Compulsory Savings and Taxes in Singapore 9789814376280

Citation preview

I5EA5 Institute of Southeast Asian Studies The Institute of Southeast Asian Studies was established as an autonomous organization in May 1968. It is a regional research centre for scholars and other specialists concerned with modern Southeast Asia, particularly the multi-faceted problems of stability and security, economic development, and political and social change. The Institute is governed by a twenty-two-member Board of Trustees comprising nominees from the Singapore Government, the National University of Singapore, the various Chambers of Commerce, and professional and civic organizations. A ten-man Executive Committee oversees day-to-day operations; it is chaired by the Director, the Institute's chief academic and administrative officer. The ASEAN Economic Research Unit is an integral part of the Institute, coming under the overall supervision of the Director who is also the Chairman of its Management Committee. The Unit was formed in 1979 in response to the need to deepen understanding of economic change and political developments in ASEAN. The day-to-day operations of the Unit are the responsibility of the Co-ordinator. A Regional Advisory Committee, consisting of a senior economist from each of the ASEAN countries, guides the work of the Unit.

Compulsory Savings and Taxes in Singapore

Antal Deutsch and Hanna Zowall

Research Notes and Discussions Paper No. 65 ASEAN Economic Research Unit INSTITUTE OF SOUTHEAST ASIAN STUDIES

1988

Published by Institute of Southeast Asian Studies Heng Mui Keng Terrace Pasir Panjang Singapore 0511 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without the prior permission of the Institute of Southeast Asian Studies.

© 1988 Institute of Southeast Asian Studies The responsibility for facts and opinions expressed in this publication rests exclusively with the authors, and their interpretations do not necessarily reflect the views or the policy of the Institute or its supporters. Cataloguing in Publication Data Deutsch, Antal Compulsory savings and taxes in Singapore I Antal Deutsch and Hanna Zowall. (Research notes and discussions paper I Institute of Southeast Asian Studies; no. 65) I. Saving and thrift ··Singapore. 2. Taxation·· Singapore. I. Zowall, Hanna. II. Institute of Southeast Asian Studies. III. Title. IV. Series. 1988 DS501 I596 no. 65 ISBN 9971-988-87-9 ISSN 0129-8828

Printed in Singapore by General Printing & Publishing Services Pte Ltd

CONTENTS

Preface I

II

The CPF: Pension Plan, Tax Shelter and Debt Instrument Inequalities in the Tax Impact of Compulsory Retirement Savings

v

1

18

III Alternative Tax Treatments of CPF Savings --An Exercise in Tax Policy

39

IV Retirement Income Changes in Singapore

65

PREFACE

The four chapters in this paper were written between November 1986 and May 1987 (before the announcement of the newly proposed long-term CPF rate), and reflect a bird•s eye view of the Singapore Central Provident Fund (CPF). The CPF is a large, important, and complex instrument of the Singapore Government, and while we may have illumi~ated some of its facets, others may have remained hidden from us. On this, our first, co-operative professional effort, we divided labour by Hanna Zowall producing the tables on the computer and Antal Deutsch doing the rest. None of our efforts would have come to fruit had we not received advice and support from a number of colleagues. Mukul Asher, Glenn P. Jenkins, Lee (Tsao) Yuan, Lim Chong-Yah, and Wee Chow Hou gave us written comments. Subbi ah Gunasekaran and Kenneth James helped to overcome computer problems. Monika Queisser co-operated in the production of Tables 2.1, 2.5, and 2.7. Celina Kiong typed it all with speed, accuracy, and unfailing enthusiasm. McGill University gave sabbatical leave to one of us to do research in Singapore during 1986-87, where the Institute of Southeast Asian Studies provided us with first class v

vi

Preface

facilities and a delightful working environment.

We are

leaving it not only with regret, but also with a deep and abiding interest in Southeast Asia in general, and Singapore in particular. Antal Deutsch Hanna Zowall

I THE CPF:

PENSION PLAN, TAX SHELTER AND DEBT INSTRUMENT

The Central Provident Fund of Singapore (hereafter CPF), may be thought of as a very unusual institution for the workinglife accumulation of funds to provide a pension. As its most basic arrangements go, there is not much to distinguish its features from many other pension plans in the world. It provides a fully vested money purchase account accumulated as a fixed percentage of pay, with an employer and an employee share. Interest is credited at something Account resembling market rates at regular intervals. balances are reported periodically to the members. There is a "normal retirement age" at 55 in the plan. CPF differs from many other pension plans by the proportion of income it absorbs, the income tax treatment of CPF savings, the exceptions to the arrangements to lock savings in unt i 1 age 55, and the forms of access to the accumulation at that age. In a most significant way, CPF appears to have a ro 1e beyond that normally expected of a pension plan, as an instrument of government policy. Singapore's savings rate, by world standards, is very high. Most of the savings arise in the public sector, but the compulsory personal savings collected by the CPF provide 1

2 The CPF a considerable addition. 1 The rate of contributions, now 25 per cent of pay by the employee and 10 per cent by the employer to a maximum total annual contribution of $27,300, was at 25 per cent until April 1986 for each employer and employee, with a total annual maximum of $39,000. Either of these magnitudes accumulated through a working life, and compounded at, say, 3 per cent over the inflation rate, should provide for a life annuity ample enough to pay for a comfortable retirement. Chapter Four provides many illustrations of the magnitudes. The accumulation of large sums by an individual is, of course, contingent on a stable lifetime record of contributions, hence of employment. This fact is the natural source of a policy dilemma for the Singapore Government. Over time, the sum of the wage and the employer CPF contribution cannot exceed the value the employer puts on the work of the employee. If the government forces the compensation package to be too high, some employees will lose their jobs, damaging both shortterm income and long-term CPF accumulation. If the package is too low, Singaporean workers will not do as well as they could. Because take-home pay is of greater immediate importance to most people, when something had to be reduced in 1986, it was naturally the employer contribution to the CPF. The consequences of this step on particular retirement incomes are estimated in Chapter Four. The CPF must be the world's best tax shelter for its contributors. When compared to other Singapore dollar denominated assets, there is no risk. On contribution, both the employer and the employee shares are excluded from the base of the personal income tax. Interest credited by the CPF to the account of the saver is not subject to the income tax.

The CPF

3

On withdrawal, the accumulations paid by the CPF are untaxed once more. As the personal income tax in Singapore has "progressive" graduated rates, the tax saving is more substantial to those in the higher brackets. Since the tax saving, or most of it, occurs in the year of contribution, and for persons in the higher tax brackets that tax saving is far greater than the annual CPF interest rate, returns on sums deposited and quickly withdrawn are much better than on long-term accumulations. (Detailed calculations are found These rate of return differentials have in Chapter Two). attracted the attention of some economists, and the use of the tax system to eliminate the differentials has been discussed. Both in the literature and in informal discussion, a number of Singapore economists have indicated vehement opposition to the taxation of the CPF. One respected observer, in a pr·ivate communication, described " ... any scheme to tax the CPF, no matter how rational, would be political suicide for the government". This may very well be true, but that does not mean that academic economists cannot speculate on ways in which the CPF may be taxed, if anyone ever decided to do so. Such a speculative exercise is found in Chapter Three. Why should one be concerned about the differences in the rates of return earned by CPF savers? Apart from the concerns of equity, there are also purely financial considerations. What is a rate of return to a saver, is the cost of borrowing those funds to the consolidated Government of Singapore. Normally, the cost of borrowing should not exceed the return on the funds to the borrower. In the case of the CPF, the borrowing cost has two elements: the CPF interest, recorded on the book as such, and the "tax

4 The CPF expenditure" of income tax revenues foregone,

w~1ic.n

by its

nature is not normally recorded in the government accounts, even though highest

in economic terms

total

it is just as real.

The

cost currently paid for CPF borrowing for

funds contributed on the first

day of the year,

is the

average CPF interest rate (3.83 per cent for 1987), plus the 33 per cent marginal tax rate of those in the highest income bracket, making a total borrowing cost of 36.83 per cent on new deposits by taxpayers with annual chargeable incomes of over $400,000. 2 Very few costs of new borrowing are this spectacular, but they are currently in the double digit range for

everyone whose annual

chargeable income

is at

least $7,500. It should be emphasized here, that for the second and subsequent years 'Jn deposit, the borrowing costs for everyone fall

to the level of the CPF interest rate.

With an on-going CPF saver, every year is an initial year in terms of the contribution.

total

What

the

does

deposits?

CPF

bonowing

Government

costs

of

his

of Singapore earn

curren~

on

CPF

Lee (Tsao) Yuan, along with a number of other

writers cited by her, points out that the net contributions to

the

CPF,

deposited

with

the

Monetary

Si1gapore (hereafter MAS), result in a" ...

Authority

of

continual drain

of domestic liquidity which is made up for by the foreign exchange intervention of the MAS in terms of buying U.S. dollars and selling Singapore dollars". 3 This conclusion is consistent with much of the theoretical literature.

If

we interpret Dr Lee literally, each time the CPF deposits S$2.12 (net of withdrawals) with the MAS, a mechanism is set in

motion

reserves.

to

add

US$1

to

Singapore's

foreign

exchange

There are, of course, simultaneously many other

The CPF

5

influences at work on Singapore's foreig·n exchange reserves. The significance of the CPF effect

is that

it makes the

reserves grow, over and above the effects of all the other forces.

One can safely say,

that with the CPF effect at

work, Singapore's U.S. dollar reserves are higher than they would otherwise be. Marginal foreign exchange reserves are usually invested in

short-term money-market

instruments.

To estimate the

current earnings on incremental U.S. dollar reserves, we use the annual yield on six-month U.S. Treasury Bills, reported 4 at 6.02 per cent. As a first approximation, there is a margin of 2.19 per cent available in excess of the average 1987 CPF rate of 3.83 per cent to pay for administration and 5 for additions to the government's reserves. If the six-month Treasury Bill rate is representative of earnings on the foreign exchange portfolio, the preceding sentence is a reasonable

reflection of the financial facts regarding 6 ongoing CPF balances. For current CPF contributions, the picture

is

changed,

drastically,

by

the

tax

expenditure

arising from the treatment of CPF contributions. We have

estimated,

Table 1.1 the

tax

using

published

sources only,

in

expenditure for 1987 at $443 million.

Details of the calculation are shown in Table 1.2.

The tax

expenditure

of

amounts

to

contributions collected.

9.5

cents

per

dollar

CPF

(Our 1987 estimate for total CPF

contributions is $4,195 million.) Because CPF contributions are made throughout the year, we can assume that the average 1987 contribution dollar will be on deposit for six months. Applying expressing

the it

1987 as

tax an

expenditure interest

rate,

on

that the

basis,

Government

and of

TABLE 1.1 Estimated Net Cost to the Singapore Government of Accepting the 1987 CPF Contributions

1987 Projections in $ Million Annual total CPF contributions, 1985 actual amount of $5,993 million x .7 scale factor for 1986 CPF cut Average amount at interest for 1987, estimated at

~

of contributions for the year

4,195 2,098

Interest earnings by the government on average amount at interest at 6.02% per annum

126

Less costs of accepting the funds CPF interest paid by the government at 3.83% per annum Income tax revenue lost by the government (see Table 1.2)

523

80 443

Net loss on accepting 1987 CPF contributions

397

Net loss as a percentage of average amount at interest

19%

Net loss as a percentage of 1987 total projected personal and corporate income taxes collected for Singapore

20%

SOURCES:

(a) (b) (c) (d)

1985 CPF collections, CPf Annual Report 1985. The 6.02% annual rate is earned by 6-month U.S. Treasury Bills, as reported in the International Herald Tribune, 5 May 1987. 3.83% is the average of the interest rates paid by the CPF in 1987. Straits Times, 16 May 1987. $1.95 billion is the estimate of 1987 income tax collections contained in the 28 february 1987 Budget presented by the Minister of Finance.

TABLE 1. 2 Projection of 1987 CPF Contributions and Tax Expenditure by Wage Levels

(l)

(2):(1)*13

(3):(2)*0.5

(4)

(5)

Number of

Yearly Wages

Estimated 1985 Total Annual Contribution

Contributors (adjusted for not specified)

( 6 )= (3 )* ( 5)

(7 )= (2)*0. 35*( 5)

Number of

Line

4 5 6

9 10 11 12 13 14 15 16 17 18 19 Total

Estimated Monthly Wages

100 250 350 450 550 700 900 1250 1750 2250 3575 3250 3750 4250 4750 5250 5750 above 6000 not spec if.

1300 3250 4550 5850 7150 9100 11700 16250 22750 29250 35750 42250 48750 55250 61750 68250 74750 above 78000 not spec if.

650 1625 2275 2925 3575 4550 5850 8125 11375 14625 17875 21125 24375 27625 30875 34125 37375 39000 not spec if.

Contributors (actual 1985)

Distributed 1985 Total CPF Contribution

Estimated 1987 Total CPF Contribution

$'000

$'000

24601 57407 56571 81329 101096 150700 101202 144911 65196 36202 20733 14904 7843 5808 3937 3652 1846 7909 3745

24705 57650 56810 81673 101523 151337 101630 145524 65472 36355 20821 14967 7876 5833 3954 3667 1854 7942

16058 93681 129243 238893 362946 688584 594535 1182379 744740 531693 372169 316178 191981 161124 122069 125151 69286 309755

11241 65577 90470 167225 254062 482009 416174 827666 521318 372185 260518 221325 134387 112787 85448 87606 48500 216829

889592

889592

6250465

4375326

TABLE 1.2 (continued) Projection of 1987 CPF Contributions and Tax Expenditure by Wage Levels

(8):(2)*0.25

Line

4 5

9 10 11 12 13 14 15 16 17 18 19

Total

(9):(2)*0.10

(10):(8)+(9)

Estimated 1987 CPF Contribution

(11 )= (2)+(9)-$3000

(12) = (2)- (8) -$3000

Estimated 1987 Chargeable Income

(13)

(14)

(15)

Estimated 1987 Income Tax Paid

Employee

Employer

Total

CPF Included

CPF Excluded

CPF Included

CPF Excluded

Difference

325 813 1138 1463 1788 2275 2925 4063 5688 7313 8938 10563 12188 13813 15438 17063 18688 19500

130 325 455 585 715 910 1170 1625 2275 2925 3573 4225 4875 5525 6175 6825 7475 7800

455 1138 1593 2048 2503 3185 4095 5688 7963 10238 12513 14788 17063 19338 21613 23888 26163 27300

0 575 2005 3435 4865 7010 9870 14875 22025 29175 36325 43475 50625 57775 64925 72075 79225 163133

0 0 413 1388 2363 3825 5775 9188 14063 18938 23813 28688 33563 38438 43313 48188 53063 135833

0 17 60 102 145 251 437 836 1539 2481 3521 4737 5977 7479 8980 10482 12110 33773

0 0 12 41 70 114 188 391 771 1200 1754 2412 3095 3880 4709 5538 6489 26413

0 17 47 61 74 137 249 445 768 1280 1768 2324 2882 3598 4271 4944 5621 7361

(16):(5)*(15) Estimated 1987 Tax Expenditure on CPF Exclusion

0 986 2691 4975 7558 20804 25320 64753 50280 46537 36804 34790 22702 20987 16886 18131 10420 58462

443087

SOURCES AND NOTES TO TABLE 1.2: Column (1) is made up of the mid-point values of the class intervals, Annex F, 1985 Annual Report, Central Provident Fund. Column (2) is the estimated annual wage, assuming 13 months pay. Column (3) is the estimated total annual CPF contribution for 1985, at 25% employer share and 25% employee share. Column (4) source Annex F, 1985 Annual Report, CPF. Column (5) the "not specified" of Column (4) are pro-rated among the specified population. Column (6) is the product of Column (3) and Column (5). Its total is $257,044,000 or 4.3% over the total of $5993,421,000 shown as the actual total, p. 32, Annex B, 1985 Annual Report, Central Provident Fund. Column (7) allows for the 1986 reduction in the employer contribution to the CPF. It is the product of 35% of Column (2) and Column (5). Column (8) and (9) are, respectively, the employee and the employer CPF shares at 1987 rates. Column (10) is the total individual CPF contribution in each case. Column (11) is an estimate of a comprehensive "Chargeable Income" for income tax purposes, on the assumption that the

Column (12)

Column (13) Column (15) Column (16)

employer contribution to the CPF is reported as taxable income, the employee contribution to the CPF cannot be used as a tax deduction, and every tax-filer enjoys $3000 "personal relief". For the open-ended interval of the highest income earners, we took the mid-point of the corresponding marginal tax rates from Budget Statement 1986 of 28.5%, and interpolated within its tax bracket to arrive at yearly wages of $158,333. is an estimate of Chargeable Income for income tax purposes on the existing tax rules governing CPF contributions: the employer contribution is not reported, the employee contribution is not deducted. As in Column (11), it is assumed that each tax-filer enjoys $3000 "personal relief". For the open-ended top income class, we used the same assumptions as in Column (11). and (14) calculate 1987 taxes payable on the basis of the rates given in Budget Statement 1986. the difference between Columns (13) and (14), represents the income taxes each Singaporean does not pay because of the tax treatment of the CPF. is the individual tax-saving in Column (13) multipled by the number of CPF contributors in the class interval, Column (5), to yield the tax-saving