Transcript (When) Should Coverage of Social Security Systems

(When) Should Coverage of
Formal Social Security Systems
Expand?
by
Estelle James
Aim is to expand coverage.
Important question: at what rate?
• Old age security traditionally provided by
informal family system: parents support
(invest in) children when young and
children support parents (return to
investment) when old
• Intergenerational households common
• Formal system needed because of failures
of family system
Advantages of family system
• No transactions, bureaucratic or tax costs
• Arrangements flexible, can vary by family
– Consumption level when old v. young and
retirement age set to maximize family utility
• Social pressures induce compliance
Family system never perfect and
is now breaking down further
• Some children don’t support parents, don’t
pay back investment
• Some children too poor to support parents
well--redistribution needed
• Movement to cities decreases social
pressures raises for compliance, raises cost
of intergenerational households
• Fertility rate declining so cost per child of
parental support is increasing
Individual action can’t always substitute
• People may save for their own old age; but
myopia, high discount rate for low earners,
makes this difficult. E.g. in Thailand over
last decade, household saving has fallen
• People may underestimate life expectancy
• Moral hazard problem--parents may expect
government to step in if they become poor,
so they don’t use appropriate safeguards
• Reliable savings/annuity tools not available
So, pressure for government
intervention
• To fill in market gaps (annuities, savings)-people would like to save but can’t
• To protect people from their own myopia
when family system breaks down--people
don’t realize the need to save
• To avoid moral hazard and social problem
of old age poverty by forced saving and by
redistributing to households whose lifetime
incomes are too low to save enough
But--problems with government
intervention
• Programs may not be sustainable
• Inflexible, doesn’t vary with family needs:
– Replacement rate may be too high for some--shifts
too much consumption from youth to old age
– Retirement age may be too early for some--costs too
much in terms of decreased consumption
– Some individuals have lower expected lifetimes,
need less savings
– Retirement savings illiquid, time pattern of benefits
fixed, unavailable for other needs
• May crowd out family transfers
• So old people may not be better off
Problems--Cont’d
• Also, coverage expansion may not be good for
society or economy if:
– Redistributions are not equitable
– Decreases personal savings
– Distorts incentives to work and employ
• Government may not have capacity to enforce
compliance if workers and employers don’t
want to pay and culture of evasion develops
Where does Thailand stand?
• So far, Thailand has covered large
enterprises and public sector employees-about 1/4 of labor force.
• Must decide whether government has
capacity to enforce collections in small
enterprises where labor is transient.
• And whether these workers (low earners),
would be better off with mandated coverage
When should coverage expand?
• Coverage should expand when
– it makes workers and economy better off
– government has capacity to enforce
compliance
– system is sustainable
• Test system to see if it meets these
criteria before expanding coverage
• Tension between expanding too fast v.
not fast enough
Sustainability criteria
• If new workers come into unsustainable
system, their benefits aren’t safe
• And they make system more unsustainable
• In Thailand new DB system isn’t
sustainable but IPD is low. If coverage
expands IPD suddenly increases and reform
becomes more difficult
• Reform system to make it sustainable
before expanding coverage
Criteria re worker welfare and
government capacity: Coverage
closely tied to per capita income
• When income is low, coverage in contributory
programs often is not desirable
• And governments in low income countries
may not have capacity to compel compliance
• Thailand is close to the fitted line--about 1/4
of the labor force is covered
• I will discuss special situations of low
Figure 1. Relation between Coverage and Income per Capita in 1990s
1990s
Contributors/Labor Force %
100
90
80
70
60
50
y = -2E-07x 2 + 0.0089x
R2 = 0.8476
40
30
20
10
0
0
5000
10000
15000
Income per capita
Source: World Bank staff analysis
20000
25000
30000
Contributory programs often not
desirable for low earners
• They have short time horizon, high discount
rate; current consumption, liquidity vital so tax
or forced saving hurts
• Other investments more important-- education,
health, house, tools; and borrowing rate is high
• Life expectancy is less for low earners. They
will lose if put into the same DB system as high
earners--this redistributes from poor to rich
• Family system may still work for them and is
more flexible, lower transactions costs; but
public system may crowd out family transfers
• So coverage may not make low earners better off
Governments often don’t have
enforcement capacity
• Difficult to collect taxes when incentive to
evade is strong
• Especially difficult with transient workers,
small employers, self-employed,
agriculture--easy to escape to informal
sector. These predominate in Thailand
• Culture of evasion develops--harmful
• As economy grows, structural changes
facilitate enforcement
Governments often don’t have
enforcement capacity
• Difficult to collect taxes when incentive to
evade is strong
• Especially difficult with transient workers,
small employers, self-employed, agriculture
• Easy to escape to informal sector
• Evasion contagious; culture of evasion bad
• As economy develops, structural changes
facilitate enforcement, decrease evasion
How to make contributory schemes
more attractive for low earners
• make system sustainable
• keep contribution rate low
• offer good rate of return
• give low earners better benefits due to
shorter life expectancy--this reflects their
different risk category
• but unless DB formula is very redistributive
low earners will be better off outside system
What should be done to prevent
poverty among low earners if
family system breaks down?
• Only general-revenue financed benefits will
reach these groups--0 pillar, social assistance
• Some high income countries offer universal
flat benefits in first pillar: redistributive to poor
(Ireland, UK, Denmark)
• Even some developing countries give small
universal benefit (Bolivia, South Africa)
• Problems: cost is low when few old but grows;
tax revenues limited and other needs great;
discourages contributory system
Should means-testing be used to
reach the poorest?
• Means-tested benefits cost less than flat, better
targeted to poor (Hong Kong, Australia)
• But means-testing difficult, high transaction
cost, crowds out personal savings
• Thailand has means-tested rural scheme
financed out of general revenues but this
reaches only 60% of poor and benefits are low
• Should this be expanded? Study transactions
costs, targeting, how to minimize leakages
• This may be better than expanding coverage in
How to cover the self-employed?
• Impossible to know their true income or
wage (part of income); underreporting
common even in industrialized countries
• They need capital for their business
• Saving for old age may put them into formal
sector with many regulations & taxes
• They may be risk-takers
• Possible solution: flat contribution for flat
benefit, upon registration (Japan)
• May not be feasible to cover now in Thailand
How to cover the self-employed?
• Impossible to know their true income or
wage (part of income); underreporting
common even in industrialized countries
• They need capital for their business
• Saving for old age may put them into formal
sector with many regulations & taxes
• They may be risk-takers
• Possible solution: flat contribution for flat
benefit, upon registration (Japan)
• May not be feasible to cover now in Thailand
What to do about women--the
oldest & largest uncovered group
• Many women don’t work enough to be
eligible due to household division of labor
• Problems: greater longevity, spouse younger
& dies sooner, retirement savings used up,
lower eligibility & benefit size
• Family system may care for women now
but problem if families break down before
women participate fully in labor market-smaller problem in Thailand
Solutions:
• DB schemes: survivors benefits important
• DC schemes: require survivors’ insurance
and joint annuities (Chile)--continues family
responsibility after death
• Design first pillar so working women will be
eligible (Chile v. Argentina)
• Universal flat benefit for very old or widows
• Avoid earlier retirement for women, since
this exacerbates problem later on
• Indexing especially important for women
Should coverage be extended on
voluntary basis?
• Safe financial instruments should be made
readily available
• Legal framework needed for employersponsored programs; tax advantages?
• Worthwhile, but voluntary insurance has
never reached majority, is mainly for high
earners--because of high discount rate, long
time periods, large amounts needed
Summary
• To expand contributory coverage make sure
– system is sustainable,
– most workers are better off when joining,
– government can enforce collections
• Equity requires higher replacement rate for
low earners to reflect their shorter lifespan
• Include non-contributory program financed
by general revenues for the poorest
• Take into account special needs of women
• Coverage grew slowly in OECD countries.
Speed-up can harm workers, system, economy