Transcript reforming social security: what can indonesia learn

REFORMING SOCIAL
SECURITY: WHAT CAN
INDONESIA LEARN FROM
OTHER COUNTRIES?
by Estelle James
Prepared for USAID workshop on social
security, Jakarta
1
Population aging and its impact
on public spending
• Proportion of world’s people > 60 will increase
from 9% to 16% by 2030
• Indonesia’s elderly will increase: from 7% to 16%
• Public spending on health and pensions increases
as population ages, so spending on health and
pensions will > 10% of GDP in Indonesia by 2030
• Young workers today will be 60+ in 2030, so
important to set good old age security system now
• This will have a major impact on economy—
quantity and productivity of labor and capital
• Many countries are reforming their social security
systems to get better economic and equity effects 2
Public Health and Pension Spending
versus Population Aging
Spending as a percentage of GDP
Austria
20
Czechoslovakia
Sweden
Poland
15
U.K.
New Zealand
Iceland
Canada
Switzerland
Spending on health and pension
10
Brazil
Trinidad &
Tobago
5
Japan
Australia
Cyprus
Spending on health
China
Jamaica
Swaziland
Zambia
S. Korea
Indonesia
0
0
5
10
15
20
25
Percentage of population over 60 years old
3
Major lessons
• Don’t rush into DB scheme without careful
actuarial analysis of long term costs—otherwise
Finance Ministry will be faced with huge pension
debt (already the case for civil service)
• Pre-funding important but don’t increase it until
you have a competitive structure that will:
– earn a high rate of return with low administrative costs
– Diversify your investment portfolio to raise return and
reduce risk
• Informal sector cannot be covered by contributory
scheme—requires finance from general treasury
4
Traditional schemes
• In past, many countries had pay-as-you-go
(PAYG) defined benefit (DB) schemes
– DB—pension based on worker’s wage and years of
contributions
– PAYG—contributions by workers (or govt) today used
to pay pensions of retirees today, no investments
– Indonesia has this in civil service but not private sector
• Some countries had defined contribution (DC)
plans—workers get back contributions + interest
(no DB). Funds accumulate but invested by public
agency (Jamsostek)—political control, low returns
• Both these plans had big problems. Recently,
many countries have switched to DC plans with
5
funds under private competitive management
PAYG schemes worked well when systems
were new and no retirees, but require large
contribution rate increases now
• Why higher contribution rate?
– Workers evade contributions so less revenue
– Workers retire early so collect benefits for more years
– Population ages so fewer workers, more pensioners
• Problems caused by high contribution rate
– Less take-home pay if worker pays
– Higher labor costs, less employment if employer pays
– Fewer workers in formal labor force (evasion, early
retirement) so less economic growth
6
Other problems caused by PAYG
DB schemes
• Redistributes to high earners (who live longer) and
first covered generations (high benefits, low
contributions)—these are better-off groups, not
poorest groups
• Government owes workers large unfunded
pension debt in return for contributions (>100%
GDP)--paying this debt may reduce public
spending on health and education
• May discourage saving, investment, economic
growth
7
Implicit Public Pension Debt, 1990
Explicit debt
Canada
Implicit public pension debt
France
Germany
Italy
Japan
United States
0
50
100
150
200
250
300
Percentage of GDP
8
Potential effects of new DB in
Indonesia
• Evasion, early retirement, low returns already
• Government would accumulate large pension debt
(which it can’t afford)
• Higher contribution rate would reduce wage for
middle-income workers, raise labor costs so fewer
jobs for minimum-wage workers—bad for growth
• If PAYG, 7% contribution rate could finance av.
pension that is only 10% of average wage in 2030
• Civil service already faces DB problems—
pensions will cost over 30% of payroll by 2030
• In private sector Indonesia has avoided problems 9
of PAYG DB so far and should continue to do so
How have countries reformed?
• Over 30 countries in Latin America, Europe, AsiaPacific have adopted multi-pillar system to avoid
these problems
• One pillar handles peoples retirement savings
– Shift to DC plan (not primarily DB)
– Shift to pre-funding rather than PAYG
– Funds are privately managed to avoid political control
• Another pillar provides safety net (redistribution)
• A third pillar is voluntary retirement savings
• Chile was first country to reform in 1981. So far:
large assets, increased saving, greater formal labor
supply, financial market development, economic
10
growth, poor are protected
The Pillars of Old Age Income Security
Objectives
Redistribution
plus
Co-insurance
Savings
plus
Co-insurance
Savings
plus
Co-insurance
Form
Flat or
Means-tested
or
Minimum
pension
guarantee
Personal
savings plan
or
Occupational
plan
Personal
savings plan
or
Occupational
plan
Financing
Tax financed
Regulated
Fully funded
Fully funded
Mandatory
Mandatory
publiclyprivatelymanaged pillar managed pillar
Voluntary
pillar
11
80
Ho ng Ko ng
70
60
In millions
50
U nit e d King d o m
40
30
Hung a ry
El S a lv a d o r
Ka z a khs t a n
P o la nd
B o liv ia
M e x ic o
A rg e nt ina
A us t ra lia
U rug ua y
C o lo mb ia
D e nma rk
P e ru
S w it z e rla nd
N e t he rla nd s
20
C hile
10
0
1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000
12
Commonalities
• Why DC?
– prevents implicit pension debt (no benefit promise)
– avoids perverse redistributions, incentives for evasion
and early retirement (penalty for early retirement)
– keeps system solvent even if workers retire early, evade
• Why fully funded?
– avoids steep future payroll tax increases as population
ages--pensions financed by own savings, not current tax
– can be used to mobilize long term domestic saving for
investment, increase productivity and output
• Why private management of funds?
– To get higher return and better allocation of capital
– improve financial markets
13
– reduce risk by diversification (including international)
Pre-funding avoids large
contribution increases, but works
well only with high rate of return.
• Monopolistic public funds have low rates of
return, inefficient allocation of capital, evasion
– Public managers required to invest in government
bonds, failing state enterprises
– Singapore CPF earned < 2% real; Jamsostek lost
money--earned 38% < inflation 1978-2000
– To get a reasonable replacement pension, rate of return
must be 2-3% higher than wage growth
– To help economy, savings should be used productively
• So reforming countries are using private
competitive markets, diversified portfolios
(including international diversification)
14
Most publicly managed funds
earned less than bank deposit rate
-1.8%
Average
Uganda
Zambia
Venezuela
Egypt
Ecuador
Sri Lanka
Guatemala
Kenya
Jamaica
Canada
Singapore
Morocco
Costa Rica
India
Malaysia
US
Sweden
Philippines
Korea
Japan
-12%
-10%
-8%
-6%
-4%
-2%
0%
gross returns minus bank deposit rate
2%
4%
15
Private pension plans have
achieved high returns
Average private schemes
Average public schemes
Switzerland (70-90)
Japan (70-87)
United States (70-90)
Canada (75-89)
Denmark (70-88)
Hong Kong (83-96)
Netherlands (70-90)
Japan (84-93)
Switzerland (84-96)
Denmark (84-96)
Australia (87-94)
United Kingdom (70-90)
Spain (84-93)
Netherlands (84-96)
Ireland (84-96)
Chile (81-96)
Belgium (84-96)
United States (84-96)
Sweden (84-93)
United Kingdom (84-96)
-10% -8%
-6%
-4%
-2%
0%
2%
4%
6%
Gross returns minus income per capita growth
8%
10%
16
Importance of high rate of return:
pension/final wage from 7% contribution
(also important to raise retirement age)
(real wages assumed to grow 3% yearly)
Real rate of return
Retirement age
r=2%
r=5%
55 (contribute 35 years)
13% replacement rate
28% replacement rate
65 (contribute 45 years)
29% replacement rate
64% replacement rate
17
Key differences among new systems—
how investment managers are chosen
• Latin American model (retail market)-workers
choose investment managers for their own
accounts—choice, but expensive
• OECD model (group market)—employers and/or
unions choose investment managers for entire
company or occupation—lower admin. costs but
agency problems (what’s best for employers may
not be best for workers)
– In UK employers can opt out of state plan and
individuals can opt out of state or employer’s plan—
maximum choice, but may be confusing
– Employer opt-out is possibility for Indonesia
18
Institutional market
• Government aggregates many small accounts into
large money blocs and negotiates group rate
through competitive bidding process (Bolivia,
Kosovo, US Thrift Saving Plan for federal govt
employees)
• Workers are given limited choice among managers
and portfolios
• Designed to keep costs low, prevent big mistakes,
get good returns
• This might work well in Indonesia with
redesigned JAMSOSTEK one part of
19
competitive scheme
What are other countries in AsiaPacific region doing?
• Singapore is allowing workers to opt out of CPF
into personal accounts (retail)
• Australia and Hong Kong require employer to
choose investment manager (group plan)
• India is requiring all new civil servants to enter
DC plan; workers choose among small group of
investment managers selected in competitive
bidding process (institutional model); existing
workers are grandfathered into old scheme.
– This could be the basis for civil service pension
reform in Indonesia.
20
New experiments with centrally
managed funds
• Canada, Ireland, New Zealand—keep DB but fund
• Most portfolios managed externally with
competitive bidding, explicit selection and
monitoring criteria, strict disclosure
• Board chosen on basis of professional
competence, not as representatives
• Object is to maximize returns with moderate risk;
no social or targeted investments (ltd govt bonds);
• Foreign investments and passive investments
dominate (reduce risk and costs)
• This would not work here, because Indonesia
21
would probably not adopt such strict restrictions.
What to do about poor workers
and informal sector?
• Can’t collect contribution from informal
sector, need non-contributory plan
• Almost all reforming countries include a
safety net financed out of public treasury—
could be minimum pension guarantee
(MPG), means-tested benefit, flat benefit or
compressed earnings-related benefit
22
The safety net
• Minimum pension guarantee (Latin America)
– pension topped up if < 25% average wage
– only for contributors (20-25 years)
– Chile has social protection for non-contributors
• Means-and asset-tested benefits (Australia, Hong
Kong, South Africa)—non-contributory
–
–
–
–
–
takes account of all income
may discourage work and saving
high transactions costs, mistargeting and bribery—leakage
hard to apply in extended family context
probably means-tested safety net is not a good plan for
Indonesia
23
Safety net: Flat benefits (Netherlands, New
Zealand, UK, Mauritius, Namibia, Botswana, Kosovo, Nepal)
• For all residents, financed by general revenues,
covers informal sector and women
• Costs more than means-tested benefits because
reaches more people
• But cheaper to administer, less corruption, doesn’t
discourage work, saving—better for countries with
low administrative capacity
• Should all old people get a small flat benefit?
Depends on whether households with old people
are relatively poor and other budgetary priorities
• Flat benefit of 20% per capita income (120,000 R)
to people > 70 would cost Indonesia .4% GDP.But
24
improving health services may be more important.
Conclusions—policy
recommendations
• Over 30 countries trying to avoid problems
of high costs, large deficits, early
retirement, evasion are moving toward
systems that use
– Funded DC plan with competitive management
– Safety net to protect lowest earners
– Avoid large PAYG DB and publicly managed
funds
25
Key questions for Indonesia
• Should new DB plan be introduced? (risky for
government, expensive for workers and
employers. If DB is used it should be small,
compressed, contain penalty for early retirement)
• What is best way to manage investment of funds?
How to avoid past problems in Jamsostek and
Taspen? (competitive management with limited
choice and diversified portfolios?)
• What kind of safety net? If wish to cover informal
sector must be financed by public treasury, not be
contributory scheme. Small flat benefit for very
old possible but depends on budgetary priorities
and whether hh with old members are poor hh. 26
(Health services may be more important).
Next steps
• Important to simulate very long run costs and
distributional effects of different policies, with
alternative assumptions, before acting
• New draft act is ambiguous on many points—
good, because careful analysis needed. Gives
Indonesia opportunity to reform its current system
and move it on a path toward greater economic
growth and old age security
27