pension reform in armenia

Transcription

pension reform in armenia
PENSION REFORM IN ARMENIA
AIPRG REPORT*
December 2005
Washington • Yerevan
www.armpolicyresearch.org
*
Prepared by Gohar Jerbashian, Ara Khanjian, Armineh Manookian, and Nerses Yeritsyan.
The financial support of the government of the Netherlands and the Netherlands Financial
Sector Development Exchange (NFX) are gratefully acknowledged. The paper was
presented at the conference on pension reform in Armenia on December 17, 2005. The
authors would like to thank Peter de Bruijne, Dan Wartonick, Mitchell Wiener, Richard
Zechter and conference participants for helpful comments. Any remaining errors are those
of the authors.
Table of Content
INTRODUCTION ...........................................................................................................................3
SECTION I. CURRENT PENSION INSURANCE SYSTEM OF ARMENIA: DEMOGRAPHIC AND
FINANCIAL PROJECTIONS ............................................................................................................5
Main Factors Affecting the Social Insurance System................................................................5
1.1-Demographic Assumptions......................................................................................5
1.2-Macroeconomic Assumptions ...............................................................................11
1.3-Labor Force, Employment and Contributors to the System ..................................13
SECTION II - FINANCIAL AND DEMOGRAPHIC ANALYSIS OF THE PENSION INSURANCE
SYSTEM.....................................................................................................................................19
2.1.
Pensioners ........................................................................................................19
2.2.
Payroll Contributions .......................................................................................21
2.3.
Dependency Ratios ..........................................................................................23
2.4.
Social Insurance Fund Surpluses and Reserves ...............................................24
2.5.
Replacement Rates and ROR on Mandatory Social Contributions .................27
SECTION III. NOTIONAL ACCOUNT SYSTEM ANALYSIS ............................................................33
SECTION IV. FUNDED ACCOUNT SYSTEM ANALYSIS ................................................................40
SECTION V. COMPARISON OF PENSION REFORM OPTIONS .........................................................46
APPENDIX: INTERNATIONAL EXPERIENCE WITH PENSION REFORM ...........................................49
REFERENCES .............................................................................................................................67
2
INTRODUCTION
In Armenia the pension system is a pay as you go, PAYG, system, where current
employees make contributions to the pension system and these contributions are used to
fund the benefits of current retirees. When current employees retire, future employees
would make contributions to the Social Security system and based on these contributions
the future retirees would collect retirement benefits. This implies that the PAYG system
is based on the solidarity principle between generations. It is also based on the solidarity
principle within generations, because high income earners subsidize the benefits of low
income earners.
It is generally accepted that the current pension system has some weaknesses and
needs improvement:
•
The benefits are very low and inadequate. After working for 30 years a retiree should
have a reasonable standard of living, instead, the average benefits are about the same
as the extreme poverty line. In 2003 the monthly average pension benefit was 7,600
Dram, while the extreme poverty line was 7, 742 Dram.
•
Currently the benefits are based on the number of years that a person worked and not
on the level of contributions. This implies that two employees who worked the same
number of years would receive the same level of benefits even if one person earns
30,000 Drams a month, while the other employee earns 300,000 Drams a month.
Especially for workers who earn above the average wage the replacement ratios,
pension benefits as a fraction of the salaries before retirement, are very low. The
result is that high income earners have little incentive to make contributions to the
pension system, which reduces the compliance rate.
In order to address these difficulties this report would focus on two models. One
model addresses the weaknesses of the existing pension system by adjusting the current
pay as you go system and adding a notional component to it. The notional accounts,
through a recordkeeping method, would link the benefits of the retiree to the
contributions that s/he made, while working. The system is not funded and the
contributions of the employees are not invested in a separate investment account. This
adjustment doesn’t alter the pay as you go nature of the system. The contributions still are
made to the Social Insurance Fund and the benefits would continue to be made through
Social Insurance Fund. Basically an accounting method would be introduced, which will
keep a record of the contributions of the employees and then when they retire; part of
their benefits would be calculated based on their contributions.
The second model would eliminate the PAYG system and introduce two pillars.
The first pillar, which is called Pillar 0 or I, is a minimum guarantee pension system
comparable to the level of current pensions. This pillar will be financed through general
taxes. The second Pillar would be fully funded and would be based on private mandatory
cumulative individual accounts. The contributions of employees would be invested in
private individual accounts and the benefits would be based and funded through these
3
accounts. This report through an actuarial simulation would focus on the contributions,
benefits and replacement ratios of these accounts. Clearly the introduction of these
private individual accounts generates many questions that require answers. The analysis
and discussion of these issues are beyond the scope of this report.
The purpose of this report is to present the main issues related with pension
reform in Armenia. At this stage the authors are not ready to make policy
recommendations.
Section II presents and discusses the results of actuarial simulation of the current
PAYG system without introducing adjustments. Section III presents the results of
actuarial simulations of the adjusted version of the current pay as you go system with a
notional account. Section IV presents and discusses the results of the actuarial simulation
of the model which represents individual mandatory cumulative retirement accounts.
Sections II, III and IV use the same demographic and macroeconomic assumptions, such
as economic growth, life span, and wages. Section I discusses and presents these
assumptions.
Section V compares the actuarial results of Sections II, III and IV. This
comparison addresses the need to provide adequate pension benefits to the retirees, based
on a balanced actuarial plan over 75 years and the need to distribute the costs of pension
reforms fairly within and among generations.
Appendix surveys pension reform experiences of Sweden, Chile, Hungary and
Canada. Before pension reforms, in each one of these countries, pension plans were
defined benefit plans and were pay as you go systems. They have each approached their
pension reforms differently and they offer a range of methods to transfer to increasingly
defined contribution and partially or fully funded systems. The experiences of these
countries are presented to provide a background for discussions on pension reform in
Armenia. This appendix doesn’t provide links between the experiences of these countries
and pension reform in Armenia
This report performs three separate actuarial simulations; first of the current
PAYG system, second of the adjusted PAYG system with notional accounts and finally
of the private individual accounts. All these simulations are done, using the same
demographic and macroeconomic assumptions and data. The main contribution of this
report is that when we use the same assumptions and data to run these different actuarial
simulations, we are able to see the weaknesses and advantages of each one of these three
systems more clearly.
PROST (Pension Reform Options Simulation Toolkit) model, developed by the
World Bank, is used for modeling the reform options presented in the paper.
4
SECTION I. CURRENT PENSION INSURANCE SYSTEM OF ARMENIA:
DEMOGRAPHIC AND FINANCIAL PROJECTIONS
Main Factors Affecting the Social Insurance System
1.1-Demographic Assumptions
The demographic situation in Armenia has changed dramatically during the recent 10-15
years. As long as 2004 is the base year for our long-term analysis, we generated the sexage distribution of the population on the basis of the 2001 Census data and other vital
statistics for 2002-2004. For estimating age-sex distribution of population for 2004, a
complementary model with base year 2001 was constructed and simulated via PROST
model. For the base year, de facto population (total 3,002,594) as per the 2001 Census
and absolute numbers of births (by age groups of mothers), deaths for each age group
(male and female separately), and total numbers of migrants for 2002-2004 as well as
sample surveys of migrants for 2002 was used in the complementary PROST model.
Fertility Rates: The fertility rates in Armenia have been dramatically low in the past
decade. But the recent strengthening of the economy has already had a positive effect on
the growing number of births and slight increase in fertility rates. Gradually more babies
born in Armenia, thus 32,066 babies were born in 2001, 32,401 in 2002, 36,079 in 2003,
and 37,863 in 2004. This raised the total fertility rate from 1.30 (in 2002) to 1.36 in 2003,
and 1.45 in 2004 still leaving it behind the rate necessary for permanent growth of the
population.
Economic growth (we assume the observed economic growth in Armenia will continue in
the long term) in Armenia has been leading to an increase in fertility. As a rule, after a
certain level of income is reached, fertility stabilizes and may even decline. The decline
in fertility as incomes grow in developed countries is the result of the rising opportunity
cost of women’s time and is observed in many developed countries. We do not foresee,
however, that Armenia will reach that point during the time period of our analysis. We
assume, instead, that in the long run the fertility in Armenia will decline as it will drop
elsewhere in the world. The United Nations (UN) envisages the world average Total
Fertility Rate (TFR) will drop below 2.1 by 20501. Taking into consideration the current
rising fertility rates as well as steady economic growth, and the fact that Armenia’s TFR
is always below the world average, we assume that a gradual increase in TFR will
continue until 2015 (reaching the level of 1.9) and remain at that level over the whole
period under consideration.
Our assumption about the temporary growth of the TFR is stipulated also by the results of
Armenian Demographic Health Survey (ADHS - 2000). Armenian women have indicated
that ideal average number of children in an ideal-sized family is 2.30, according to
ADHS, whose survey took place under more severe economic conditions in Armenia
than those currently in existence. According to the same survey, the number of desired
children positively correlates to the number of actual children, i.e. women having no
1
see http://www.lifesite.net/ldn/2003/feb/03022703.html
5
children desired to have on the average 2.3 children, women having one child desired to
have 2.4 children, women having two children desired to have 2.7, etc. In general the
ideal average number of children for all women was 2.7, for married women - 2.8.2
Hence, our assumptions are rather realistic and can even be assumed to be pessimistic.
The next argument favouring our assumption about fertility-rate growth is stipulated by
the young average age of the population: 31.5 for male and 34.9 for female (in 2004).
This provides prospectus for fertility rate escalation for the period of economic growth.
For the purposes of the forecast, it is assumed that the TFR will grow to 1.9 by 2015 and
then remain at that level by 2080. We presume rapid growth of TFR initially and its
stabilization after 2015. Our assumption relies on macroeconomic trends for the coming
ten years and afterwards and child bearing preferences of Armenian families.
Expectations of people on improvement of economic wellbeing over that period would
encourage them to have more babies. However, after 2015 we assume TFR stabilizes as
macroeconomic situation soothes and women’s economic activity rises. The TFR level
1.9 (besides being below the envisaged world average level 2.1, as indicated above) fits
pretty well also with the United Nations projections of fertility rate growth trends (see
Figure A-2 in Appendix ). Table 1-1 shows the TFR used in the PROST model.
Table 1-1. TFR in Armenia for 2004-2080
Total Fertility Rate (%)
2004
2005
2010
2015
2020
2030
2040
2050
2060
2070
2080
145.0
160.0
181.7
190.0
190.0
190.0
190.0
190.0
190.0
190.0
190.0
Sex ratio at birth: Table 1-2 shows that for the recent years the sex ratio at birth (boys
per 100 girls) in Armenia was well above the world average level 1.06 recorded by
United Nations Population Division for 1995-2000.3 This is partially due to after war
period syndrome4 (observed in many countries) as well as to the practice of offspring sex
selection that “…can be found in a large variety of historical cultures from all continents.
In virtually all cases, the selection was in favor of male infants”.5
Table 1-2. Sex ratio at Birth in Armenia for 1998-2004
Boys
Girls
Total Births
Sex ratio at Birth
1998
21176
18190
39366
1.16
1999
19758
16744
36502
1.18
2000
18699
15577
34276
1.20
2001
17286
14780
32066
1.17
2002
17280
14949
32229
1.16
2003
19038
16469
35507
1.16
2003
19293
16786
36079
1.15
2004
20247
17616
37863
1.15
2
Armenia Demographic Health Survey - 2000, NSS, MOH, ORC Macro, Yerevan, 2001, p. 138
See http://esa.un.org/unpp/p2k0data.asp
4
Sex ratio at birth increases during after war periods, this phenomena was observed in many countries
3
5
Hudson, Valerie M.; Den Boer, Andrea, “A Surplus of Men, A Deficit of Peace: Security and Sex Ratios
in Asia's Largest States”, http://it.stlawu.edu/~govt/361F02Hudson.html
6
For the purposes of projections we assume that the well observable decline of the sex
ratio at birth during 2000-2004 (see Table 1-2) will continue further. However, taken that
ratios within the range from 105 to 107 are considered normal across racial groups6, and
the male preferences in Armenian society that result in widespread selective abortions for
having a boy baby in Armenia, we assume that the sex ratio at birth will not drop below
the projected for Asian countries average 1.067 and will stabilize on the level 108 after
2020 (see Table 1-3).
Table 1-3. Sex ratio at Birth in Armenia for 2004-2080
Sex ratio at birth (%)
2004
115
2005
114
2006
113
2010
111
2015
109
2020
108
2080
108
Mortality Rates: In 2004, the infant mortality rate was 11.5 per 1,000 live births as per
the Statistical Yearbook (see Table 1-4). This figure is substantially less than the DHS2000 indicator 36.18 (DHS uses indirect methods for assessing infant and child mortality
rates). During the last 10-15 years, among those households that left the country, the
highest share comprised those with less (than the average number) children, as they had
higher mobility. Due to this aspect of emigration, the probability that the women
interviewed had more children on average (than would have been the case in the absence
of emigration), and hence the probability that they would have had experienced child
deaths during the observed period was higher than in average. Nonetheless, for our
analysis and projections we take smaller mortality indicators (1.58 for infant boys and
1.21 for girls) calculated on the basis of the average number of deaths for 2001-2003, as
the smaller mortality indicators contributes to higher life expectancy and hence are
“supportive” factor for the pessimistic scenario for pension system projections to which
we would like to adhere.
Table 1-4. Infant mortality rates (1999-2003)
1999
2000
2001
2002
2003
2004
572
540
497
450
422
408
Infant mortality rate - total (per 10000 live births)
154.4
155.6
153.5
139.6
117.9
114.6
0-4 MR – male (per 1000)
20.8
22.9
21.9
17.8
14.9
15.2
0-4 MR – female (per 1000)
17.3
16.4
15.2
Source: Statistical Yearbook of Armenia – 2004, 2005 (pp. 37, 40)
15.2
11.9
10.4
Deaths (under 1)
With the assumed rates of fertility and mortality, the total population of Armenia is
projected to fall after reaching 3,446,822 in 2043 to 3,031,633 by 2080, as a result of the
aging of the population. In 2080, the average age for males is foreseen to be 40.5 years
(versus 31.5 years in 2004) and the average age for females is foreseen to be 43.0 years
(34.9 years in 2004). This is due to the aging of the population.
Life Expectancy is quite high in Armenia: in 2004 it was 69.3 for men and 75.8 for
women, and 72.3 for both genders. For the purpose of the simulation, a gradual decline in
6
IBID
See http://esa.un.org/unpp/index.asp?panel=2
8
IBID, Table 9.2, p. 115
7
7
mortality rates for both genders was assumed, so that over the time period under
consideration (i,e., over 75 years) the life expectancy for female will increase from 75.8
to 78.8 (that is increases by 3.0 years) and for male from 69.3 to 72.5 (that is increases by
3.2 years), which implies increase in 0.40 years for female and 0.42 years for male in
each 10 years. The difference in growth rate is natural as life expectancy grows at a
higher rate for males compared to females. For both genders, life expectancy is expected
to be 75.5 in 2080 compared to 72.3 in 2004. As life expectancy increases by 1 year
every 10 years in countries with lower life expectancy and by 1 year every 15 years in
countries with the highest life expectancy,9 the envisaged increase pattern put into the
model is closer to those of highest life expectancy nations and hence is pretty realistic.
Mortality tail (mortality rates for ages over 95) was adjusted with a best fit exponential
function to avoid peculiar to higher ages mortality volatility.
In 2004, an Armenian man who attained retirement age 63 could expect to live another
14.4 years; a woman who was 63 could expect to live further 17.3 years. In 2004 life
expectancy at retirement for female was 20.1 years as they retired at 59.5 (see Table 15). In 2004, life expectancy at 63 for both genders was 15.9. By 2080, life expectancy at
retirement is projected to reach 16.2 for male and 19.4 for female. In 2080, life
expectancy at birth would reach 75.5 years and 17.9 at retirement for both genders
together. Provided with the presumed age-sex mortality rates (for the period of 20052080), PROST generates life expectancy data for intermediate ages, 20, 60, 65, and at
retirement.
Table 1-5. Life Expectancy at Birth, Retirement, and other Ages
2004
2005
2010
2015
2020
2030
2040
2050
2060
2070
2080
Life Expectancy: At Birth
69.3
69.5
69.6
70.2
71.0
71.1
71.2
71.3
71.6
71.8
72.0
At Age 20
At Age 60
At Age 65
At Retirement
Retirement Age
50.9
51.0
51.1
51.6
52.2
52.3
52.4
52.5
52.7
52.9
53.1
16.4
16.5
16.6
16.9
17.3
17.4
17.5
17.5
17.7
17.8
18.0
13.2
13.3
13.4
13.7
14.0
14.1
14.2
14.2
14.4
14.5
14.7
14.4
14.5
14.6
14.9
15.3
15.4
15.4
15.5
15.6
15.8
15.9
63.0
63.0
63.0
63.0
63.0
63.0
63.0
63.0
63.0
63.0
63.0
75.8
76.0
76.1
76.5
77.1
77.3
77.4
77.5
77.8
78.0
78.3
57.1
57.2
57.3
57.7
58.2
58.3
58.4
58.5
58.8
59.0
59.2
19.7
19.8
19.9
20.2
20.6
20.7
20.8
20.9
21.1
21.3
21.5
15.8
15.9
16.0
16.3
16.7
16.7
16.8
16.9
17.1
17.3
17.4
20.1
19.8
18.7
17.9
18.2
18.3
18.4
18.5
18.6
18.8
19.0
59.5
60.0
61.5
62.8
63.0
63.0
63.0
63.0
63.0
63.0
63.0
73.2
73.9
74.1
74.3
74.5
74.8
75.0
75.3
75.5
Male
Female
Life Expectancy: At Birth
At Age 20
At Age 60
At Age 65
At Retirement
Retirement Age
Life Expectancy: At Birth – Both Genders
72.3
72.5
Migration: Recent statistics on migration point to a substantial reduction in emigration in
2002-2004, and according to the information by the Department of Migration and
9
“PROST: Preparing Input” Manual (p. 15) “… life expectancy will increase by 1 year every 10 years in
countries with lower life expectancy today and by 1 year every 15 years in countries with the highest life
expectancy today”.
8
Refugees, in 2004, for the first time in the last 13 years, arrivals exceeded departures by
2,060 (see Table 1-6). Figures for 2005 reflect information available for the nine month
period of January to September. Keeping in mind that, in 2003 and 2004, during the
October-December period arrivals exceeded departures by 40 and 50 thousand
respectively,10 we believe that the arrivals will outnumber the departures and final balance
for 2005 will be positive as well. Nonetheless, we presuppose it will be negligible and,
for the purposes of our projections, we disregard the factor of migration for 2005 and
over the whole period under consideration. It should be noted, however, that the factor of
migration was taken into account for arriving at the population estimates for 2004.
Table 1-6. Migration in Armenia for 2000-2005
2000
2001
2002
2003
2004
2005∗
Arrivals
399,663
508,211
590,654
618,348
739,886
608,587
Departures
457,162
568,600
593,373
628,509
737,826
642,510
Balance
-57,499
-60,389
-2,719
-10,161
2,060
-33,923
∗
NOTE: Figures for 2005 reflect information only for January-September
Source: State Department for Migration and Refugees
(http://www.dmr.am/ADMR/IRAVIJAK/Iravic%7E1.htm)
Table 1-7 presents the population numbers in total as well as for different age groups. It
summarizes some other demographic indicators for the period 2004 – 2080. As per
Census-2001 data the sex ratio was 92.3 percent for de jure population and 88.2 for de
facto population. As de facto population served for generating the sex-age distribution of
population for 2004, we arrive at 88.4 percent for 2004 and with the presumed trends in
sex ratio at birth and mortality end up at 98.8 percent in 2080.11 This is a quite reasonable
rate taken widely accepted male child preference and improvements in life expectancy. It
is worth recalling that in Armenia before World War II the sex ratio was 103.9 as per
Census-2026, 102.3 (Census-1939) and dropped substantially to 91.5 by 1959. Since then
this indicator grew to 97.8 percent in 1989 as per Census-1989.12
10
11
See the State Department for Migration and Refugees site: http://www.dmr.am
Below are presented sex ratios for the following selective countries (for 2000):
Armenia
88.8
Ireland
98.7
Romania
95.5
Lebanon
Georgia
90.2
Netherlands
98.3
Germany
95.2
India
94.8
Pakistan
106.1
Iraq
102.6
106.6
Turkey
101.9
Iran
103.3
Azerbaijan
95.3
France
Latvia
84.6
French Guiana
96.1
105.2
Source: United Nations Population Division website: http://esa.un.org/unpp/index.asp?panel=2
12
Collection of Demographic Data – Armenia (1940-2000), NSS (Statistics Sweden, SIDA), Yerevan,
2002, p. 10
9
Table 1-7. Population of Armenia for 2004-2080
2004
2005
2010
2015
2020
2030
2040
2050
2060
2070
2080
Total Population
3,017.9
3,030.2
3,136.3
3,259.5
3,350.8
3,401.6
3,446.1
3,399.2
3,299.6
3,185.4
3,039.7
Male
1,415.8
1,423.0
1,481.8
1,549.7
1,601.3
1,636.6
1,672.0
1,665.4
1,627.1
1,577.7
1,510.4
Female
1,602.1
1,607.2
1,654.5
1,709.8
1,749.5
1,765.0
1,774.1
1,733.8
1,672.5
1,607.7
1,529.3
687.1
663.4
614.6
668.8
736.7
618.7
604.1
605.9
546.2
553.0
516.2
1,913.0
1,952.9
2,158.1
2,204.6
2,172.2
2,194.2
2,229.9
2,071.8
1,955.7
1,932.6
1,806.2
Retirement Age +
417.8
413.9
363.6
386.1
441.9
588.7
612.0
721.5
797.7
699.9
717.2
Males per 100 Females (%)
88.4
88.5
89.6
90.6
91.5
92.7
94.2
96.1
97.3
98.1
98.8
22.8
21.9
19.6
20.5
22.0
18.2
17.5
17.8
16.6
17.4
17.0
59.4
0 - 14
15 - Ret. Age
Share (%)
0 - 14
15 - Ret. Age
63.4
64.4
68.8
67.6
64.8
64.5
64.7
60.9
59.3
60.7
Ret. Age +
13.8
13.7
11.6
11.8
13.2
17.3
17.8
21.2
24.2
22.0
23.6
Pop. Dependency Rate
57.8
55.2
51.6
45.3
54.3
55.0
54.5
64.1
68.7
64.8
68.3
Age 0-14 / Age 15-Ret. Age
35.9
34.0
31.9
28.5
33.9
28.2
27.1
29.2
27.9
28.6
28.6
Ret. Age + / Age 15-Ret. Age
Support Ratio: Age 15-Ret.
Age / Ret. Age +
Total Fertility Rate (%)
21.8
21.2
19.7
16.8
20.3
26.8
27.4
34.8
40.8
36.2
39.7
4.6
4.7
5.1
5.9
4.9
3.7
3.6
2.9
2.5
2.8
2.5
145.0
160.0
181.7
190.0
190.0
190.0
190.0
190.0
190.0
190.0
190.0
Figure 1-1 shows the dynamics of Armenia’s population under the assumptions made in
this section regarding fertility, mortality, and migration. To attain the level of 3,501.9
thousand until the end of the considered period, the TFR should grow from 1.9 (2015) to
2.0 (in 2020) and then to 2.3 in 2080. If the presumed trends (1.9 TFR) hold for another
25 years then the population will drop further to 2,744,179 until 2105.
Figure 1-1. Total Population of Armenia in 2004-2080 (thousand)
Total Population (000)
3,500
3,400
3,300
3,200
3,100
3,000
2,900
20
79
20
74
20
69
20
64
20
59
20
54
20
49
20
44
20
39
20
34
20
29
20
24
20
19
20
14
20
09
20
04
2,800
10
1.2-Macroeconomic Assumptions
Macroeconomic conditions significantly affect the financial stability of the pension
system and the well being of the population and pensioners among them. Therefore, the
relevant indicators that present these conditions are part of the background or endogenous
parameters of the model we discuss in the paper. For the purposes of the analysis the
main macroeconomic indicators were derived from Socio-Economic Conditions of the
Republic of Armenia for 1999-2004, Economic Development and Research Center
(EDRC) website, Armenian-European Policy and Legal Advice Center (AEPLAC)
information bulletins, and whenever applicable they were compared with the indicators
provided by the PRSP final document.
The overall macroeconomic developments remain favorable for Armenia. The Armenian
economy has been steadily growing for the last decade. Since 2001 GDP has been
growing by a double digit number and the year 2004 was not an exception, as GDP grew
by 10.1 percent. Inflation remains subdued and labor market indicators have been
improving year after year.
GDP, Inflation, and Productivity Dynamics: As Table 1-8 shows, GDP numbers have
demonstrated severe seasonality during 1996-2004. However, our analysis show that total
annual GDP quotient to semiannual GDP has been stable around 3 in the last 9 years,
which has been used to forecast (nominal) GDP in 2005.
Table 1-8. Total GDP over half year GDP for 1996-2004
Q1
Q2
Q3
Q4
Total
Total GDP /
(Q1+Q2) GDP
1996
82,490.5
133,606.4
219,061.0
226,051.1
661,209.0
3.1
1997
93,334.8
159,224.0
266,218.9
285,557.9
804,335.6
3.2
1998
124,683.7
198,020.6
308,304.3
324,376.2
955,384.8
3.0
1999
130,215.7
216,904.2
325,718.9
314,604.9
987,443.7
2.8
2000
130,015.8
217,449.7
329,062.3
354,810.5
1,031,338.3
3.0
2001
151,149.3
237,113.5
389,072.5
398,541.5
1,175,876.8
3.0
2002
162,803.9
275,280.1
440,877.3
483,510.4
1,362,471.7
3.1
2003
189,849.7
330,657.9
532,357.5
571,777.6
1,624,642.7
3.1
2004
216,743.9
386,503.5
641,121.7
652,072.5
1,896,441.6
3.1
Source: Statistical Annex Q1 2005, AEPLAC, own calculations
Nominal GDP in the first half year, 704.713 billion AMD, was multiplied by a factor of
3.1, which provides 2,166.2 billion AMD of annual GDP for 2005. Thereafter, annual
real growth rates (and inflation) were used to project long term annual GDPs. Table 1-9
summarizes the main macroeconomic indicators that are put into the model for all the
scenarios developed in this paper. We assume that after 2015 the real GDP growth rate
gradually declines to 4.0% in 2020 and then to 2.0% in 2080. These assumptions are
consistent with the current trends: Armenia has the highest GDP growth rate among CIS
countries and the assumed rates may even be rather conservative.
13
Socio-Economic Situation of Armenia, NSS, January-June, 2005, p. 7 and
http://www.edrc.am/project.html?cat_id=38
11
Table 1-9. Main Macroeconomic Indicators (in percent)
2004
2005
2006
2010
2015
2020
2025
2030
2040
2050
2060
2070
2080
Real GDP (in billion
AMD)
1,896.4
2,063.3
2,187.1
2,718.0
3,491.0
4,329.5
5,242.2
6,296.6
8,867.7
12,092.1
15,963.7
20,401.6
25,237.5
GDP Growth Rate
10.1
8.8
6.0
5.3
5.0
4.0
3.0
2.9
2.7
2.5
2.4
2.2
2.0
Real Interest Rate
6.0
5.8
5.5
4.8
4.0
3.0
2.5
2.4
2.2
2.0
1.9
1.7
1.5
Nominal Interest
Rate
13.4
8.9
8.7
8.0
7.1
6.6
6.1
6.0
5.8
5.6
5.4
5.2
5.1
Inflation Rate
7.0
3.0
3.0
3.0
3.0
3.0
3.0
3.0
3.0
3.0
3.0
3.0
3.0
Price Index
1.0
1.0
1.1
1.2
1.4
1.6
1.9
2.2
2.9
3.9
5.2
7.0
9.5
Wages: For 2001-2004, the share of midyear average wage (AW) in the annual average
wage was 92-95 percent (see Table 1-10). Hence, the annual estimated average wage is
rounded to 52,100 drams for 2005 as the semiannual average wage was 49,474 drams.
Table 1-10. Annual and Mid-Year Average Wages (2002-2005)
Annual Average Wage (AMD)
Average Wage at Mid year (AMD)
2002
2003
2004
2005
27,324
25,029
32,518
29,977
41,976
40,014
52,100.0
49,474
0.92
0.92
0.95
0.95
Proportion of Mid-year average wage in annual AW
Source: Socio-Economic Situation of Armenia, NSS, 2002, 2003, 2004, 2005 (January-June, JanuaryDecember)
Historical data on the structure of GDP based on income methodology demonstrate that
the wages average to 41.1 of GDP for the period of 2000-2003 (see Table 1-11). In
Armenia, during the recent 3-5 years the wages grew at much higher rates than GDP14,
which is common for countries in transition where the wages comprise small share of
GDP (it is worth to mention that in developed countries the wages comprise almost 70
percent of GDP).
Table 1-11. Structure of GDP by Income (in percent)
GDP
Wages
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
57.4
53.0
38.4
41.8
41.5
39.3
40.0
41.8
41.3
42.9
42.7
41.8
39.5
40.3
15
GP&MI
31.3
41.1
53.2
53.1
53.9
55.7
53.6
49.3
48.7
46.8
46.6
46.9
49.1
49.0
Net Taxes
11.2
5.9
8.4
5.1
4.6
5.0
6.5
8.9
10.0
10.3
10.6
11.3
11.4
10.7
Source: http://www.edrc.am/project.html?cat_id=67
14
Wage and GDP growth rates (%) in Armenia in 1999-2003:
Wage Growth Rate
GDP Growth Rate
1999
2000
2001
2002
2003
113.6
103.3
109.1
106.0
109.7
109.6
115.0
112.9
115.1
113.9
Growth Rate of Money Income of population and GDP real growth rate for 1999-2003
Source: Socio-Economic Situation in Armenia, NSS, 1999, 2000, 2001, 2002, 2003
15
GP&MI - Gross Profit and Mixed Income
12
The share of all wages in economy in GDP was 40.3 percent of GDP in 2003 (see Table
1-11). In our exercise the wages of self-employed and agricultural workers are not taken
into account, which is incorporated in the Economic Development and Research Center
(EDRC) website data. Taken that almost half of the working people work in the sphere of
agriculture and taking into account the substantial increase in real wage (20.6 %) in 2004
we come to 23.0 percent of GDP effective wage bill in 2004.
Table 1-12. Wage Bill
2004
2005
2006
2010
2015
2020
2025
2030
2040
2050
2060
2070
Wage Bill: Nominal
Contributors (in
billion AMD)
436.9
553.7
684.7
1,235.0
2,063.3
3,206.0
4,609.3
6,429.9 12,375.1 21,672.9 36,698.6 63,710.9 104,271.2
Average Nominal
Wage of Effective
Contributors (in
thousand AMD)
503.7
607.3
713.9
1,048.8
1,381.2
1,773.9
2,168.2
2,559.5
3,460.1
4,666.1
6,121.9
7,856.6
9,947.4
Productivity Growth
20.6
20.5
17.5
5.5
5.1
4.1
3.1
3.0
2.8
2.6
2.4
2.2
2.0
Wage Bill of Nominal
Contributors as a
Percent of GDP
23.0
26.2
29.9
39.5
45.7
51.5
53.5
55.2
58.9
58.2
56.7
57.5
56.0
Effective Wage Bill
as a Percent of
Nominal Wage Bill
53.9
54.1
54.3
55.1
56.2
57.2
58.3
59.4
61.5
63.6
65.7
67.9
70.0
GDP Growth Rate
10.1
8.8
6.0
5.3
5.0
4.0
3.0
2.9
2.7
2.5
2.4
2.2
2.0
Effective Wage Bill
as a Percent of GDP
12.4
14.1
16.0
21.0
24.0
26.4
28.2
29.9
33.6
35.0
35.8
38.2
39.0
For the purposes of assessing the effective wage bill the growth of real wage is assumed
to follow the pattern presented in Table 1-12 (see the line: Productivity Growth). The
envisaged trend will produce nominal wages constituting 56.0 percent of the Nominal
GDP in 2080. We also assume that Effective Wage Bill as a Percent of Nominal Wage
Bill grows from 53.9 percent (in 2004) to 70.0 percent (in 2080). These assumptions are
stipulated by the idea that the share of agriculture will decline as compared to industrial
and service sectors of economy where the wages will grow at higher rate than GDP will
do.
The noted above trends will result in Effective Wage Bill comprising 12.4 percent of
GDP in 2004 and 39.2 percent in 2080 (see Table 1-12). We assume that after 2006 the
productivity growth rate gradually declines (in parallel to decline in GDP growth rate)
from 20.6 percent (in 2004) to 5.5 percent in 2010, 5.1 percent in 2015, 4.1 percent in
2020, 3.1 percent in 2025, and afterwards gradually drops to level out with the GDP
growth rate 2.0 percent in 2080. Recall that we assumed GDP growth rate to be 5.3
percent in 2010, 5.0 percent in 2015, 4.0 percent in 2020, 3.0 percent in 2025, and 2.0
percent in 2080.
1.3-Labor Force, Employment and Contributors to the System
PROST model was used to project the number of working age people in the population.
Figure 1-2 demonstrates the proportion of population for the following three cohorts:
children 0-14 years old, people of working age (from the age of 15 to retirement age), and
13
2080
those above retirement age. The figure shows that the proportion of retirement age people
slowly increases after 2011. Meanwhile, the proportion of working age population
gradually decreases. This raises the population dependency ratio, meaning that in order to
support larger cohort of retirees the contributors will have to make higher contributions in
the future.
Figure 1-2. Distribution of Population of Armenia by Age Groups (2004-2080)
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0 - 14
15 - Ret. Age
20
80
20
75
20
70
20
65
20
60
20
55
20
50
20
45
20
40
20
35
20
30
20
25
20
20
20
15
20
10
20
05
20
04
0%
Retirement Age +
For estimating the number of contributors to the system we need to assess the number of
economically active people in the population and share of unemployed among them.
Economic activity rates presented in Table 1-13 were used to estimate the economically
active population. Taking into consideration that, according to the Labor Market sample
survey 80.7% of male and 55.0% of female are economically active we come to the
number of economically active population 1,287.7 thousand of which: 881.4 thousand are
employed (68.4%) and 406.3 thousand unemployed (31.6%). The difference between our
data and the official data on the number of economically active people, 1,225.7
thousand16 can likely be explained by how the official information reflects information on
working age (i.e., until the retirement age) population, while our estimate includes
economically active population of all ages.
Recently conducted several Labor Force sample surveys provided us with sufficient and
reliable data for understanding the real situation in labor market. Table 1-13 presents the
economic activity rates for different age groups of men and women.
16
Socio-Economic Conditions of Armenia, NSS, 2003, p. 102
14
Table 1-13. Economic Activity Rates by Age and Sex Groups – 2003 (in percent)
Total
15-19
20-24
25-29
30-34
35-39
40-44
45-49
50-54
55-59
60-64
65+
male
80.7
26.8
81.3
91.9
93.5
94.7
95.0
93.8
92.6
87.7
62.4
36.7
female
55.0
25.0
49.5
51.8
64.6
66.4
72.4
67.2
67.4
57.5
31.4
19.8
total
66.0
25.7
62.5
69.0
76.5
78.9
82.3
78.7
78.3
72.1
43.9
27.7
Source: Labor Market in the RoA, NSS, 2004, p. 17
The Labor Sample Surveys suggest 40.9% unemployment rate in 199917, which declined
to 31.6% in 2004 (see Table 1-14) due to economic growth. The actual level of
unemployment is far for being considered at a natural level (some 8-11% in developed
countries). Needless to recall that approximately 95 percent of out migration of the last 35 years is attributable to job factor: lack of workplaces (58.0%), impossibility to earn
enough money to provide satisfactory living standards (31.3%), and lack of work by
specialty (5.3%)18. Therefore, creation of new jobs is the most effective means to hold
back out-migration and retain the decline of the population in the long run. For the
purposes of our projections we assume that the total unemployment rate will decline to
9.0 percent for male and 15.0 percent for female in 2080.
Table 1-14. Unemployment Rates by Age and Sex Groups – 1999-2003 (in percent)
1999
42.6
39.1
40.9
Male
Female
Total
2001
36.8
40.2
38.4
2002
31.4
39.7
35.3
2003
24.9
38.2
31.2
2004
26.6
37.6
31.6
Source: Labor Market in the RoA, NSS, 2004, p. 11, 18
Labor Force and Child Labor Survey, NSS, 2004, p. 15
The unemployment rates peculiar to men and women of different age groups, presented
in the Table 1-15, serve as base year data in the PROST model for this work. It is
interesting to observe that female unemployment rates are close to or higher than those of
men for almost all age groups. As a result of this phenomenon total female
unemployment rate (38.2%) is more than 1.5 times higher than the male unemployment
rate (24.9%). In 2003, unemployment totaled to 31.2 percent. In 2004, the unemployment
rate slightly increased to 31.6 percent: among men 26.6 percent, and among women –
37.6 percent.19
Table 1-15. Age-sex Distribution of Unemployment Rates (2003)
Total
15-19
20-24
25-29
30-34
35-39
40-44
45-49
50-54
55-59
60-64
65-69
70+
Male
24.9
71.7
49.3
28.8
22.4
18.0
20.5
16.7
12.2
21.5
19.2
31.5
5.9
Female
38.2
70.7
63.9
41.2
40.6
34.8
35.5
28.5
29.1
18.5
27.6
30.3
0.0
31.2
70.9
56.1
34.1
31.4
25.9
28.0
22.4
20.5
20.3
22.8
31.0
2.9
Total
Source: Labor Market in the RoA, NSS, 2004, p. 17
17
Labor Market in Armenia, NSS, 2004, p. 11
Sample Survey of Passenger Turnover (Migration) at the Border Guarding Posts of the RoA, 2003, NSS,
p. 70
19
Labor Force and Child Labor Survey, NSS, 2004, p. 15
18
15
For the purposes of the projections we assume that concurrent to economic growth the
economic activity will increase and unemployment will decline. In addition, we assume
that economic activity and the employment rate of women will grow at higher pace than
that of men. This assumption is stipulated by the current trends worldwide.
Table 1-16. Distribution of Employed in Different Age and Sex Groups (in percent)
Male
Female
Total
0-14
0.1
0.0
0.1
16-19
1.9
1.5
1.8
20-24
8.7
5.6
7.4
25-29
12.2
9.6
11.1
30-34
10.3
9.7
10.0
35-39
11.7
11.5
11.6
40-44
14.9
16.9
15.7
45-49
14.1
15.1
14.5
50-54
9.4
13.0
10.9
55-59
6.7
7.8
7.1
60-64
3.8
3.9
3.8
65-69
4.2
4.1
4.1
Source: Labor Force and Child Labor in Armenia (Labor Force Sample Survey), NSS, 2004, p. 47
The share of employed by age groups are shown in Table 1-16 as per sample survey of
Labor Force and Child Labor. This table shows that the female aged 40-64 show higher
than or equal employment rates as men did in 2004 (analogous situation was registered
also in 200220). Figure 1-3 shows the share of employed in different age groups as per
Table 1-16.
Figure 1-3. Share of Employed in Different age Groups - 2003
70+
65-69
60-64
55-59
50-54
45-49
40-44
35-39
30-34
25-29
20-24
15-19
0-14
0.0%
2.0%
4.0%
6.0%
8.0%
10.0%
Total
Male
12.0%
14.0%
16.0%
18.0%
Female
Since few babies were born during the last ten years of economic transition, the labor
force will decline after 2020, a period when those born in 1990s will join the labor force.
With the strengthening of the Armenian economy, more jobs will be created (the PRSP
forecasts the creation of more than 245,000 new jobs in the next 10-12 years), and these
rates will change: economic activity will grow and unemployment rates will fall. We
20
Socio-Economic Situation of Armenia, NSS, 2002, p. 106
16
70+
2.2
1.5
1.9
assume also that many of the gray, or shadow, economy jobs will merge into the regular
economy as a result of anti-corruption measures taken by the Government of Armenia.
These trends are taken into account in our calculations as a gradual increase in economic
activity and a steady decline in unemployment. It should be noted that well designed
pension system will itself encourage increased participation of working people in the
formal economy.
55.1% of all employed male and 44.9 % of all employed female work in the sphere of
agriculture, which means that almost half of the working people make no contribution to
the system. This pretty well fits with the information on the number of effective
contributors to the system, namely some 430.0 thousand people21 (48.8% of all the
employed), which indirectly indicates on improvement in compliance rate as compared to
the historical data. It should be noted that the number of contributors to the system was
decreasing until recently: according to SSIF information the number of contributors was
441.7 thousand in 2000; in 2002 the number of effective contributors was 417.9
thousand22 and 410.0 thousand in 2003. The reduction in the number of contributors is
apparent. The ROA Law on Social Security Cards (adopted in September 24, 2003)
requires all Armenians to have social security cards after July 1st, 2004 and to use their
social security numbers for all financial transactions. Social cards along with personified
reporting (mandatory after October 1st, 2005) system, stricter regulation to register the
employees, and transfer of contribution collection to the State Agency for Revenues led
to shrinking of the shadow economy and improvement in estimates of effective
contributors. The compliance rate is assumed to reach 90.0 percent in 2015 and remain at
that level over time. This is partially also due to removing the most undisciplined
contributors (self employed) from the system and introducing social security card and
personified reporting systems in pension insurance system.
To assess the number of working people, the number of unemployed is subtracted from
the number of economically active. However, the number of employed include those in
both the formal and the informal economy – and informal economic activity entails
paying little or no taxes and social contributions. We derived the proportion of effective
contributors in the cohort of nominal contributors using current data and assumed that
this proportion would increase over time. Table 1-17 shows that the share of effective
contributors in nominal contributors (which includes also self-employed and agricultural
workers) grows from 53.0 percent in 2004 to 70.0 percent in 2080.
21
22
Employer-Employee, “Hayastani Hanrapetutyun”, No 191 (3815), November 2, 2005
“Hayastani Hanrapetutyun”, No 225 (3358), November 28, 2003
17
Table 1-17. Labor Force, Nominal and Effective Contributors (000)
2004
2005
2010
2015
2020
2030
2040
2050
2060
2070
2080
Total Population
3,017.9
3,030.2
3,136.3
3,259.5
3,350.8
3,401.6
3,446.1
3,399.2
3,299.6
3,185.4
3,039.7
Total Labor Force
1,287.7
1,315.7
1,464.6
1,587.2
1,633.6
1,639.9
1,684.0
1,587.6
1,503.6
1,475.3
1,397.5
Labor Force as % of Total
Population
42.7
43.4
46.7
48.7
48.8
48.2
48.9
46.7
45.6
46.3
46.0
Unemployment Rate (%)
31.6
31.1
28.6
25.6
21.1
21.0
20.5
20.3
20.7
20.3
20.7
Nominal Contributors (NC)
881.4
906.1
1,045.3
1,180.9
1,289.1
1,295.9
1,338.7
1,265.6
1,191.7
1,175.3
1,108.7
Effective Contributors (EC)
467.4
478.9
543.5
606.0
644.3
691.4
758.8
758.6
752.9
782.4
776.1
Share of EC in NC (%)
53.0
52.8
52.0
51.3
50.0
53.3
56.7
59.9
63.2
66.6
70.0
Special attention should be paid on the factor of hidden employment, which according to
the Labor Market sample surveys constituted 22.6 percent of all the employed (in 2001),
26.7 percent in 2002, and 26.9 percent in 2003.23 These figures demonstrate the share of
workers working on oral agreement as well as unregistered employers and self-employed
in the number all employed. Another form of shadow economy, i.e. widespread underrecording of the wages, should be taken into account for estimating the real volume of
potential contributors to the pension system. Taken into account that on the average the
“gray” wages are twice higher than the officially registered wages then it would be clear
that a substantial amount (more than half of currently collected contributions) is left out
of the system. We assume that the segment of shadow economy will decrease as a result
of which the compliance (contribution collection rate) will improve.
23
Labor Market in the RoA, NSS, 2004, p. 11, 18
18
SECTION II - FINANCIAL AND DEMOGRAPHIC ANALYSIS OF THE PENSION
INSURANCE SYSTEM
2.1. Pensioners
The proportion of beneficiaries in the total population will decline over the next 10 years
as the generation born during World War II retires, which is a small cohort. The number
of privileged pensioners from World War II is also gradually declining, and no new
entrants under this category will join the strata of the beneficiaries. This refers also to
multi-children mothers and mothers caring for disabled children. This is the largest
cohort of privileged pensioners (47,344: 65.9 percent of all privileged pensioners and
12.3 percent of all pensioners in 2003). When, starting in 2003, no new privileged
pensions are allocated to this category of beneficiaries, their number dropped to 40,859
(66.3 percent of all privileged pensioners and 10.7 percent of all pensioners in 2004).
Figure 2-1 shows that, after a consistent decline to 363.6 thousand, the number of
retirement-age population is to increase after 2010 and, after reaching its maximum of
803.2 thousand, drop again in 2057. This phenomenon is mostly due to the fact that, in
2060s, the second generation of people born after World War II reaches retirement age.
Figure 2-1. Retirement Age Population (000)
900.0
800.0
700.0
600.0
500.0
400.0
300.0
200.0
100.0
20
80
20
76
20
68
20
72
20
60
20
64
20
52
20
56
20
44
20
48
20
36
20
40
20
28
20
32
20
20
20
24
20
12
20
16
20
08
20
04
.0
Retirement Age Population
The second reason for the initial decline in the retirement age population is the increase
in retirement age for almost all categories of retirees. The old-age retirement age is set 63
for both genders. It has already attained 63 for men and will gradually approach 63 for
women by 2011. The increase in retirement age reduces the number of new beneficiaries
19
and increases the size of the labor force -- thus playing a positive role in growth of
contribution revenue to the SSIF.
Table 2-1 presents the breakdown of below working-age, working age, and retirement
age population over time. It can be easily observed that the share of retirement-age
population is increasing while the working-age population declines. Unsurprisingly, the
population dependency rate grows from 57.8 percent (in 2004) to 64.8 percent (in 2080).
Table 2-1. Distribution of Population (in percent)
2004
2005
2006
2010
2020
2030
2040
2050
2060
2070
2080
Share
100.0
100.0
100
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
0 – 14
22.8
21.9
21.0
19.6
20.5
22.0
18.2
17.5
17.8
16.6
17.4
15 - Ret. Age
63.4
64.4
66.0
68.8
67.6
64.8
64.5
64.7
60.9
59.3
60.7
Ret. Age +
13.8
13.7
13.0
11.6
11.8
13.2
17.3
17.8
21.2
24.2
22.0
Pop. Dependency Rate
57.8
55.2
51.6
45.3
47.8
54.3
55.0
54.5
64.1
68.7
64.8
Age 0-14 / Age 15-Ret. Age
35.9
34.0
31.9
28.5
30.3
33.9
28.2
27.1
29.2
27.9
28.6
Ret. Age + / Age 15-Ret. Age
21.8
21.2
19.7
16.8
17.5
20.3
26.8
27.4
34.8
40.8
36.2
Age 15-Ret. Age / Ret. Age +
4.6
4.7
5.1
5.9
5.7
4.9
3.7
3.6
2.9
2.5
2.8
The total number of pensioners entitled to different types of pensions changes over time.
The model incorporates the parametric changes in the pension system (i.e., retirement age
change and revision of proportion of population entitled to privileged, disabled,
survivors’ and long service pensions). We assume that the proportion of disabled in total
population would steadily decline due to change in disability category entitlement rules
and health care improvements. The latter will also result in gradual decrease of the share
of survivors and orphans (see Table 2-2). As long as self employed including agricultural
workers don’t cumulate credits for entitlement of insurance pensions, the share of disabled beneficiaries as well as survivors’ pensioners will drop further over the period under
consideration. These assumptions along with the others discussed in Section I on
changes in entitlement of privileged pensions will result in decline of the percent of total
beneficiaries of pension insurance system in retirement age population from 111.2
percent in 2004 to 89.1 percent in 2080 (see Table 2-2).
Table 2-2. Number of Beneficiaries and Share of Different Category Beneficiaries in Total
2004
2005
2006
2010
2020
2030
2040
2050
2060
2070
2080
Retirement age Population
417.8
(000)
413.9
396.8
363.6
441.9
588.7
612.0
721.5
797.7
699.9
717.2
Total Number of
Beneficiaries (000)
464.6
479.9
482.4
460.8
475.6
529.4
531.3
613.5
678.6
632.3
639.3
Number of Beneficiaries
as a Percent of Retirement 111.2
age Population (%)
116.0
121.6
126.7
107.6
89.9
86.8
85.0
85.1
90.3
89.1
Share (in percent)
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
Old Age Pensioners
78.1
78.5
78.3
76.2
78.2
82.0
83.1
88.1
89.6
89.2
90.2
Disabled
15.7
15.4
15.6
17.0
16.0
12.9
11.1
7.9
6.7
6.8
6.4
Survivors & Orphans
3.2
3.0
3.0
2.9
2.4
1.7
1.7
1.4
1.2
1.3
1.2
Long Service
3.0
3.1
3.2
3.9
3.4
3.4
4.1
2.6
2.5
2.7
2.3
20
2.2. Payroll Contributions
According to the draft Law on Amendments and Additions in the Law on Mandatory
Social Insurance Contributions24 (that has passed first reading and is expected to come
into force on January 1, 2006) the payroll-based mandatory social contribution scale will
be changed: the flat amount of 5,000 drams will be increased to 7,500 drams and leave
the rates untouched. Payroll-based mandatory social contributions sustain pensions old
age, long service, privileged, disability, and survivors’ pensions and other social
insurance programs, i.e. maternity leave benefit, temporary sickness benefit. The
contribution formula is presented in Table 2-3.
Table 2-3. Social Contribution Rates (will be effective from January 1, 2006)
Wages
Contribution
Employee
Less than 20,000 AMD
Employer
3%
7,500 AMD
20,000 – 100,000 AMD
3%
7,500 AMD + 15% of pay exceeding 20,000 AMD
100,000 + AMD
3%
19,500 AMD + 5% of pay exceeding 100,000 AMD
This innovation will increase the average contribution rate from 22.8 percent in 2004
(21.8% projected for 2005) back to higher than 25.0 percent in 2006 (see Table 2-4).
Argumentation about the size of average wage in 2005 and assumptions on its growth
rate afterwards are given in Section I.
Table 2-4. Average Contribution rates (2002-2007)
2001
2002
2003
2004
2005
Average Wage (AMD)
24,483.0
27,324.0
32,518.0
41,976.0
52,100.0
63,100.0
74,400.0
Employer contribution
5,672
6099
6,877.7
8,296.4
9,815.0
13,965.0
15,660.0
Employee contribution
2006
2007
734
820
975.5
1,259.3
1,563.0
1,893.0
2,232.0
Total contribution
6,407
6919
7,853.2
9,555.7
11,378.0
15,858.0
17,892.0
Contribution rate (percent)
26.2
25.3
24.2
22.8
21.8
25.1
24.0
Contribution Rates: With the change in contribution rates applicable after January 1st,
2006, the combined contribution rate for average wage grows. Table 2-5 shows that the
employers and employees in 2006 will pay much higher rates than in 2004.
Table 2-5. Contribution Rates for Different Wage Levels (projection for 2006)
Contributions / Contribution rate by new formula
Percent
of
Average
Wage
Wage
(AMD)
Employer
Contribution
Employee
Contribution
Combined
Employee
and
Employer
Contribution
Contributions / Contribution rate by old formula
Combined
Employee and
Employer
Contribution
Rate (%)
Employer
Contribution
Employee
Contribution
Combined
Employee and
Employer
Contribution
Combined
Employee and
Employer
Contribution
Rate (%)
50
31,550
9,233
946.5
10,179
32.3
6,732.5
946.5
7,679
24.3
100
63,100
13,965
1,893
15,858
25.1
11,465
1,893
13,358
21.2
24
See Î-684-15.09.2005-üì,ê²-010/1 on the http://www.parliament.am/drafts.php?sel=onagenda&lang=arm
page
21
150
94,650
18,698
2,839.5
21,537
22.8
16,197.5
2,839.5
19,037
200
126,200
20,810
3,786
24,596
19.5
18,310
3,786
22,096
20.1
17.5
250
157,750
22,388
4,732.5
27,120
17.2
19,887.5
4,732.5
24,620
15.6
300
189,300
23,965
5,679
29,644
15.7
21,465
5,679
27,144
14.3
500
315,500
30,275
9,465
39,740
12.6
27,775
9,465
37,240
11.8
Table 2-5 shows that for an average earner the total employee and employer contribution
will be almost 4 percent (25.1% instead of 21.2%) more than it would have been under
the old contribution scheme. In general the Combined Employee and Employer
Contribution Rates for the years from 2006 to 2080 are assumed to follow the presented
in Table 2-6 pattern.
Table 2-6. Combined Employee and Employer Contribution Rate (in percent)
Combined Contribution from
Employees and Employers
2004
2005
2006
2007
2008
2009
2010
2015
2020
2080
22.8
21.8
25.1
24.0
23.3
22.7
22.0
20.0
20.0
20.0
Most SSIF revenues come from direct monthly contributions of employees and
employers. To project future revenues, we used initial data and trends for the number of
contributors, their average wage, the contribution rate, and the rate of compliance. The
cohort of contributors itself consists of effective contributors that regularly make
contributions and those who evade and/or are exempt from paying contributions officially
and meanwhile accumulate years of service (i.e. mothers caring for babies).
In 2004, the average wage for employees working full-time in the formal sector was
41,976 AMD25. The social contribution, which is a flat proportion of wage for employees
and progressive-regressive for employers, are paid each month no later than 20th day of
the next month. The average contribution rate as a percent of the wage bill was
approximately 23 percent in 2004. The model suggests a gradual increase in the number
of nominal contributors. Initially, the real wages may increase substantially, however,
with time passage, the rate of increase will level out with the GDP growth rate. Table 2-7
presents the wage bill of contributors and the expected payroll contributions to the SSIF
for the selected years.
Table 2-7. Wage Bill of Effective Contributors and Contributions Collected
2004
2005
2010
2020
2030
2040
2050
2060
2070
2080
Wage Bill: Nominal
Contributors (billion AMD)
436.9
684.7
1,235.0
3,206.0
6,429.4
12,368.6
21,641.8
36,596.3
63,419.1
103,560.5
Average Nominal Wage
(thousand AMD)
503.7
757.4
1,252.3
2,846.5
5,519.4
10,022.9
18,148.6
31,956.4
55,017.3
93,404.5
Total as Percent of
Nominal GDP
23.0
29.5
38.1
46.1
50.4
54.7
54.9
54.3
56.0
55.3
Wage Bill: Effective
Contributors (billion AMD)
235.5
371.8
680.6
1,834.1
3,815.9
7,605.3
13,767.8
24,060.1
43,044.0
72,492.4
Effective Wage Bill as % of
Total Nominal Wage Bill
53.9
54.3
55.1
57.2
59.4
61.5
63.6
65.7
67.9
70.0
Payroll Contributions Due
(billion AMD)
53.6
93.4
149.6
366.8
763.2
1,521.1
2,753.6
4,812.0
8,608.8
14,498.5
Payroll Contributions
50.9
84.1
134.7
330.1
686.9
1,369.0
2,478.2
4,330.8
7,747.9
13,048.6
25
Socio-Economic Conditions of Armenia, NSS, 2004, p. 10
22
Collected (billion AMD)
Since early 2003, SSIF has required employers to keep records on hiring, firing, and
shifting employees from one position to another. These records (made in special journals)
track the employment position, administrative unit of employment, and, most
importantly, wages. In 2003, the practice of employment record-keeping increased the
number of effective contributors by 6,000. The penalty for keeping the journal
improperly is 200 times the amount of the computational minimum wage (currently 1,000
AMD).
Penalties, fees, and interest on deposits of reserves that accumulate in banks are the other
sources of SSIF revenues. In 2004, these sources contributed 0.4 percent to the revenue
accumulation. For our projections, we assumed that “Other Income as Percent of
Employee, Employer & Pensioner Contributions” would be 0.5 percent throughout the
projection period.
2.3. Dependency Ratios
Pension-system dependency ratios show how many retirees are supported by each
working-age individual (the population dependency rate) and how many actual
pensioners are supported by each worker paying contributions (the nominal system
dependency rate and effective system dependency rate). Over time, the population
dependency rate will fall from 24.4 percent in 2002 to 20.1 percent in 2013 and then increases to 34.7 percent in 2060 as the population ages. Then it grows further to 35.4
percent by 2080. Figure 2-2 highlights the dynamics of dependency ratios: it presents the
nominal and effective dependency ratios, as well as population dependency rate.
Figure 2-2. Dependency Ratios
120.0%
100.0%
80.0%
60.0%
40.0%
20.0%
Nominal System Dependency Rate
Effective System Dependency Rate
20
80
20
75
20
70
20
65
20
60
20
55
20
50
20
45
20
40
20
35
20
30
20
25
20
20
20
15
20
10
20
06
20
05
20
04
0.0%
Population Dependency Rate
The effective system dependency rate also changes over time. In 2001 it was 111 percent
implying that each effective contributor supported at least one pensioner. After removing
23
social pensions from the pension insurance system (according to the Law on State
Pensions), this indicator has improved; it was 102.9 percent in 2002 and 99.4 percent in
2004. The system dependency ratio is projected to fall to 82.4 percent by 2080 (see Table
2-8).
Table 2-8. Dependency Ratios (in percent)
Nominal System Dependency Ratio
Effective System Dependency Ratio
Population Dependency Ratio
2004
2005
2010
2020
2030
2040
2050
2060
2070
2080
52.7
99.4
24.3
53.4
100.2
24.6
46.2
84.8
21.4
42.0
73.8
21.9
45.4
76.6
24.1
43.1
70.0
23.8
51.5
80.9
29.6
59.3
90.1
34.7
54.8
80.8
32.7
57.7
82.4
35.4
The fluctuations are the result of the pension system changes discussed in Section I and
first part of this section, presumed improvement in employment, as well as incongruent
dynamics of population of different age groups (see Figure 2-3).
Figure 2-3. Dynamics of Working and Retirement Age Population (2004-2080)
2,500.0
2,000.0
1,500.0
1,000.0
500.0
Working Age Population
20
79
20
74
20
69
20
64
20
59
20
54
20
49
20
44
20
39
20
34
20
29
20
24
20
19
20
14
20
09
20
04
.0
Retirement Age Population
2.4. Social Insurance Fund Surpluses and Reserves
SSIF’s total expenditure depends on the number of pensioners in a given year and on the
way the pensions are indexed. Projections of the number of beneficiaries is discussed in
Section 2.1. The current Law on State Pensions defines that the new rate of main pension
benefit and supplemental pension may not be less than the previously defined rates but
offers no rules to guide pension indexation beyond stating that this will be defined by RA
legislation. For our calculations, we assumed that pensions would increase in step with
wages so that the replacement rate for all categories of pensions approaches 25.0 percent
in 2010 and 40.0 percent in 2020 and maintains that level afterwards. It should be noted
24
that the internationally acceptable level is at least 40 percent, with preferable level being
60-70 percent.
Most SSIF expenditures are for old age, privileged, disability, long service, and
survivors’ benefits and for some social safety net payments (temporary sickness,
maternity leave, etc.). Besides supporting pensioners and working people in various
circumstances, the SSIF incurs administrative costs, membership fees, mailing services
and other items. To calculate total spending on social insurance programs, we first
calculated total funds spent on social insurance and social security programs then
allocated all other expenses proportionately to total expenditures on social insurance
(sustained by social contributions) or social security programs (sustained by the state
budget).
The projections show that SSIF will experience a surplus until 2022 (see Table 2-9). The
current balance of the pension system under existing regulation will become negative in
2019. The reserves accumulated during the previous years will exhaust by 2022. The
reserve is the accumulated “current account balance,” assuming surpluses are deposited
in a bank account (dynamics of interest rates are discussed in Section I) each year, and
deficits are paid from balance of this bank account.
Table 2-9. Total Revenue, Current Balance and Fund Reserve as Percent of GDP
2004
2005
2006
2010
2020
2030
2040
2050
2060
2070
2080
Total Revenue
2.7
3.0
3.6
4.3
5.0
5.4
6.1
6.3
6.5
6.9
7.0
Revenue from
Employers and
Individuals
2.7
3.0
3.6
4.1
4.8
5.4
6.1
6.3
6.4
6.9
7.0
Total
Expenditures
2.6
3.0
3.3
3.8
5.3
7.5
8.3
10.4
12.1
11.6
12.3
Pension
Payments
2.5
2.9
3.2
3.7
5.1
7.3
8.1
10.2
11.9
11.4
12.0
Current Balance
.1
.0
.3
.4
(.3)
(2.1)
(2.2)
(4.1)
(5.6)
(4.7)
(5.2)
Fund Reserve
.1
.1
.4
1.6
2.5
Along with other data discussed in this section and Section I, the structure of SSIF’s
revenues and expenditures for 2002-2004 were used for our projections.26 Figure 2-4
presents the Contribution Revenues, Pension Payments, and Current Balance as percent
of GDP.
26
Socio-Economic Situation in Armenia, 2002, NSS, p. 128.
25
Figure 2-4. Contribution Revenues, Pension Payments, and Current Balance as % of GDP
10.0%
8.0%
6.0%
4.0%
2.0%
20
04
20
05
20
10
20
15
20
20
20
25
20
30
20
35
20
40
20
45
20
50
20
55
20
60
20
65
20
70
20
75
20
80
(2.0% )
(4.0% )
(6.0% )
(8.0% )
Total Revenue
Current Balance
Fund Reserve
Calculations show that, to keep the current balance at the “0” level, much higher contribution rates should have been imposed on wages than were assumed in the model. For
example, in 2040 instead of the assumed 20.0 percent of combined social contribution
rate, 27.4 percent should have been applied. In 2080, 34.8 percent of wages should be
collected to balance the expenses instead of the 20.0 percent currently in the model. Table
2-10 presents the required rates for social contributions, which are necessary for keeping
the SSIF current balance at zero level.
Table 2-10. Assumed and Required Rate of Social Contributions
2004
2005
2010
2020
2030
2040
2050
2060
2070
2080
Assumed
22.8
21.8
22.0
20.0
20.0
20.0
20.0
20.0
20.0
20.0
Required (for Zero Current
Balance)
22.2
21.5
19.7
21.3
27.9
27.4
33.1
37.5
33.8
34.8
If the contribution rates follow the pattern used in the model, the replacement ratios will
reach the level of only 21.1 percent in 2080 instead of assumed on-going average 37.3
percent (see Table 2-11).
Table 2-11. Replacement Rate for Pensioners
2004
2005
2010
2020
2030
2040
2050
2060
2070
2080
On-Going Average
20.7
20.4
20.8
26.2
32.0
34.4
36.1
36.8
36.9
37.3
Affordable (for Zero Current
Balance)
21.2
20.7
23.2
24.5
22.8
24.9
21.5
19.3
21.5
21.1
26
The first year with a negative current balance is projected to be 2019, which is 16 years
from the base year (2004). The first year with a negative fund reserve is projected to be
2023, 20 years from the base year (2004). The contribution rate required to balance the
fund over the simulation horizon is 31 percent. The average replacement rate required to
balance the fund over the simulation horizon is 21 percent. The present values of Total
Revenue, Total Expenditure, and Current Balance from 2004 to 2080 are presented in the
Table 4-12.
Table 4-12. Present Values of Total Revenue, Total Expenditure and Current Balance
PV of Total Revenue:
11,087,414.3
PV of Total Expenditure:
16,809,743.4
PV of Current Balance:
(5,722,329.1)
2.5. Replacement Rates and ROR on Mandatory Social
Contributions
Replacement rates (RR) are assumed to be good estimates of pension system efficiency
valuation. Pensions should be big enough (at least 40 percent of the individual wage) to
ensure economic independence and a dignified life (i.e., close to the standards prior to the
retirement) to the pensioner. In many developed countries, the RR is some 60-80 percent
of the individual average wage recorded during the work life of the retiree. In Armenia,
only prosecutors, judges, and military servants are entitled pensions comprising 70
percent of their pre-retirement wages.
The pensions in 2004 had no link to the wages paid and the contributions made, so that
people with the same years of service received the same pension regardless of the
contributions made. The RRs for different years of service for 2004 are presented in
Table 2-13a. The average wage was 41,976 AMD in 2004.
Table 2-13a. Replacement Rates for Different Years of Service (2004)
Years of Service
Base Benefit
Suppl. Benefit
Total Pension
RR (%)
0
5
10
15
20
25
30
35
40
4,000
4,000
4,000
4,000
4,000
4,000
4,000
4,000
4,000
160
640
1,440
2,560
4,000
5,280
6,720
8,320
4,000
4,160
4,640
5,440
6,560
8,000
9,280
10,720
12,320
9.5
9.9
11.1
13.0
15.6
19.1
22.1
25.5
29.4
Table 2-13b illustrates that the RR increases with the years of service. However, as long
as the pension benefit formula does not directly link the contributions and pensions, the
RR will not vary for different wage levels.
27
Table 2-13b illustrates that the RR drops with the increase of wage: the higher the wage
the lower the coefficient of replacement of individual wages.
Table 2-13b. Replacement Rates (Individual Wages) for Different Level of Wages (2004)
Wage
Percent of Average Wage
Replacement Ratio (%)
20,988
41,976
62,964
83,952
125,928
209,880
419,760
50
100
150
200
300
500
1000
44.2
22.1
14.7
11.1
7.4
4.9
3.7
The average wage in 2005 is projected to be 52,100AMD (see the justification in Section
I: Wages). The old base and supplemental pension amounts are in force also in 2005. The
replacement rates for different years of service for 2005 are as presented in Table 2-14a.
Decrease in replacement rates as compared with those in 2004 is apparent.
Table 2-14a. Replacement Rates under the Current Benefit Scheme for Different Years of
Service (projection for 2005)
Years of Service
Base Benefit
Suppl. Benefit
Total Pension
RR (%)
0
5
10
15
20
25
30
35
40
4,000
4,000
4,000
4,000
4,000
4,000
4,000
4,000
4,000
160
640
1,440
2,560
4,000
5,280
6,720
8,320
4,000
4,160
4,640
5,440
6,560
8,000
9,280
10,720
12,320
7.7
8.0
8.9
10.4
12.6
15.4
17.8
20.6
23.6
In Table 2-14b RR drops with increase of wage, as in previous years, and even decreases
as compared to the replacement rates recorded in 2004: 17.8 percent (in 2005) vs. 22.1
percent of the average wage in 2004.
Table 2-14b. Replacement Rates under the Current Benefit Scheme for Different Levels
of Wages (projection for 2005)
Wage
Percent of Average Wage
Replacement Ratio (%)
26,050
52,100
78,150
104,200
156,300
260,500
521,000
50
100
150
200
300
500
1000
35.6
17.8
11.9
8.9
5.9
4.0
3.0
28
The following tables (2-15a and 2-15b) show that the replacement rates drop for all levels
of years of service and different levels of wages in 2006, when the base pension is
proposed by the Government to be 4,200.0 AMD and the supplemental pension 180
AMD for each year of service. The average wage is projected to be 63,100 AMD in 2006.
The replacement rate despite the proposed increase of benefit drops further: for a person
with 30 years of service RR is 16.1 percent now against 22.1 percent in 2004 and 17.8
percent in 2005.
Table 2-15a. Replacement Rates under the Purposed Benefit Scheme for Different Years
of Service (projection for 2006)
Years of Service
Base Benefit
Suppl. Benefit
Total Pension
RR (%)
0
5
10
15
20
25
30
35
40
4,200
4,200
4,200
4,200
4,200
4,200
4,200
4,200
4,200
180
720
1,620
2,880
4,500
5,940
7,560
9,360
4,200
4,380
4,920
5,820
7,080
8,700
10,140
11,760
13,560
6.7
6.9
7.8
9.2
11.2
13.8
16.1
18.6
21.5
The situation is even worse than in 2005 with RR for different level of wages.
Table 2-15b. Replacement Rates (individual wages) under the Purposed Benefit Scheme
for Different Level of Wages (projection for 2006)
Wage
Percent of Average Wage
Replacement Ratio (%)
31,550
63,100
94,650
126,200
189,300
315,500
631,000
50
100
150
200
300
500
1000
32.1
16.1
10.7
8.0
5.4
3.6
2.7
Today, pension beneficiaries receive inadequate pensions relative to their contributions
made to the SSIF. To support very low pensions, employees together with their
employers pay close to 21.8 percent of their salaries to the SSIF (see Table 2-16).
Table 2-16. Social Contributions by Wage Level (2005)
Wage as % of
AW
Wage (AMD)
min
50%
100%
150%
200%
250%
300%
500%
1000%
13,000
26,050
52,100
78,150
104,200
130,250
156,300
260,500
521,000
Employee
contribution
390.0
781.5
1,563.0
2,344.5
3,126.0
3,907.5
4,689.0
7,815.0
15,630.0
Employer
contribution
5,000.0
5,907.5
9,815.0
13,722.5
17,210.0
18,512.5
19,815.0
25,025.0
38,050.0
Total
contributions
5,390.0
6,689.0
11,378.0
16,067.0
20,336.0
22,420.0
24,504.0
32,840.0
53,680.0
29
Percent of wage
41.5
25.7
21.8
20.6
19.5
17.2
15.7
12.6
10.3
Table 2-16 presents the rates of the mandatory social contributions for different wage
levels. It highlights that the total contribution as a percent of wage declines as the pay
increases. Table 2-14b and Table 2-16 highlight that a person earning 10 times more than
the average wage should pay to the system more than 10 percent of wage while might
expect to receive benefits replicating his/her wage only by 3.0 percent.
Regardless of the higher contributions made on behalf of highly paid workers, the latter
are eligible for benefits that are the same as those of lower paid workers with the same
years of service. The size of contributions paid over an employee’s working lifetime
plays no role in benefits received. A priori, this means that the higher paid receive a much
lower RR on their contributions than the lower paid. Table 2-17a demonstrates the share
of actual pensions as a percent of individual wage for different levels of earnings and
different duration of the service. Throughout the paper in the analogous tables the row
presenting RR for average wage and different years of service and the column presenting
RR for average years of service (30 years) and different wage levels are highlighted with
bold text.
Table 2-17a. Pensions as a Percent of Individual Wages (2004)
Years of service
5
10
15
20
25
30
35
40
RR for 50 of AW
19.8
22.1
25.9
31.3
38.1
44.2
51.1
58.7
RR for 100 of AW
9.9
11.1
13.0
15.6
19.1
22.1
25.5
29.4
RR for 150 of AW
6.6
7.4
8.6
10.4
12.7
14.7
17.0
19.6
RR for 200 of AW
5.0
5.5
6.5
7.8
9.5
11.1
12.8
14.7
RR for 250 of AW
4.0
4.4
5.2
6.3
7.6
8.8
10.2
11.7
RR for 300 of AW
3.3
3.7
4.3
5.2
6.4
7.4
8.5
9.8
RR for 500 of AW
2.0
2.2
2.6
3.1
3.8
4.4
5.1
5.9
17.9
20.0
23.4
28.3
34.5
40.0
46.2
53.1
Wage Level producing 40 % RR
for average years of service: 55.3
percent
Our calculations show that a person having 30 years of service (average years of service
for all beneficiaries) and earning 55.3 percent of average wage may be provided a 40
percent of individual wage replacement (see the last row presented in Table 2-17a).
Meanwhile it should be noted that in Conventional PAYG system (as it is now) pensions
replace the average wage on the same rates regardless of the wage level, that is for all
wage levels pensions replaces the average wage at rates presented for average wage (see
Table 2-17a and Table 2-17b rows and columns in bold)
Table 2-17b. Pensions as a Percent of Average Wage (2004)
Years of service
5
10
15
20
25
30
35
40
RR for 50% of AW
9.9
11.1
13.0
15.6
19.1
22.1
25.5
29.4
RR for 100% of AW
RR for 150% of AW
RR for 200% of AW
RR for 250% of AW
9.9
9.9
9.9
9.9
11.1
11.1
11.1
11.1
13.0
13.0
13.0
13.0
15.6
15.6
15.6
15.6
19.1
19.1
19.1
19.1
22.1
22.1
22.1
22.1
25.5
25.5
25.5
25.5
29.4
29.4
29.4
29.4
30
RR for 300% of AW
RR for 500% of AW
9.9
9.9
11.1
11.1
13.0
13.0
15.6
15.6
19.1
19.1
22.1
22.1
25.5
25.5
29.4
29.4
Table 2-17b demonstrates that in the current pension system the replacement rates are the
same for the same years of service and are not influenced by the wage level.
Rate of Return on Contributions: To calculate the rate of return on social contributions
we continue holding all the assumptions made in the previous sections. A financial
investment is attractive if it provides participants with a rate of return comparable with
other types of financial investment. The “rate of return” is a good indicator of the
“profitability” of a pension system to its participants. When the rate of return is in line
with other possible options for the investment, the employees will be willing to
participate in the system. Otherwise they will be unsatisfied and will try to evade paying
contributions.
It is assumed that the average salary increases at a rate envisaged in Section I, which
considers the existing pension system. The unisex life table at retirement (in 2037) year
is generated by PROST model on the basis of initial data for 2004 and prescribed trends
of age and sex specific mortality rates. As per constructed by PROST unisex life tables
for 2004-2080, unisex life expectancy at birth was 72.3 in 2004, and 74.4 in 2037, life
expectancy at retirement will be 17.1 in 2037.
Table 2-18a presents the rate of return (ROR) on mandatory contributions according to
the contribution schemes effective since January 1, 2001, made over 30 years, and
assuming that the pension benefit (calculated on the basis of the average pension in 2006)
is indexed (over the whole period covering contribution period and the period of
receiving benefit) 100 percent to inflation, 50 percent to inflation and 50 percent to
nominal wage growth, and 100 percent to nominal wage growth. All tables on ROR
reflect real value of the rate.
Table 2-18a. ROR on Mandatory Contributions without Minimum Guarantee Pension
(PAYG)
Indexed to inflation
Indexed 50 % Inflation and
50% to Nominal Wage Growth
Indexed to Nominal Wage
Growth
RR for 50% of AW
-4.7
-3.9
-3.2
RR for 100% of AW
-8.1
-7.2
-6.4
RR for 150% of AW
-10.0
-9.1
-8.3
RR for 200% of AW
-11.4
-10.5
-9.6
RR for 250% of AW
-12.5
-11.6
-10.7
RR for 300% of AW
-13.4
-12.4
-11.5
RR for 500% of AW
-15.8
-14.8
-13.9
This shows that the actual ROR is negative for all categories of participants. Given this
situation, low compliance in contributions is not a surprise.
Minimum Guarantee Pension equal to 25 percent of previous year average wage will
improve the situation, however, again leaving those earning above 150 percent average
wage with negative ROR (see Table 2-18b).
31
Table 2-18b. ROR on Mandatory Contributions with Minimum Guarantee Pension
(PAYG)
Indexed to inflation
Indexed 50 % Inflation and 50%
to Nominal Wage Growth
Indexed to Nominal Wage
Growth
RR for 50% of AW
5.0
5.1
5.1
RR for 100% of AW
2.0
2.1
2.2
RR for 150% of AW
0.2
0.3
0.4
RR for 200% of AW
-1.1
-1.0
-0.9
RR for 250% of AW
-2.1
-2.0
-1.9
RR for 300% of AW
-2.9
-2.8
-2.7
RR for 500% of AW
-5.3
-5.2
-5.1
Indexing by 50 % inflation rate combined with 50 % by wage growth rate produces
intermediate results as compared to indexing 100 percent by inflation and 100 percent by
wage growth rate.
Given this situation, low compliance in contributions is not a surprise. In any pension
system the relationship between contributions paid and benefits received is crucial for the
system’s financial sustainability and social fairness. And if high contribution rates and
low benefits provide room for financial stability of the system, these do little for social
justice. Comparable to other financial instruments return to the social contributions will
augment the fairness of the system
Thus, it is essential to create a pension insurance system that promotes voluntary
participation of employees and provides reasonable value for their contributions. Future
reforms should ensure increase of pensions so that to provide for a positive rate of return
on medium-income as well. Improved living conditions may play a critical role in
improving the ROR by means of increasing the life expectancy of participants upon
retirement and, thus, prolonging their period of receiving benefits.
32
SECTION III. NOTIONAL ACCOUNT SYSTEM ANALYSIS
Many countries now consider reforming their PAYG system to create a direct link
between contributions and benefits, yet preserve elements of inter-and intra-generational
redistribution (aiming at protecting the poor and other vulnerable groups). The notional
accounts are not funded accounts per se, though they represent individualized claims on
future public pension resources. The notional accounts keep track of all individual
contributions (combined contribution by employer and employee), notional interest
“earned” on the “balance”, and at the retirement convert the notional balance (taking into
account the current and prospective demographic and productivity changes) into a life
pension-annuity, thus creating an explicit link between the “cumulated” account and
establishing a kind of defined-contribution pension system while maintaining
redistributive pension system as the current employed pay for current pensioners.
The introduction of notional accounts into the current PAYG system will require some
adjustments to the current system. Hypothetically, the reform starts in 2006 and the first
group of pensioners will be entitled to notional pensions as soon as they retire (although
it is recommended establishing at least 5 years of insured service for accruing pension
rights). For the purposes of the projections, the following assumptions were made:
•
Notional accounts are applied to all those covered, i.e. all shift to the new system,
Retirement age is the same as in the conventional PAYG system,
Demographic and Macroeconomic assumptions under analyzed above conventional
PAYG system are in force,
We assume that all the parameters involved in this projection change over time
according to the macroeconomic assumptions presented in Section I. According to
our projections the average wage will be 63,100AMD in 2006.
For the purposes of convenience we present the indicators for some selective years:
•
•
•
•
Macroeconomic indicators used in the exercise (in percent)
2006
2007
2008
2009
2010
2020
2030
2040
2050
2060
2070
2080
Real GDP Growth
Inflation Rate
6.0
3.0
5.8
3.0
5.7
3.0
5.5
3.0
5.3
3.0
4.0
3.0
2.9
3.0
2.7
3.0
2.5
3.0
2.4
3.0
2.2
3.0
2.0
3.0
Real Interest Rate
Real Wage Growth
Rate
Contribution rate
Total Contribution
Rate Credited to
Switcher's Account
5.5
5.3
5.2
5.0
4.8
3.0
2.4
2.2
2.0
1.9
1.7
1.5
17.5
14.5
11.5
8.5
5.5
4.1
3.0
2.8
2.6
2.4
2.2
2.0
25.1
24.1
23.3
22.7
20.0
20.0
20.0
20.0
20.0
20.0
20.0
20.0
6.3
6.0
5.8
5.7
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
•
•
•
In the accumulation phase the contributions and balances increase at a nominal
interest rate minus 1 percent,
In the decumulation phase annuitization rate is lower than nominal interest rate for
1.5 percent
Pensions are indexed to inflation. Although in a rapidly growing economy linking it
to the wage growth rate is more appropriate, we make this assumption in order to be
33
•
•
consistent with the pessimistic approach in modeling the pension system for the longterm used for analyzing and simulations of the current pension system (see Section I
and Section II, and Appendix ).
In 2006 the base pension is 4,200AMD and supplemental pension is 180AMD for
each year of service, as suggested by the government of Armenia.
Reforms start in 2006 and at least five years of service under the new plan is required
for paying the notional part of the pension. Five percent of the total Employee and
Employer Contribution (20 percent) serves the notional accounts, which is only for
accruing notional account rights and at retirement converting the “cumulated” amount
into the notional part of the pension (life pension annuity under pure endowment
scheme). In reality, the whole 20 percent contribution will serve for paying the
current pensioners’ pensions. Each year, five percent of this amount is used for asset
management and administrative expenses of the SSIF.
Replacement rates under the notional account system for different years of service and
benefit schemes: For calculating the plausible replacement rates under the notional
account scheme, we assume the above noted macroeconomic indicators hold over the
whole “accumulation” period notional amounts. Main indicators for the retirement year
(2037) are as follows:
Main Indicators on the Retirement year:
Inflation (in percent)
Real GDP Growth Rate (in percent)
Real Interest Rate (in percent)
Nominal wage growth rate (in percent)
Nominal Rate of Interest (in percent)
Interest Rate for Pension Life Table (in percent)
Contribution Rate recorded in the individual notional account (in
percent)
Average wage in 2006 (AMD)
Base Benefit (AMD)
Supplemental Benefit Per Year of Service (AMD)
Hire Age (year)
Retirement Age (year)
3.0
2.8
2.3
6.7
5.4
4.4
5.0
63,100.0
4,200
180
33
63
Rules for Annuitization:
Decrease in Interest at the Decumulation Phase of Annuitization (in
percent )
1.5
Type of Mortality Tables Used for Annuitization
Unisex
We suppose that the average wage will be 63,100 AMD in 2006. Actuarial life table for
6.7 percent rate of return was generated (for the plausible retirement year 2037) on the
basis of survival rates as per life expectancy table produced by PROST model under the
same assumptions on mortality that was used in the conventional PAYG system.
34
As it is seen in the Table 3-1, a person having 30 years of service will be entitled to a
pension at 27.0 percent of average wage, while the old conventional system could at best
guarantee 16.1 percent replacement rate. In Notional PAYG system 40 percent level of
the RR is achievable in 44 years.
Table 3-1. Replacement Ratios for Notional Account System (projection for 2006)
Years
of
Service
Base
Benefit
Supplement
al Benefit
Total
PAYG
Pension
RR of
PAYG
(%)
Total
Notional
Pension
RR of
Notional
PAYG (%)
0
4,200
-
4,200
6.7
4,200.0
6.7
5
4,200
180
4,380
6.9
1,235.2
5,615.2
8.9
10
4,200
720
4,920
7.8
2,294.1
7,214.1
11.4
15
4,200
1,620
20
4,200
2,880
5,820
9.2
3,297.8
9,117.8
14.4
7,080
11.2
4,309.3
11,389.3
18.0
25
4,200
4,500
8,700
13.8
5,293.9
13,993.9
22.2
30
4,200
5,940
10,140
16.1
6,244.2
16,384.2
26.0
35
4,200
7,560
11,760
18.6
7,161.6
18,921.6
30.0
40
4,200
9,360
13,560
21.5
8,048.4
21,608.4
34.2
45
4,200
11,340
15,540
24.6
8,906.3
24,446.3
38.7
50
4,200
13,500
17,700
28.1
9,737.3
27,437.3
43.5
Notional
Pension
With introduction of the minimum guarantee pension (25 percent of previous year
average wage), the situation improves substantially for those with less years of
service (see Table 3-2). However, the 40 percent limit is not attained even for those
with 50 years of service. In this paper the notional pensions are presented at their
present values in order to be comparable with the current wages and pensions.
Table 3-2. Replacement Ratios for Notional Account System (Average Wage) with
Minimum Guarantee Pension (projection for 2006)
Years of Service
Minimum Guarantee
Pension
0
13,025
5
13,025
10
Notional
Pension
Total Pension
Total Replacement
Rate (%)
13,025.0
20.6
1,235.2
14,260.2
22.6
13,025
2,294.1
15,319.1
24.3
15
13,025
3,297.8
16,322.8
25.9
20
13,025
4,309.3
17,334.3
27.5
25
13,025
5,293.9
18,318.9
29.0
30
13,025
6,244.2
19,269.2
30.5
35
13,025
7,161.6
20,186.6
32.0
40
13,025
8,048.4
21,073.4
33.4
45
13,025
8,906.3
21,931.3
34.8
50
13,025
9,737.3
22,762.3
36.1
35
For providing a person with 30 years of service a pension amounting to 40 percent of the
replacement rate, either an increase in the base benefit, supplemental benefit, or
contribution rate to the notional accounts is necessary. Calculations show that, for
example, an approximately 8.5 percent contribution to the notional accounts is required to
ensure 40.0 percent of replacement rate to a person with 30 years of service.
Meanwhile, leaving the contribution rate to the notional accounts at the level of five
percent but increasing the base pension to 5000 AMD and supplemental pension to 500
AMD gives us the promised 40.0 percent replacement rate, even with 28 years of insured
service (see the Table 3-3). In the notional accounts system, a person having 30 years of
service will be entitled to a pension providing a 44.0 percent replacement rate of average
wage versus 34.1 percent under the conventional PAYG scheme.
Table 3-3. Replacement Ratios for Notional Account System
(high base and supplemental pensions)
Years
of
Service
Base
Benefit
Supple
mental
Benefit
0
5,000
-
Notional
Account
Benefit
Total
Benefit
5,000
Replacement Ratio (in percent)
Suppleme
Base
Notional
Total
nt
7.9
0.0
7.9
5
5,000
500
1,235
6,735
7.9
0.8
2.0
10.7
10
5,000
2,000
2,294
9,294
7.9
3.2
3.6
14.7
15
5,000
4,500
3,298
12,798
7.9
7.1
5.2
20.3
20
5,000
8,000
4,309
17,309
7.9
12.7
6.8
27.4
25
5,000
12,500
5,294
22,794
7.9
19.8
8.4
36.1
30
5,000
16,500
6,244
27,744
7.9
26.1
9.9
44.0
35
5,000
21,000
7,162
33,162
7.9
33.3
11.3
52.6
40
5,000
26,000
8,048
39,048
7.9
41.2
12.8
61.9
45
5,000
31,500
8,906
45,406
7.9
49.9
14.1
72.0
50
5,000
37,500
9,737
52,237
7.9
59.4
15.4
82.8
Replacement ratios for different level of wage for average years of service: In the
current pension system the replacement ratios are very similar regardless wage level.
With introduction of notional accounts a worker earning 250 percent of the average wage
might expect to get 34.6 percent replacement of average wage compared to 22.0 percent
for a worker earning average wage. This option is very different than the relationships
under the current plan where workers earning 250 percent of the average wage have a
replacement ratio of less than 13.6 percent. As wages increase, so do the portions of the
total benefit coming from notional accounts (see Table 3-4).
36
Table 3-4. Replacement Ratios as Compared to Average Wage for Notional Account
System at Different Levels of Wages and 30 Years of Service
Replacement Ratios (in percent)
Base and
Supplem
Supplem Notional
ental
ental
Pension
Pension
Pension
8.0
13.6
4.2
Percent
of AW
Base
Pension
Supple
mental
Pension
50
4200
5940
3,122.1
13,262
5.6
100
4200
5940
6,244.2
16,384
5.6
8.0
13.6
8.4
22.0
150
4200
5940
9,366.3
19,506
5.6
8.0
13.6
12.6
26.2
200
4200
5940
12,488.4
22,628
5.6
8.0
13.6
16.8
30.4
250
4200
5940
15,610.4
25,750
5.6
8.0
13.6
21.0
34.6
300
4200
5,940
18,732.5
28,874
5.6
8.0
13.6
25.2
38.8
500
4200
5,940
31,220.9
41,363
5.6
8.0
13.6
42.0
55.6
Notional
Pension
Total
Pension
Base
Pension
Total
Pension
17.8
Comparison of the benefits with the wages produces above 40 percent replacement of
earned wage for those earning 50 percent of average wage (see Table 3-5). The portion of
notional benefit is constant (9.9%) for all wage levels. However, as wages increase, the
base plus supplementary pension portion drops, and in combination with notional benefits
produce diminishing rates of replacement.
Table 3-5. Replacement Ratios for Notional Account System at Different Levels of
Wages and 30 Years of Service
Replacement Ratios (in percent)
Base and
Supplem
Supplem Notional
ental
ental
Pension
Pension
Pension
18.8
32.1
9.9
Percent
of AW
Base
Pension
Supple
mental
Pension
50
4200
5940
3,122.1
13,262
13.3
100
4200
5940
6,244.2
16,384
6.7
9.4
16.1
9.9
26.0
150
4200
5940
9,366.3
19,506
4.4
6.3
10.7
9.9
20.6
200
4200
5940
12,488.4
22,628
3.3
4.7
8.0
9.9
17.9
250
4200
5940
15,610.4
25,750
2.7
3.8
6.4
9.9
16.3
300
4200
5940
18,732.5
28,873
2.2
3.1
5.4
9.9
15.3
500
4200
5940
31,220.9
41,361
1.3
1.9
3.2
9.9
13.1
Notional
Pension
Total
Pension
Base
Pension
Table 3-6 presents the RR for different years of service and wage levels on notional
accounts plus guarantee pension (25 percent of AW). In this and analogous tables the
pensions are compared with the average wages of respective years in 5, 10, 15, etc. years
after the reform year 2006. Note the low replacement rate for 50 percent AW and five
years of service. It is important to keep in mind that the guarantee pension is calculated
on the basis of the previous year’s average wage and that as a result the RR might be less
than 25 percent.
37
Total
Pension
42.0
Table 3-6. RR for Different Wage Levels and Years of Service on Notional Accounts
Years of service
5
10
15
20
25
30
35
40
RR for 50% of AW
24.0
25.0
26.0
27.0
27.8
28.5
29.3
30.0
RR for 100% of AW
27.0
26.8
28.6
30.4
32.0
33.5
35.0
36.4
RR for 150% of AW
26.0
28.6
31.2
33.8
36.2
38.4
40.6
42.8
RR for 200% of AW
26.9
30.4
33.8
37.2
40.4
43.4
46.3
49.2
RR for 250% of AW
27.9
32.2
36.4
40.6
44.6
48.3
52.0
55.5
RR for 300% of AW
28.9
34.0
39.0
44.0
48.7
53.3
57.7
61.9
RR for 500% of AW
32.8
41.3
49.5
57.7
65.5
73.1
80.4
87.4
Rate of Return on Social Contributions under different plans: According to our
assumptions 1/4 of contributions serve as a basis for accruing notional pension rights and
“converting” those into the notional part of the pension. In reality, the whole 20 percent
contribution will serve for paying the current pensioners’ pensions. Table 3-7 presents the
Rate of Return on Social Contributions under different plans and different wage levels.
Cash flow constructed from the (annual) contributions and (annual) pensions yield 5.1
percent ROR for 50 percent of average wage level if we consider notional plan plus
minimum guarantee pension. The execution of mixed system (conventional and notional),
produces a much higher ROR as compared to Conventional PAYG system.
Table 3-7. Rate of Return on Notional Accounts with Minimum Guarantee Pension (%)
Notional Pension Indexed
100% to Inflation
Notional Pension Indexed 50% to
inflation and 50% to wage growth
Notional Pension Indexed 100%
to Nominal Wage Growth
RR for 50% of AW
5.1
5.2
5.3
RR for 100% of AW
2.7
2.9
3.0
RR for 150% of AW
1.4
1.6
1.9
RR for 200% of AW
0.6
0.9
1.1
RR for 250% of AW
0.0
0.3
0.7
RR for 300% of AW
-0.4
-0.1
0.3
RR for 500% of AW
-1.5
-1.0
-0.6
Table 3-7 presents ROR on mandatory social contributions calculated with the rate as in
conventional PAYG system, and under the same indexing schemes that were considered
for Conventional PAYG. Note, however that for high wages (i.e. 300% of average wage)
the real ROR is negative for the options of indexing to inflation and 50 percent to
inflation and 50 percent to nominal wage and gets positive for indexation by nominal
wage growth rate.
SSIF Revenue, Current Balance and Fund Reserve as Percent of GDP: It is well
observable that in the long-term the financial prospectus of the Social Insurance Fund is
worse off than the conventional PAYG system.
38
Table 3-8. Total Revenue, Total Expenditure, Current Balance and Fund Reserve (as %
of GDP)
2006
2007
2008
2010
2020
2030
2040
2050
2060
2070
2080
Total Revenue
3.6%
3.8%
4.0%
4.0%
5.7%
6.2%
6.1%
6.3%
6.4%
6.8%
7.0%
Total Expenditure
3.5%
3.2%
3.0%
2.6%
3.1%
6.9%
10.7%
19.8%
27.3%
26.1%
29.2%
Current Balance
.1%
.6%
1.0%
1.4%
2.5%
(.7%)
(4.7%)
(13.5%)
(20.9%)
(19.3%)
(22.2%)
Fund Reserve
.2%
.8%
1.7%
4.0%
20.5%
16.8%
Table 3-8 shows that the financial stability of the pension system runs into solid deficit
after 2027.
39
SECTION IV. FUNDED ACCOUNT SYSTEM ANALYSIS
In this section we analyze a pension system that introduces mandatory private fullyfunded accounts (funded pillar) and changes the current state pensions into a minimum
guarantee pension system for everyone. It does not discuss all the design issues of the
new system and mainly concentrates on the long-term projections of a specific reform
scenario, which is presented below.
The main assumptions of the proposed Funded Account System scenario that are needed
for modeling it by PROST are the following:
•
According to our hypothesis, the reform starts again in 2006, and the first group of
pensioners will be entitled to get benefits from the funded pillar started in 2020.
People under the age of 30 will be moving to a fully funded system that only
provides pensions on the basis of defined contributions and the accumulated
contributions are privately managed. Those between the ages of 30 and 49 will have
a one-time choice at the beginning of the reform either to move to the new system
or to stay in the old system. This latter group that remained in the old system would
eventually move to the new system in 10 years. Without ruling out other options,
we assume that 50 percent of the 30-49 age group will stay in the old system and 50
percent will switch to the funded system. The table below represents the anticipated
pattern of different age groups switching to the new system.
Table 4-1. Designed Switching Pattern to the Funded Pillar
Contributor Groups:
From Age
To Age
New Entrants
14
15
24
25
29
30
44
45
49
50
62
63
74
75
105
Recognized Accrued Rights
Switchers
Non-switchers
0.0%
100.0%
40.0%
100.0%
60.0%
100.0%
80.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
Switching Pattern
Switchers
Non-switchers
100.0%
0.0%
100.0%
0.0%
100.0%
0.0%
100.0%
0.0%
50.0%
50.0%
50.0%
50.0%
100.0%
100.0%
100.0%
•
Everyone above 50 years old at the outset of the reform will remain in the old
system. However, the Notional PAYG system is introduced for these members of
the society,
•
Retirement age is the same as in the conventional and Notional PAYG system,
•
Demographic and Macroeconomic assumptions are also the same as in the
conventional and Notional PAYG systems,
Combined employer and employee contribution rate is 20% applied to all levels of
wages (particularly, no wage cap is considered).
In the accumulation phase the contributions and balances increase at a nominal
interest rate presented in the Macroeconomic Assumptions part in Section I.
•
•
40
•
In the decumulation phase annuitization (that is discount) rate is lower the nominal
interest rate in 1.5 percent.
•
Pensions are indexed to inflation.
•
In 2006 the base pension is 4,200AMD and supplemental pension is 180AMD for
each year of service, as suggested by the government of Armenia.
•
Combined Employee and Employer Contribution (20%) cumulate in the individual
mandatory accounts. At retirement the balance will be converted into the funded
benefit (life pension annuity under pure endowment scheme) through dividing the
balance of the account on actuarial factor (see the calculation in the Section I).
Survival rates for the year of retirement of the hypothetical employee are generated
via PROST model holding all the assumptions on mortality rates as in the previous
scenarios.
•
The issue of the contribution share of employers and employees after 2010 is not
discussed in this scenario.
Rules for accumulation phase: We assume that a person earning the average wage starts
working in 2006, works for 30 years, and retires at 63 in 2036. If the wage growth-rates
follow the trends presumed in Section I, then, in 2006, the average wage will be 63,100
AMD. After 2006, the proposed new scheme of contributions is used for calculating the
social contribution (employer and employee combined), which will be 25.1 percent due
to the proposed increase of employer contribution. We assume also that the 20 percent
contribution rate would be applied on the wages covered by the funded-account system
from 2010 on. In general, we assume that all the parameters involved in this exercise
change over time according to the macroeconomic assumptions presented in Section I;
however, for the purposes of convenience, we present the indicators for some selective
years also here:
Macroeconomic indicators used in the exercise (in percent)
Real GDP
Growth
Inflation
Rate
Real Interest
Rate
Real Wage
Growth Rate
Contribution
rate
2004
2005
2006
2007
2008
2009
2010
2020
2030
2040
2050
2060
2070
2080
10.1
8.8
6.0
5.8
5.7
5.5
5.3
4.0
2.9
2.7
2.5
2.4
2.2
2.0
7.0
3.0
3.0
3.0
3.0
3.0
3.0
3.0
3.0
3.0
3.0
3.0
3.0
3.0
6.0
5.8
5.5
5.3
5.2
5.0
4.8
3.0
2.4
2.2
2.0
1.9
1.7
1.5
20.6
20.5
17.5
14.5
11.5
8.5
5.5
4.1
3.0
2.8
2.6
2.4
2.2
2.0
22.8
21.8
25.1
24.1
23.3
22.7
20.0
20.0
20.0
20.0
20.0
20.0
20.0
20.0
Particularly, in the assumed “retirement” year (2037), the macroeconomic indicators are
supposed to be as follows:
Real GDP Growth (in percent)
Inflation Rate (in percent)
Real Interest Rate (in percent)
Real Wage Growth Rate (in percent)
Contribution rate (in percent)
2.8
3.0
2.3
2.84
20.0
41
Rules for Annuitization: For calculating the duration of receiving pensions, a unisex life
expectancy table was generated through the PROST Model. The survival rates of that
table, along with above presented macroeconomic assumptions, were used for calculating
the actuarial factor for the retirement year. It is assumed that the pensions will be indexed
to inflation (although other scenarios may be considered as well).
We assume that “Asset Management and Administration Costs” will be five percent of
contributions and interest earned on them. According to the Law on Mandatory Social
Insurance from Temporary Disability, which has passed its second reading, several
programs are to be covered by the social contributions. We also assume that, in total, 95
percent of the end-balance will be used for conversion into life pension annuity. This
exercise assumed also that the person worked without brakes. The Actuarial Factor turns
out to be 11.82 under accepted assumptions.
Rules for Annuitization:
Decrease in Interest at the Decumulation Phase of Annuitization (in percent )
Type of Mortality Tables Used for Annuitization
1.5
unisex
Hypothetical Retirement History: Suppose that a person who switched to the Funded
Accumulation system may expect to be entitled funded benefits after five years of
service, although according to the switching patters described in Table 4-1 the first cohort
of switchers will retire not earlier 2020.
Replacement Rates: Tables 4-2a and 4-2b, which present replacement rates for 50, 100,
150, 200, 250, 300, and 500 percent of average wage levels for different levels of years of
service without and with guarantee pensions, nicely demonstrates that the replacement
rates (compared to the average wage of respective year in 5, 10, 15, etc. years after the
reform year 2006) escalate with the insured years of service and level of wages.
Comparison with replacement rates of pensions under notional accounts points on
substantial improvement: higher wages combined with higher years of service provide
much higher replacement rates.
Table 4-2a. Replacement Rates for Different Wage Levels and Years of Service without
Guarantee Pensions (in percent)
Years of Service
RR for 50% of AW
RR for 100% of AW
RR for 150% of AW
RR for 200% of AW
RR for 250% of AW
RR for 300% of AW
RR for 500% of AW
5
3.9
7.8
11.7
15.7
19.6
23.5
39.2
10
7.3
14.5
21.8
29.1
36.4
43.6
72.7
15
10.5
20.9
31.4
41.8
52.3
62.7
104.5
20
13.7
27.3
41.0
54.6
68.3
82.0
136.6
25
16.8
33.6
50.3
67.1
83.9
100.7
167.8
30
19.8
39.6
59.4
79.2
99.0
118.7
197.9
35
22.7
45.4
68.1
90.8
113.5
136.2
227.0
40
25.5
51.0
76.5
102.0
127.6
153.1
255.1
In Table 4-2b, note that 5 years of service now promise above 25 percent replacement as
compared to 24.0 percent in notional account system (see Table 3-6).
42
Table 4-2b. Replacement Rates for different Wage Levels and Years of Service with
Minimum Guarantee Pension (in percent)
Years of service
5
10
15
20
25
30
35
40
RR for 50% of AW
26.9
30.4
33.8
37.2
40.4
43.4
46.3
49.2
RR for 100% of AW
32.8
37.7
44.3
50.9
57.1
63.2
69.0
74.7
RR for 150% of AW
34.8
45.0
54.7
64.5
73.9
83.0
91.7
100.2
RR for 200% of AW
38.7
52.2
65.2
78.2
90.7
102.8
114.4
125.7
RR for 250% of AW
42.6
59.5
75.6
91.9
107.5
122.6
137.1
151.2
RR for 300% of AW
46.5
66.8
86.1
105.5
124.3
142.3
159.8
176.7
RR for 500% of AW
62.2
95.8
127.9
160.1
191.4
221.5
250.6
278.7
Note that the replacement rates presented in Table 4-2a and Table 4-2b would have been
different also depending on the period of pension rights accrual. If a person cumulates the
pension rights (gets insured years of service) just prior to the retirement then the benefit
will replace his/her last wage at the close to those presented in the table rates. However, if
the person worked long ago his/her retirement then the benefit would be higher reflecting
the interest earned on the balance of account during the period until the retirement.
Rate of Return on Social Contributions on funded accounts: Table 4-3a presents the
results of Rate of Return for fully-funded contributions without adding guarantee
pensions. The table demonstrates that the Rate of Return exercise on pure funded
accounts produces the same results for all levels of wage. However, depending on the
option of the future pension indexation we get different results: indexation to nominal
wage growth being the most favourable for pensioners.
Table 4-3a. ROR on Funded Contributions for Different Levels of Wages without
Minimum Guarantee Pension (in percent)
50% of average wage
100% of average wage
150% of average wage
200% of average wage
250% of average wage
300% of average wage
500% of average wage
Indexed to inflation
Indexed 50 % Inflation and 50%
to Nominal Wage Growth
Indexed to Nominal Wage
Growth
2.6
2.6
2.6
2.6
2.6
2.6
2.6
3.2
3.2
3.2
3.2
3.2
3.2
3.2
3.8
3.8
3.8
3.8
3.8
3.8
3.8
It is obvious that adding the minimum guarantee pension to funded pension will produce
a much higher rate of return, which is actually illustrated below in Table 4-3b. Although,
in the case of funded accounts plus guarantee pension ROR decreases with the wage
level, it remains much higher than the ROR on pure funded accounts, which was constant
for all levels of wages as we saw above.
43
Table 4-3b. ROR on funded contributions for different levels of wages (with Minimum
Guarantee Pension)
Indexed to inflation
Indexed 50 % Inflation and 50%
to Nominal Wage Growth
Indexed to Nominal Wage
Growth
6.6
5.1
4.4
4.1
6.8
5.4
4.8
4.5
7.0
5.7
5.2
4.9
3.8
3.6
3.3
4.3
4.1
3.8
4.7
4.6
4.3
50% of average wage
100% of average wage
150% of average wage
200% of average wage
250% of average wage
300% of average wage
500% of average wage
Comparison with rates of return in other scenarios points that ROR on funded accounts is
much higher than in other systems that were considered in this work.
Funded Pillar Revenue, Current Balance and Fund Reserve as Percent of GDP: The
PAYG system with introductions of Fully Funded system will face shortage of finances,
which is seen in Table 4-4a. This is mainly because of sharp reductions in the number of
effective contributors to the PAYG system, as all under the age of 30 and half of those
between 30-49 years old join the new system, which causes the number of effective
contributors to the PAYG system to shrink. In 2020, 50 years old people will reach
retirement age and the first cohort of pensioners will retire with fully funded benefits.
The average wage earner may expect almost 45 percent replacement of his/her earning
(see Table 4-2). 30 yeas old will attain their retirement age in 2040, which is in effect the
last “possible” year for Conventional PAYG system.
Table 4-4a. All Pillars: Financial Flows (transition age group 30-49)
As % of GDP
2006
2007
2008
2010
2020
2030
2039
2040
2050
2060
2070
2080
Total Revenue: PAYG
Total Expenditure:
PAYG
Current Balance: PAYG
1.9
1.9
1.9
1.6
1.1
.4
.1
3.7
3.5
3.3
3.0
3.7
4.9
5.2
(1.7)
(1.5)
(1.4)
(1.4)
(2.7)
(4.4)
(5.1)
1.8
2.1
2.4
2.8
5.6
8.1
10.2
10.4
11.4
11.3
11.7
11.8
.3
.3
.3
.4
1.0
2.3
3.3
3.5
5.7
6.4
6.1
6.7
1.5
1.8
2.0
2.4
4.5
5.8
6.9
6.9
5.6
4.9
5.6
5.2
1.5
3.2
5.0
8.7
31.6
57.9
81.9
84.3
97.9
96.8
100.5
103.9
Fund Reserve: PAYG
Total Revenue: Funded
Total Expenditure:
Funded
Current Balance:
Funded
Fund Reserve: Funded
The next scenario assumes that people under the age of 30 will join the new system, those
over 40 will remain in the old system, and 50 percent of those in the age group 30-39 will
stay in the old system while the other 50 percent join the funded system. The results of
this scenario are presented in Table 4-4b.
44
Table 4-4b. All Pillars: Financial Flows (transition age group 30-39)
As % of GDP
2006
2007
2008
2010
2020
2030
2039
Total Revenue: PAYG
Total Expenditure:
PAYG
Current Balance: PAYG
2.6
2.6
2.5
2.2
1.5
.5
.1
3.7
3.5
3.4
3.1
4.1
5.6
5.6
(1.1)
(.9)
(.8)
(.8)
(2.7)
(5.1)
(5.5)
1.1
1.4
1.6
2.0
4.8
7.8
10.2
.2
.2
.2
.3
.8
1.6
1.0
1.2
1.4
1.7
4.0
1.0
2.1
3.3
6.0
25.6
2040
2050
2060
2070
2080
10.4
11.4
11.3
11.7
11.9
3.2
3.4
5.7
6.4
6.1
6.7
6.2
7.1
7.1
5.7
4.9
5.6
5.2
55.7
82.4
85.0
98.8
97.7
101.2
104.5
Fund Reserve: PAYG
Total Revenue: Funded
Total Expenditure:
Funded
Current Balance:
Funded
Fund Reserve: Funded
The presented next scenario of transition to the funded system is illustrated in Table 4-4c.
It assumes that people under the age of 30 will join the new system and those who are
older will remain in the old system.
Table 4-4c. All Pillars: Financial Flows (no intermediary transition age)
As % of GDP
Total Revenue: PAYG
Total Expenditure:
PAYG
Current Balance: PAYG
2040
2050
2060
2070
2080
9.9
10.2
11.4
11.4
11.8
11.9
1.3
2.3
2.6
5.6
6.4
6.1
6.7
3.3
5.8
7.6
7.6
5.8
5.0
5.7
5.2
19.9
47.9
78.9
82.2
99.7
98.9
102.3
105.5
2006
3.0
2007
3.0
2008
3.0
2010
2.7
2020
2.0
2030
.9
2039
.2
3.7
3.5
3.4
3.1
4.1
5.8
5.3
(.7)
(.5)
(.4)
(.4)
(2.2)
(4.9)
(5.1)
.7
.9
1.1
1.4
3.9
7.0
.1
.1
.1
.2
.6
.6
.7
.9
1.2
.6
1.3
2.1
3.9
Fund Reserve: PAYG
Total Revenue: Funded
Total Expenditure:
Funded
Current Balance:
Funded
Fund Reserve: Funded
For a successful transition to the new system, some funds from Family Poverty Benefit
system might be transferred to balance the deficit in the pension system. Moreover, those
remaining in the old system would be entitled to only minimum guarantee pension equal
25 percent of the average wage of the previous year.
In all cases, the drawback of the transition is its cost: some generations will have to
contribute (at least for a while) both for their own retirement and for the current unfunded
liabilities of the PAYG system.
45
SECTION V. COMPARISON OF PENSION REFORM OPTIONS
So far, we have concentrated on the current situation of Conventional PAYG, discussed
some aspects of Notional PAYG and Private Fully-Funded Pillar, as well as considered
several options of switching patters for transition age groups in Funded Pillar. Yet many
questions urge further clarification and analysis.
Replacement Rates in PAYG, Notional PAYG, and Private Fully-Funded Pillar:
Replacement rates measure the extent to which the pension system enables the
participants to maintain their previous personal standard of living when moving from
employment to retirement. In this respect, competition of replacement rates under
different portfolios will give better understanding of “Which of the scenarios makes
employees and pensioners better off?”
Table 2-17b visibly demonstrates the redistributive character of PAYG system. It
provides the same replacement rate for all wage levels given the same years of service.
The notional pension cumulative component was calculated on the assumption that 1/4
out of social contributions (see Macroeconomic indicators used in the notional account
calculations) “cumulate” in the notional accounts on the basis of which cumulative
component of the notional benefit will be allocated to a retiree. Notional accounts, while
providing a higher RR than conventional PAYG explicitly links the benefits to
contributions, yet they are not funded accounts. Comparison of notional RR with funded
benefits reveals that the “funded benefit produces the highest replacement rate. As shows
Table 5-1a, the discrepancy magnifies with years of service: RR may reach a 75 percent
replacement rate for those with 40 years of service in the fully-funded system.
Table 5-1a. RR for Average Wage in Different Systems for Different Years of Service (in
percent)
5
10
15
20
25
30
35
40
Conventional PAYG
system
9.9
11.1
13.0
15.6
19.1
22.1
25.5
29.4
Notional PAYG system
27.0
26.8
28.6
30.4
32.0
33.5
35.0
36.4
Fully Funded system w/o
Guaranteed Pensions
7.8
14.5
20.9
27.3
33.6
39.6
45.4
51.0
Fully Funded system with
Guarantee Pensions
32.8
37.7
44.3
50.9
57.1
63.2
69.0
74.7
Note that Notional Account option incorporates minimum guarantee pension.
Table 5-2b. RR for Average Years of Service (30 years) and Different Levels of Wage (in
percent)
Average Wage
Level
Conventional
PAYG
Notional
PAYG
Funded Pillar without
Minimum Guarantee Pension
Funded Pillar with Minimum
Guarantee Pension
RR for 50 of AW
22.1
28.5
19.8
43.4
RR for 100 of AW
22.1
33.5
39.6
63.2
RR for 150 of AW
22.1
38.4
59.4
83.0
46
RR for 200 of AW
22.1
43.4
79.2
102.8
RR for 250 of AW
22.1
48.3
99.0
122.6
RR for 300 of AW
22.1
53.3
118.7
142.3
RR for 500 of AW
22.1
73.1
197.9
221.5
Table 5-2b presents replacement rates for average years of service (30 years) and
different levels of wage under current pension system (Conventional PAYG) and
different reform options.
SSIF Revenue, Current Balance and Fund Reserve for Conventional PAYG, Notional
PAYG, and Private Fully-Funded Pillar: The financial prospectus of the SIF becomes
worse off with notional PAYG as compared with the conventional PAYG system.
Meanwhile, the funded account system, while showing surpluses over time, affects the
conventional PAYG system so that it experiences permanent deficits. Taking the
favourable replacement rates and other preferable features that funded system entails,
some special measures should be taken to overcome this drawback. Particularly, as
already discussed, a guarantee pension payable from the general taxes can contribute to
meeting the needs. For example some part of other social welfare programs budget might
be partitioned for this purpose.
Table 5-2. SSIF Total Revenue, Current Balance and Fund Reserve of PAYG system
under different course of reforms (as percent of GDP)
2006
2007
2008
2010
2020
2030
2040
2050
2060
2070
2080
3.6
3.9
4.1
4.3
5.0
5.4
6.1
6.3
6.5
6.9
7.0
3.3
3.6
3.8
3.8
5.3
7.5
8.3
10.4
12.1
11.6
12.2
.3
.4
.3
.7
.3
.9
.4
1.6
(.3)
2.5
(2.1)
(2.2)
(4.1)
(5.6)
(4.7)
(5.2)
3.6
3.8
4.0
4.0
5.7
6.2
6.1
6.3
6.4
6.8
7.0
3.5
3.2
3.0
2.6
3.1
6.9
10.7
19.8
27.3
26.1
29.2
.1
.2
.6
.8
1.0
1.7
1.4
4.0
2.5
20.5
(.7)
16.8
(4.7)
(13.5)
(20.9)
(19.3)
(22.2)
1.9
1.9
1.9
1.6
1.1
.4
.1
3.7
3.5
3.3
3.0
3.7
4.9
5.4
(1.7)
(1.5)
(1.4)
(1.4)
(2.7)
(4.4)
(5.4)
1.8
2.1
2.4
2.8
5.6
8.1
10.4
11.4
11.3
11.7
11.8
.3
.3
.3
.4
1.0
2.3
3.5
5.7
6.4
6.1
6.7
1.5
1.8
2.0
2.4
4.5
5.8
6.9
5.6
4.9
5.6
5.2
1.5
3.2
5.0
8.7
31.6
57.9
84.3
97.9
96.8
100.5
103.9
CONVENTIONAL PAYG
Total Revenue: PAYG
Total Expenditure:
PAYG
Current Balance: PAYG
Fund Reserve: PAYG
NOTIONAL
Total Revenue: PAYG
Total Expenditure:
PAYG
Current Balance: PAYG
Fund Reserve: PAYG
FUNDED PILLAR
Total Revenue: PAYG
Total Expenditure:
PAYG
Current Balance: PAYG
Fund Reserve: PAYG
Total Revenue: Funded
Total Expenditure:
Funded
Current Balance:
Funded
Fund Reserve: Funded
47
Rate of Return on Social Contributions under different reform options: Table 5-3
assures that the rate of return on contributions is improving as we move from
conventional to notional and then to fully-funded PAYG system.
Table 5.3. ROR on Mandatory Contributions in all Systems
Conventional PAYG System
Conventional PAYG System with Guarantee
Pension
Notional PAYG System with Guarantee Pension
Fully-Funded System without Guarantee
Pension
Fully-Funded System with Guarantee Pension
Indexed to
Inflation
Indexed 50 % Inflation and 50%
to Nominal Wage Growth
Indexed to Nominal Wage
Growth
-8.1
-7.2
-6.4
2.0
2.7
2.1
2.9
2.2
3.0
2.6
5.1
3.2
5.4
3.8
5.7
The quick comparison shows that introduction of fully-funded system provides much
better pensions, however causes the PAYG system to suffer deficits.
Minimum Guarantee Pension: Guarantee Pensions (25 percent of average wage of
previous year) for old age pensioners would compose 2.6 percent of GDP in 2006 and
gradually increasing reach 7.0 percent of GDP in 2080 (see Table 5-4). This is mainly
due to much higher growth rate of wages than GDP grows. Provision of guarantee
pensions to all pension insurance system beneficiaries will require 3.5 percent of GDP in
2006 and 8.0 percent of GDP in 2080. The issue of sources of funding for this benefit is
not considered in this paper, although as an option the employer might be required to pay
for the privileged pensioners and their dependents, the state budget might provide funds
for the rest).
Table 5-4. Minimum Guarantee Pensions as percent of GDP (2006-2080)
2006
2007
2008
2009
2010
2020
2030
2040
2050
2060
2070
2080
All Beneficiaries
482.4
481.5
477.0
468.4
460.8
475.6
529.4
531.3
613.5
678.6
632.3
639.3
Guarantee Pensions
(bln AMD)
73.2
85.9
97.5
106.9
114.3
201.3
328.2
445.8
694.2
1,009.5
1,206.4
1,544.1
Guarantee Pensions
as % of GDP
3.5
3.9
4.2
4.4
4.4
4.8
5.7
5.9
7.0
8.0
7.7
8.0
Old Age Pensioners
377.6
375.2
369.5
359.6
351.2
371.9
434.2
441.7
540.3
608.0
564.2
576.4
Guarantee Pensions
(bln AMD)
57.3
67.0
75.6
82.1
87.1
157.4
269.1
370.7
611.3
904.5
1,076.7
1,392.1
Guarantee Pensions
as % of GDP
2.6
2.9
3.1
3.2
3.2
3.6
4.5
4.7
6.0
7.0
6.7
7.0
48
APPENDIX: INTERNATIONAL EXPERIENCE WITH PENSION REFORM
There are many countries that have changed their pension plans and face the challenge of
improving the adequacy of pension provisions to cope with the problem of rising
dependency ratios (the ratio of pensioners to those of working age in a given population).
Advanced economies are confronted with a variety of retirement challenges associated
with an aging population. Demographic changes are critical issues not only in developed
countries, but also in lesser developed countries. While the increased dependency ratio is
the driving factor behind pension reforms in the developed countries, in emerging
countries the inefficiency of existing social security schemes, mismanagement, and
financial imbalances are additional motivators for reform.
Apart from the economic conditions and the degree of development in each country,
demographic changes are the main factors that cause the gradual increase in the
dependency ratio. Increasing longevity and a low, declining fertility rate are the main
causes that make pension provision inadequate for current retirees. The system will face
more serious problems in the future if there is no reform.
Increasing longevity: In recent decades, life expectancy at birth has consistently
increased in all developed economies, from an average of about 66 years in the postwar
period to 76 years
today, and is projected
Figure A-1. Life Expectancy at Birth
to reach 82 years or
Life expectancy at birth
more by 2050. Data
(in years)
1950-2050
from World Population
Prospects (UN) on
world life expectancy
shows a steeper trend
for the same period
(from 47 in the
beginning of the 1950’s
to 66 today and
projected to reach 75
by 2050). With regard
to pension costs, it is
more important to
consider life
expectancy after age 60 or 65. World data Indicates that remaining life expectancy for
both sexes after age 60 will rise from 6 years currently, to a projected 15 years or more
in 2050. (For Armenia this trend will be from 12 to 17 years for the same period. The
figure above shows the trend of increase in life expectancy at birth (scale: after age 60)
for selected countries, among them Armenia.
90
Sweden
85
Canada
Chile
80
Hungary
Armenia
75
70
65
60
1950- 1955- 1960- 1965- 1970- 1975- 1980- 1985- 1990- 1995- 2000- 2005- 2010- 2015- 2020- 2025- 2030- 2035- 2040- 204555
60
65
70
75
80
85
90
95
2000
05
10
15
20
25
30
35
40
45
50
Sources: United Nations, World Population Prospects: The 2004 Revision Population Database
Low and declining fertility rates. Between the early 1950s and the late 1990s, fertility
rates in developed economies dropped from approximately 2.8 to 1.6 children per
woman. Data indicate a reduction from 5 to 2.8 in the world average fertility rate since
49
the beginning of this century, and it is predicted to decline to 2.1 by 2050. Therefore,
while the aging population is a global phenomenon, it is more serious problem in some
countries, especially in advanced ones, where it is closer to replacement rate.
Nevertheless, some emerging countries also see this
Figure A-2. Total Fertility Rate
as a potential threat for
the future of their
pension systems. The
declining fertility rate in
the selected countries is
shown in the figure
below. It indicates that,
regardless of the fertility
rate each country had in
the mid-twentieth
century, the declining
trend will lead to a birth
rate of 1.85 by 2050,
which is below the
replacement rate. 27
6
Total fertility rate
(children per woman)
1950-2050
5.5
chile
Armenia
Canada
Sweden
5
4.5
4
3.5
3
Hungary
2.5
1.85
2
1.5
1
1950- 1955- 1960- 1965- 1970- 1975- 1980- 1985- 1990- 1995- 2000- 2005- 2010- 2015- 2020- 2025- 2030- 2035- 2040- 204555
60
65
70
75
80
85
90
95
2000
05
10
15
20
25
30
35
40
45
50
Sources: United Nations, World Population Prospects: The 2004 Revision Population Database
IMF Global Financial Stability Report (GFSR)28 concludes that, a direct implication of
the increase in longevity and declining of the fertility rate is the continued increase in the
dependency ratio. A higher dependency ratio implies greater demand for retirement
income relative to contributions from working income and more pressure on pension
plans. Policy-makers have started to address these challenges and to rethink their pension
systems and implement different solutions and reforms. Thus far, pension reforms have
been aimed at reducing the shortcoming of existing systems in various ways: reducing
benefits, increasing contributions (e.g., taxes to pay for state pensions), redefining risksharing between sponsors and beneficiaries, increasing the retirement age, and increased
funding of pension obligations by both the public and private sectors as well as greater
retirement savings by individuals.
This chapter will focus on the pension-reform implementation in the certain selected and
relevant developed and developing countries and will try to summarize their experiences
and highlight their useful lessons. The selected countries are Sweden, which has a
partially privatized pension system, Chile, whit mandatory private fully-funded accounts,
Hungary, a multi-pillar system, and Canada, which institutes a mix of private and public
accounts.
Sweden
27
28
Data Sources: United Nations, World Population Prospects: The 2004 Revision
IMF, GFSR, September 2004, p 84.
50
Background and the old Pension System
Sweden has a long tradition of social insurance. Sweden was the first nation that
implemented a mandatory governmental retirement-system for all citizens. Also Sweden
is one of the pioneers in transferring from public towards a private pension system.
Sweden’s former pension system (which still determines benefits for older workers and
retirees) was a tax-financed, pay-as-you-go program. The system had two components.
The first part was the basic pension, which was available to everyone regardless of
contributions paid. In addition, a second, earnings-related supplementary pension was
available. The supplementary pension provided old-age benefits linked to each worker’s
earnings history. To finance these benefits, the government imposed two separate payroll
taxes on workers: a tax of nearly 6 percent for the basic benefit and a tax of 13 percent
for the earnings-related portion of the program. The government also used general
revenues to help finance the basic benefit. 29
Prior to the 1960 reform, workers were only covered by a low flat-rate benefit. The
national supplementary pension system was introduced in 1960. It was a PAYG system,
with a buffer fund being gradually built up for the purpose of evening out contribution
revenues and expenditures over time. 30
Why the reform was undertaken
The motivation for Swedish policy-makers to reform their pension plan was awareness of
the fact that the country’s working defined-benefit system (a combined flat-rate benefit
with an earnings-related scheme) was not well adapted to meet future challenges and that
the system was financially unstable. Among the main factors that caused instability
within the system and threatened a large increase in the rate of contribution and
unfairness of the redistribution were:
• long-term unfunded pension system liabilities and deficits;
• already high tax rates and possible increases in the future;
• low long-term rates of real economic growth (2.5 percent on average in the last 5
years31); and
• demographics (increasing dependency ratio due to increase in longevity and
decline in fertility rate).
Thus, the most important reason for reforming the old system was to achieve long run
financial stability and to promote greater justice, both within and between generations, as
well as to comply with economic and demographic developments. In the new system, the
goal is to provide a framework that will promote mandatory saving through the pension
system – but with privately managed assets.
29
-Internet Source: http://www.heritage.org/- (The Heritage Foundation)
Appendix A, National Strategy Report on the Future of Pension Systems – Sweden, P 28
31
IMF, Sweden: 2005 Article IV Consultation—Staff Report
30
51
In a series of steps in the 1990s, Sweden converted a two-tier defined benefit scheme into
a combination of notional defined contribution (NDC) pay-as-you-go and financial
defined contribution (FDC) schemes. 32 These two accounts are also called Notional
Individual Accounts (NIA) and Private Individual Accounts (PIA)33 respectively and will
hereinafter be referred to as such. The Swedish pension system includes contractual
benefits as third pillar.34 The Swedish parliament, the Riksdag, reached an agreement to
reform the system in 1994 and approved implementing legislation to do so in 1998. Then,
it started gradual implementation of the policy until it came into full effect in 2003. 35
The New Pension System
Under the new Swedish old-age pension system, each worker’s future pension will be
based on the amount of money accumulated in two separate individual accounts. The
retirement income consists of a notional account maintained by the government on behalf
of the individual (NIA) and from a completely private individual account (PIA).
The overall contribution rate for the two schemes together is 18.5 percent of earnings,
with a split of 16 percent and 2.5 percent between the notional and private account
schemes, respectively. Accounts in the NIA system earn an economic rate of return,
whereas accounts in the PIA scheme earn a financial rate of return.36 The employer and
employee split the 18.5 percent contribution, with each paying 9.25 percent of eligible
income into the system.37 This money is then divided, based on the above-referenced
percentages, between the government-maintained account and the private account.
Beside the two main account-features, which will be discussed in more detail below, the
system also has two important considerations worth mentioning. One is the safety net to
protect the poor, in that the government will continue to guarantee a minimum pension
funded by general tax revenues. Also, to mitigate the negative effects on retirees and
older workers during the transition period, current retirees and older workers continue to
receive retirement income based on the old program. Workers born between 1938 and
1953 will receive benefits from both the old and the new systems as the new system is
gradually introduced.
The Notional Individual Accounts (NIA)
As described above, the larger part of 18.5 percent payroll taxes-16 percent-goes to the
government portion of the program. As is the case with pay-as-you-go systems, the
money is used to fund benefits for existing retirees. Each worker’s 16 percent payroll tax
is credited to an individual account, although the accounts are notional. These imaginary
accounts are credited each year with "earnings" based on Sweden’s per capita wage
32
Palmar, The New Swedish Pension System, p.1.
-Internet Source: http://www.heritage.org/- (The Heritage Foundation)
34
Palmar, 2001, p.6.
35
National Strategy Report on the Future of Pension Systems – Sweden, p.2.
36
Palmar, The New Swedish Pension System, p1.
37
Same, p.7
33
52
growth. Beginning at age 61, a worker can retire, and the government uses the money in
these notional accounts to calculate an annuity (annual retirement benefit) for the worker.
Retirees under age 65 will receive a benefit that is72 percent of the benefit they would
receive by waiting to age 65 to start taking their pension. If they wait until age 70, their
benefit would be 157 percent of the benefit available at age 65.38 It is also important to
note that, since the calculations are based on life expectancy, the longer a worker stays in
the workforce, the larger the annuity received.39 This reform is expected to discourage
workers from retiring early.
The new pension system will adjust retirement benefits for inflation, although the
adjustment may vary according to real wage growth. This provision creates a direct link
between the growth of labor-force income, the taxes paid on that income (contribution),
and the benefits paid to retirees. This ensures financial sustainability of the system.40
There are also other characteristics of the notional part of the Swedish pension system
which are worth mentioning. These characteristics are quoted below:
•
•
•
•
NIA will be charged a modest administrative fee to cover the government’s cost
of running the program.
Account balances are not transferable to heirs.
Workers who leave the workforce to care for children are credited with
contributions to their account.
The full tax is imposed on all income up to a defined amount. Above that amount,
the government collects only the employer’s share of the tax. The revenue
generated by this extra employer’s contribution is not, however, credited to the
worker’s account. 41
The Private Individual Account (PIA)
Two and a half percent of the payroll contribution allocated to a private individual
account. This is the funded part of the system. The money collected in this account is
invested in real assets that can then be sold to generate income during retirement. The
38
Schieber, 2001, p13.
An annuity is calculated by dividing an individual’s capital balance at retirement by unisex life
expectancy for men and women together at that age and in that year:
Annuity = Capital/Life expectancy
Note: In the Swedish scheme, the life expectancy factor used in calculating the annuity is not a forwardlooking cohort life expectancy estimate, In practice, this generosity comes at the expense of coming
generations, and could be avoided by calculating the annuity factor with a cohort-based forecast like
Latvian version of the Swedish reform. (Palmer, The New Swedish Pension System, p.14 )
39
40
Based on a study on Sweden’s pension System, Sweden’s NIA pension system is financially stable if:
(i) the figure for life expectancy used in computing (NIA) annuities is on average correctly estimated and,
(ii) the rate of return in the account scheme follows the rate of growth of the contribution base. (Palmar,
2001, p.17)
41
The Heritage Foundation , Policy research and Analysis website
53
government has set up a centralized agency, the Premium Pension Agency (PPA), to keep
records for the new system.
There are five funds. The PIA allows workers to invest in these funds and create their
own portfolios by allocating to these five funds.
Other aspects of the PIA system quoted below:
•
•
•
•
The system has been structured to discourage funds with high administrative
fees.
Workers can choose survivor’s insurance that will provide income to heirs in the
case of early death.
The private-pension funds must follow prudent principles and are largely free of
new regulations.
Workers will be able to switch funds but will be charged an annual
administrative fee to do so.42
Transition from the Old to the New System
How to make the transition from one to another system is one of the most difficult issues
in pension-system reform. During the process, it is critical to decide which birth cohort
would be the first to be covered in the new system and whether the transition should be
gradual for some birth cohorts, using a weighted average benefits computed according to
the old and new rules.43 The pension-system reform took into account three population
cohorts, and there is a gradual implementation of the full system for whole society. The
three cohorts are defined as follows:
i) people born in 1937 or before are covered solely by the old system. (e.g. 61
years old and over in the year of implementation of the new system);
ii) the intermediate generation, born between 1938 and 195344 (45-60 years old in
the year of implementation of the new system), earn a combination of benefits,
toward supplementary pension in the old system (from 1960 at the earliest) and
the premium pension in the new system. The extent to which a person belonging
to this intermediate generation is affected by the new or old system depends on
year of birth. People born in 1938 would receive 4/20 of their pension from the
reformed system and the remainder (16/20) according to the old rules. Those born
in 1939 will receive 5/20 from the new system, and so on; and
iii) people born in 1954 or later are covered entirely by the reformed pension
system.45
The starting point for the payment of PIA contributions was 1995. Thus only those who
entered the labor force after 1995 are completely covered by the private individual
42
Ibid
Palmar, The New Swedish Pension System, p.4.
44
Originally in 1994 this group includes people born between 1935-1953. As the reform process drew out
in time, the first age cohort chosen in the legislation (in 1998) was the birth cohort of 1938.
45
Palmar, The New Swedish Pension System, p.5.
43
54
account system,and the inclusion of older birth cohorts in the PIA scheme was mainly
symbolic. Palmar46 explains that:
It would have been logical to start the financial scheme with a younger age cohort and
perhaps to have made it optional for persons who, due to their age at the time of
introduction of the scheme, would not be able to accumulate large account values. This
feature considered in the more recent, similar pension reforms that have taken place in
countries such as in Latvia and Poland.
The major part of the system NIA scheme (the pay-as-you-go system) has a buffer fund
that generated due to fluctuations in the sizes of birth cohorts. The reserves accumulated
in this fund will help to finance the transition period especially when the large cohorts
born in the 1940s become retirees. 47
Shortcoming of the Reforms and Challenges Ahead
While the Sweden’s experience confirms that the individual accounts are a good solution
for existing problems before reform, but should be mentioned that Sweden has not made
a perfect transition to the new system. The individual accounts are not expected to be
fully operational until a long time after the reform was undertaken. Therefore the
government collects the money, holds them and invests in low-risk assets, until the
individual accounts are ready. After that, the government will distribute all of the funds
and allocate earnings to the appropriate individual personal account.48
Chile
Background and the old Pension System
Based on a study in World Bank Group49, Latin America has a relatively long tradition of
formal, institutionalized social security among developing regions. Since the early 1900s,
various levels of government, unions, and trade associations have administered different
social insurance and retirement schemes. In Latin America, national public-pension
systems were first established in Chile in 1924.
In the past, Chile was a pioneering representative of the new system and has served as a
model for other countries with regard to its radical reforms. Two decades ago, Chile’s
government radically altered its approach to old-age income security and changed public
pensions from a pay-as-you-go defined benefit system into individual accounts that are
mandatory and fully-funded by private management.
46
Ibid.
Palmar, 2001. p.2.
48
The Heritage Foundation , Policy research and Analysis website
47
49
Perry, 2003, p.15.
55
The old Chilean system was a-pay-as–you-go, defined-benefit system. It is important to
describe the system’s structure, because its inefficiencies and mismanagement were the
reasons that forced Chile to develop a totally new pension system instead of following the
much easier path of “improving” the existing one. 50
Acuña and Iglesias study the Chile’s old system and its problems as follows:
In 1980, Chile had 32 social security funds51 and over 100 different pension
programs, many of them with such a small number of contributors. The scale of
operation for some of them was extremely inefficient and, since the affiliation to
each institution was tied to the economic sector in which the worker was
employed, the lack of competition between the funds and pension programs did
not create any incentive to manage the funds efficiently.
On the other hand, contribution rates were high and differed from one fund to
another. Thus the total contribution rate for all social security programs reached
its maximum level (between 59 and 69 percent) in March 1974, which encouraged
evasion and resulted in considerable distortions in the labor market. There were
also great differences in benefit levels and in the requirements that had to be met
to obtain them. Different interest groups lobbied to obtain special privileges from
the system. In fact, the history of Chilean social security has been described as
an uninterrupted sequence of schemes specially created for groups with political
power and of resources collected by pension program being used to finance other
types of benefits. There were no automatic mechanisms to index pensions to
inflation. 52
Why the reform was undertaken
The idea that Chile’s pension program needed a radical change existed for a considerable
time before the 1980 reform. The old Chilean pension system faced some serious
problems and obstacles which made it by and large inefficient and unfair and forced a
radical reform. There were, therefore, several reasons for Chile’s pension reform.
One reason was the magnitude of the financial imbalances forecast. Acuña and Iglesias in
their paper referenced a study,53 based on which the social security deficit forecasted to
grow significantly and reach between 10.3 percent and 16.1 percent of GDP by the year
2000. Changes in the form of reduced benefits and increased contribution rates seemed,
therefore, inevitable, though these would have caused great political resistance. The
mismanagement of financial resources, the use of social security contributions to finance
other types of benefits, contribution evasion, the creation of unfunded new benefits, and
the provision of special benefits to some groups along with aging population were among
the main factors that cause financial deterioration of the old social security system.
50
Acuña and Iglesias, 2001, p.21.
The management institutions of the old system are known as “Cajas de Previsión”
52
Acuña and Iglesias, 2001, p.22.
53
Wagner H. Gert: Estudio de la Reforma Previsional: Previsión y Reforma, Efectos en la Industria y en el
País. Institute of Economics, Catholic University of Chile (May 1983).
51
56
Second, since in the mid-1970s there was a general economic and social development in
Chile, which drastically reduced the size of the state and aimed to shape a new marketoriented model that eliminated the monopolies both in businesses and in trade unions.
The social security system could not be excluded from all these reforms.
The other reason for the reform was to find a particular design that would reduce political
risks faced by the traditional social-security system, which was threatened by the constant
pressures of different interest groups trying to obtain special privileges from the system.
There was a need for a new system that would provide some protection from the
pressures of these groups. 54
The New Pension System
The revolutionary 1980 pension reform created a new system based on mandatory
contributions to individual accounts with private fund management. The new pension
system gives those workers covered by the scheme the right to choose between different
pension providers and between different forms of payout after their retirement. The new
system is known as the “AFP System”55 and was implemented in May 1981. The AFP
system offers pensions for old age, common disability, death (or “survivorship”), and a
single funeral grant. In this new system contributions are capitalized in individual
accounts. The rate of contribution is defined in the law as a proportion of an individual’s
wage.56 The reformers also created a social safety net that would provide a minimum
pension guarantee (MPG) to all participants who contributed to the system for at least 20
years. Currently, the MPG is set at around US$140 a month and is indexed to inflation to
help protect retirees’ purchasing power.57
Old-age benefits-The amount of payment of the old-age pensions depends on the balance
accumulated in each worker’s individual account; his or her life expectancy, in addition
to that of the beneficiaries; and on the discount rate. To be entitled to this benefit, a man
must be 65 years old, while a woman must be 60 years old.
Disability58 and survivorship benefits59-. The law defines disability and survivorship
benefits as a percentage of the “base wage”, which is an average of the wages received by
a deceased contributor over the past ten years, updated by the change in the Consumer
Price Index during this period.60
54
Acuña and Iglesias, 2001.
“AFP” stands for Administradoras de Fondos de Pensiones (Pension Fund
Management Companies).
56
Acuña and Iglesias, 2001.
57
IMF Survey. October 17, 2005.
58
Include all contributing members who are not retired and members who are unemployed (up to twelve
months).
59
Include surviving spouse, legitimate, biological adopted children and the mother of the biological
children.
60
Acuña and Iglesias, 2001.
55
57
Some specific features of the system are:
• The worker is free to choose among different AFPs, which are registered, singlepurpose, pension management institutions. The AFPs are private and competitive
firms whose purpose is to invest the funds in the capital market on behalf of its
members.
• The contribution rate for the old age, disability and survivorship program
managed by the AFPs is 10 percent of the wage. In addition to the 10 percent
contribution, workers must pay insurance and management charges to the AFPs.
Contributors may also make voluntary contributions up to a maximum limit,
which are not subject to income tax, to increase the level of benefits or bring
forward their retirement age.
• The pensioner is free to choose the way in which he or she can receive the
pensions from three different pension modes: i) scheduled withdrawal (by keeping
his or her funds in the personal AFP account and withdraws annually), ii) life
annuity (by purchasing from the life insurance companies), and iii) temporary
income with deferred life annuity (by signing a contract with life insurance
companies). All are indexed to prices.
• The government issued “recognition bonds” for workers who made the switch
from the old system to the new to recognize the contributions that they have
already made to the old system. The bonds pay four percent annual interest, and
may be cashed in by the workers upon their retirement.
• The State plays mainly a “subsidiary role” manifested by its responsibility to
regulate and supervise the system, finance minimum pensions and provide certain
guarantees.61
Transition from the Old to the New System
The transition was faced two important problems: i) how to persuade the workers to
transfer to the new system and ii) how to handle closing down the old program.
To handle these problems smoothly, it was decided that the new system should be
mandatory only for those workers who entered the workforce after January 1, 1983. Thus,
for workers who entered the workforce before that date and for the self-employed, the
new system was voluntary. There is, therefore, a long transition period, which will
continue until the last pensioner of the old system receives its pension. This period is
projected to be approximately 50 years after the commencement of the pension in 1980.62
Although workers lacked confidence in the old system, the impression that the new
system offered unknown opportunities, the potential to increase wages for those who
changed over and the security of the Recognition Bond, all served to make the process of
transferring workers to the AFP system extraordinarily rapid. In only eight months,
therefore, 65 percent of all civilian workers covered by mandatory pension program had
61
62
Ibrd.
Ibrd, p.16.
58
transferred to the AFP system. By the end of 1983, this figure had reached 77 percent but
fell to 51 percent by the end of 2000. 63
Pension reform had commenced prior to 1981. After 1974, the government implemented
gradual changes in regulations (regarding minimum pensions, contribution rates, age
requirements, and minimum contributions) aimed at rationalizing the operation of the old
program and improving financing. Regulation of the AFP system has also undergone
numerous changes since 1980. For example, between 1981 and 1996, 412 modifications
to the main law of the AFP system were introduced, by means of 33 different laws.
One point should be noted that in order to eliminate old system completely, having a
well-developed market for life and disability insurance it is a necessity. This in turn
required a well-developed long-term fixed income market.
Shortcoming of the Reforms and Challenges Ahead
Nearly 25 years after the institution of pension reform, as the first generation of workers
under the new pension system begin to retire, Chileans face serious shortcomings of the
new system, and there are challenges ahead.
As mentioned above, the system’s coverage is less than had been expected. A significant
share—roughly half—of the Chilean work force, remain outside the system, because they
earn much of their income in the underground economy, are self-employed, or are
seasonal workers. Thus, only half of Chilean workers are captured by the system.
Other evidence indicates that many middle-class workers who contributed regularly are
finding that their private accounts fail to deliver as much in benefits as they would have
received if they had stayed in the old system. 64
Other growing pains of the system is related to insufficient competition of the funds. The
IMF Survey65 concluded that:
Currently, there are only six AFPs, and they dominate the financial
system. Also, the costs of administering the system have remained quite
high.
In the end it should be noted that Chile’s pension reform has many advantages in terms of
simplicity, transparency, ease of supervision and reduction of the government deficit.
The privatization of the funds has had a positive impact in the creation of a modern
capital market and generates cheaper credit for companies that formerly could turn only
to banks when they wanted to expand. But still there is an uncertainty and debate
regarding whether the privatized system has been less successful in ensuring sufficient
retirement income for the elderly. In other words, the new system could be a good system
for the Chilean state, but it may be unable to provide an adequate pension for most
Chileans.66
63
Ibrd, Table 2.
Rohter, 2005.
65
IMF survey, October 2005.
66
Acuña and Iglesias, 2001.
64
59
Hungary
Background and the old Pension System
Hungary’s pension system was organized originally on the basis of principles
implemented in the Austro-Hungarian period. After World War II, along with other
countries belonging to Central and Eastern Europe, the Hungarian pension system was
based on key communist goals and guidelines. In the communist era, the Hungarian
public pension system was a defined benefit scheme financed on a pay-as-you-go basis.
Hungary, like Kazakhstan, was among the first former socialist countries to implement a
systemic change from a purely pay-as-you-go, defined-benefit system to a partially
funded, defined-contribution system.67
Why the reform was undertaken
Hungary’s pay-as-you-go system matured rapidly during the post-war period. In addition
to the inevitable impact of an aging demographic, the deterioration of pension finances in
Hungary was strongly related to the policies that prevailed under the socialist regime
(such as a low retirement age68) and to certain strategies followed during the transition to
a market-oriented economy. Thus, the old system was characterized by high systemdependency ratios, large expenditures relative to GDP, and high contribution rates. The
high payroll tax-rates and the diminishing confidence in the public pay-as-you-go scheme
drove economic activity into the informal sector, contracting the tax base further. A
shrinking tax base and growing pension expenditures would inevitably have resulted in
either larger general government deficits or higher payroll taxes. It was, therefore,
necessary to reform the pension system and to design and implement a new system that
could restore long-run financial viability and reduce economic distortions while still
protecting the most vulnerable groups during the transition.
In May 1996, the Hungarian Government announced its intention to introduce a multipillar system in Hungary. The law was passed by Parliament in July 1997 and was fully
implemented beginning in 1998. 69
The New Pension System
The new legislation replaced the old structure with a multi-pillar70 one for all workers.
67
Natali, 2004
After World War II, the retirement age was reduced from its pre-war levels of 65 to 60 and 60 to 55 for
men and women, respectively.
69
Palacios and Rocha, 1997. p.1.
70
Based on World Bank terminology: the first pillar refers to a pay-as-you-go, publicly-managed pillar, the
second pillar is mandatory, fully-funded, and privately-managed, and the third pillar is voluntary, also
fully-funded and privately-managed. The same terminology is also used in this paper.
68
60
The mixed system will be mandatory for all workers who enter the workforce as of July
1, 1998. For those who have a right to a pension under the old system, participation in the
second tier is optional.71
The first (public and mandatory) pillar consists of two different tiers. The first tier is
consistent with the pay-as-you-go mechanism and defined-benefit logic, but it is a
“modernized” form of the old pay-as-you-go system. The main changes are:
• A decrease in employers’ contributions from 24 percent to 22 percent of gross
wages, while employees’ contributions now comprise up to 9 percent (a
contribution of 31 percent in total). Since 2003, employers’ contribution rate has
decreased to 19.5 percent, and there is a diversion of 8 percent of the employees’
contribution from the pay-as-you-go system to private, second pillar accounts;
• A shift in the indexation from net wage to a combination of price-wage
indexation;
• A gradual increase in the retirement age until it reaches 62 for both men (as of
2000) and women (in 2009);72
• A new benefit formula that can gradually eliminate some of the redistributive
elements of the system and proportionally reduce contribution rates.73
The second (public-private and mandatory) tier is represented by supplementary funded
pensions, based on a fully-funded method of financing. In this tier, as noted above, there
is an 8 percent employee contribution rate. These contributions are collected and
managed by private institutions, which also pay benefits. Benefits are indexed following
the same rules adopted for the first tier. Pension funds officially take the form of mutual
savings associations managed exclusively by their members.74 The State Financial
Supervisory Authority, which issues fund licenses, supervises the funds. The
supplementary tier comprises approximately 50 percent of the economically active
population. In 2001, 21 licensed private funds were operating in financial institutions,
large employers or groups representing employees’ interests, and accounting firms or
other private associations.
The third pillar (voluntary schemes), which had already developed in the late 1990s,
supplements the mandatory pension program. Employers can participate in financing
voluntary funds, which allows them tax deductions). 75
The shift of an 876 percent contribution rate from the pay-as-you-go system to private,
second pillar accounts, implies a loss of revenue that would increase over time due to
coverage expansion in the new system and would result in an increased public deficit in
71
Natali, 2004
Natali, 2004, Annex 2.
73
Palacios and Rocha, 1997. p.21
74
This is a Hungarian construction, differing from the trusts in the Anglo-Saxon countries (whose boards
either consists of employer appointees, or have a split representation, as in Australia), or the Swiss
foundations (whose boards contain employer and employee representatives in equal proportions).(Palacios
and Rocha, 1997)
75
Natali, 2004.
76
Natali, 2004, Annex 2
72
61
the absence of offsetting measures. The increase in the deficit would be roughly
equivalent to the increase in private savings, meaning that there would be no change in
national savings (except interest-rate effects or other effects on private savings), so that
the reform implies a replacement of explicit for implicit debt. Thus, it is estimated that
the potential growth effects of this strategy would be limited, and the burden of the
transition will shift entirely to future generations. With regard to Hungary’s financial
market, there is evidence of the improvement of Hungary’s credit rating in the months
that followed the implementation of pension-reform legislation.77 The study on market
valuation due to liabilities is limited, however, to private firms and does not extended to
the public sector, so it is not possible to generalize from these observations.
Shortcoming of the Reforms and Challenges Ahead
Population ageing remains the greatest challenge for the future stability of the pension
system (with a particularly high system dependency ratio in the next few decades),
however, the 1998 reform is expected to limit the future deficit of the public pension
program. It is too soon to determine whether the changes have had the expected effect.
Canada
A world Bank research on Canada’s Pension future stated that:
The Canadian public pension system has been characterized as one of the best
in the world. Over time, it has succeeded in decreasing poverty among the
elderly through the institution of a series of targeted, flat-rate benefits that
redistribute income to pensioners.78
The distinguishing feature of the Canadian pension system is the defined-benefit logic
that continues to exist but is improved by some funding. Like Japan and France, Canada’s
some public-sector plans are (at least partially) funded. Before the system’s 1997 reform,
it was run largely on a pay-as-you-go basis through which the accumulated fund
supporting the plan was equivalent to two years of benefit payments.79
Why the reform was undertaken
As is true for many developed countries, Canada’s aging population threatened the
financial sustainability of the system. The 1997 reforms increased the financial
sustainability of the first pillar, pay-as-you-go, earnings-related element to the public
system, and some measures have been implemented to make the plan partially funded. As
a consequence, there appears to be a general sense of increased confidence for the future
of public pensions, at least among the Canadian population.80
The New Pension System
77
Palacios and Rocha, 1997. p.35
Pozzebon, WB 2004-15
79
Schieber, 2001.
80
Pozzebon, WB.2004-15, p.1.
78
62
Canada’s multi-pillar approach, with both funded and unfunded components, provide a
solid foundation for retirement income. After the 1997 reform, Canada’s public pension
system has both an unfunded component, the Old Age Security (OAS) system, and a
funded element, the Canada Pension Plan (CPP) Investment Fund. The latter was adopted
as a measure to improve public pension retirement security. The public system also
embodies a balance between poverty reduction (the OAS program) and retirement
income support (CPP). The private pension system is generally considered the funded
component of Canada’s retirement system. It has proved successful in providing income
replacement.
Canada’s pension system has two pillars: i) a public pension system and ii) private plans.
First Pillar-Public Pension System
The public pension system consists of two main tiers: the OAS and the CPP.
i)
Old Age Security (OAS) program
The OAS program includes three sub-components: a) the OAS pension, b) the
Guaranteed Income Supplement (GIS), and c) a spouse’s allowance. The OAS
pension is a flat benefit available to all individuals aged 65 and over, who have been
resident in Canada at least 10 years.81 The OAS pension is taxable and indexed
quarterly to the Consumer Price Index (CPI). The other two sub-components, the GIS
and the spouse’s allowance, are means-tested, flat-rate benefits. The GIS targets low
income OAS recipients, while spouses of OAS recipients who have been resident in
Canada a minimum of 10 years are eligible for the spouse’s allowance. Both are nontaxable, and, like the OAS pension, indexed quarterly to the CPI.
The maximum first pillar benefit, including the base, represents 31 percent of the
average earnings of a single person or 50 percent of the average of a couple’s
earnings. The government covers the entire cost of the universal and means-tested
pension.
ii)
Canada Pension Plan (CPP)
Employed residents of the Canada (with the exception of Quebec residents)82 are
covered by the federally-administered CPP. The CPP is a compulsory, pay-as-you-go
earnings-related scheme. This Plan also offers disability, survivor, and death benefits.
Some important features of the CPP are:
• CPP retirement benefits are set at approximately 25 percent of average
pensionable earnings over the contribution period, excluding the lowest 15
percent of years.
81
To receive a full pension, which represents approximately 14 percent of average earnings, 40 years of
residency after the age of 18 are required; the OAS pension is reduced proportionally for each year of nonresidency.
82
Quebec residents contribute to the provincially administered Quebec Pension Plan (QPP).
63
•
An individual’s contribution period spans the years between his or her 18th and
65th birthday. The normal retirement age is 65, though benefits can be obtained as
early as 60 or as late at 70.83
• The CPP is financed by contributions split evenly between employees and
employers. The 1997 reform implemented a rapid increase in total contributions
from 5.6 percent in 1996 to 9.9 percent in 2003, where they are projected to
remain for some years.
• The CPP Investment Board manages CPP surpluses. It was set up to oversee the
investment of CPP reserve assets “to maximize return without undo risk” in a
diversified portfolio of securities including stocks, real estate, bonds, and private
equities. Professional investment managers have been employed to assist the CPP
Investment Board in managing the private equity components of the reserve fund.
Though independent, the CPP Investment Board is held to high standards of
accountability; it is subject to external audits and has a wide-reaching disclosure
policy. 84
The CPP is designed to provide a secure, modest, base retirement income. Public
pensions are not intended to meet all the retirement income needs of Canadians. The
Canadian government has taken the position that it is each person's responsibility to look
at their own circumstances to decide what level of income is right for them and develop
their own retirement plan.85
Second Pillar-Private Pension System
The second major pillar of the Canadian retirement income system consists of
voluntary pension plans that are sometimes provided by employers or labor unions.
Under Canadian law, registered pension plans (RPP) are the most common type of
employer-sponsored plans. These are subject to minimum standards prescribed by federal
and provincial pension legislation. Contributions for both employers and employees are
tax-exempt and subject to annual limits. The vast majority of Canadian RPP members
have historically belonged to defined-benefit plans, however, recently there has been
some debate over whether RPPs have shifted from defined-benefit plans to definedcontribution plans.
Shortcoming of the Reforms and Challenges Ahead
The private sector component of the Canadian pension system appears to have weakened
substantially during the last decade. The gradual shift from defined-benefit to definedcontribution plans, with a parallel move away from retirement saving arrangements
covered by pension regulation, increased insecurity for future retirees. As a result, it is
necessary to improve Canada’s occupational pension system. In short, it will require
some effort to strengthen the multi-pillar foundation of Canadian retirement income.86
83
A six percent annual reduction (increase) in benefits applies for those who opt to receive a pension before
(after) age 65.
84
Pozzebon, WB.2004-15
85
Internet source: Canada's Retirement Income System.htm
86
Pozzebon, WB.2004.
64
Summary and Concluding Remarks
Pensions systems around the world face the same problems of demographic aging and
financial instability of their pension systems. This section has described Sweden’s,
Chile’s, Hungary’s and Canada’s pension systems and the reforms that these countries
have undertaken. The starting point of each country was a defined-benefit plan funded on
a predominantly pay-as-you-go basis. Through a different range of solutions and different
types of reforms, each country has tried to solve its existing system’s shortcomings. The
following chart summarizes the general information and framework of reform in the
countries discussed in this section.
Table A-1 General Information and Framework of Pension reforms in selected
Countries
Date of Reform
Date of implementation
of new regulations
Reasons for reform
Pension System
Contribution rate
Retirement age (years)
Logic of Payments
Sweden
Chile
Hungary
Canada
1998
Gradual- full effect since
2003
1980
1981
1997
1998
1997
1997
a) financial instability
b) demographic aging
Multi-pillar:
1-Notional Individual
Accounts (NIA)
2- Private Individual
Accounts (PIA)
3-Quasi-mandatory
contractual, employee
insurance
Social safety net: minimum
pension guaranteed by
government
1-NIA- 16 percent
2-PIA- 2.5 percent
for both sexes:
65-100 percent
61-72 percent
70- 157 percent
Defined Contribution (DC)
a) financial imbalances
b) mismanagement of
financial resources
c) general trend to marketoriented socio-economic
model
d) political risk
e) demographic aging (less
important)
Mandatory contributions in
individual accounts with
private management of the
funds
Social safety net: minimum
pension guaranteed (MPG)
10 percent
(and voluntary up to 20
percent)
65 for men
60 for women
Defined Contribution (DC)
Financing
1-NIA-pay-as-you-go
2-PIA-Fully funded
Fully funded
Indexation
1-NIA- wage growth rate
2-PIA-financial rate
Inflation
a) demographic aging
b) deterioration of
pension finances due to
policies prevailing under
the socialist regime
c) lack of long-run
financial viability
a) demographic aging
b) financial instability
Multi-pillar:
1-Public and mandatory
1-1-Modernized payas-you-go system
1-2- supplementary
funded pensions
2-Private Mandatory
funded pensions
3-Voluntary schemes
Multi-pillar:
1-Public Pension System
1-1-Old Age Security(OAS)
1-2-Canada Pension
Plan (CPP)
2-Private Pension System
1-Public Mandatory-19.5
percent
2-Public/Private
Supplementary- 8 percent
62 for both sexes
men: since 2000
women: since 2009
9.9 percent
1-Public Mandatory-DB
2-Supplementary -DC
1- Public mandatory
1-1-Pay-as-you-go
1-2-Funded
2-Private Mandatoryfunded
3-Voluntary-Funded
combination of
price/wage growth rate
Normal age 65
A six percent annual
reduction (increase) for
receiving pensions before
(after) age 65
Defined Benefits (DB)
1-Public Pension System
1-1-OAS-pay-as-you-go
1-2- CPP-funded
2-Private Pension Systemfunded
CPI
65
To present a visual understanding about the range of solutions that these countries have
chosen to overcome the problems of their pension systems, the figure below presents the
this section’s selected country’s approaches to funding (from fully-funded to pay-as-yougo) and the logic and basis of their pension payments (either defined-benefits, or definedcontribution).
Figure A-3. Location of Selected Countries in the Pension reform Options
Full Funding
Chile
Hungary
Canada
Sweden
Pay-As-You-Go
Pure Defined Benefit (DB)
Pure Defined contribution (DC)
To conclude, the features of each system are summarized below:
Sweden: A system with some funding, 2.5 of the total 18.5 percent contribution rate,
though retirement benefits are provided mainly on a defined-contribution basis. There is
also a social safety net provided through the government’s minimum-pension guarantee.
Chile: Almost fully funded and with benefits provided on a defined-contribution basis,
excepting the minimum-pension guaranteed payment.
Hungary: A mixed system with a first-tier providing benefits on a defined-benefit basis.
It is partially funded, 8 percent supplementary contribution of the total 27.5 percent
contribution rate.
Canada: A defined-benefit plan with some funding.
66
REFERENCES
Acuña, Rodrigo R. and Iglesias Augusto P. December 2001. “Chile’s Pension Reform
After 20 Years”. Social Protection Discussion Paper Series, No. 0129. Social
Protection Unit, Human Development Network The World Bank.
Armenia Demographic Health Survey—2000, NSS, MOH, ORC Macro, Yerevan, 2001.
Bonnar, Steve and Service, David. May 2004. “Renovate to Rejuvenate: Canadians Need
a 21st Century Pension Plan”. Towers Perrin.
Perry, Guillermo, November 2003. “Keeping the Promise of Old Age Income Security in
Latin America”, World Bank Group.
Hayastani Hanrapetutyun, “Employer-Employee ,” No. 191 (3815), November 2, 2005.
Hayastani Hanrapetutyun, No. 225 (3358), November 28, 2003.
Hudson, Valerie M.; Den Boer, Andrea, “A Surplus of Men, A Deficit of Peace: Security
and Sex Ratios in Asia's Largest States,”
http://it.stlawu.edu/~govt/361F02Hudson.html.
IMF Survey. October 17, 2005. “Chile’s Pension System faces Growing Pains,” pp. 312314.
IMF, “Global Financial Stability Report”(GFSR), September 2004. Chapter 3: Risk
Management and the Pension Fund Industry.
Public Interest Institute, “Chile’s Private Pension System,” Institute Brief, Number 10
August 1996, Volume 3.
Labor Market in the Republic of Armenia, National Statistical Service of Armenia, 2004.
Labor Force and Child Labor Survey, National Statistical Service of Armenia, 2004.
Natali, David. 2004. “ Hungary: The Reformed Pension System”. Research Project,
Service Public Fédéral Sécurité Sociale.
National Strategy Report on the Future of Pension Systems, Sweden.
Normann, Göran and Mitchell, Daniel J. “Pension Reform in Sweden: Lessons for
American Policymakers,” The Heritage Foundation, Policy Research and
Analysis.
67
Palacios, Robert and Rocha, Roberto. November 1997. “The Hungarian Pension System
in Transition,” World Bank’s Central European office in Budapest.
Palmer, Edward, “The New Swedish Pension System.”
Palmer, Edward, 2001, “Restructuring the first pillar: Financial Stability and Individual
Benefits in the Swedish Pension Reform Model”, National Social Insurance
Board and Uppsala University Sweden. 21-22 November 2001.
Pozzebon, Silvana, 2004, “The Future of Pensions in Canada”. Pension Research Council
Working Paper. PRC WP 2004-15. The Wharton School, University of
Pennsylvania.
PROST: Preparing Input, Manual.
Rohter, Larry, 2005. “Chile's Retirees Find Shortfall in Private Plan,” New York Times.
Sample Survey of Passenger Turnover (Migration) at the Border Guarding Posts of the
RoA, National Statistical Service of Armenia, 2003.
Schieber, Sylvester, 2001, “Social Security: Budgetary Tradeoffs and Transition Costs”
Testimony before the President’s Commission to Strengthen Social Security.
Socio-Economic Conditions of Armenia, National Statistical Service of Armenia, 2002,
2003, 2004, 2005.
Sundén, Annika, 2000, “How Will Sweden's New Pension System Work?” Center for
Retirement Research at Boston College.
United Nations, World Population Prospects: The 2004 Revision
Internet sources
http://www.sdc.gc.ca/en/home.shtml
http://www.heritage.org/about/sitemap.cfm
Draft Amendment to the Law on Mandatory Social Contributions,
http://www.parliament.am/drafts.php?sel=onagenda&lang=arm.
UN warns of below replacement fertility levels, By 2050 75% of developing countries
will be under 2.1 birth rate (http://www.lifesite.net/ldn/2003/feb/03022703.html).
Sex ratio at birth, United Nations Population Division
http://esa.un.org/unpp/p2k0data.asp, http://esa.un.org/unpp/index.asp?panel=2.
68
State Department for Migration and Refugees
http://www.dmr.am/ADMR/IRAVIJAK/Iravic%7E1.htm)
Statistical Annex Q1 2005, AEPLAC, http://www.aeplac.am.
Macroeconomic data, Economic Development and Research Center,
http://www.edrc.am/project.html?cat_id=38.
69