Anda di halaman 1dari 4

2tional saving will increase as a

tial part of a privatization is the


:fits, saving during the transition
[changed, the transition involves
which private saving is reduced
2onse to any other tax increase.
strained may choose to smoo.th
2ars and therefore save less than
already liquidity-constrained will
that are saving primarily for
heir behavior at all. The impact of
,1 therefore be on the order of
nposed.
nsition that achieves a bit more in
illing to give up a bit of national
2stigated the possibility of making
ystem voluntary in a way that is
g-out" in the United Kingdom
.on to be voluntary, we can be sure
; to make households that do enter
of that work thus far is that the
ing is critical to the success of any
goals of high voluntary participa-
saving rate will in general require
!s of an optimal income tax.
lesson to emerge from the work by
2conomy such as the United States,
velfare, not saving. When discuss-
stem, the two concepts are simply
mprove welfare is to remove the
ng ot Social Security from future
ltbout 18 percent of covered payroll
est for future generations. When
urity that might achieve that end, it
important element of reform is the
rl otherwise be incurred. The saving
straightfoward. When the reform
tion but a change in the promised
'ecome more ambiguous.
1997. "The Economics of Prefunding Social
ling in 8. Bernanke and J. Rotemberg, eds.,
1.4: National Bureau of Economic Research.
'eterogeneity and Social Security Reform."
1
Agustin G. Carstens"
Mex~co, like many other countries in Latin America, has recently
embarked on a major reform of its social security system. The main
element of this reform is a transition from a pay-as-you-go pension
system to a fully funded system with individual accounts and a mini-
mum pewion guarantee. In more general terms, this is basically a
transition from a defined-benefit to a defined-contribution system. Mex-
ico's prer-ious pension system was operated by two public institutions:
LMSS for private sector employees and ISSSTE for federal government
employees. The IMSS pension system, which covers over 70 percent of the
formal sector work force, was the focus of the recent reform. The most
immediate motivation for the reform was the increasing financial diffi-
culties the present pension system faced, rendering it unsustainable.
Beneficial side effects expected from the reform include a boost to
domestic 5aving, the development of stable, long-term sources of fmanc-
ing, and a reduction of distortions in the labor market.
Of the many factors underlying the pension system's present diffi-
culties, the most salient are the following three. First was a complete
dissociation between contributions and benefits. In a pay-as-you-go
system, one can expect that contributions will not have a rigid relation-
ship to benefits, particularly in extreme cases, that is, those of the
lowest-pad or the highest-paid workers. Such a system often has some
redistributive aspects. In Mexico, however, the imbalance between con-
tributions paid in and benefits received became excessive. For example, a
worker might contribute for only 10 years and thereby obtain the
minimum pension of minimum wage for his life plus the life of his
'Director General, Department of Economic Research, Banco de Mexico.
154 Agustin G. Carstens THE REFORM OF SOCIAL SECURITY IN h/
dependents. This is equivalent to the Mex~can government issuing
internal debt at a real rate of return of more than 50 percent. No doubt t hs
is an extreme example. Nevertheless, for the vast majority of cases, the
participation of workers in the previous pension fund system was
equivalent to their investing in government debt that paid a multiple of
the market's real rate of interest.
A second major problem the pension systcam faced was that reserves
that should have been accumulated under the pay-as-you-go system had
been depleted to subsidize deficits in other programs in the social security
system, such as health and maternity benefits. A third problem was the
relatively high contribution rates required by !he social security system,
31.5 percent of nominal salary, which generafed incentives for massive
under-invoicing and outright evasion. (This 31.5 percent accounted for
disability, housing, and health insurance in addition to retirement insur-
ance.) As a result, enrollment in the system was less than what was
projected by pure demographics. This history ct f low enrollment, together
with the aging of the Mexican population, imvlied a rapid deterioration
of the elderly dependency ratio, compounding the system's financial
difficulties.
In my view, these three factors alone justified the reform of Mexico's
pension system. The actuarial deficit of tht. previous IMSS pension
system was estimated to be more than 100 percent of GDP. Had the
reform not been implemented, contributions M-ould have had to increase
by more than four times by the year 2020 to avoid a cash flow deficit.
Clearly, such an enormous increase in contrib~ltions to the pension fund
system would have generated unacceptable distortions in the labor
market.
What then are the major features of our pf.nsion fund reform, which
will begin operating in July 1997? The reform basically involves improv-
ing and strengthening the Sistema de Ahorro para el Retiro, the "SAR"
system, a fully funded pension system that I.vas initiated in 1992 as a
complement to the IMSS system. The former 8.5 percent of salary
contribution to the previous IMSS pension sy5tem (which included both
disability and life insurance) is now divided into two parts: 4.5 percent to
be accumulated in individual accounts and 4 percent to go directly to the
IMSS for the provision of life, health, and disability insurance for retirees.
In addition, contributions to individual accounts will include the 2
percent contributions to the SAR system along with a new "social
contribution" by the government, which will be a fixed amount equiva-
lent to 5.5 percent of the minimum wage as of January 1997, indexed to
inflation. Therefore, for a worker earning minimum wage, total contri-
butions to his individual account will amount to 12 percent of his salary.
For a worker earning an average wage, total cc;ntributions will amount to
8.5 percent of salary. The new regulations also allow for voluntary
contributions to the individual accounts.
To be eligible for a pension, a 1
The tax-free benefits depend on the
vidual's working life, that is, the
earned on his account. And the mir
to qualify for a minimum pension h;
The management of pension
pension fund administrators called
vised by a special government agen
AFORE of their choice. As of this 1
already have affiliated. Workers t
account to another AFORE once per
operate several pension funds, so
resources in one or several of th
AFORE.
AFOREs will be allowed to ck
fees as a percentage of the con
individual account balance, or an.
responsible for the collection of coi
why commissions in Mexico are c
One drawback in the current regul;
is the imposition of a maximum
individual AFORE, starting in 19
years. Until July 1997, this restricti
A crucial determinant of the 1
the nature of the transition from a 1
capitalizatio~~ scheme. This process
redistributive effects, with long-te
the reform, the government n0.c
pensions of current retirees and o
contributed to the previous pensic
ment expenditures in the future, 14
reductions in other government '
debt.
How these fiscal requiremenf
crucially influence the net impa
Debt-financing implies that in t'r
capital stock will be only margin
!
tudes that depend on the net effil
well as on voluntary saving. On
deficits are financed through tz
1
pension reform with tighter fiscz
transfer of resources from curren
I 1
higher rate of saving and, thus, c,
follow a combination of both optic
i
1
reform to have a positive impact
[exican government issuing
I
I
han 50 percent. No doubt this
i
le vast majority of cases, the
!
I
pension fund system was
t debt that paid a multiple of
i
I
y'stem faced was that reserves
he pay-as-you-go system had
xograms in the social security
fits. A third problem was the I
by the social security system,
erated incentives for massive
lis 31.5 percent accounted for
I addition to retirement insur-
tem was less than what was
)ry of low enrollment, together
. implied a rapid deterioration
~ndi ng the system's financial .
ustified the reform of Mexico's
f the previous IMSS pension
100 percent of GDP. Had the
,ns would have had to increase
10 to avoid a cash flow deficit.
~tributions to the pension fund
table distortions in the labor
)ur pension fund reform, which
'orm basically involves improv-
lorro para el Retiro, the "SAR
that was initiated in 1992 as a
former 8.5 percent of salary
In system (which included both
led into two parts: 4.5 percent to
ld 4 percent to go directly to the
disability insurance for retirees.
1 accounts will include the 2
2m along with a new "social
will be a fixed amount equiva-
as of January 1997, indexed to
; minimum wage, total contri-
3unt to 12 percent of h s salary.
-a1 contributions will amount to
ons also alIow for voluntary
THE R!:FORM OF SOCIAL SECURITY IN MEXICO
155
TI) be eligible for a pension, a worker must be at least 65 years old.
The tax-free benefits depend on the balance accumulated over an indi-
vidual s working life, that is, the contributions paid in plus returns
earned on his account. And the minimum number of contributing years
to qualify for a minimum pension has been increased from 10 to 25 years.
The management of pension funds will be entrusted to private
pension fund administrators called AFOREs, to be regulated and super-
vised b? a special government agency. Participants may affiliate with the
AFORE of their choice. As of this writing, more than 8 million workers
already have affiliated. Workers will have the right to transfer their
account to another AFORE once per year. Each AFORE will be allowed to
operate several pension funds, so workers will be able to invest their
resources in one or several of the pension funds managed by their
AFORE.
AFC)REs will be allowed to charge freely determined management
fees as a percentage of the contribution flows or the outstanding
individu.tl account balance, or any combination thereof. IMSS will be
responsible for the collection of contributions. This is one of the reasons
why commissions in Mexico are currently the lowest in Latin America.
One draw-back in the current regulations on the operation of the AFOREs
is the imuosition of a maximum market share of 17 percent for any
individua! AFORE, starting in 1997 and continuing for the next four
years. Until July 1997, this restriction had not been binding, however.
A crucial determinant of the pension reform's long-term success is
the nature of the transition from a pay-as-you-go system to a fully funded
capitalization scheme. This process will have important fiscal and income
redistribut:ve effects, with long-term implications. As a consequence of
the reforrr, the govemment now faces the problem of paying the
pensions OF current retirees and of honoring the claims of workers who
contributed to the previous pension system. This implies higher govem-
ment expenditures in the future, which could be financed by more taxes,
reductions in other govemment expenditures, or the issuance of new
debt.
How these fiscal requirements are financed over the transition will
crucially intluence the net impact of the reform on national saving.
Debt-financing implies that in the short run, national saving and the
capital stock will be only marginally affected, and primarily by magni-
tudes that depend on the net efficiency gains of the financial system as
well as on .ioluntary saving. On the other hand, if these future fiscal
deficits are financed through taxation, this will essentially combine
pension reform with tighter fiscal policy. This strategy would imply a
transfer of re3ources from current to future generations, encouraging a
higher rate 01 saving and, thus, capital stock formation. Mexico plans to
follow a combination of both options, which in principle should allow the
reform to have a positive impact on national saving.
The fiscal costs of the reform in net pr-esent value are estimated at
roughly 80 percent of GDP. The incidence 0;' this cost ~7ill be between 0.8
and 1.0 percent of GDP per year in the comng years, but it will increase
through time, reaching a peak of about 3.C percent of GDP in approxi-
mately 35 years. Several studies have analj zed the transition under the
assumption that it will be partly debt-fin.~hced. The results obtained
predict increases in net saving of between 2 to 3 percent of GDP in the
near future, independent of the impact of voluntary saving. The latter
could make a sigruficant contribution to sal ing, given the fact that with
the reform a large portion of the populatioll will for the first time have
access to saving services in the formal financial system, whichshould
translate into more efficient intermediation of saving in the economy as a
whole.
In addition, the pension reform will prcduce a significant long-term
increase in the availability of financial rescuces to the economy. Ths
should promote further financial specialization and the creation of new
financial instruments, in particular, mediun- and long-term securities.
These are urgently needed in Mexico, give11 that most lending is now
concentrated in loans tied to short-term interest rates, with maturities of
less than a year in most cases. The pension fund reform should also
promote the development of the stock market in Mexico.
Estimates indicate that within 25 year;, the reform could double
financial savings in Mexico. The availabil-ty of these resources will
promote financial deepening and long-term investment, which should
have a significant impact on growth. For the Mexican economy to grow
at a minimum of 5 percent annually, it requi-es an investment rate of 25
percent of GDP. For this to be possible without excessive reliance on
foreign saving, domestic saving ~7ill need to :ncrease in the coming years
by between 5 and 6 percent of GDP. The pl?nsion fund reform should
provide half of that requirement, which is a 1 ery significant contribution
indeed.
References
Diamond, Peter. 1995. "Economic Support in Old Age" Annual Bank Conference on
Development Economics, The World Bank. May.
Feldstein, Martin. 1974. "Social Security-Induced Retirement and Aggregate Capital Accu-
mulation." ]oilriznl o f Political Ecoiloiiry, vol. 24, pp. 935-20.
Martinez, Gabriel. 1996. "Ahorro y Seguridad Social er Mexico." Paper presented at the
International Seminar on Social Security for the 21s' Century, Mexico City, March.
Noriega, Carlos C. 1997. "Domestic Savings and Per sion Reform in Me;ico." Paper
presented at the International Seminar on Pension Re forms and Transition Fiscal Costs,
Cancun, April.
Sales, Carlos, Fernando Solis-Soberon, and Alejandro V.llag6mez. 1996. "Pension System
Reform: The Mexican Case." NBER Working Paper KO. 5780. September.
Solis-Soberbn, Fernando. 1995. "Description del Sistema de Ahorro para el Retiro."
CONSAR, Docuriwrrto de Trnbajo hro. 3, September.
Richard Disney*
The United Kingdom, c
does not appear to have a f i ~
has accomplished a substantial
without apparent political contrc
of social security without vocife
achieved, and what have been
provides a brief exposition of 1
future financial projections, and
saving and on retirement behavi
It is important first to note
privatization of social security in
Chile. I have argued elsewhere
Johnson 1997) that the history of
mid 1980s in fact involved a
arrangements: for example, the (
ance organized by workers into
Insurance and, much more rec
provided pension plans througl
lished in the 1975 Social Securit
1986 saw an abortive experinr
following the introduction of the
(SERPS), and it was not until tl-
from public to private pension F
A second preliminary point i
'Professor of Economics, Queen hl i
and Research Fellow, Institute for Fiscz!

Anda mungkin juga menyukai