporate funding from general revenues.

-2-
Argentina has made extraordinary progress in extending coverage to the elderly by expanding
its non-contributory programs and establishing automatic adjustment mechanisms. Halfway
through the 2000s, a few measures were introduced to make the pension system universal. Though
this allowed expanding coverage to almost one hundred percent of the elderly, some equity issues
persist, and they need to be pointed out. In terms of adequacy, a pension indexation mechanism has
been implemented that allows updating the value of benefits periodically, thus avoiding significant
losses in actual value due to inflation. This generated benefits that exceeded the average of several
other countries in the region.
However, a serious long-term equity and sustainability issue prevails. On the one hand, this is
connected to the intra-system heterogeneity where various differential regimes coexist, each featuring
very different replacement rates and formulas to calculate benefits. On the other hand, the number of
benefits granted to each older adult is, in many cases, more than one. This situation makes the pension
system a very generous scheme and its main objective becomes blurred. Finally, in a population aging
context, the sustainability of the system is threatened as the active-passive ratio will naturally tend to
decrease, thus rendering the contributory nature of the system unfeasible.

Figure 1.1 - Pension indifference curves. Benefit generosity ratio, dependency ratio and pension
spending as % of GDP (year 2015)
0.90 50
BRA
0.80 45

40

Pension spending as % of GDP

0.70
35
Benefit generosity ratio

0.60
ARG
30
0.50
25
VEN
0.40
PAN 20
NICECU
0.30 MEX
URY
COLBOL CHI OECD 15
0.20 HND
PERCRI 10
PRY
0.10 SLV 5
GTM DOIM
- 0
5 15 25 35 45
Older adults dependency ratio
Source: Author’s own elaboration based on Rofman and Apella (2020) and United Nations, Population Division.

Argentina, together with Brazil and Uruguay, with a less advanced degree of aging compared
with developed countries, faces higher spending on pensions than the OECD average, in
terms of GDP. Figure 1.1 features the three dimensions that characterize pension systems
simultaneously for a group of countries in the region and the average for OECD member countries,
through indifference curves (see Annex 1). That is, the total public spending on pensions, represented
by the location of the level curve, resulting from the combination of benefit generosity, defined as the
product of the replacement rate and coverage and the dependency ratio of the elderly. More precisely,

-3-
pension level curves break down public spending into pensions in terms of the dependency ratio of
the elderly and the benefit generosity rate for the 18 Latin-American counties and the OECD average.
On the one hand, Argentina is, after Uruguay, one of the countries facing the highest elderly
dependency ratio. The x-axis shows the demographic dependency ratio for older adults, which is
defined as the population over the age of 65 compared with the number of those of working age (15
to 64). But the degree of population aging of the country, as in the case of its peers in Latin America,
is still below the average recorded in OECD countries. The demographic dependency ratio in Latin
America was of only 12 adults of non-working age for every 100 adults of working age; while across
OECD nations the value accounted for 30 adults of non-working age for every 100 adults of working
age.
On the other hand, Argentina and Brazil have the most generous pension systems in the
region. The y-axis represents the public policy decision: coverage and level of pension benefits. Both
countries, though they face a dependency ratio that is more favorable than that of OECD countries,
maintain a level of coverage and replacement rate that is much higher than the average for the other
group of countries, and this places them on a level curve that is farther away from the origin, with
pension spending being 3.4 percentage points above that of developed countries.
These dimensions will be analyzed in more depth in the next part of this section.

2.1 Coverage
Given the declining trend in coverage analyzed by different studies (Rofman et al., 2008;
Secretaría de Seguridad Social, 2005; among others), in 2005 (with subsequent extensions in
2014, 2016 and 2019), a Moratoria Previsional (Social Security Moratorium) program was
implemented, designed as a regime that granted easy payment terms to workers who owed
contributions to the pension system. This initiative provided a possibility for any citizen that met
minimum age eligibility requirements but not those relative to years of contributions, to declare having
a debt corresponding to those years in the self-employed workers’ scheme, and thus be included in a
payment scheme that would be rolled out in parallel to the collection of benefits. Likewise, access to
pension benefits was extended to the rightful claimants of deceased workers, so that they could receive
a survivors’ pension. The moratorium was originally approved by the end of 2005, in the case of
periods of contribution prior to 1994. In 2014, a new law allowed for the inclusion of contributions
corresponding to the period between 1994 and 2013 in the debt, setting a two-year limit to apply. In
2016, however, (and, once again, in 2019) this period was extended, though only in the case of women.
Finally, in 2016 a universal benefit was established, the Universal Pension for the Elderly (PUAM).
PUAM ensures older adults over the age of 65 who have no contributory coverage to receive a
monthly income equivalent to 80% of the minimum pension received by SIPA beneficiaries. Thus,
the income protection system for older adults reached an almost universal coverage, with this
expansion targeting the lowest income quintiles (Figure 1.2).

-4-
Figure 1.2 - Percentage of adults over the age of 65 who receive a pension benefit (years 1992-2018)
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%

Quintil I Quintil V Total

Source: Author’s own elaboration based on household surveys EPH, INDEC.

Due to this coverage expansion, the country ranks among those with the highest coverage for
people over 65 years old (Figure 1.3).

Figure 1.3 - Percentage of adults over the age of 65 receiving a pension benefit, by country (year 2018)
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
Argentina
El Salvador
Perú
Paraguay

Ecuador

Costa Rica

Bahamas

Trinidad y Tobago
Saint Kitts and Nevis

Belize
Barbados
Uruguay

Chile

Ucrania
Estados Unidos
Guyana
Honduras

Jamaica
Grenada

Venezuela
México

Rusia

Canadá

Noruega
Brasil

Antigua y Barbuda
Guatemala

Santa Lucia

Panamá
Dominica

Saint Vincent

Bolivia
Aruba

España
Reino Unido
India

Colombia

Tailandia
Haiti

República Dominicana

Luxemburgo
Nicaragua

Source: World Social Protection Report 2017-2019 - International Labour Organization.

The expansion of non-contributory benefits led to the reduction of horizontal inequity issues
that were generated in the past. In this sense, the contributory scheme was funded, not only with
contributions from workers’ wages, but also with resources coming from general revenues, which

-5-
generated a situation of inequality as people with no coverage were funding the benefits of those
covered by the pension system. When coverage was expanded, many individuals who were funding
the system through payment of their taxes became eligible to receive a pension benefit.
However, the attempts to correct the coverage issue did not take into account the labor record
of workers and their past partial contributions. The low coverage of the contributory scheme
originates in the setting of an eligibility condition based on a minimum number of years of
contributions (in this case, thirty years). However, the labor history files of most workers include
periods of informality, unemployment or lack of activity (sometimes for very long periods), so in many
cases they are not eligible to receive benefits. A low average coverage may derive from a segmented
population, with a group of workers that contribute frequently and another group that does not
contribute, or due to the fact that many workers contribute only during a certain part of their labor
history. In the first scenario, those workers that are contributing to the pension system are expected
to receive a contributory pension, while the other group is excluded. Conversely, in the second case,
there may be a considerable number of workers (more workers than those excluded in the first
example) that do not meet the eligibility requirement and, therefore, do not have coverage despite
having recorded some periods of contribution.
Accordingly, it is not only the average number of contributors to the system that matters, but
the frequency of transitions between the contributory and non-contributory status of workers.
This opens the way not only to a pension assistance initiative -through a non-contributory pension
funded by general revenues- but also to examine the level of rigidity that complying with a certain
number of periods of contribution implies.
In fact, the density of workers’ contributions in Argentina is not homogeneous and, far from
showing a bimodal distribution, it shows quite a uniform distribution. Figure 1.4 shows an
approximation to the problem described.6 This was based on information provided by the Muestra
Longitudinal del Empleo Registrado (Registered Employment Longitudinal Sample) prepared by the
Ministry of Labor and Social Security, that includes longitudinal information on a sample of wage
earners from the private sector regarding their contributions to the social security system between
January 1996 and December 2015.

6 These results must not be considered final but only approximate, as this database does not contemplate the totality of
the working population. In particular, it excludes public sector wage earners and self-employed workers.

-6-
Figure 1.4 - Distribution of total contribution density, men and women (years 1996-2015).

Source: Author’s own elaboration based on Labor Histories Database, Ministry of Labor and Social Security.

Though, in average, workers only contribute to 34% of their total labor records, this average
value hides a very significant heterogeneity. The left side of the distribution concentrates 50% of
the total workers who have contributed less than 25% throughout the whole window considered.
Additionally, on the right margin, 2.7% of the workers observed in the sample were able to reach
100% of contributions (Table 1.1). However, the distribution of the working population does not end
with these two groups. On the contrary, the remaining 46.9% of workers show a contribution density
ranging between 25 and 99%, and all of them receive equal treatment from the standards that regulate
both the contributory and the non-contributory pension systems.
The problem is that a considerable number of workers have contributed different amounts
throughout their labor history, yet they do not reach the minimum threshold required. Even
though the universal pension benefit corrects the coverage gap by protecting those who do not meet
the eligibility requirements, this evidence suggests a need to treat workers differently, according to the
efforts made during labor life in order to improve coverage equity, as those individuals who were
excluded because they did not meet minimum requirements have nevertheless funded—though
partially—the system with their past partial contributions.

-7-
Table 1.1 - Contribution densities according to characteristics (years 1996-2015)
% of workers with contribution density (d)
Characteristics Average
d<25% 25%<d<50% 25%<d<50% 75%<d<100% d=100%

Total 33.6% 50.3% 20.5% 15.3% 11.1% 2.7%

Sex
Men 36.2% 46.6% 21.0% 16.6% 12.6% 3.2%
Women 29.0% 56.5% 20.0% 13.2% 8.5% 1.8%
Income Quintiles
I 15.5% 77.7% 14.9% 5.5% 1.7% 0.2%
II 30.7% 50.8% 25.0% 16.5% 7.0% 0.7%
III 44.0% 33.3% 23.8% 23.3% 17.1% 2.6%
IV 54.5% 24.2% 19.7% 22.6% 26.8% 6.8%
V 56.4% 26.2% 17.4% 16.6% 24.7% 15.1%
Cohorts
1985-1995 28.3% 54.6% 23.7% 15.3% 6.2% 0.2%
1975-1985 31.4% 50.4% 22.1% 18.7% 8.6% 0.3%
1965-1975 37.7% 47.7% 18.2% 13.5% 15.9% 4.7%
1955-1965 40.4% 46.7% 16.4% 12.1% 16.4% 8.5%
Age
20-30 32.4% 52.6% 18.5% 13.8% 10.3% 4.8%
30-40 40.8% 49.0% 11.1% 9.8% 13.2% 16.9%
40-50 41.0% 50.6% 9.6% 7.6% 10.2% 22.1%
50-60 41.1% 50.7% 9.2% 7.6% 9.8% 22.8%
Source: Author’s own elaboration based on Labor Histories Database, Ministry of Labor and Social Security

2.2 Adequacy
In Argentina, the combined effect of having different social security systems operating in
parallel, plus the volatility in the value of the national currency make it difficult to present a
time series that clearly reflects the trends in benefits. Figure 1.5 shows the value of average
benefits of the national system, pointing the difference between retirement benefits and survivors’
pensions. Notably, there are some dramatic peaks in the series resulting from monetary instability
periods (especially in 1975 and 2002), but the trend seems to decline in the value of benefits between
the early 70s and late 80s, with a recovery in the 90s; after the fall due to the 2001-2002 crisis there is
a more stable situation between 2003 and 2008, and a sustained rebound up until 2013 when values
became stable once again. Finally, the series ends with high volatility in recent years, as a result of
inflation, change and counterchange of the value adjustment mechanism.

-8-
Figure 1.5 - Average pension benefit in the national system (years 1971-2020)
- constant in 1997 pesos -
700

600

500

400

300

200

100

0
1971
1973
1975
1977
1979
1981
1983
1985
1987
1989
1991
1993
1995
1997
1999
2001
2003
2005
2007
2009
2011
2013
2015
2017
2019
Total Old age pension Survivor´s pension

Source: Author’s own elaboration based on Social Security Report – ANSES and INDEC.

Changes in the trends observed are due to modifications in regulations over time. The rapid
growth of pension amounts observed between 2008 and 2013 (when pensions increased by 43% in
real terms) and their later stabilization, were caused by the application of automatic benefit adjustment
mechanisms. This benefit indexation formula (fórmula de movilidad) combines the increase in salaries
and the collection of revenues which, due to its design, generated a highly pro-cyclical mechanism. By
the end of 2017, a reform was introduced whereby the benefit indexation started to take into
consideration the weighted variability of inflation (70%) and the salaries of the formal economy (30%).
Later on, in December 2019 the Social Solidarity and Productive Reactivation Act was passed; in its
article 55, the law suspended the application of the pension indexation. During its suspension, the
Executive Branch established, in a discretional manner, a quarterly increase of the pension amounts
corresponding to the general regime while a commission integrated by representatives from the
Ministry of Economy, Ministry of Labor, Employment and Social Security and members of the
different commissions of the National Congress, was appointed to propose a new pension indexation
law. Finally, in December 2020 Act Nº27609 was passed and a new benefit indexation formula was
established. This mechanism adjusts benefits on a quarterly basis, taking into account a 50% variability
of formal salaries and 50% in terms of ANSES revenues.
Inflation, fiscal deficit, and the relative burden of social security spending have pushed
pension indexation towards the center of policy and macroeconomic debate, breaking away
from its role within the group of parameters that govern the triad of coverage-adequacy-
sustainability. The real value of the average pension amount has been increased, in real terms, by
almost 25% between 1994 and 2020, with a dramatic fall in 2002. Despite the fact that the increased
purchasing power of pensions was accompanied by higher salary levels of formal workers, this
increased purchasing power was slightly lower and caused an increase in the replacement rate.

-9-
In the course of the last three decades, the replacement rate7 reached its historic high by mid-
2020, when the average pension amount accounted for 46% of the average salaries in the formal sector
(Figure 1.6). This historic high was reached in the course of the COVID-19 pandemic and social
mobility restrictions. This phenomenon explains the fall of the purchasing power of formal salaries.
The level reached by the replacement rate of the pension amount in 2020 is even higher than the levels
for 2001 and 2005, when this indicator reached 42%, on both occasions. On the other hand, the
historic minimum of this series is observed in the second half of the 90s when the pension amount
merely exceeded 30% of the average salary.

Figure 1.6 - Replacement rate and salary versus minimum/average pension amounts ratios (years 1994-
2021). (Mobile average over 12 periods)
100%

90%

80%

70%

60%

50%

40%

30%

20%

10%

0%
May-96
Apr-97

May-07
Apr-08

May-18
Apr-19
Sep-03
Aug-04

Sep-14
Aug-15
Nov-01

Nov-12
Jun-95

Mar-98

Dec-00

Jun-06

Mar-09

Dec-11

Jun-17

Mar-20
Jul-94

Feb-99

Oct-02

Jul-05

Feb-10

Oct-13

Jul-16

Feb-21
Jan-00

Jan-11

12 per. Mov. Avg. (Replacement rate) 12 per. Mov. Avg. (Miminum benefit / Maximum benefit)

12 per. Mov. Avg. (Minimum benefit / wage (RIPTE))

Source: Author’s own elaboration based on Social Security Report – ANSES and INDEC.

Not all benefits paid by the pension system in the country are adjusted according to the
pension indexation mechanism in force and this limits the capacity to estimate the real
adequacy of the system. On the one hand, benefits paid and adjusted automatically only represent
one part of the total social security spending of the country, as multiple systems, regimes and benefits
coexist which are not directly affected by these values. For example, a change in the value of benefits
paid by the provinces to their retired civil servants will not be reflected in Figure 1.5; however, it will
have an impact on the average income of retirees in Argentina. Similarly, different regulations allow
one person to be entitled to more than one benefit (both in the case of two or more pensions if the
individual generated entitlements under different systems, as could be the case when combining a
retirement benefit plus a survivors’ pension). The necessary information to analyze these effects in
more detail is scarce. In the case of SIPA, the number of benefits vis-à-vis the number of beneficiaries
7 Estimated here as the quotient between the average pension amount of the general scheme and the average salary of
formal employees. RIPTE is used as a proxy of the evolution of formal salaries. This indicator estimates the average salaries
of workers registered (either in SIJP or SIPA, according to sworn statements by their employers, submitted monthly to
the AFIP or Federal Tax Authority) during 12 consecutive months.

- 10 -
increased in around 10 percentage points between 2004 and 2020, pointing to the fact that the number
of persons receiving more than one benefit might have increased in that same proportion (Figure 1.7).

Figure 1.7 - Benefits and beneficiaries of the national system (years 2001-2020)
8000 1.36
1.33
7000
1.30
6000 1.27
5000 1.24
1.21
4000
1.18
3000 1.15
2000 1.12
1.09
1000
1.06
0 1.03

Beneficiaries Benefits Benefits/Benefiacires (Right axis)

Source: Social Security Report – ANSES

The implementation of the Social Security Moratorium not only allowed many older adults to
receive a benefit, but also helped many of them to access double benefits. This hike in the
benefit/beneficiaries ratio that is observed between 2005 and 2007 is noteworthy. Up until 2005, the
number of benefits granted (retirement and pensions) exceeded the number of beneficiaries only by
10%. From the moment the Social Security Moratorium was implemented, this ratio more than
doubled and kept growing over the years. That is, the pace at which benefits grew was greater than
the growth pace of beneficiaries. This phenomenon is explained by the fact that the moratorium
granted a possibility to a large number of individual to access their own pensions, even though they
did not meet eligibility requirements but were already receiving a survivors’ pension after the decease
of their spouses.
With the expansion of coverage and the automatic updating of benefits, the Argentine pension
system played a role of ever-growing importance as an instrument to alleviate poverty,
particularly in the case of the elderly. Initiatives implemented over the past fifteen years have
resulted in alleviating poverty incidence levels, among individuals aged 65 and over, that are well under
the average values of the society (Figure 1.8).

- 11 -
Figure 1.8 - Percentage of persons in poor households, by age groups, before and after pension
transfers (year 2020)
80%

70%

60%

50%

40%

30%

20%

10%

0%

Age group

Pensiones < Pensión mínima Observado

Source: Author’s own elaboration based on household surveys EPH, INDEC.

In 2020 poverty affected 37% of the population; however, without transfers from the social
security system it would have affected 46.8% of the population, with its hardest impact on
older adults. In fact, the incidence of poverty among the older adults is of about 11% in average,
while this percentage would rise to approximately 67.8% if there were no transfers from the pension
system.

2.3 Sustainability
The achievements of the pension system in terms of coverage, adequacy and impact on
poverty incidence brought about some effects on costs incurred. In 2020, Argentina allocated
almost an 11.8% of its GDP to fund pension benefits in the different modalities, and ranked among
the countries with the highest spending worldwide.8 The evolution of spending can directly relate to
the various policies implemented. In this regard, over the last 25 years we can observe six very clearly
defined periods and changes in the trends (Figure 1.9):
1. Between 1994 and 2000: a stable trend with a slight drop as a result of a reduction in
coverage, as stricter eligibility requirements were implemented.
2. Between 2001 and 2004: a sharp fall of actual spending associated with the lack of
indexation (except the minimum) in a context of inflation.
3. Between 2006 and 2009: rapid increase in costs due to a rise in the number of beneficiaries
after the implementation of the Social Security Inclusion Program (Moratorium).

8This number includes 1% of the GDP that goes to payment of non-contributory pensions (such as mothers with seven
or more children, Falkland Islands (Malvinas) War veterans; disability pensions and other smaller groups.

- 12 -
 4. Between 2008 and 2012: spending increases due to the application of the indexation
scheme relative to Law N°26417, within a context of economic growth and further
volatility following the economic cycle.
5. In 2017, the Ley de Reparación Histórica (Reconciliation Act) allows for an increase in pension
benefits to compensate the lack of pension indexation until 2008.
6. In 2018 there is another fall in actual spending after the implementation of the new
pension indexation scheme, in a context of accelerated inflation and drop of actual wages.

Figure 1.9 - Public spending in retirement and pension benefits, by level of government (years 1980-
2020)
(in % of GDP)
14.0% Reparación
Second histórica
12.0% Moratorim
Indexation
Moratorium
Coverage decreaseReduction
due of the real benefit
10.0%
due to the lack of
indexation
8.0%

6.0%

4.0%

2.0%

0.0%
1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 2020

Nacional Provincial

Source: Author’s own elaboration based on data from the Ministry of Public Economy and Finance, Economic Policy
Bureau; Savings-Investment account of ANSES; INDEC.

Worth noticing in this process is the fact that changes at the aggregate level of spending have
almost no connection with the demographic trends, the effect of which is slower and
somehow covered by the volatility of short-term regulations. In Argentina, public spending on
pensions is similar to that of developed countries where population aging is considerably higher. In
fact, with 12% of its population over the age of 65, Argentina’s spending on pensions is similar to that
of countries like Japan, Germany, or Finland, where the share of older adults’ population almost
doubles (26%, 21%, and 20%, respectively) (Figure 1.10).

- 13 -
Figure 1.10 - Public spending on pensions and percentage of population over the age of 65 (year
2020)
18.0
Greece, 16.9
16.0 Italy, 16.2

14.0 France, 13.9

Austria,Portugal,
13.3 13.3
Pension spending (% of GDP)

12.0 Brazil, 12.0

Argentina 2020, 11.0Poland, 11.1 Finland, 11.4
Uruguay, 11.0Slovenia,
Spain, 11.1
Belgium, 11.0
10.7
10.0 Germany, 10.1
Hungary, 9.2 Japan, 9.4
Luxembourg, 8.4 Denmark,
Czech 8.18.1
8.0 Turkey, 7.719 OECD, 8.0Republic,
Slovak Republic,
United 7.3 Sweden, 7.2
States, 7.1 Latvia, 7.0
Norway, Lithuania,
6.6 6.7
6.0 United Kingdom, 6.2
Argentina 2005, 5.1 Netherlands, 5.4
New Zealand, 4.8
Colombia, 4.5 Israel, 4.8 Canada, 4.7
4.0 Australia, 4.2
Ireland, 3.6
Mexico, 3.1 Chile, 2.8 Korea, 3.0
2.0 Iceland, 2.1
Peru, 1.6

0.0
0 5 10 15 20 25 30

% population 65 or older

Source: Author’s own elaboration based on IMF GFS, Ministry of Economy ad INDEC.

At present, almost half of ANSES’ spending on pensions is covered by funding from general
revenues. For many decades now, funding of the pension system complements its own contributory
resources with the support of funds from general revenues. After transitioning a natural period of
“initial surplus” between the 40s and 60s, the pension system arrived at a virtual equilibrium in the
early 70s (Cetrángolo and Grushka, 2004). Already by the late 70s, the system started to reveal growing
deficits, resulting not only from its own maturity, but also from the weakness of the contributions
density. In the first place, the deterioration of the labor market itself, alongside the recurrent
macroeconomic crises, affected the volume and stability of the formal labor market and, in addition,
its capacity to contribute to the system. In the second place, the recurrent use (i.e., reduction) of the
social security contribution rate as a tool to alleviate the loss of competitiveness (i.e., “fiscal
devaluation”) generated by the subsequent overvaluation schemes, eroded the density of
contributions. The sum of all these factors generated the need to supplement contributory financing
with funding from general revenues of the Treasury.9
One structural concern is the financial sustainability of the system in a context of population
aging. On this topic, different authors have pointed to the potential rise of spending on pensions in
a variety of scenarios due to the increase in the older adults dependency ratio (Grushka 2016; Rofman
and Apella, 2014; inter alia). Pay-as-you-go systems seek to maintain the usual financial balance
(Equation 3). Thus, contributions by active workers today are directed to pay benefits to the passive
population. The first of these concepts depends on some variables that are exogenous to those

9
This will be discussed in more detail in Chapter 5.

- 14 -
responsible for social security policy, such as the number of workers registered in the economy
(𝐿 × 𝑓), where 𝐿 and 𝑓 represent the number of workers and the percentage of formal workers,
respectively, the average wage (𝑤), and the contribution rate (𝑐) which is part of the system’s design.
In terms of disbursements, the latter depend not only on the number of beneficiaries(𝐵), but also on
the salary for which the initial pension amount is calculated (𝑤) and the replacement rate or
adequacy(𝑟).

𝐿×𝑓×𝑤×𝑐 =𝑟×𝑤×𝐵 (3)

Social security systems based on pay-as-you-go regimes in the long run suffer from deficit
pressures when the contributor-beneficiary ratio falls (ceteris paribus). When systems are
created and are young, the number of contributors is considerably higher than that of beneficiaries, as
the latter are much fewer. As the social security scheme matures, the number of elderly beneficiaries
grows and the contributor-beneficiary ratio, that is to say, the pension support ratio, starts to decrease.
Finally, and due to an aging population, the pressure on the pension support ratio increases, and the
system naturally falls into a higher deficit.
In order to better understand the scope of this challenge, some simulations on the flow of
revenue from contributions, spending and financial results of SIPA are presented below. Long-
term projections are necessary to assess the system’s reaction if confronted to expected changes in the
economic, political and demographic conditions. This is not a forecast to “predict the future” but
rather a means to evaluate long-term outlooks for the system, given certain reasonable hypotheses on
the evolution of determinant variables. Moreover, this exercise allows to evaluate the impact of
alternative hypotheses and/or parametric changes that are often present in public debate.10
In the long run, the support ratio of the pension system would show a declining trend, going
from 3 contributors for every beneficiary in 2020 to 1.8 in 2050 and to 1 in 2100. This result is
typical of the population aging process the country is experiencing, and this process not only implies
a much higher participation of the elderly population but also a lower percentage of active population.
According to current parameters, the number of beneficiaries in the SIPA contributory scheme would
go from 3.4 million in 2021 to 6.5 million in 2040 and 10.5 million in 2100 (Figure 1.11), whereas the
number of contributors would also show a growing trend, though of a smaller magnitude, up until the
2050s, when it would become relatively constant around 12 million.

10
In this exercise we used the PROST (Pension Reform Options Simulation Toolkit) model developed by the World Bank
that has been used in more than 90 countries and provides a unified exercise methodology and an integrated analysis,
incorporating elements relative to coverage, sustainability and sufficiency of benefits. Data and demographic assumptions
required by the PROST model —fertility and mortality rates— were taken from calculations by the United Nations
Population Division. Demographic information required by the program includes: initial population, by age and sex;
simulation of fertility and mortality rates, by age and sex. Regarding the future evolution of some relevant macroeconomic
variables, slightly optimistic assumptions were held: a long-term per capita GDP growth rate of 1.5%, an increase in
productivity rate of 1%. All values are expressed in real terms. Finally, information relative to contributors, beneficiaries,
labor income, benefit amount and rules of the game, was obtained from information published by ANSES through the
Social Security Report. Chapter 6 presents information in greater methodological detail.

- 15 -
Figure 1.11 - SIPA: Number of contributory beneficiaries, contributors and pension support ratio
(year 2018-2100)
14,000.0 3.5

12,000.0 3.0

10,000.0 2.5

8,000.0 2.0

6,000.0 1.5

4,000.0 1.0

2,000.0 0.5
2018
2021
2024
2027
2030
2033
2036
2039
2042
2045
2048
2051
2054
2057
2060
2063
2066
2069
2072
2075
2078
2081
2084
2087
2090
2093
2096
2099
Contributors Beneficiaries Support ratio (right axis)

Source: Author’s own estimation based on ANSES

Consequently, the purely contributory social security spending managed by ANSES would
show an upward trend, transitioning from 5.7% of GDP in 2021 to 10% in 2050 (Figure 1.12.a).
If on top of these values we add payment corresponding to Moratorium benefits and those associated
with the Universal Pensions for the Elderly, the social security spending would maintain the same
trend, yet starting at a higher level: it would go from the current 8.5% to 12.1% in 2050 and could
even soar to 15.7% by 2100.11
In this context, SIPA would maintain a balanced financial result until the end of the 2020s
(Figure 1.12.b), however, from then on it would reveal a deficit. When looking at all the
contributory and non-contributory programs (Moratorium and PUAM), SIPA experiences a deficit
situation today and this deficit might increase as the country goes through its demographic transition.

11It is worth remembering that, in addition, Argentina spends around 3% of its GDP in retirement and pension benefits
through other programs that are not under the management of ANSES; therefore, the total projected spending should be
increased in at least that value.

- 16 -
Figure 1.12 - SIPA – Spending on pensions and financial result (years 2018-2100)
(as % of GDP)
a. Pension spending b. Financial result
18% 0.02

16% 0

2018
2022
2026
2030
2034
2038
2042
2046
2050
2054
2058
2062
2066
2070
2074
2078
2082
2086
2090
2094
2098
14%
12% -0.02

10%
-0.04
8%
6% -0.06
4%
-0.08
2%
0% -0.1
2018
2023
2028
2033
2038
2043
2048
2053
2058
2063
2068
2073
2078
2083
2088
2093
2098
-0.12
Contributory Contributory
Contributory and moratorium Contributory and moratorium
Contributory, moratorium and PUAM Contributory, moratorium and PUAM

Source: Author’s own estimation based on ANSES

The need for resources from general revenues to finance the pension system that will keep
growing in the future, poses specific challenges not only from the fiscal deficit point of view
but also in terms of conflicting public policy preferences regarding the provision of funding
to other public goods and services (education, health, infrastructure, etc.). As a consequence
of future weaknesses in the sustainability of the pension system and a greater need to resort to funding
from general revenues, some parametrical reform ideas that often appear around SIPA, which are
directed to contain pension spending, are usually linked to some of the main rules in pension systems:
demographic rules (like retirement age or the required periods of contribution to access the benefit,
which affect the number of beneficiaries) and financial rules (the contribution rates in the case of
resources and the expected benefits). The options that are typically considered include: i) setting an
equal retirement age for both men and women; that is, 65 years of age; and ii) establishing an automatic
mechanism to increase the age of retirement for both men and women, to accommodate increases in
life expectancy.
Regarding the first policy option mentioned, scenario 1 in Figure 1.13, its potential impact on
the level of financial deficit is not significant. In fact, the deficit might even be reduced by 1
percentage point annually.

- 17 -
Figure 1.13 - SIPA: SIPA financial result according to public policy scenario (years 2018-2100), as %
of GDP
2018
2021
2024
2027
2030
2033
2036
2039
2042
2045
2048
2051
2054
2057
2060
2063
2066
2069
2072
2075
2078
2081
2084
2087
2090
2093
2096
2099
(2.0%)

(4.0%)

(6.0%)

(8.0%)

Base scenario (contributory, moratorium and PUAM)

(10.0%) Scenario 1
Scenario 2
Scenario 3
(12.0%) Scenario 4
Scenario 5

Source: Author’s own estimation based on ANSES

Note: Scenario 1: retirement age being equal for both men and women in 2025
Scenario 2: increase of retirement age by 1 year every ten years up until 2068
Scenario 3: gradual elimination of survivors’ pensions (2050 =0)
Scenario 4: all contributory salaries taken into account to calculate benefit
Scenario 5: scenario 4 + increase of retirement age

The very low impact that the increase in the statutory pensionable age of women until it
equals that of men at the age of 65 would generate is associated to the low percentage of
women between the ages of 60 and 64 that receive a contributory pension benefit from SIPA
(Figure 1.14). In average, only 9.9% of women between 60 and 64 meet the eligibility requirements
to access a contributory benefit. Therefore, any change in the parameters to modify the eligibility
requirements in this population group would not produce a significant effect.

- 18 -
Figure 1.14 - Percentage of individuals receiving a contributory pension benefit from SIPA, by gender
and age (year 2020)
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0% 9.9%
50 52 54 56 58 60 62 64 66 68 70 72 74 76 78 80 82

Age
Men Women

Source: Author’s own elaboration based on ANSES and INDEC.

Endogenizing life expectancy within the parameters of the system by increasing the statutory
retirement age by one year every ten years would allow an annual reduction of the deficit of
approximately 1.5 percentage points in terms of the GDP, as from the second half of this
century. This scenario evaluates the impact of establishing a gradual increase in the retirement age of
all workers, in line with the expected increases in life expectancy by the age of 65 (of approximately 1
year every 10 years). The idea underpinning this assumption is to endogenize post-retirement life
expectancy into the pension system and keep it in the current levels so as to maintain the total benefits
payment period relatively constant.
It would be interesting to evaluate a third scenario that entails certain political complexity
and that is the gradual elimination of survivors’ pensions. In labor markets where the distribution
of household chores tends to be more equitable, the participation of women in the labor market grows
incessantly and access to pension benefits is universal, the concept of survivors’ pension somehow
loses its original purpose. These benefits were created in a context where the distribution of household
chores and the economic activity were clearly shaped according to the sex of the spouses. In case of
death of the male spouse, the system granted a pension to the widow to compensate for the loss of
income generated by the event. In a society where such duties are no longer so clearly differentiated,
and in economies where the labor participation of women is growing, the risk would be covered by
the entitlement to a retirement pension generated by each spouse.
The elimination of the pension benefit would have a significant impact on the total spending
and, therefore, on the financial results of SIPA. The deficit may well be reduced up to 3 GDP
percentage points when compared with the baseline scenario. Of course, this type of reform should
take into consideration some relevant exceptions (such as pension benefits received by surviving

- 19 -
minors), and also the total income of the couple before the event,12 yet the magnitude of the impact
is quite clear.
In addition, two reform scenario simulations are presented that, apart from attempting to
improve financial sustainability, seek to enhance access to the system: to take into account
the average of all salaries throughout the labor history of the worker to calculate the benefit.
Two alternative scenarios are simulated: scenario 4, which gradually includes all contributory salaries
when calculating the pension benefit, and scenario 5, which results from combining scenario 4 and
the increase in retirement age.
Combining these measures can help reduce the deficit, yet none of the scenarios described
above make the system sustainable. The required rate of contribution to reach a balance must
rapidly increase to 35%. Going beyond the issue of sustainability, it will be important to consider
accessibility to the system (how many resources will be needed to cover the deficit under the different
scenarios).
The system becomes fairer as the ratio between contributions and pensions increases.
Scenario 4 is relevant, not only because it allows reducing the deficit, but also because it makes the
system more accessible. As contributions are designed on the basis of all salaries, benefits could also
be calculated on the basis of all annual salaries and not only on the basis of the last ten annual salaries,
as is the case in the current system.
Calculating pensions according to a limited number of years might lead to unfair benefits,
abuse or even perverse incentives in terms of economic behavior. For example, some individuals
may declare higher salaries only during the last 10 years. Likewise, basing pensions on a limited number
of years tends to be regressive, as workers whose last salaries are substantially over their average
working life salaries are more likely to receive higher benefits. In consequence, this group of
workers/beneficiaries obtains a very high internal rate of return from the system.
Many countries have reformed this measure, in particular OECD countries, where the
majority calculate pensions according to all years of contribution. Improvements in
administrative information systems have facilitated the change from taking into account the final
salaries at the end of the working life to taking into account the average of all labor life salaries, in
order to calculate pension benefits.
With the exception of specific measures to combat short-term inequity, the outcomes of any
reform that implies changing the rules of the game will not materialize in the short term, but
rather have an impact in subsequent decades. The steady expansion of spending on pensions
without an adequate actuarial programming in the 2000s often considers the containment of pension
spending as a key component in any type of medium-term fiscal program. However, reforming the
system parameters would not only lack an immediate effect in the financial results but would, in
addition, increase the level of spending by incentivizing the retirement of workers that would have
otherwise wished to remain in the labor market. In any case, to initiate gradual and orderly processes

12 The scenario hereby described is extreme. A less drastic alternative that could also be taken into account is the partial
elimination of the survivors’ benefit by taking into account the total household income before the event, as the full
elimination of the pension could leave the claimant spouse in a situation of poverty. For example, if the wife receives a
pension of $100 and the husband one that amounts to $1000, being these benefits the result of their labor record,
eliminating the survivors’ pension upon the husband’s death would imply a very sharp reduction in the household income
and the surviving spouse.

- 20 -
in terms of feasible mid-term reforms may provide short term benefits by building investors’ trust and
improving access to financing.
The Argentine pension system is at a crossroads that requires redefinition, both in terms of
its objective and funding, as well as in taking into consideration all difficulties in addition to
financial sustainability. Funding pension benefits is the responsibility of the State and can be
satisfied in the same way as with other public policy components (that is, through different types of
taxes or debt). The challenge lies in considering, on the one hand, whether the necessary taxes or
indebtedness to fund certain levels of spending are acceptable in terms of their impact on other sectors
of the economy and, on the other, even when it is possible to generate such resources, there may be
other public policies with similar or even higher urgencies or priorities. Of course, the answers to
these questions are not merely technical but depend on policy preferences and should therefore be
decided in that context.

2.4 Explicit and implicit inequities

There is one characteristic that affects across the three dimensions we have analyzed, the
resulting fragmentation and inequity. The expression “Argentine pension system” is mostly used
when referring to the retirement and pension regime known as SIPA, managed by ANSES pursuant
to Law 24241 and subsequent amendments. This association is justified due to historical reasons, as
SIPA includes the main schemes created throughout the 20th century, and to the fact that it is
responsible for the largest share of pension spending and coverage in the country. However, there are
many other schemes and institutions that operate in parallel to SIPA, thus giving a high level of
fragmentation to the system as a whole. In addition, the National Constitution grants the provinces
the right to organize pension schemes for provincial and municipal civil servants. All the provinces
created these schemes during the last century, and their funding and access regulations differ from the
national ones and have varied fiscal impacts. Between 1994 and 1996 the national government
accepted the transfer of provincial pension regimes from eleven jurisdictions: Catamarca, Jujuy, La
Rioja, Mendoza, Salta, San Juan, San Luis, Santiago del Estero, Río Negro, Tucumán and the City of
Buenos Aires. Nevertheless, there are still thirteen provinces that maintain their own pension schemes
for provincial civil servants. On the other hand, at the national level there are several special retirement
and pension schemes for the Army and law-enforcement forces, as well as special regimes for
university faculty and teachers in general, scientific researchers, magistrates, foreign services staff,
among others) that, despite operating within the SIPA framework, maintain different parameters when
compared to the general regime.
Institutional fragmentation and current regulations, apart from creating equity issues, affect
reforms that might have had a significant impact but are, in fact, highly limited. Figure 1.15
shows the composition of the total pension spending in Argentina in 2020. According to numbers
published by the Ministry of Economy, the consolidated spending accounted for 11.8% of the GDP
in 2020. This includes 2.3% of the GDP corresponding to the provincial regimes, 1% to non-
contributory pensions and 0.6% to the Army and law-enforcement forces scheme, so SIPA’s spending
was of approximately 7.9% of the GDP. Out of this figure, 3% was originated in moratorium benefits
and 1% in special regimes (teachers, researchers, magistrates, etc.). Consequently, only 4% of the GDP
was geared to pay retirement benefits in the general regime of SIPA in Argentina (this percentage
includes ordinary retirements and disability pensions, and it is not possible to discriminate spending
in each case).

- 21 -
Figure 1.15 - Public spending on pensions, by system and benefit types, as percentage of the GDP in
2020.
14%

12%

Provincial
10%
Others
Non-contributive
8%
With moratorium
6% 2.85%
National
9.5% SIPA Special regimes
4%
7.86%
Without moratorium
5.01% General regime
2%
4.0%

0%

Source: Author’s own elaboration based on data from the Ministry of Economy and ANSES.

In Argentina there are three large groups of pension regimes, according to the scope of their
administration: SIPA, the special and differential schemes managed by the National
Government and the provincial regimes. Each of them has a specific scope, level of benefits and,
consequently, different levels of spending, all of which point to the degree of fragmentation of the
system. Figure 1.16 shows a first source of fragmentation in the system. While the main pillar, SIPA,
that covers the totality of workers employed by the private sector, self-employed workers, civil
servants working for the national government and civil servants in the provinces that transferred their
provincial schemes to the national system, accounts for 73% of the total spending of the system, its
share in the total benefits granted is higher, equivalent to 87.4%. In turn, the share of provincial
regimes in the total pension spending accounted for 21.2% (2.31% of GDP), while the total benefits
granted by these provincial regimes only accounted for 10.3% of the total.

- 22 -
Figure 1.16 - Vertical distribution of spending and benefits granted according to administration (year
2020)

Spending Beneficiaries
10.3%
2.3%

21.2%

5.6%

73.2% 87.4%

SIPA ANSES Other national schemes SIPA ANSES Other national schemes
Provincial schemes Provincial schemes

Source: Author’s own elaboration based on ANSES

The rest of spending corresponds to other national schemes that account for 0.61% of the
GDP, and implies a share of total spending of 5.6%, and with a population of beneficiaries
reaching 2.3% of the universe. These schemes are traditionally called differential regimes and they
provide insurance schemes for law-enforcement workers (Armed Forces, Federal Police, Gendarmería
or Security Forces, and staff in correctional institutions). The creation of these regimes responded to
the specificity of the tasks undertaken, which require very qualified skills in the different fields. Though
there are many activities that require very specific skills that cannot be transferred, as is the case of
professional sports, there are some sectors that could be granted certain types of differential treatment.
This is the case of military activities and security forces, where age is a decisive factor for participation
and where jobs available become scarce as individuals age. Accordingly, people who occupy these
positions must retire when they grow older.13
The greater generosity of provincial, special and differential regimes explains their higher
relative spending compared to their share in the number of beneficiaries. As the rules of the
game in the provincial regimes (regarding retirement age and periods of contribution) have been
gradually homogenized in the course of the last fifteen years, the promised replacement ratios are
much higher than those of SIPA (Figure 1.17).

13
The case of military staff is well known, but similar activities are performed by ballet dancers, professional football
players or rock musicians (with very few exceptions like the Rolling Stones).

- 23 -
Figure 1.17 - Average monthly benefit according to province and national regime (year 2020)

Average benefit (in monthly $ - 2020)

120000
100000
80000
60000
40000
20000
0

Source: Author’s own elaboration based on ANSES

Fragmentation, however, does not occur only across systems but also within the main pillar
itself. It is possible to observe different exception regimes within SIPA that are typically called
“special” and that relate to university professors, non-university teachers, scientific researchers,
magistrates, foreign service staff, among others. These exceptions are usually connected to particularly
difficult or physically exacting conditions in certain jobs that would result in, for example, “premature
aging” of the worker. Also included in this category are jobs which, due to some special “merit”, justify
having different access to pension benefits under the form of award or compensation, as is the case
with former presidents, judges, and Olympic medalists, among others.
This type of regime grants a generosity level that can be twice the one provided by the general
regime. According to an estimate by Bertín (2019), the internal rate of return (IRR) generated by each
scheme for its beneficiaries is very heterogeneous (Figure 1.18). University professors generate an IRR
of 5% and 3.7% in the case of women and men, respectively, while this indicator accounts for 2.7%
for women and 1.3% for men in the general scheme.

- 24 -
Figure 1.18 - Internal rate of return of the national special pension regimes and SIPA

Univeristy Professor 5.0%

3.7%

Magistrates 3.9%
3.3%

Scientific researchers 3.9%

2.7%

Foreing service 3.8%

3.0%

SIPA 2.7%
1.3%

Women Men

Source: Bertín (2019).

Differences originate in the amounts of benefits granted. The special pension regimes have
minimum retirement ages and contribution requirements similar to those of the general regime, but
the amounts of the pension benefits can be much higher due to two factors. On the one hand, they
establish higher initial replacement rates, as the basic salary on which they calculate the pension benefit
tends to be the highest in the labor record. For example, the base salary used to calculate the retirement
amounts for the staff in the Foreign Service of the Nation corresponds to the highest hierarchical
position. In the case of university faculty, it is calculated on the basis of the salary at the time they stop
working. On the other hand, indexation rules may also differ as there is an automatic adjustment when
the salaries of active workers are raised.
In that context, it is necessary to follow a strategy aimed at harmonizing the rules across
different schemes, eliminating inequities and management problems, so as to build a model
where the core of the system is the beneficiary, and not his/her occupation. This would allow
approaching a more equitable pension model and, at the same time, a more sustainable model. Having
said so, there may be activities that still enjoy a differential retirement age and are not more dangerous
than others that follow the general regime parameters. Therefore, the existence of special indexation
regimes undermines the idea of harmonization. Going beyond calculation of pension indexation or
the age to access pension benefits, the sole existence of special pension regimes poses multiple
challenges from an equity perspective.

3. A reform proposal that is becoming less and less Bismarckian

Though the Argentine pension system has made some progress in terms of coverage and
adequacy of benefits paid, it still confronts many difficulties. Some are relevant in the short term,
relating to coverage and equity of the system, while other longer-term difficulties refer to financial and
economic sustainability.
The issue of coverage should be analyzed, though not through the traditional lens that
identifies whether a worker is eligible or not to access a contributory benefit, but rather
through the observation of the number of contributions the worker has made during his/her

- 25 -
working life. Addressing the coverage issue by incorporating this dimension would not only allow
for improving equity in terms of contributions and benefits, but also to generate some appropriate
incentives to extend the labor life of older adults, which has clear implications on the degree of
difficulty in the long run, and on the financial and economic sustainability.
The coexistence of different pension schemes is a clear inequity generator within the national
system, where provincial schemes and special and differential regimes coexist, each of them
with its own rules and, in particular, with a benefit generosity ratio that is higher. Moving
towards a strategy that aims to harmonize the rules across different schemes, eliminate inequities and
management issues, and build a model that focuses on the beneficiary -and not on the benefit- as the
center of the system, would not only bring the model closer to a more equitable system but at the
same time, a more sustainable one.
Concern about the growing need for general revenues, both to fund SIPA and provincial
regimes, is once again in the limelight of current debates. Expansion of coverage, together with
the demographic transition towards an aging population structure, reveal the need of extraordinary
resources to keep promises and honor commitments.
The following are some reform ideas aimed at contributing to an informed debate on an
eventual pension reform. Far from modifying some parameters, the proposed approach seeks to
redefine the objectives of the system and suggest instruments that help attain them.

3.1 Actions to minimize fragmentation and inequity

First, and in order to minimize inequities stemming from the coexistence of multiple regimes,
it is possible to identify two areas where public policy can engage: actions directed to
provincial schemes and actions relative to special and differential regimes.
1. From the viewpoint of provincial schemes, though some progress has been achieved in
harmonizing the parameters that rule subnational pension systems, several differences persist
and, therefore, provincial schemes grant more generous benefits than the national regime.
Full harmonization of parameters becomes a relevant dimension in the social security policy
of the country, as workers in the provincial public sector risk the loss of income in old age,
similar to those covered by SIPA and, therefore, from an equity perspective, differences are
not justified in the actual value of retirement and pension amounts received in these two
regimes. This becomes more relevant in a context where the national government grants
resource transfers to the provinces in order to fund imbalances in the special schemes.
An alternative proposed since 2016 seemed to facilitate reaching a second-best balance: the
national government funds the pension deficit of the special schemes that have not been
transferred, but only the part that results after applying the SIPA parameters. Though under
this process equity is achieved in the transfer of national public resources, this does not
eliminate inequity in the generosity of benefits within the system.
A better alternative would be to implement the transfer of the provincial regimes that are still
not transferred into the orbit of SIPA. In this case, the provinces would have to transfer both
their resources as well as their payment commitments to ANSES, and the latter would be in
charge of dealing with any associated deficits.
2. Regarding special and differential regimes, their existence in multiple forms does not respond
to a comprehensive strategy, built on the basis of a thorough analysis of the conditions of

- 26 -
 different groups of workers so as to identify those with more vulnerabilities or merits, but
rather to the specific success of different stakeholders which have managed to have their
respective regimes approved over the years.
3. To classify economic activities or occupations on the basis of their risk exposure and, above
all, according to “merit”, is complex, and solving such problems from the pension system does
not seem to be the more equitable mechanism. On the one hand, in order to classify
occupations or activities according to health risks it is necessary to identify the risks workers
are exposed to and define policies that protect workers. If such risks and/or merits could be
identified during the active life of workers, using insurance schemes –such as health insurance
or labor risk insurance, either for individuals or groups of workers– would provide much
better protection to the workers.
In the case of special regimes, due to aging in the activities, another possibility would be to
replace such regimes with a scheme that includes individually defined benefits, that evaluates
workers’ health conditions and, on the basis of such evaluation, provide temporary benefits
that allow workers to abandon the activity earlier than with the general regime, if affected by
the tasks carried out. Thus, a worker whose labor capacity or life expectancy is in fact reduced
by his/her specific activity, can request these benefits at any point in time, and granting such
benefits would depend on the existence of confirmed effects on his/her health. These benefits
could be for life, as is the case at present, or temporary, until the worker reaches the statutory
retirement age of the general regime. The difference between using the disability retirement
scheme and the labor risk scheme implies defining who will be responsible for the cost of
financing the benefits: in the first case, it will be the society as a whole, through the social
security system; while in the second case, it will be the employers, and incentives could be
generated for them to reduce factors causing premature aging.
4. Regarding special regimes, those cases where the rationale is merely based on merit, as is the
case with former authorities, researchers, university faculty and others, it is difficult to make a
case for them when public policy seeks equity and efficiency. Within this framework, there
seems to be no reason why certain groups of workers receive higher pension benefits or enjoy
more generous pension indexation regimes compared with the rest of the system’s
beneficiaries. Merits for serving society should not be acknowledged through promises of
future monetary payments but rather through public demonstrations of appreciation and,
when applicable, economic compensations during service.

3.2 Redesigning a multi-pillar system

Given the composition of the current social security system and its multiple objectives, and
in order to consolidate all programs into one single scheme, redesigning the system on the
basis of a two-pillar (or maybe a three-pillar) model is suggested. This multi-pillar scheme must
respond to two very well-defined objectives: one contributory objective associated with that of income
substitution and a non-contributory one related to the protection against the risk of ending up in a
situation of poverty.
Pay-as-you-go systems should maintain a current financial balance in terms of the parameters
that define their revenues and spending. The “purely contributory” component of the Argentine
pension system, that is to say, the one that does not include payment of benefits through a moratorium

- 27 -
nor through PUAM, inevitably moves towards a deficit situation as –due to the reduction of the
pension support ratio caused by population aging and the maturity of the system itself– disbursements
to pay for contributory retirement and pension benefits will, in the medium term, exceed revenues
from contributions.
Using revenues from the National Treasury to fund a contributory pension scheme creates a
horizontal inequity problem. This happens because the population that is inactive, unemployed or
in the informal labor market contributes –for example, when paying their taxes (i.e., VAT when
conducting commercial transactions) – to funding the benefits of insured workers, without generating
their own future pension entitlements. This rationale justifies the intervention of the State to grant
pension benefits (semi- or non-contributory), to acknowledge the indirect contribution of both
informal and unemployed workers, by awarding them a basic old age pension that gives them decent
livelihoods and reduces the risk of poverty.
The current configuration of SIPA is similar to a two-pillar scheme. One pillar is contributory
and based on a pay-as-you-go financial regime with defined benefits that provides coverage to formal
workers that meet the total periods of contributions (though the benefits are not exactly proportional
to the contributions). The second pillar is non-contributory (PUAM) and is directed to all the older
adults who are not eligible for contributory benefits and is funded through contributions and general.
The natural trend of the system is toward a greater demand for resources from general
revenues. The expected increase in the demand for resources from general revenues to fund a benefit
directed to the elderly, both contributory and non-contributory, implies that the objective of
preventing poverty situations will become more relevant in the future. In other words, if the largest
share of funding for the system comes from general revenues, then the non-contributory benefits
must become more important when defining priorities.
In view of this scenario, and in order to achieve higher horizontal equity and more clarity in
the rules of the game, a two-pillar pension system is proposed (Diagram 1.1).14 This would
imply a basic pillar, providing a universal benefit to all the older adults, regardless of their labor history,
which ensures an income that will protect them against the risk of poverty. This also implies using
some type of poverty indicator as reference when defining the benefit; for example, the value of the
Total Basic Basket per adult equivalent.15
In addition, we suggest redesigning the second pillar, starting with a definition of a
contributory benefit that is proportional to the density of contributions and salary of reference;
this mechanism could be an inclusive, as well as a sustainable, alternative from a fiscal
standpoint. Given that labor records of the majority of workers include periods of informality and
unemployment (often too long periods) so that in many cases they are not eligible to receive benefits,
having duly paid their contributions to the system without receiving a benefit that corresponds to what
they contributed.

14 The question that comes up after this proposal relates to the cost that a system with these characteristics might imply.
The level of spending brought about by this design would depend not only on the basic universal pension, but also on the
replacement rate to be established according to the density of contributions.
15 The need to limit the target population of this non-contributory benefit could be discussed, and this could be done

through some kind of focalization mechanism. However, the objective of this proposal is to present a rationale for a full
reform scheme that could be used to inform debates.

- 28 -
Therefore, we suggest considering the average of all incomes during the working life (and not
only the last ten years) to calculate the benefit. This would not only help reduce the deficit, but
would also make the system more accessible.

Diagram 1.1 - Universal benefit plus share for contributions

Replacement Rate

Universal
Benefit

0 Salary / Density of contribution

Source: Author’s own elaboration

With this type of rules and definition of benefits, enhanced horizontal equity could be
achieved and incentives could be generated so that workers postpone their retirement from
the labor market. This does not mean a statutory increase of their retirement age but rather to design
a structure of incentives that provides flexibility to workers either to delay their retirement from the
labor market or to facilitate the retirement of those who find it difficult to stay, in this case by
considering the contribution density in the definition of the benefit.
However, discussions relative to potential changes or reforms to the pension system take
place within a framework of high political tension, where stakeholders with opposing interests
are confronted. This institutional challenge usually requires a long period of discussions until
consensus is reached and then implementation takes place. The longer a reform is postponed, the
higher the need to have it; therefore, creating opportunities for dialogue and agreement across the
political sector, the private sector, trade unions and civil society is of utmost importance to consolidate
a pension system that ensures the protection of the elderly in the medium and long terms, with
reasonable levels of coverage, adequacy and financial sustainability. Creating a commission that
involves all the above-mentioned stakeholders in the analysis and proposal of solutions is instrumental
to promote dialogue and reach the required consensus.

- 29 -
 4. References
Apella, I. 2014. “La Relación Nación-Provincias en el Financiamiento del Sistema Previsional
Argentino” in Rofman. R (ed.) La Protección Social en Argentina. El Rol de las Provincias. World Bank, Buenos
Aires, Argentina.
Apella, I., 2010. “Historias Laborales y Frecuencia de Contribuciones a la Seguridad Social en
Argentina”. XLV Reunión Anual de la Asociación Argentina de Economía Política, Buenos Aires, Argentina.
Apella, I. and L. Casanova, 2008. “Función de Costos de la Industria Argentina de AFJP. Un Enfoque
Semiparamétrico”, XLIII Reunión Anual de la Asociación Argentina de Economía Política, Córdoba,
Argentina.
Bertín, H., 2019. Hacia Una Historia de la Previsión Social Nacional en Argentina: 1904-2018. Universidad
Nacional de La Plata. Facultad de Ciencias Económicas, 1a ed., La Plata, Buenos Aires, Argentina.
Ferro, G., 2003. “Regulación y Costos Variables Endógenos en el Mercado de Fondos de Jubilaciones
y Pensiones Argentino”. Universidad del CEMA, Working document Nº 231.
Grushka, C., 2016. “Perspectivas del Sistema Integrado Previsional Argentino y de ANSES, años
2015-2050”. Working document. Dirección de Estudios de la Seguridad Social, ANSES, Ciudad de Buenos
Aires.
Grushka, C., 2019. “The Within-system Redistribution of Contributory Pensions Systems: Conceptual
Framework and Empirical Method of Estimation”, in Lustig, N. (ed) CEQ.
Rofman, R., 1994. “Diferenciales de Mortalidad Adulta en Argentina”. Notas de Población 22(59): 73-
91, CEPAL.
Rofman, R, and I. Apella, 2014. “La Protección Social Argentina en un Contexto de Transición
Demográfica”. in Gragnolati, M., R. Rofman, I. Apella, and S. Troiano (editores): Los Años No Vienen
Solos. Oportunidades y Desafíos Económicos de la Transición Demográfica en Argentina. pp:143-
170. World Bank. Buenos Aires.
Rofman, R. and I. Apella, 2020. When We Are 64. Opportunities and Challenges for Public Policies in a
Population-Aging Context in Latin America. International Development in Focus. Washington, DC.

- 30 -
Annex 1. Indifference curves in the pension system

The total spending of a pay-as-you-go system is the result between the average benefit received by
each individual and the number of individuals who are beneficiaries. From this, the share of GDP
allocated to fund spending on pensions can be divided into two multiplicative components, as shown
in equation (1).
PS𝑡
PSt N+65,𝑡 N+65,𝑡
= GDP ∗ (1)
𝐺𝐷𝑃t 𝑁15−64,t
𝑁15−64,t

Where:
PS𝑡 represents the total pension spending at moment t
GDPt is the gross domestic product in t
N+65,𝑡 identifies the number of individuals over the age of 65 in t

The total public spending on pensions expressed as a percentage of GDP is the product of two factors:
one is economic and the other is demographic. The first one is represented by the average spending
on each individual at retirement age (over 65). The second is the ratio between the size of the
population at retirement age and the working age population.
In equation (1) the economic factor is represented by the first scalar unit. According to Miller et al.
(2011), such factor is usually known as the pension benefit generosity ratio (BGR in the diagram),
which expresses the generosity of average pension benefits in relation to GDP per adult individuals
of working age. Standardization of GDP per adult person of working age is useful for international
N
benchmarking of benefits. The second scalar unit 𝑁 +65,𝑡 represents the older adults dependency ratio.
15−64,t

A higher benefit generosity does not necessarily imply a higher transfer per beneficiary. This variable
reflects spending on pensions both in terms of the level of benefits and in terms of the system
coverage, that is, the actual number of eligible individuals that have effective access to the program.
For example, a higher BGR could be associated either to a higher benefit or a broader coverage, or
both. Equation (2) illustrates this breakdown. In this equation, Ct represents the actual number of
beneficiaries. As clearly shown in (2), BGR is equal to the average benefit per eligible person when
coverage is universal, that is to say, equal to one.

PS𝑡 PS𝑡
PSt N+65,𝑡 N+65,𝑡 𝐶𝑡 𝐶t N+65,𝑡
= GDP ∗ = 𝐺𝐷𝑃𝑡 ∗N ∗ (2)
GDPt 𝑁15−64,t +65,𝑡 N15−64,t
𝑁15−64,t 𝑁15−64,t

After demonstrating that BGR is the product of the average benefit expressed in terms of the average
product per worker (approximation to the concept of replacement rate) and the level of pension

- 31 -
coverage, for simplicity’s sake it is convenient to go back to equation (1). Defining a level of pension
spending as constant, it is possible to transfer terms in 1 and represent the relation between benefit
generosity and the dependency ratio through a level curve (Diagram A.1.1). This curve indicates the
level of spending as percentage of the GDP, which is assumed by the combination of BGR and the
dependency ratio. While the demographic dependency ratio is a given exogenous variable, BGR is the
public policy variable that can be defined on the basis of the former and the desired level of spending.

Diagram A.1.1. Indifference curves in the pension system

BGR

Pension spending

Dependency rate

Diagram 1 shows that if the level of spending remains the same, as the country advances with respect
to its demographic transition, the BGR should go down (movement on the curve). Conversely, if the
policy decision were to maintain the generosity of the system, the population aging would necessarily
imply confronting a higher level of spending and the system would have to jump toward a level curve
that moves farther away from origin.

- 32 -