Download CF AT 2012_final reworked v.2

1
Federal Ministry of Finance
January 2012
Country fiche on
pensions - Austria
For the attention of the Economic Policy Committee‘s Ageing Working Group
1
In cooperation with the Federal Ministry of Labour and Social Affairs and Statistics Austria
1. Overview of the Austrian pension system
1.1. Description
The public pension system in Austria is predominantly based on a pay-as-you-go (PAYG)
scheme. Public pension benefits are by far the primary source of income for retirees (more
than 95%). In order to harmonise the different schemes of blue and white collar workers,
farmers, self-employed and civil servants, a standardised, more actuarially-oriented pension
account system was introduced for all employed under 50 years in 2005, established in the
Act on Harmonisation of Austrian Pension Systems. This new pension system will
gradually replace those different pension schemes over the long run.
The defined-benefit formula “45-65-80” is central in shaping the actual individual
replacement rate. A person contributing 45 years to the public pension system and retiring at
the statutory retirement age of 65 years is entitled to receive a gross public pension
amounting to 80% of his average life-time insured earnings. This results in an effective
annual accrual rate of 1.78%. The assessment basis is being extended gradually from the
best 15 to the best 40 years of work life income until 2028 (2011: best 23 years). In the public
sector it is also being extended to the best 40 years, starting from the best year in 2003 (cf.
graph 1). Entitlements for a regular old-age pension arise with a minimum of 15 insurance
years (thereof at least 7 contribution years) and when the statutory retirement age has been
reached.
Graph 1: Convergence and extension of assessment bases
Due to the establishment of the Act on Harmonisation of Austrian Pension Systems a system
of parallel accounting in the transformation process from the old to the new law applies. For
those aged 50 years or older by 2005 old law regulations hold. Otherwise, benefits are
calculated both according to the old and the new law; subsequently the sum of the two
numbers is drawn, weighted by the shares of the contribution periods before and after 2005
(e.g. if the contribution period ranges from 1970 to 2010, 1/8 comes from the new law, 7/8
from the old law regulations).
1
The statutory retirement age is 65 years for men and 60 years for women. In consequence
of a Constitutional Court ruling in the 90ies, the female retirement age will be gradually raised
to 65 years in the period from 2024 to 2033 (by ½ years steps). Generally,
bonuses/deductions for late/early retirement amount to 4.2pp per year (max. 15pp for
deductions, max. 12.6pp for bonuses).
According to the new law, there are two possibilities for early retirement:
• From 62 years onwards, a person might be eligible for corridor pension
(“Korridorpension”). The minimum requirement of insurance years is 37.5 years. For
women this gets relevant only by 2028 with the phasing in of the harmonisation of
retirement ages.
• If work has been conducted in the area of “hard labour” for at least 10 years within 20
years before retirement, men can exit at the age of 60 years (“Schwerarbeitspension”).
The minimum requirement of insurance years amounts to 45 years; the annual deduction
is reduced to 1.8pp. For women this scheme becomes relevant from 2024 onwards.
Beyond that, several early retirement regulations exist in the old law:
• Early old-age pension due to a long period of insurance (“Vorzeitige Alterspension bei
langer Versicherungsdauer”), with a minimum requirement of 37.5 insurance years. This
option is expiring since 2004 by a stepwise increase of the entry age and will be fully
phased out by 2017 (current entry ages: 63.5/58.5 years).
• In case of a corridor pension, there is an additional annual deduction of 2.1pp for old law
pensions, which is deducted without regard to the “loss ceiling” of 15pp, resulting in an
annual deduction of 6.3pp.
• Pension settlement for long-term insured (“Langzeitversichertenregelung” or
“Hacklerregelung”): early retirement with 55/60 years (w/m) after having contributed
40/45 years. From 2014 onwards, the entry age steps up to 62 years (for men
instantaneously, for women gradually starting at 57 years). Eventually, this regulation will
be encompassed by the new law corridor pension. Furthermore, in case of “hard labour”,
deductions are reduced to 1.8pp like in the new law.
Table 1: Statutory retirement age, earliest retirement age and penalties for early retirement
year
M
2010
2015
2020
2025
statutory retirement age
65
65
65
65
with 20
earliest retirement age
65
65
65
65
contribution
penalty in case of earliest
year
ret. age
statutory retirement age
65
65
65
65
with 40
earliest retirement age
62
62
62
62
contribution
penalty in case of earliest
year
10,5%
7,4%
16,4% 15,6%
ret. age
F*
statutory retirement age
60
60
60
61
with 20
earliest retirement age
60
60
60
61
contribution
penalty in case of earliest
year
ret. age
statutory retirement age
60
60
60
61
with 40
earliest retirement age
55
57
60
61
contribution
penalty in case of earliest
year
0%
10,5%
ret. age
* For female civil servants the statutory retirement age is 65 over the whole period.
2
2030
65
65
2040
65
65
2050
65
65
2060
65
65
-
-
-
-
65
62
65
62
65
62
65
62
14,9%
13,2%
12,6%
12,6%
63.5
63.5
65
65
65
65
65
65
-
-
-
-
63.5
62
65
62
65
62
65
62
7,4%
13,2%
12,6%
12,6%
Source: MoF, MoSA.
Due to the “co-existence” of old and new law regulations, the Austrian pension system has
become rather complicated, with various rules interacting in a complex way. This is illustrated
in table 1, where statutory and earliest retirement ages and penalties for early retirement are
shown. With only 20 contribution years the case is rather simple, because early retirement
options do not exist. For men, retiring after 40 contribution years is possible with the corridor
pension. In 2010, the penalty is still computed with regard to the entry age for the early oldage pension due to a long period of insurance (“Vorzeitige Alterspension bei langer
Versicherungsdauer”), which was 63 years in 2010; hence
(65-63)*4.2% + (63-62)*2.1% = 10.5%.
From 2020 onwards, parallel accounting applies, where according to the old legislation the
penalty is 6.3% per year and it is 4.2% according to the new legislation. The share of old law
is 60%, the share of new law is 40%, hence
0.6*18.9% + 0.4*12.6% = 16.4%.
For 2030/40, the relative shares are 36/64 and 10/90, respectively. For 2050 and 2060, only
new legislation applies anymore.
For women, in 2010, 40 contributions years were still sufficient to be eligible for the pension
settlement for long-term insured, which allowed them to retire at the age of 55 without
deductions (men would have needed 45 years for retiring at the age of 60). Later, when
women become eligible for corridor pension due to the rise of their statutory retirement age,
penalties are calculated in the same way as for men.
Table 1.a: Early retirement regulations
Early retirement regulations according to the
... old law
Early old-age pension due to a long period of
insurance (“Vorzeitige Alterspension bei langer
Versicherungsdauer”)
Earliest possible
retirement age
Penalty
Required
insurance years
2004: 61½/56½
...
2011: 63½/58½
...
2017: 65/60
currently: 60/55
2016: 62/57
2018: 62/58
2020: 62/59
2022: 62/60
2024: 62/61
2026: 62/62
4.2% p.a. (max.
15%)
37.5
currently: none
as of 2014: 4.2%
p.a. (max. 15%)
45/40
45/42
45/43
45/44
45/45
45/45
45/45
... new law
Corridor pension
62
Hard labour
60
4.2% p.a. (max.
15%)
1.8% p.a.
Pension settlement for long-term insured
(“Langzeitversichertenregelung” or
“Hacklerregelung”)
hard labour: 1.8%
p.a.
37.5
45 (of which 10 y.
in "hard labour")
The public pension system comprises also disability and survivors’ pensions. To be
entitled to a disability pension, a medical certificate is required documenting the invalidity.
Moreover, the status of disability must have prevailed for at least 6 months and one must
apply for rehabilitation before applying for disability pension. The entitlement condition for a
survivor pension is the death of the husband/wife. The deceased must have contributed for a
3
certain period to the public pension system (this depends on the age at which the spouse
died).
Pension benefits are adjusted to consumer price inflation, the assessment bases are linked
to the average insured wage, both with some small discretionary room for manoeuvre for the
government.
If pensions claims are below legally defined thresholds (€ 793 for singles, € 1.190 for couples
in 2011), the gap is closed by federal budget contributions in order to guarantee a minimum
pension (in form of minimum pension allowances, “Ausgleichszulagen”). Currently,
approximately 12% of all pensioners are eligible to such social allowances. The total amount
is about 0.35% of GDP.
The public pension system is financed mainly through compulsory contributions. These are
levied on gross salaries and deducted from these before personal income tax. The present
contribution rates are uniformly set at 22.8% in the private social insurance sector, whereof
the employer bears 12.55% and the employee 10.25%. There are no contributions of the
employers in the civil service sector. There, the employee’s contribution ranges from 12.55%
to 10.25%. The contribution rates of farmers and self-employed amount to 15,25% (2011)
and 17.5% respectively, the former increasing to 16% by 2014. The differences to the
standard contribution rate (22.8%) are borne by federal government transfers. Contributions
are levied up to a maximum assessment base of € 4.200 p.m. for employees (2011) and €
4.900 for farmers and self-employed. The federal budget also covers the deficits in most
public pension schemes in the case of their actual emergence (“Bundesbeitrag”). These
deficits are, thus, financed by general taxation. Contributions to public old-age provisions in
Austria are exempt from taxation, but pension benefits are subject to income tax and health
care contributions.
Generally, private pensions in Austria (both occupational and private) are still of much less
quantitative importance than public pensions. Estimates from 2008 show that private pension
benefits paid in 2007 correspond to less than 5% of overall pension benefits. Nevertheless,
the volumes of private pensions have increased rapidly in recent years.
The Austrian Occupational Pension Act (“Betriebspensionsgesetz”) contains all regulations
for occupational old age provisions (2nd pillar). This Act regulates primarily following firmrelated retirement provisions: 1) pension provision funds (“Pensionskassen”), 2) occupational
collective insurances, 3) direct provisions allowed by a company to an employer and 4) life
insurances.
The implementation of a new severance payment (“Abfertigung neu”) in 2002 increased the
relevance of the second pension pillar, as it made occupational pensions mandatory
(however, disbursement due to occupational change is still possible). Since then, employers
are obliged to transfer 1.53% of the monthly salary of their employees to a staff provision
fund (“Mitarbeitervorsorgekasse”), set up especially for this purpose. In view of old-age
provision, retiring employees can choose to receive the payout in form of the total sum (taxed
4
with 6%), a monthly paid additional pension (tax exempt) or a reinvestment in a pension
investment fund (which is tax exempt).
Private pension provisions made by individuals form the third pillar of the Austrian pension
system. Like in the occupational sector, also in the private sector individuals can choose
between a multiple range of investment products fulfilling directly or indirectly the purpose of
old-age provision. Hence, in the private sector one can generally distinguish between
concrete pension directed provisions and a general accumulation of savings over the lifecycle. Concrete pension directed provisions are aided by the state in order to strengthen the
development of the third pension pillar. Traditionally, life insurances play a significant role in
long-term savings. Private life insurance contracts have continued to show a major upward
trend over the past years. While in general a private life insurance leads to a one-off
payment, private pension insurance contracts are usually concluded for the purpose of
obtaining a life-long pension.
The most popular private old-age provision represents the premium-aided pension savings
scheme (“Zukunftsvorsorge”). This was introduced in 2003 as a kind of life insurance (incl. a
capital guarantee) subsidised by the state with a tax premium. After a minimum investment
period of 10 years, the taxpayer may dispose of his entitlements. If the entitlements are,
however, paid out, half of the allowed state bonuses must be paid back, together with a
retroactive tax of 25% on the capital gains, and the capital guarantee will be lost. If the
entitlements are transferred or used for pension payments, no tax will be due. This scheme
has been recording strong growth since its launch in 2003.
At the end of 2010 more than 1.5 million contracts have been held by insurance companies
and investment firms. Contributions in 2010 were about 900 million Euros und the total of
assets increased to 5 billion Euros.
5
1.2. Recent reforms of the pension system included in the projections
The Austrian Pension Reforms of 2003 and 2005, together with the long-term harmonisation
of the statutory retirement ages of women and men, have already been taken into account in
the previous rounds of long-term public pension projections. Apart from that, the federal
government introduced the following measures in its budgetary consolidation package of
December 2010, which are also included in the current projections:
• Rehabilitation should be intensified in order to keep people in the workforce and reduce
disability pensions. Specifically, a person is now strictly obliged to go for rehabilitation
before applying for disability pension. Furthermore, occupational protection will be lifted
substantially. During rehabilitation, payments are higher than unemployment benefits,
and unemployment benefits would also be paid out longer, if one does not find a job after
rehabilitation.
• The pensionsable age of the “Hacklerregelung” will be increased by 2 years (men: to
62 years, women: to 57 years) as of 2014. By introducing annual deductions, this old law
scheme will be phased out with the new law corridor pension getting fully effective in the
long run.
• On top of that, purchasing study and schooling contributory years has been
rendered more expensive since 2011 and will be fully abolished by 2014. This will make
the current early pension schemes much more difficult to be eligible for.
2. Pensions projection results
The projection results comprise public pension expenditures for private and public sector
employees, self-employed and farmers, which – according to ESSPROS – accounts for more
than 95% of total pension expenditures. They do not include occupational and private
pensions as in the previous projections rounds.
Table 2: Eurostat (ESSPROS) vs. Ageing Working Group definition of pension expenditure (%
GDP)
Eurostat total pension expenditure
Occupational pensions
Eurostat public pension expenditure
Minimum pension allowances (paid by the federal budget as social
assistance)
Injured person's pensions (paid by the accident insurance)
Temporary allowances (paid by the unemployment insurance)
Supplementary survivor pensions
Public pension expenditure (AWG)
2008
13.90%
0.55%
13.35%
0.35%
0.19%
0.04%
0.04%
12.73%
In 2008, a main difference between ESSPROS public pension expenditure data and AWG
(public) expenditure data arose from public expenditures spent by institutions other than the
pension insurance (such as injured person’s pensions, temporary allowances). The other
main discrepancy stem from the exclusion of minimum pension allowances in the 2008 AWG
data set. As noted in chapter I, minimum pension allowances “top up” pension claims that are
under a certain threshold in order to ensure a minimum level pension benefit for singles or
couples. They are financed by the federal budget and they are captured in the current
projection exercise for the first time.
1
2.1. Overview of projection results
Table 3: Projected gross pensions spending and contributions (% of GDP)
Gross public pensions expenditure
Public pensions contributions
2000
13.8
9.1
2010
14.1
8.4
2020
15.1
8.5
2030
16.7
8.5
2040
16.5
8.6
2050
16.4
8.6
2060
16.1
8.6
Peak year
2032
2048
Overall, gross public pension expenditures in Austria are projected to rise from 14.1% of
GDP in the year 2010 to a high of 16.7% in 2032 reflecting the steep rise of the old-age
dependency rate and the phasing-in of reforms at the same time. A moderate decline to
16.1% of GDP in 2060 will follow, since the decline of pension expenditure in the public
2
sector slightly outweighs the increase in the private sector . Although the number of
contributors stagnates or even slightly goes down over time, contributions are projected to
step up somewhat from today 8.4% of GDP to 8.6% in 2060. While the number of
contributors develops broadly in line with employment (cf. table 7), the average contribution
3
per employee advances with the average wage . The slight increase of total contributions
can be ascribed to changes in the structure of contributors. This alludes to a shift from
contributors with smaller assessment bases (farmers, unemployed) to those groups with
relatively higher ones (employees, self-employed).
Table 4: Projected gross public pension spending by scheme (% of GDP)
13.8
14.1
15.1
16.7
16.5
16.4
16.1
Peak
year
2032
Private Sector employees
9.9
10.6
11.7
13.4
14.1
14.8
14.7
2054
Public sector employees
3.9
3.5
3.4
3.3
2.4
1.6
1.4
2010
9.7
10.8
12.4
12.6
12.8
12.9
2055
7.3
8.5
10.2
11.0
11.7
11.8
2055
Pension scheme
Total public pension
of which
Old age and early pensions
Private Sector employees
Public sector employees
Disability pensions
Private Sector employees
Public sector employees
2000
2010
2020
2030
2040
2050
2060
2.3
2.3
2.3
1.6
1.1
1.0
2025
2.4
2.4
2.4
2.2
2.0
1.9
2026
1.8
1.8
1.9
1.8
1.8
1.7
2026
0.6
0.5
0.5
0.4
0.3
0.2
2010
2.0
1.9
1.9
1.8
1.6
1.3
2011
Private Sector employees
1.4
1.3
1.4
1.4
1.3
1.2
2038
Public sector employees
0.6
0.6
0.5
0.4
0.2
0.1
2010
Others (survivors)
“Old age and early pensions” comprise all pensioners aged 62 or above, while “disability pensions”
comprise pensioners under the age of 62.
2
Administrative costs of the pension insurance are not captured in the projections. They sum up to
around 0.2% of GDP.
3
For civil servants, contributions per employee increase more slowly than average wage, since (i) the
contribution rate is going to decrease from today’s 12.25% to 10.25% and (ii) there is going to be a
maximum assessment base according to the new legislation (similar to the private sector), which did
not exist in the old civil servants’ regulations.
2
The projections are classified according to two schemes: the public sector scheme
encompasses all civil servants from the federal, state, municipal levels and other (formerly)
public entities (such as railways, postal services), and the private sector scheme comprises
4
private employees, public employees with private contracts , self-employed and farmers. The
increase of private sector pension spending goes in line with the changing demographic
structure of the population, while accounting for the various pension reforms phasing in,
predominantly until 2034.
For the purpose of the projection exercise, the categories “old age and early pensions” and
“disability pensions” are defined according to an age threshold: ≥ and < 62 years,
respectively. While the functional aspect of the pension obviously is not fully respected
(“disability pensions” are regarded as “old age and early pensions” as soon as the pensioner
reaches 62 years) mainly due to data issues, this definition comes nevertheless rather close
to the current and even more to future actual practice, with (early) pensionable age at 62
years. Against this background, spending for “disability pensions” will even drop slightly in
relation to GDP, while spending for old age pensions surges even until 2055.
In the public sector, the reasons for the sharp expenditure decline are twofold: Firstly, last
decade’s pension reforms will lead to a strong cutback of the average pension entitlement of
civil servants by a substantial amount (in the terms of the benefit ratio more than twice as
much as that of private sector employees). Secondly, the number of civil servants’ pensions
will plummet substantially, to around 60% of the current level. This is partly due to a marked
reduction of the public workforce planned by the governments during the next years, but
more importantly through the substitution of civil servants by public sector employees with
private sector contracts. This has already been practice in public administrations since the
end 90ies. However, the fall of civil servants’ pensions does not compensate for the stronger
rise of private sector pensions, which closely mirrors the change in the demographic
structure of the population.
4
Public sector employees with private sector contracts are captured in the private sector pension
scheme, their share in total public sector employment will increase from 56% today to 69% in 2060.
Put differently, their share in total private sector employment will increase from today’s 10% to 12% in
2060.
3
2.2. Description of main driving forces behind the projection results
By the following arithmetic decomposition pension expenditure dynamics is decomposed into
its various “driving forces”:
Ratio
Ratio
64Dependency
4474
4 48 64 4Coverage
4474
44 4
8
Pension Exp .
Population 65 +
Number of Pensioners
=
×
GDP
Population 20 − 64
Population 65 +
Employment Rate
Benefit Ratio
64 41 /4
474 4 4 48 64 4 4
474 4 44
8
Population 20 − 64
Average Pension
×
×
×
GDP
Working People 20 − 64
HoursWorke d 20 − 74
WorkingPeo ple 20 − 64 HoursWorke d 20 − 64
×
×
HoursWorke d 20 − 64 HoursWorke d 20 − 74
1 4 4 42 4 4 43 1 4 4 4
424 4 44
3
1 / labour in tensity
Re sidual
Table 5: Factors behind the change in public pension expenditures between 2010 and 2060 (in
percentage points of GDP)
Public pensions to
GDP
Dependency ratio effect
Coverage ratio effect
Employment ratio effect
Benefit ratio effect
Labour intensity effect
Residual
201020
202030
203040
204050
205060
201060
Average
annual
change
1.0
1.6
-0.2
0.0
-0.4
2.0
0.040
1.9
-0.1
-0.1
-0.6
0.0
-0.2
4.6
-1.6
-0.1
-0.8
0.0
-0.6
3.2
-1.5
-0.4
-1.3
0.0
-0.2
0.6
0.5
0.1
-1.2
0.0
0.0
0.8
-0.3
0.0
-0.7
0.0
-0.1
11.0
-2.9
-0.6
-4.5
0.1
-1.1
0.211
-0.061
-0.011
-0.092
0.001
-0.009
Overall, public pension spending is expected to increase significantly in 2060, intimately
associated with the strong rise in the dependency ratio of old-aged people to working-age
people. While this ratio amounts to 0.29 today (2 elderly per 7 younger people), it is expected
to almost double to 0.55 until 2060 (4 elderly per 7 younger people). If no other effect offset
these adverse dynamics, pure ageing would boost public pension spending by 11.0 pp of
GDP. However, declining coverage and benefit ratios and improving labour market
participation are projected to limit the increase of pension spending to 2.0 pp of GDP.
The reduction of the coverage ratio goes back predominantly to the enacted legal changes
assuming a marked increase of exit ages from the labour market in the coming decades.
After some reforms stepping in during the subsequent years (e.g. phasing out of old law early
pension opportunities), the harmonisation of the statutory retirement age of women from 60
to 65 years between 2024 and 2033 is expected to have the largest impact on exit ages.
Since this phasing-in is fully implemented by the mid 30ies, the coverage ratio is expected to
remain broadly stable during the last two decades of the projection horizon. The declining
number of pensions in relation to elderly people is also linked to the reduction of the relative
4
share of survivor pensions (currently 24% of all pensions, in 2060 only 18%). This results
from emerging changes in family structures, converging life expectancies of women and men
and fading out of pensions for WW II victims or veterans.
The benefit ratio falls over the whole projection period, which in general is due to the
gradual substitution of the more generous old regulations by the new law (“parallel
accounting”). In particular, this implies an annual accrual rate of 1.78% compared to the 2%
in the old regime and much longer actuarial assessment periods. Beyond that, the adoption
of the private pension scheme for civil servants implies an even stronger decline of average
pensions, since their pension entitlements were overall more generous under the old law.
While the benefit ratio for private sector employees drops by 4 pp until 2060, that of civil
servants’ is presumed to fall by 13 pp.
Labour market developments also help to counteract demographically induced spending, but
they play a more minor role: The employment rate is presumed to increase by almost 3 pp
until 2060 (employment ratio effect).
Table 6: Replacement rates (in % of average insured wage)
Public pension scheme
2010
48
2020
47
2030
46
2040
42
2050
40
2060
37
Private sector employees
45
45
44
42
39
36
Public sector employees
79
69
73
57
57
61
Replacement rates are the ratio between newly awarded pensions (old age, early and disability
pensions) and the economy-wide average insured wage (assessment basis). The average insured
wage is higher for public sector employees than for private sector employees, but it is assumed to
increase at a pace of only 2/3 of that of the private sector (which is - by historical means - even a
generous assumption), implying a convergence of the wage levels in the very long run (beyond 2060).
The reason for this assumption is that life-time earning patterns in the public sector have become
flatter and more similar to those of the private sector, these patterns will fully unfold in the future.
While the benefit ratio measures the relative size of the average pension (including survivors’
pensions) to average wage developments, the replacements rates shown in table 6 indicate
the relative size of the first pension (and, thus, exclude survivors’ pensions). Replacement
rates in the private sector will drop by 9 pp on average over the projection horizon.
Consequently, lower relative first pensions feed through ultimately to the pension stock, with
the respective dampening effects on benefit ratios, as explained above. Apart from parallel
accounting, dampening effects stem also from the increasing share of cross-country
pensions: While only 12% of the current pension stock is transferred to pensioners living
abroad, it is around 17% of new pensions. Since these cross-country pension entitlements
are on average smaller, replacement rates keep decreasing. The replacement rate of civil
servants’ first pensions will even decline twice as much, though its pattern over time is more
volatile. The large drops until 2020 and between 2030 and 2040 can be attributed to
relatively abrupt terminations of transition arrangements.
5
Table 7: Number of pensioners and contributors in the public scheme (in 1000), population
over 65 and total employment (in 1000) and related ratios (%)
Number of pensioners (I)
Number of people aged 65+ (II)
Ratio of (I) and (II)
Number of contributors (III)
Employment (IV)
Ratio of (III) and (IV)
Ratio of (III) and (I)
2005
2 069
1 333
1.6
3 574
3 818
0.9
1.7
2010
2 216
1 478
1.5
3 778
4 124
0.9
1.7
2020
2 553
1 712
1.5
3 901
4 241
0.9
1.5
2030
2 904
2 159
1.3
3 822
4 164
0.9
1.3
2040
3 051
2 485
1.2
3 770
4 108
0.9
1.2
2050
3 224
2 545
1.3
3 707
4 041
0.9
1.1
2060
3 213
2 584
1.2
3 612
3 939
0.9
1.1
In 2010, the number of pensioners amounted to 2,215.614 (45% men, 55% women), of
whom 1,912.504 (86%) received only one pension, 300.713 (almost 14%) received 2
pensions and 2.397 (approx. 0.1%) received three or more pensions. The overwhelming
majority of the “multi-pensioners” (84% of them being women) receive an old-age (81%) or
an invalidity pension (17%) in combination with a survivors’ pension.
From the total sum of 2,535.385 pensions, 1,813.020 (72%) were granted to people aged 65
years upwards. Due to the reforms and the higher life longevity in the future, this share is
expected to go up to 85% by 2060. While still 12% of all pensions are granted to civil
servants today, this share will go down to 5% in 2060. In the public pension sector, the share
of persons aged 65+ years is slightly higher (currently 74%, 91% in 2060), which is probably
due to the fact that the statutory retirement ages for female civil servants are fully
harmonised with men, i.e. the legal age is 65 years already by now.
On the revenue side, the number of contributors will slightly decrease, developing broadly in
line with total employment.
Table 8: Pensions* (all schemes) to inactive population ratio by age group (%)
2010 2020 2030 2040 2050 2060
Age group -54
7
7
6
6
6
4
Age group 55-59
80
69
74
72
69
69
Age group 60-64 118
113
100
112
107
101
Age group 65-69 117
133
116
116
127
125
Age group 70-74 128
130
122
102
118
118
Age group 75+ 135
131
133
127
122
121
* Ratios may increase beyond 100%, since some people receive more than one pension.
With regard to the decomposition in table 8, there are three different patterns underlying
these developments: Firstly, within the age groups -54, 55-59 and 60-64 years both the
number of pensions and the number of inactive people will decline, the former at a faster
pace than the latter. Secondly, this trend is reversed in the age group 65-69 years: Both
indicators show an upward trend, but the number of pensions rises faster than the
corresponding inactive population, this brings about an increasing ratio. This illustrates the
“reshuffling” of pensioners from the former groups (-54, 55-59, 60-64 years) to the latter (6569 years), which is a consequence of increasing retirement ages. The third pattern can be
6
observed in the age groups above 70 (70-74, 75+ years), where both the number of pensions
and the size of the inactive population are growing continuously, but the latter stronger than
the former (e.g. due to the lower share of survivors’ pensions).
Table 10: Projected and disaggregated new public pension expenditure
New pension
I. Projected pension expenditure (m. Euro)
2010
1 716
2020
2 814
2030
3 591
2040
4 467
2050
6 594
2060
8 552
private sector employees
1 370
2 254
3 113
4 216
5 950
7 685
public sector employees
346
560
478
251
644
867
II. Average contributory period
36.0
37.2
37.6
37.5
37.7
37.7
private sector employees
35.7
36.8
37.2
37.4
37.4
37.4
public sector employees
39.4
41.0
42.6
42.1
42.3
42.6
III. Monthly average pensionable earnings
2 475
3 489
4 912
6 974
9 875
13 944
private sector employees
2 403
3 419
4 853
6 892
9 789
13 902
public sector employees
3 399
4 615
5 965
8 515
11 494
14 706
IV. Average accrual rates (in %)
1.33
1.27
1.32
1.13
1.07
0.99
private sector employees
1.25
1.22
1.19
1.12
1.05
0.96
public sector employees
2.01
1.68
1.70
1.36
1.34
1.43
100
118
112
108
118
117
private sector employees
91
105
104
104
111
110
public sector employees
9
13
8
4
7
7
14
14
14
14
14
14
1 663
2 724
3 544
4 471
6 556
8 530
private sector employees
1 370
2 254
3 113
4 216
5 950
7 685
public sector employees
346
560
478
251
644
867
53
90
47
-5
38
22
private sector employees
0
0
0
0
0
0
public sector employees
0
0
0
0
0
0
V. Sustainability / Adjustment factor
VI. Number of new pensions (in 1.000)
VII. Average number of months paid the
first year
VIII. Product of II – VII (m. Euro)*
I-VIII
* Differences may arise due to rounding errors
Table 10 shows a breakdown of the projected new pension expenditure. Unlike the
development on the civil servants’ side, spending on new pensions in the private sector
moves rather smoothly. In the civil service, regular recruiting will resume in the coming years
after some years of stagnation. Furthermore, the number of new pensions will peak around
2020 and will decline steadily afterwards. Once the number of new pensions will have
reached its low at around 2040, it will start to rise again.
7
2.4. Sensitivity tests
Given the uncertainty surrounding the assumptions of long-run projections, it is necessary to
carry out a number of sensitivity tests so as to quantify the responsiveness of projection
results to changes in key underlying assumptions. Pension expenditure dynamics could be
dampened by higher labour productivity growth and higher employment rates, especially,
higher employment rates of older workers. On the other hand, additional pressures will be
5
put on pension expenditures assuming a higher life expectancy and lower migration.
Table 12: Public and total pensions expenditures under different scenarios (deviation from the
baseline)
Public pensions expenditure
Baseline
Higher life expectancy
Higher lab. productivity (+0.1 pp.)
Lower lab. productivity (-0.1 pp.)
Higher interest rate (+0.5 pp.)
Lower interest rate (-0.5 pp.)
Higher emp. rate (+1 pp.)
Higher emp. rate of older workers (+5 pp.)
Lower migration (-10%)
2010
14.1
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
2020
15.1
0.0
0.0
0.0
0.0
0.0
-0.1
-0.2
0.0
2030
16.7
0.0
-0.1
+0.1
0.0
0.0
-0.4
-0.7
+0.1
2040
16.5
+0.1
-0.2
+0.1
0.0
0.0
-0.5
-0.9
+0.2
2050
16.4
+0.2
-0.2
+0.2
0.0
0.0
-0.6
-0.9
+0.4
2060
16.1
+0.2
-0.3
+0.1
0.0
0.0
-0.7
-0.7
+0.4
As already shown before, pension spending is projected to increase from 14.1% today to
16.1% of GDP in 2060 in the baseline scenario. The annual average GDP increase amounts
to 1.42% in real terms, while annual average real pension spending increases by 1.69%.
Higher life expectancy: A scenario with an increase of life expectancy of one year
by 2060 compared with the baseline projection.
The assumption of higher life expectancy will raise expenditure in 2060 by 0.2 pp
of GDP. The increase develops gradually in line with the higher longevity, annual
average real GDP growth is the same as in the baseline scenario, the corresponding
figure for pension spending rises to 1.72%. Higher life expectancies leaves average
pensions unchanged, but increases the years spent in retirement and, hence, the
number of pensions.
Higher/Lower labour productivity: Labour productivity growth is assumed to
converge to a productivity growth rate which is 0.1 pp higher/lower than in the
baseline scenario. The increase/decrease is introduced linearly during the period
2016-2025, and remains 0.1 pp above/below the baseline thereafter.
If the average productivity growth will increase/decrease by 0.1 pp, public pension
expenditure will be reduced/raised by 0,3/0,1 pp of GDP in 2060, respectively. Annual
average real GDP growth will go up/down to 1.50/1.34% (i.e. ±0.08pp compared to
5
As interest rates do not influence the projections for the social security pension expenditures in
Austria, the scenario of „higher/lower interest rate“ is irrelevant for the Austrian case.
8
the baseline), while annual average real pensions spending amounts to 1.74/1.62%
(+0.05/-0.07pp). The number of pensions is almost unchanged compared to the
baseline, the average pension spending per pensioner changes: With gross
replacement rates being equal or even slightly above/below with higher/lower
productivity, benefit ratios decrease/increase. This is owed to the fact that the
assessment base of first pensions is adjusted to wage developments (hence,
“neutral” to productivity shocks), whereas the benefit ratio reacts inversely to wageprice differentials (due to inflation adjustment of existing pensions), the latter
rising/falling in the case of higher/lower productivity.
Higher employment rate: A scenario with the employment rate being 1 pp higher
compared with the baseline projection. The increase is introduced linearly over the
period 2016-2025 and remains 1 pp higher thereafter. The higher employment rate is
assumed to be achieved by lowering the rate of structural unemployment (NAIRU).
A rise in the employment rate compared to the baseline scenario is projected to
cause a reduction of pension expenditures by 0.7 pp at the end of the projection
period. Considering annual average real GDP and pension spending growth rates, ¼
of the overall dampening effect stems from the denominator effect (enhanced GDP
growth, 1.44% on average) and ¾ is due to the numerator effect (decreased pension
spending, 1.63% on average). The denominator effect comes from higher
employment raising (potential) output. The numerator effect is due to the reduction in
the number of pensions, slightly compensated by higher pension entitlements.
Higher employment rate older workers: A scenario with the employment rate of
older workers (55-64 years) being 5 pp higher compared with the baseline projection.
The increase is introduced linearly over the period 2016-2025 and remains 5 pp
higher thereafter. The higher employment rate of this group of workers is assumed to
be achieved through a reduction of the inactive population.
An increase of the employment rate of the elderly (55-64 years) in relation to the
baseline scenario will result in lower pension expenditures by 0.7 pp in 2060.
However, other than in the previous scenario, spending reductions are higher in
former years from 2030 onwards, where they peak at 0.9 pp. If higher employment is
concentrated among the older workers, the impact on the number of pensions is
stronger and occurs earlier.
Lower migration: A scenario with 10% less migration compared to the baseline
projection.
If migration is 10% lower than in the baseline, pension spending to GDP will
increase by 0.4 pp in 2060. This result draws heavily on the migrant population to
become, to a large extent, part of the working age population until 2060, therefore
only a small fraction adds to the retirees. Hence, lower migration decreases
employment and output, whereas pension spending fall only marginally (i.e. therefore
the “denominator effect” dominates the “numerator effect”).
9
2.5. Description of the changes in comparison with the 2006 and 2009
projections
Table 13: Change in public pension expenditure to GDP during the projection period under the
6
2001, 2006, 2009 and 2012 projection exercise
2001 (2001-2050)
2006 (2005-2050)
2009 (2007-2060)
NEW (2010-2060)
Public
pension
to GDP
Dependency
ratio
Coverage
ratio
Employment
effect
2.4
-1.0
0.9
2.0
10.5
11.3
9.9
11.0
-3.0
-5.8
-2.6
-2.9
-2.2
-1.3
-0.5
-0.6
Benefit
ratio+ labour
intensity
effect
-2.9
-4.3
-5.0
-4.4
Residual
0.1
-0.8
-1.0
-1.1
Table 13 presents the effects of the different underlying factors on total pensions spending
across the current and previous projection exercises. However, one has to be cautious with
comparing the differences over time due to the varying magnitudes of the residuals.
The demographic outlook has worsened since the last round, i.e. the dependency ratio has
stepped up. The employment effect is rather stable compared to 2009, the expected drop of
the coverage ratio is somewhat higher, the decline of the benefit ratio somewhat lower than
in the previous projection. The changes of the coverage and benefit ratios stem from
changed assumptions about the future development of cross-country pensions: In the 2009
projections the share of (on average smaller) newly awarded cross-country pensions was
supposed to be around 20% of all new pensions, in the current projections it is assumed to
be only around 17%. This means that there are presumed to be fewer, but on average larger
pension entitlements than projected in 2009.
But most importantly, GDP assumptions are significantly lower (1.4% on an annual average)
compared to 2009 (1.6%). Adding the worse demographic outlook (the work force is
supposed to decline faster than assumed in 2009), this explains the increase of projected
pension spending in % of GDP.
Table 14: Decomposition of the difference between 2009 and the new public pension projection
(% of GDP)
Ageing Report 2009
Change in assumptions
Improvement in the coverage
Change in the interpretation of
constant policy
Policy related changes
New projection
2000
13.4
2005
13.2
0.3
2010
12.7
1.0
0.4
2020
13.0
1.7
0.4
2030
13.8
2.5
0.4
2040
13.9
2.2
0.4
2050
14.0
2.0
0.4
2060
13.6
2.1
0.4
0.4
13.8
13.5
14.1
15.1
16.7
16.5
16.4
16.1
6
The new labour intensity effect is added to the benefit ratio, in order to be comparable to former
calculations.
10
3.
Description of the pension projection model and its
base data
3.1. Institutional framework for long-term pension projections
Traditionally, medium-term pension projections, covering at least five future years, are
contained in the yearly report submitted by the Austrian Pension Commission (APC) to the
federal government in preparation of annual pension adjustments. This consultative body
represents the main forum for periodic policy discussions. It is composed of experts,
academics, government and social partner representatives.
Initially, these medium-term projections, which are limited to the private social insurance
schemes ("gesetzliche Sozialversicherung"), have been the central policy instrument for
assessing pension developments. However, the tendency towards a more frequent use of
quantitative analyses and external advice was intensified during past reform efforts. As a
result, long-term pension projections based on demographics by Statistics Austria were
presented as a complementary tool to clarify the need for adjustment and to assess the
impacts of the major past reform efforts initiated by the federal government. This has proven
to be a very helpful and transparent instrument. This is why, with the aim to have long-term
pension projections constantly available and to ensure long-term financial sustainability of
the Austrian pension system, the federal government set up a permanent monitoring
mechanism as of 2007. The APC reviews financial developments in the pension system
every three years and in particular with regard to the sustainability factor newly established in
2005. This sustainability factor does not operate automatically. The analysis of the financial
sustainability of the Austrian pension system by the APC is based on recent demographic
projections of Statistics Austria, in particular projections of life expectancy at the age of 65. If
life expectancy exceeds the reference value as defined in the law by more than 3% the
committee is obliged to put forward respective proposals to offset potentially higher pension
expenditures (e.g. through changes in the contribution rate, retirement age, benefit
adjustment). The APC concluded work on the sustainability factor the first time in April 2011
with a set of recommendations to the federal government. The APC will now puts its
emphasis on the monitoring of the implementation of measures to increasing actually the
effective retirement age during the next decade. However, a new and more effective
monitoring mechanism together with a reform of the APC is also foreseen by the coalition
agreement.
11
3.2. Assumptions and methodologies applied
The Austrian pension projections within the given EU framework are based on two
autonomous models, covering the private social insurance sector and the civil service
schemes, respectively. They include all benefits and contributions to old-age, earlyretirement, disability and survivor schemes. The pension projections, therefore, include all
public pension expenditure and , for the first time, also cover additional social assistance
benefits (minimum pensions allowances). The pension projections contain the effects of all
existing major pension reforms.
Both models consist of partial equilibrium models and comprise deterministic elements only.
In order to achieve consistency in the results, the two basic models for the private social
insurance and the civil service sectors are consolidated, both as to macroeconomic
developments and to expected shifts of contributors from one to the other category of
schemes. For instance, the developments in civil service sector employment are captured by
the private social insurance sector model; vice versa the macro scenario of the private social
insurance sector schemes forms an important input into the civil service projections. Hence,
though the two models are autonomous, they have been made fully consistent with regard to
employment and wage developments.
The private social insurance sector model, accounting for nearly three quarters of total
public pension expenditure is central to simulate the financial effects of population ageing. It
covers all relevant social insurance schemes, for blue and white collar employees (ASVG)
incl. public employees with private contracts (“Vertragsbedienstete”), self-employed and
farmers, among others. The model is composed of two major blocs that are intimately linked
together. The macro part is made up of ten modules, reflecting economic, labour market,
public finance and pension insurance developments. In effect, most single parameters are
endogenously determined with the exception of participation and inflation rates, which fit in
as exogenous inputs. The pension-specific micro part relies on inputs from the macro side on
employment and on the payroll, from demographics and from age-related time series
describing past pension contributions and benefits. These micro modules are designed so as
to incorporate already enacted reforms with their effects in the near and distant future and to
simulate reform options. These pension modules permit to calculate the bulk of already
existing pensions, the number of new pensions and of exits, average pension benefits and
replacement rates as well as aggregate figures in a given (future) year. On the other hand,
pension contribution rates and the level of the social insurance pension deficit covered by the
federal budget feed back into the macro modules.
Secondly, the civil service model takes into proper consideration the fact that these pension
benefits are fully financed out of the federal, Länder and the various communal budgets. The
federal sector clearly dominates by size. In this vein, the federal segment comprises all
pension and survivors’ benefit payments to civil service retirees of the federal government,
the postal, telecom and railway services and specific groups of regional governments, such
as primary and secondary school teachers. However, the model also takes account of all
vital developments at the other government levels.
12
Ongoing structural reforms in the civil service sector aim at enhanced application of
private-sector-based labour contracts for employees in the public sector
(“Vertragsbedienstete”). As a general trend, civil service developments are assumed to be
much more exposed to the present age-structure in the civil service and the future internal
reforms rather than to demographics and economic developments, which are nonetheless
taken into adequate consideration. These reform measures will dwell upon the
comprehensive efforts to harmonise private social insurance and civil service sector pension
systems, raising effective retirement ages and contribution rates as well as pursuing
restrictive recruitment in the civil service sector in general and into the civil service status in
particular.
13
3.3. Data input
3.3.1. Demographic developments
According to the EUROPOP2010 population projection, the Austrian population is expected
to increase from 8.4m persons today to a peak of 9.0m in 2043, before it starts to decline
again to a level of 8.9m by 2060. The overall size of the Austrian population is projected to
be larger by about 480.000 inhabitants by then, but also much older than it is now. According
to the projections, the working-age population (aged 15-64) will continue to expand modestly
from 5.67m to 5.70m people until 2019, then it will go down to a level of 5.1m by 2060
(despite continuous positive net immigration of 30.000 people on average). Over the whole
projection period, the total labour supply will drop by 5%. The young population (aged 0-14)
will decline by 3% over the projection horizon, while the elderly population (aged 65 and
above) will increase by 75%, rising from 1.5m today to 2.6m in 2060. The very old population
(80+ years) is projected to rise even stronger, from about 405.000 today to over 1 million by
2060.
The old-age dependency ratio (the ratio of persons 65+ years in relation to the age cohort
15-64 years) almost doubles from 26% at present to 51% in 2060 due to the baby-boom
generation reaching the retirement age and life expectancy increasing by more than 6.5
years. The economic dependency ratio (i.e. the ratio of people not in the labour force to the
people in the labour force) will step up from 97% to 125%, as the fall in the young population
will not compensate for the much stronger rise of older people. The convergence scenario
approach employed in the EUROPOP2010 projection by Eurostat assumes a process of
convergence in key demographic determinants (fertility rates, mortality rates/life expectancy,
migration) across Member States to that of the forerunners over the very long-term (by the
year 2150). For Austria, in consequence, the total fertility rate is projected to rise from 1.39 in
2010 to 1.46 by 2030 and further to 1.56 by 2060. Life expectancy at birth of men is
projected to increase by 7.2 years over the projection period, from 77.6 in 2010 to 84.8 in
2060. Life expectancy of women is expected to go up by 6.1 years, from 83.0 in 2010 to 89.1
in 2060. Annual net migration inflows are projected to increase from about 19,000 people in
2010 to 35,000 by 2020 and then decline again to 26,000 people by 2060. In an additional
calculated “lower migration scenario” the assumption of 10% less migration would lead to a
drop of the working-age population (15-64) of 5.7m today to 4.9m by 2060, which is
approximately 170.000 less than in the baseline scenario.
3.3.2. Labour force and employment developments
The labour force projection (based on a cohort component methodology initially defined by
the OECD) shows the outcome for the labour force by extrapolating recent trends in rates of
entry to and exit from the labour market. This base case projection reflects the working
assumption of “no policy change” and, therefore, does not account for more or less likely
future developments. Since the common macroeconomic projections of the Ageing Working
Group already account for the pension reforms of the last years, in particular the effects of
14
raising and harmonising legal retirement ages and enhancing financial incentives to remain
longer at work, their effects on employment are reflected by the Commission macro
assumptions, accordingly.
The labour force over the next 50 years is projected by combining the projections of
population and of rates of participation by gender/age group (based on the EU labour force
concept). The overall participation rate (for the age group 15 to 64) in Austria is anticipated to
increase by 2.5 pp over the period 2010-2060 (from 75.0% in 2010 to 77.6% in 2060). The
projected upward shift in the overall participation rate is mainly due to the increase of
participation rates for women and the elderly. While the participation rate for men is
presumed to decrease by 1.0 percentage point over the projection horizon (from 80.8% in
2010 to 79.7% in 2060), the participation rate for women will be boosted by 6.0 pp (from
69.3% in 2010 to 75.3% in 2060). For the total age-group 15-74, the projected participation
rate oscillates around the current rate of 66.7%. Apparently, due to the enacted pension
reforms, the biggest raise in participation rates is projected for older workers (55-64 years);
around 21.4 pp for females and 4.0 pp for males within the projection horizon. Yet, the
relative improvements on the labour market cannot outweigh the changes of the overall
demographic structure of the population, so that the total labour force (aged 15 to 74) is
projected to drop by 5% from 2010 to 2060, whereby the female labour supply drops by 1.1%
and the male labour supply decreases by 8.3% within the projection horizon.
Unemployment rates are expected to converge to the estimated NAIRU in 2015, based on
the Spring 2011 economic forecasts by the European Commission and the medium term
extension of the production function method, afterwards they converge to the country specific
minimum (capped at 7.3% for countries with high rates). For Austria, these assumptions
imply an initial unemployment rate of 4.5% in 2010, decreasing to 4.1% until 2015 and
staying at this level thereafter. Given the population projection, the unemployment rate
assumptions and the labour force projection, the overall employment rate (of people aged 15
to 64) in Austria is projected to increase from 71.7% in 2010 to 72.8% in 2020, and to reach
74.4% in 2060.
The old-age employment rate (55-64) is expected to increase from the initially low level of
around 42% at present to 55% in 2060. This relates to the effective increase of the retirement
age by about 2.1 years for women (from 60.2 to 62.3 years) and 1.3 years for men (from 61.3
to 62.5 years) over the projection horizon. Women's employment rate (15-64) is expected to
rise by 6.1 pp from 66.3% in 2010 to 72.4% in 2060. The expected increase in overall
employment rates is assumed to result in a further slight employment growth in the period up
to 2018. Then employment will start to decline by around 0.2% per year on average until
2060, thereby steadily contributing negatively to potential GDP growth.
3.3.3. Long run growth
As in previous projection rounds, a production function approach for projecting potential
output growth has been applied.
15
The annual average GDP growth rate in Austria is projected to decline from 2.0% in 2010 to
1.3% in 2060. Over the whole period 2010-2060, real GDP growth rates in Austria comply
with those in the EU-27 area, amounting to 1.4% on average. Driving factors of GDP growth
are labour input and labour productivity, whereby economic growth up to 2060 is strongly
influenced by a shrinking labour supply. Labour input in Austria is projected to increase up to
the 2020s. Thereafter, the changes of the demographic structure act as a drag on growth.
Henceforth, labour productivity will be the sole source of economic growth. For Austria,
labour productivity growth (which is based on assumptions about total factor productivity
growth and capital stock developments) is projected to remain fairly stable throughout the
projection period close to 1.5%.
16