World Bank Reprint Series: Number 144 Parthasarathi Shome and Katrine W. nato The Impact of Contractual Savings on Resource Mobilization and Allocation: The Experience of Malaysia Social Security Funds in Singapore and the Philippines: Ramifications of Investment Policies Investments of Social Security Funds in India and Sri Lanka: Legislation and Experience Reprinted with permission from The Malaaian Economic Review, vol. 23, no. 1 (April 1978), pp. 54-72; Labour and Society, vol. 5, no. 1 (January 1980), pp. 19-30; and Th7e Indian Journal of Economnics, vol. 60, part 3, no. 238 (January 1980), pp. 349-60. THE IMPACT OF CONTRACTUAL SAVINGS ON RESOURCE MOBILIZATION AND ALLOCATION: THE, EXPERIENCE OF MALAYSIA PARTHASARATIII SHCME and KATRINE W. SAITO* I. INTRXIZZ-ODUCTION Contractual savings schemes, such as pension and provident funds, social security schemes, and life insurance, serve a significant economic as well as social objective in many developed and developing countries. These schemes provide security to individuals in a variety of risk-ridden situations, and also serve as an important source of resources which can be used to finance econo- mic development. The experience of developed countries such as the U.S. and U.K. bear witness to the significant contribution which these schemes can make, at a critical development stage, to the industrial and financial develop- ment of the country. 1 In both these countries, pension flinds and insurance companies exerted a stabilizing influence on the long-run development of the financial markets, and also acted as an innovative investor, being a major source of industrial finance.2 Although in many developing countries contrac- tual savings schemes are still limited in operation and coverage, in some countries they have already developed into significant financial institutions3 Malaysia is such a country, and the objective of this study is to assess the role which the Malaysian contractual savings institutions have played in the mobili- zation of savings, and the extent to which the savings of these institutions have contributed to the financial and economic development of the country. The layout of the paper is as follows. The existing contractual savings institutions in Malaysia are reviewed, and their scope and coverage examined. Their success in mobilizing funds is then assessed, and finally the allocation of these funds is examined, paying particular attention to the contribution made to the socio-economic development of the country. * We arc extrerncly grateful to Dr. V.V. Bhatt, Chief of Division of Public ano Privatc Financc, I)cveloprnent Economics Dcpnrtrmcnt. the World Bank, for long conversations and subsLantial suiggestions on the paper, and the general educative impilc' we have had froni them. 1. OECI) emphasised this point: "The efficiency of certain markets, in particular thosc in thz United States, and the United Kingdom, the Netherlands and in Switzcrland seems to stem to a r.nsidcrable extent from the part played in these markets by intermediaries that collect funds on1 a contractual basis and invest them priniipally in long-term forms." OECtl), Committee for Invisible Transactions, Cpiteal Markets Study, General Report, Paris 1967, p. 12. 2. In both these countries nnsion funds and insurance companies held substantial amounts of industrial securitics, both equitics and long-tcrm loans, in their portfolios; see ident Chapter II and D.K. Sheppard, The Growvth and Role of U.K. Finantcial Inistitutions, London: Metlinen & Co., 1971. 3. This was emphasised in a survey paper by P. Shome, "Contractual Savings and the Mohilization of Resources: A Preliminary Strvey" Studies in JJomestic Finance, No. 29, World Bank, November 1976, RESOURCE MOBILIZATION AND ALLOCATION 55 11. COVERAGE AND SCOPE OF CONTRACTUAL S'AVINGS INSTITUTIONS The contractual savings institutions covered in this paper includ.: both the life insurance business as well as other formws of contracts such as the pro% iLlent fund, social security schemes etc. There is a reasonable nicalUre of douht regarding the definition of and any disaggrcgatcd taxononoy oIn contractual savings due to the wide nature and forms of tliecs institUtioln. Some schemes may be voluntary, while others mandatory, depending on, inconle scale or occupation factors. This sometimes may lead to a definition of Contractual savings cxcludinig mandatory schemes. Hlowever, voluntary schenie' mav ako be considered 'compulsory' by the ccntributors. onee the insurance conyract is entered into. It is for this reason all forms of coilracts, voluntary and m11anda- tory, are included in this study. Apart from the prijate life insurance business, rNial yia has 1rogral1rn1mes for retirement and death, covered by the Emnployees Pi idenit F und (1052) under the jurisdiction of the Ministry of Finance, and the TeachersN Pro' iLidlnt Fund (1962) under the Public Services Departneint as well as acec ient injtury, medical care, disablement and dependenIt bemnelii<, covered hy the rinploek-e'.S Social Security Act (1969) under the juridiction of the Ministry of 1.11h1olir. The latter is further subdivided into the ElnpEoyllient Injury Tlmislrancie Slchlelme (tor employment injury only) and the Invalidity IPensioni ScIlelme (h'Ioni0 \Nhiate%er cause), becoming effective from 1971 and 1973 respectikely. Ani oullilm of ilie major social insurance programnmies in M11alaysia is prresenited in Tiible 1. TABLE I MAJOR SOCIAL I\SL RA%NCE PRO(iR \NINi S IN MALAYSIA Covered Social Insurance Year Person Nli1or Administer -'onrribEuu Adoptedl In m3 in1gA \1 "Cn 1. Employees Provident 1952 ,880* Retirement .Ministry of FnimZlo e2s! Fund death I .itiance 1:nlvtr cr5 2. Teachers' Provident 1962 750* RetirenLnt L Public I lniplo%es Fund .Services I mptlo%rs .Employment Injury 1971 226t Fniployrnent Ministry of i niplo\ceY Insurance Scheme , Iniusa-,.nd Iahour I'nipl,.w;o Invalidity Pension 1973 1,000t \ccideat Mizuniiry, of Enirploeci' Scheme Injury Labour I rniplo",ers from any source * 1975 figures. t Early 1973 figures. Source: 1. Annual Report of the Sociail Seruritv Orgunistutim, 1972, Government of Malaysia, Kuala Lumpur. 2. Exiployees Provident Fund Ortianie&', 19 ;I, Reprint No. 3 of 19,)s, ItKala Lumpur. 56 RESOURCE MOBILIZATION AND ALLOCATION Most contractual savings institutions have to operate within clear guidelines regarding the inv,,stmnnnt of surplus funds. The Insurance Act (1963), for example. was amended in 1970 to regulate the investment of insurance funds: at least 10 percent of the total assets of insurance companies were to be in Malaysian government securities by the end of 1970, 15 percent by the end of 1971, and 20 percent by the end of 1972. A 1975 Amendment did not change this percentage. The investment of the surplus of EPF is similarly regulated by the Trustee Ordinance (1949) and the Amended Trustee Investment Act (1965), whereby the Fund is required to invest at least 70 % of total investible funds in government securities and the remainder in high quality approved assets determined by the Ordinance. Smaller private provident funds are required to maintain 55 percent of all investments in local assets. Finally the investments of the Social Security Organization are primarily in government securities. The rates of contribution are 1 and 1I % of wages for invalidity and employment injury, respectively. With regard to the EIS, the total contribution is paid solely by the employer. For the IPS, the contribution is shared cqually by employer and employee. The government does not contribute to either scheme except as an employer and towards administration costs. Benefits in the form of pension are paid only if death or invalidity occurs after 3 years of enrollment; otherwise only the contributions plus interest are paid back. The investments of the Social Security Organization have been primarily in Govern- ment securities, the remainder being in fixed deposits with the Central Bank. Tables 2 to 4 provide information regarding the coverage and scope of different contractual savings institutions in Malaysia in terms of the percentage of population covered, total sums insured or contributed to these institutions, these sums as a percentage of national income, the number of foreign vis-a-ris domestic companies, the extent of benefit payments made, etc. In terms of most of these indices it emerges that the coverage and scope of contractual savings institutions in Malaysia have grown significantly since their inception. The percentage of population covered by the EPF has grown from 14.8 % in 1960, to 16.5% in 1965, 19.3% in 1970, reaching 23.9% in 1975 (Table 3). Data on working population are available on a consistent basis only from 1966, and the figures may well understate the actual working population (for example, the published data includes only full-time employees.) They do however, provide a clear indication that IEPF has steadily been grow.ing in its coverage (the share of working population covered in 1966 was 77 percent, in 1970 was 85 percent, and in 1974 was 97 percent). The share of the total population covered by life insurance companies has also been growving, increasing from 2.5% in 1971 to 3.7% in 1974, (Table 2). The annual precenitage increase in the total sums insured by life insurance companies has been impressive, being in the range of 15 %-24 %. From 1965 to 1975 the sum insured per-capita increased from M$78 to M$334.6; from 1971 to 1974 the sum insured per policy increased from M$6752.2 to M$8050.8. More impressive and more indicative of the increase in coverage are the figures for total sums insured as a percentage of national income, reaching 18.5% in 1975 (Table 2.) RES-OURCE, MOBILIZATION AND ALLOCAtION 57 TABLE 2 COVERAGE OF LIFE TNSUR\NCF, LTS!JNFSS Total Sumn'- Insuireci in Force Total NewSum 'Sums % f Insured j As% ci'f Insured Population M$m M$m Q inl,nrcase perM$p National i per 1(olicLy Covered i Incomte I M$ 1965 203-9 I 758.3 - 78.0 (8.8) n.a. n.a. 1966 224.2 875.2 1 (15.4) 90.0 (9.5) n.a. n.a. 1967 259.8 1002.2 (14.5) 99.9 i 0.4) n.a. n.a, 19fs ' 23.6 1172.4 ( 17.0) 113.6 (16.4)' n.a. n.a. 1969 365.1 1365.6 (16.5) 1I5.8 (12.4) n,a. n.a. 1970 431.1 i 1570.0 i (15.0) 144.3 (12.9) n.a n,a, 1971 545.4 i16"'.3 1(l.9) 167.2 (14.9) 6752' 2.5 1972 710.5 21315.4 i (24.1) 0' 02- (17.0) 7425.9 9 .7 1973 861.6 2881.0 (24.4) 245.2 (16.5) 7719.7 3.2 1974 1084.2 3 485. 0 Z 21.O 297.8 (16.4) 8050.8 3.7 1975 n.a. 402, 7 (I5.5) 334.6 (18.5) n.a n.a. Source: Director G ,neral of Insurance, Anrual Report for year rndinr, December 31 1975; National Income data (GNP at market fruiii) irvin 'ank Nega'ra Malaysia, (uartrrly L.'0h,rnii BDidilen various issues; )OidPlation d.lta froni World Bank, WfAvrld rables. C0..oitributions to the EPF have also gro% n rapidhl , but at a less 'Ateadv r.tc than contributions to life in'lurance conipanies. For x,xatnplc, in ol') the annual increase in contrilUltions was 6.60 and in 1975 was 20.6 '; 1etwcen these two dates the annual increase oscillated considerably-, reaching a high of 29.4o in 1970 wN-hen the EPF legislationi was amenided to cover all em- ployees. irresfeLtive of salary. The number of contributors' accounts with the EPF has steadilv grown from 1,143,000 in 1960 to 2,800.000 in 1975, while total contributi(ons as percentage of national inconmc has risen steadily from 1.30 to 1.8o- in the sarne time period.4 From Table 4 some conclusion-i can be drawn conecerninlg the operation of the insurance business in Malaysia. Thc annual premia rc-ceivei by the life insurance business have grown sienificantly over the last fifteen years. Foreign- OWned companies have received the major portion of the premia (on average, almost 80('o), although in recenit years there is some indication that this share is declining. Benefit payvnents have also expanded with coverage, with a few ups and downs during the vears in question.5 One interesting aspect of the life insurance business is the "ihome service" type of enterprise. This business has grownM1 effectively over the years. The sumns insured as well as the annual prernia have inicreased seven-fold over he period 1965 to 1973.6 The criticisml has frequlently been levied at conitri-actulii saving,' institutioni that they serve to benefit the nmiddle to upper income groups and not those beloITginig to lower inconme strata. This clearly is not the case in Mvalaysia, wlherc 4. In 1971 and 1972, the perceniaees were as high as .20 and 2.1. 'This inight have been because total contributions rosc %ignilfieantl dulring thcse years whereas the rate of rro%th in national income was less inipreviive than earlier years. 5. While death and disability, surrcnders, policy dividcnds and lonuses have all grown, maturity payments have expanded itie most r.apidl'y, reflc.ting the overall aging of polic- holders, 6. In 1965, sums insured were MSlO.1 million, the annual nremia being IvIS1.2 million, The corresponding fipures for 1973 were MS77.1 million and MS7.0 millinn resPecLti%ly. 58 RESOURCE MOBILIZATION AND ALLOCATION TABLE 3 COVERAGE OF EMPLOYEES PROVIDENT FUND Number of Total Contributions Contribu- Percent of Percent of tors' % of Population Working Accounts M$ million % increase National Covered* Population ('000) i ncome Covercdt 1960 1,143 84.9 - 1.3 14.8 n.a. 1961 1,232 90.5 6.6 1.4 14.7 n.a. 1962 1,250 94.4 4.3 1.4 14.5 n.a. 1963 1,430 105.5 11.8 1.4 16.0 n.a. 1964 1,488 114.6 8.6 1.5 16.3 n.a. 1965 1,554 126.4 10.3 1.5 16.5 nma. 1966 1,631 137.6 l 8.9 1.5 17.3 76.9 1967 1,717 1.48.0 7.6 1.5 17.1 82.6 1968 1,830 152.7 3.2 1.5 17.7 88.6 1969 1,949 159.9 4.7 1.5 18.3 88.3 1970 2,104 206.9 29.4 1.7 19.3 85.0 197! 2,217 245.9 18.8 2.0 19.7 87.4 1972 2,365 284.4 15.8 2.1 20.7 90.9 1973 2,546 291.0 2.2 1.7 21.7 93.3 1974 2,710 315.9 8.6 1.5 23.2 97.2 1975 2,880 381.0 20.6 1.8 23.9 n.a. * Contributors' Accounts as ° of population. * Comprises population engaged in following activities: agriculture, including forestry and fishing, mining and quarrying, Government and quasi-government, manufacturing, "pioneer industries." Source: Bank Negara Malaysia, Quarterly Economic Bulletin, various issues; World Bank, 'World Tables; The Treasury of Malaysia, Economic Report. TABLE 4 ANNUAL PREMIA AND BENEFIT PAYMENTS OF LIFE INSURANCE BUlSINESS Domestic Cos. Foreign Cos. Total A-smllio otal ot f Benefit M$ million As of M$ million Totas M$ million Increase Paym.nts 1965 0.5 I (6.9) 6.7 (93.1) 7.2 (22.5) 3.5 1966 2.3 (41.1) 3.3 (58.9) 5.6 (-22.2) 0.9 1967 1.3 (22.0) 4.6 (78.0) 5.9 (13.2) 0.7 1968 1.2 (16.7) 5.9 (83.1) 7.1 (14.0) 4.6 1969 1.5 (19.7) 6.1 (80.3) 7.6 (13.2) 3.0 1970 1.6 (21.6) 5.8 (78.4) 7.4 (-2.6) 4.8 1971 2.2 (20.8) 8.4 (79.2) 10.6 (14.6) 2.5 1972 3.5 i (26.1) 9.9 (73-9) 13 4 (16.1) 0.7 1973 5.1 (24.4) 15.8 (75.6) 20:9 1 (21.6) 0.7 1974 3.2 (14.3) 19.2 (85.7) 22.4 (19.0) 8.4 SQurrec: Director G;eneral of Insurance, A1iny!al Report, various issues, RESO)URCEh MOBIIl AATVION ANI) .WINl.L()C'U Ic both the EPF and the life insurance comnparnies have attempItedL to include in those they serve the lower income icginents of the population. The homne service scheme of life insurance, for e. iple, was d0sig'ned to pro. ile a life in-suir inice service specifically for thoSCwith ies>, inconni, apilpawetlv witlh c I1%hlLr.le suci cess.7 The EIPF fronm its very inception was aimed at hm.cri-.conLe inil1oxee.l. It covered employees whose salary was M$400l per month or less jincre.tawse to M$500 in 1963, and the salary ceiling was abolklied entirely in 1 '70). and the initial constraint that the establishmiienit should cmploy at least 5 workers was abLandioned in 1964. By 1970 all enmployees, with few excepinus. cre .Sered by the Fund. Such mlepsJures cleairly inidicate a e-e;ii tt) inciltude thosev ith ilo l%en incomes in the coverage of these schemes, and in the case of the LPI i until P')70). of excluding those who%e incomiie exceeded a cer.tain limit. Furthermcie, jnThing bv the extremely high percentage of the working popultion co%ercd, at least by the EPF, these measures appear to have been successfuil in this obhjcci%, IIT. RESOURCE MOBILIZATION TIIHROUG(JH 0C' FP. CJTt . SAVINGS INSTITUTIONS Saving in contractual ftrm. suchi as through pros id.nt funds or penioni schlemes, or through the purcliase of life insurance, h1.i^ the adhan,nta., e of regularity which helps to stabilize the savings rate. Sa%ini.l thus becom"ies a1 cornmi rnent on the part of the lhousehold and is Ino lon&!or ain err1atic resihiual of ince-nme less consumption. While all contractual savings institutions haive this stahiliiina infl0lencc On aggregate sa\ings. their potential for -ener,tting sa'ingtai can titker depzndin upon the type of institution. Life insurance contracts ha.,ve consideraible ptlten- tial for mobilising savings, since they cover not only the risk of iuture cointin- -ency, but also have an annuity element for the insurledi o.en allnd above the ri.sk Loverage.8 By contrast, the savings potential of social insuraiic progranmies, such as provident or pension scheme,. depend-s eN:,entiallv on the principle of funding on which they are based. In Nalayiyia both the EP Eand TPF are based on the "funded" principle of benefit payment herleca the EIS and the IPS follow the "pay-as-you-go" principle. In the fill funding (or general a%.eiatte premium) system denmarcated funds are set a.ide for contributors, creatinat a reserve and earning investment income (intere,t antd apprecidtion of value). These fuLnids are then used at retirement etc., to pay the clainmed benefits. In the pay-as-you-go (or annual assessmnenit) systen. henulits ire paid from the allnual re'venuc usually with no prior reserve formnation except may be a precautionary fund. Short-term progranimes are usually COVercL by thle pav-as-v-gi system. but lonig-termiii progranmmes may be coxcieed by eithler. 'I'lh fully-flunded 111ethid is generally asISUmed to have a gtreatter c;pacity to generate h.urpluses beL ause of 7. In terms of sums insMred, the honme -e,r ice business has l rown at a faster rate than the total life insurance business (26 percent compared to I') percent compound rate of gro%% Lih over the the years 1965 to 1973). 8. In this sense, group property and term inmurance do not mnobilize persoxial ~.'s ir1g.. I or further elaboration, see K. Saitw, "An Estiinationi of 11uiusehold Savigrm in N;ri l anka", $tudies in Domestic Fintance, No. 27, World Bank, Atisiist 1976, 60 RESOURCE MOBILIZATION AND ALLOCATION the creation of reserves. A programme based on the pay-as-you-gD principle may also generate savings through contingency reserves,9 though this is likely to be to a lesser extent than the fully-funded principle. WN'hile plans are newly established, deferred benefits and broader or expanded coverage may bring in added revenues. As the economy grows, if the benefits are controlled and the retiremiient age is not early, then the period of savings generation can be extended. If claims are illiquid in the sense that it is difficult for employees to borrow on the basis of their contributions, or that employers do not get back their contrib'.tions if an enterprise is liquidated, then again the savings potential is high. Household sector savings may also increase through pensions if the poorer employees would not save except through compulsion. Workers may increase other forms of savings after becoming participants in the plans since now they are more aware of the need to save and believe that a comfortable retired life is not impossible to achieve. Also if contributions are illiquid, people may feel the need to save in more liquid forms. Cash contribu- tions and benefits payments of social insurance plans are, by and large, regulated and, therefore, predictable. Thus there is a greater freedom in mobilizing these savings on a long-term basis. IO The extension of social security schemes has been actively encouraged in many LDC's on the argument that this would increase aggregate savings. It is quite true that such an extension would increase the total contractual savings which would have the beneficial effect of reducing the erratic element in aggre- gate saving as N_ell as, under certain conditions, stimulating the deve opment of the capital market. It is not all clear, however, that contractual savings schemes do have a positive effect on the average savings rate. To conclude that aggregate savings will increase by the total amount of savings generation through these contractual schemes ignores the reactions of the subscriber to the scheme and also in the case of pension or provident funds, of the employer. The reaction of the saver to his participation in contractual savings schemes is extremely difficult to gauge. A rational response would be to fully recognize that his contributions, as well as those of his employer, comprised a part of his total sai ings and if he was motivated by some notion of target wealth, he may regard saving in contractual form as a substitute for holdings of other financial assets and reduce his savings by an equivalent amount. It should be recognized, however, that substitutability is limited by the attributes possessed by contrac- tual savings whirlh, in manyrespects, are quite distinct from other financial assets held by households. He may also be encouraged to save less in non-contractual form since his contractual saving guarantees certain future benefits and so weakens the nrecautionary motive for saving. On the other hand, saving may -.1 9. For a more detailed analysis on the two principles, refer to Franco Reviglio, "Social Security: A Mcans of Savings MNobili?ation for) Economic Development", IMP SmJf N-Pers, July 1967. 10. It can also be argued that these schcmes may have a negative impact on savings, both in other forms ailv other sectors. For sorme lower income grouips it may be that their consump- tion cannot be 'urlher reduced, and their participation in contractual savings schemes may force them to decrease their savings in otlher forms. 'Tlhere may also be a negative impact on business savings unless prices or productivity increase. Pension deductions allowable in income tax rates may lower government savings. A detailed analysis of the pros and cons of savings generation by pension funds may be found in M.A. Olde, Pcnzsioni FuinSds in Labour Sturpltus Economics, Institute for Social and Economic Research, Univcrsity of West Indies, 1974, RESOURCE MOBILIZATION AND ALrLOCATION 61 be stimulated through reducing the volitional element in the decision process, or in the case of provident and pension funds, through "setting a pattern for retirement rather than a life-time of work," I 1 An examination of the Malaysian experience over the last fifteen years reveals that saving in contractual form has been growing steadily (Table 5). Gross saving in contractual form consists of gross saving through life insurance (estimated as the annual change in the life insurance funds of all registered life insurance companies) plus gross saving through the EPF (estimated as the annual payments received less beniefits paid out). Of the two, saving through the EPF has been much larger, averaging approximately four times the saving through life insurance. Total gross contractual saving has averaged 10 percent of gross private savings over the period 1961-1975, with the share decreasing slightly over the period, with saving through the life insurance increa-ng at a slightly faster rate (an annual compound rate of 10.2 percent compared to 9.3 percent). Unlike saving through life insurance saving through the EPE has not been steady. In 1968 and 1969, it declined, and in the next two years it rose sharply (by 53 and 32 percent respectively). This rapid increase can be attributed to a change in the EPF regulations in 1970 which, to include all employees, widened its coverage. The total gross saving through these institutions grew at an impressive annual compound rate of 11.2 percent. Net saving through these institutions is derived by subtracting the credit which they extend to policy holders/contributors from the gross saving figures. Since the EPF does not extend any credit to its contributors, only loans made by life insurance to its policyholders need to be taken into account (colutnn 4, Table 5). In fact, these loans have been substantial, so that net saving through life insurance is much lower than the gross figures. Gross contractual saving as a percentage of gross private savings and gross national savings has been impressively high, the former ranging from 6 to 13 percent, and the latter from 5 to 11 percent. Over the fifteen years, the annual average of gross contractual savings to gross private savings was 8.7 percent and gross contractual savings to gross private savings was 7.5 percent. On the important question concerning the impact of contractual savings on aggregate savings, the Malaysian experience indicates that sav., in contractual form has not had an adverse impact on aggregate savings, and may have had a positive effect. Offhand, reaction of the saver to his participation in contractual savings schemes is extremely difficult to gauge. In order to test the Malaysian experience, the marginal propensities to save out of private disposable income, both including and excluding contractual savings, are estimated and compared. The following functions are estimated: (1) GPS a+b PDY (2) NCS c+d PDYl where GPS- gross private savings, PDY private disposable income, NCS - non-contractual forms of private savings (i.e currency holdings, bank deposits bond and equity holdings etc.) and PDY =that part of PDY from Nvhich non-contractual forms of savings are made, The rationale for this procedure is 11, Alicia H. Munnell, The E]fect of Social Security on Personal Saving, Cambridge, Mass: Ballinger Publishing Co., 1974. TABLE 5 GROSS AND NET CONTRACTUAL SAVING (M$ Million) Gross Contractual Net Contractual I Gross Saving Net Saving* GoSaving Saving Through Through Through As % of I As Y. of As % of j As % of M Life Employees Total Life Total Gross Gross Grcss Gross Insurance Provident (1)+(2) Insurance (4)-+ (2) Private National Private National 0 - Fund i Savings Savings Savings Savings > (1) (2) (3) (4) (5) (6) (7) (8) (9) 1960 n.a. 72.6 72.6 n.a. 72.6 n.a. 5.1 n.a. 5.i O 1961 n.a. 77.1 i 77.1 n.a. 77.1 1 11.9 6.7. 1962 n.a. 80.6 80.6 J n.a. 80.6 11.0 7.8 11.0 7.8 1963 n.a. 86.6 86.6 n.a. 86.6 11.5 8.1 11.5 8.1 1964 n.a. 90.7 90.7 n.a. 90.7 10.4 8.0 10.4 8.0 > 1965 19.0 99.2 118.2 10.7 109.9 10.1 7.9 9,4 7.4 1966 ' 31.3 106.3 137.6 20.8 127.1 10.5 8.9 9.7 8.3 0 1967 18.7 111.2 129.9 5.7 116.9 9.8 8.2 8.8 7.4 Z 1968 28.8 85.6 114.4 11.1 96.7 8.0 6.5 6.8 5.5 1969 30.9 78.3 109.2 11.0 89.3 5.9 4.8 4.9 4.0 Z 1970 1 33.6 133.8 167.4 1 10.2 144.0 9.9 6.7 8.5 5.8 1971 37.3 I 182.9 220.2 8.8 191.7 12.7 10.7 11.0 9.3 1972 41.7 217.1 258.8 7.7 224.8 12.6 11.1 10.9 9.7 1973 92.7 211.8 304.5 17.6 229.4 9.4 7. 7.1 5.4 0 1974 50.7 220.4 271 1 1.4 221.8 7.9 5.1 6.5 4.2 C 1975 81.6 275.8 357.4 | n.a. 275.8 I 8.7 7A 6.7 5.7 > * No credit is extended by E.P.F. to subscribers. 0 Source: 1. World Tables, World Bank (several issue). 2. Quarterly Economic Bulletin, Bank Negara Malaysia (various issues). 3. Annual Report, Director General of Insurance (various issues). RESOURCE MTOBILIZATION AND ALLOCATION 63 based on the hypothesis that households regard CS as a tax, i.e. as a reduction in their disposable income, and not as a substitute for NCS. Under such circum- stances, CS will not be expected to have an adverse impact on GPS. It is assumed that households regard CS as a reduction in their disposable income by an amount PDY2, where PDYj+PDY2=PDY, Since households regard CS as a tax, then the marginal propensity to save in the form of CS from PDY 2 will be equal to one. In other words, CS=PDY2. It follows that the coefficient b, which is a weighted average of the propensities to save in contractual form plus in non-contractual forms, would be greater than or equal to d, 12 if households dlo regard CS as a tax. Rewriting equation (2) as follows: (GPS-CS) = c+d (PDY-PDY2) or (GPS-CS) = c+d (PDY-CS) (2) and comparing (1) and (2), it can be stated that if b ,d, households regard contractual savings as a tax. The marginal propensity to save in non-contractual forms will not, therefore, be adversely affected by the presence of contractual savings institutions. This analysis can be carried further in as much as it is possible to test the relative tax elements in savings through the Employees Provident Fund (EPF) vis-a-vis savings through life insurance companies (LIC). This is done by estimating: (GPS-EPF) e+f (PDY-EPF) (3) and (GPS -LIC) g+h (PDY-LIC) (4) Following the above argument, if households perceive a greater tax element in their savings through EPF than through LIC, the tax element in CS (which is a weighted average of the tax elements of EPF and LIC) will be between these two, with the result: f4< d Sh. 1 3 The regression coefficients are presented in Table 6 below. The fits are good. Both the hypotheses b d, and f( d < h are confirmed. 14 Of further interest is the result that f=d; therefore it may be expected that the tax element in the whole of CS is the same as that in EPF alone. Since EPF is the major component of CS, the tax element in EPF carries over in its entirety to CS. 15 In conclusion, it is noted that the impact on aggregate savings by contractual savings institutions has not been adverse, much of the savings through these institutions having been treated as a tax by households. 12. This, of course, incorporates the assumption that d, thc mnarginal propcnsity to save in non-contractual form, is less than one. 13. Following the above argument, the greater the tax element in a particular form of saving, the ess will be the marginal propensitv to save when that element is removed, Thus if EPF has the highest tax element, tlhen r, the marginal propensity to save when EPF is removed, is the least. 14. Furthermore, as is to be expected in any savings function, note that the intercepts a, c, e and g are negative. 15. This indirectly substantiates the assumption of CS PDY2 made in the estimation of the equations. 64 RESOURCE MOBILIZATION AND ALLOCATION TABLE 6 REGRESSION RESULTS Regressien (1) (2) (3) (4) a -1097.83 c 1052.99 c 1080.07 g 1071.31 (-9.20) (-9.25) (--9. 1 3) (9,33) b .33 d1*-2 .31 f- .31 IIh .32 (25.33) (24.54) (24.14) (25.59) R2 .98 IR - .98 R2 .98 R 2 .98 D-W 2.99 D-W - 2.93 D-W 2.97 D)-W 2.96 (t- statistics in parcnthescs). IV. INVESTMENT PATTERN OF CONTRACTUAIL SAVINGS INSTITUTIONS The impact which saving in contractual form has on the development of a country depends critically on how these funds are allocated. Certain general criteria guiding the investment of such funds can be specified. First, because Ilte obligations of these institutions are of a long-termll nature and are, to sonic extent, accurately determinable, it is not necessary for tlic,e institutions to maintain large liquid and short-term reserves. This is certainly the case for provident funds since withdrawals are permitted only on deaith, peltl1lahent disability or retirement at a specific age, and so can largely be pre-LCtIcrniined Estimation of the emergence of claims for life insurance companies is also possible with considerable accuracy and there is small possibility of unantici- pated withirawals. Secondly. funds should be invested in assets wvhich are botl safe and readily marketable to ensure the continued confidence of policyllolders and to be able to fulfill contractual obligations specdily. In the long-ruln there is a relationship between risk anid the subsequent rate of return of a portfolio, and by being aware of this relationship, the appropriate "risk-level" for the portfolio can be decided. 16 In Malaysia, the major contractual savintzs institutions are government- owned, so that the problem of maintaining adequate liquidity in the asset portfolio should not arise. However, the investment of these institutiolns has been conistrained, and they have been required to invest a major share of their funds in government or scmi-governnment securities. These constraints have channelled the funds of the contractual savings inistitutions into government- favouredl socio-econonmic development. This was in preference to industrial development through the direct purchase of securities of iec private sector, which was the role played by these institutionis in both the U.K. and tlic UJ.S. It can, however, be argued that had there been no such con.straints, the invest- ment pattern of these institLutions in Malaysia may lhave beeni siniilhr to thlat. of 16. Schwimmer and Mlalca (1976) have expanded oni this idSea in ihc context ol the portl'olio management for pension funds. "This risk le\ el is developied throulih a joint decision of the client and the portfolio manager. . . Ihfact. . . dilTerent types o companiies should have different risk levels dependinig upon the "maturitv" of ihe pen,ion, the Si7e of the company, and the composition of the company's la'oour force' (p. 132). On thiis isste, see also Harbrecht (1959): Martin J. Schwimmer ann Edward Malca, Pensiorl citic n tIstitu- tionalPortfolio Management, New York, Washington, London: Praeger, 1976: Paul P. flarbrecht, Penision Fituzds and Econotmtie Power, New Y ork: The Twentieth Centwry Fund. 1959. 'RESOURCE MOBILIZATION AND ALLOCATION 65 the commercial banks in that cotuntry. In view of the fact that the Malaysian commercial banks invested hea-vi!- abroad, a strong case could be mfade that the existence of these constraint . iifketively, channel these funds into uses ahicliT stirmulated domestic development. The allocation patterni of investments of the contractual savinlgs institutions in Malaysia is summarized in Tables 7 and 8. This pattern conforms to the regulations detai!e' in Section II, which primarily stipulate the minimum reg.ulations detailed in Section II, which primarily stipulate the minimum percen- tage of assets which must be in government or semi-government securities. Table 7 shows the distribui tioni of assets of life insurancecompanies. It isapparent that the insurance companies have, o)er the entire period for which data are available, allocated amaJor portion of their assets to investment in variOus types of assets to investment in various types of bCcurities (an annual aN erage of 54 percent over tle period 1963-- 1975). Holdings of securities were alniost equally ,]i.bl,eld between tho-se ikiied by N,!rioLS h[brnche.s of g(,%ernnment and those i by corporaftions (the annual average hol(dings over ihepriod (lN erec.sactlIy the same; 27.7 percent of total assets were government securities and the same viercentave were corporate securities). Holdings of gvernmment secutrities cmnprise those, issued by the F ederal Government of Mfalaysia, bv State and Local Gervcumnenlt' of &Ialaysyia. and by I r.eign, Gov'ernmiemt.s. Of tliese, "'ldinpc: of State and ILocal Go\eminient secrijil ics are e\ treiinelv small. 1- oldi n 1 1s .t forieo n nn-xci mjen t securit ic, have decliim(l' steadfily bothlini absolute termils and as a share of taLl aXssets fromii 23 pelcent in 1003 to less thaln 3 percent in 1975. By contrast, holdings of Mahlysiain Fedleral Governnment -ccurities lhave hliarply increased over the period in absolute amounts., and have rather more modestly increased as a share of total assets (from 15.5 percent to 21.9 perent'). Assets held in the form of corporate securities have risen marlkedly both absolutely and as a percentage of total assets. Unfortunately, no information is availaole as to tnc kind of business in which. the stock was purcha.. 'd. The remainder of the asset porrtfolio consists of various kinds of loans, and fixed and other assets. On average, loans have comprised 22 percelnt of total assets, most of these being policy loans. It is initeresting to note that insurance companies have held a larger share of their total assets in Mvlalav:lian ioxemrinient securities than statutorily reuilc%d. For examnlple, by the end of 1970, 1971 and 1972 they were required to inivest 10 percent, 15 percent, and 20 percenit, respectively of total assets in Malav1ian Go CrI11ent securities; the actual figures were 15 percent, 19 per- cent andl 23 percent, respectively. This dexelopmient can, to some extent, be aittributed to the interest rates olffcred oni Government securities (rancging bet%%een 5 and 8 pericenit on 3 to 20 year securities over the period 1956-74) xx idic appear to compare ravouirably with thosc available on corporate secuiri- ties. No data are available oin the latter, but an indication can be deriv(el from the fact that the average rates of intere.t earncd (both gross antd net of tax) oni ;assets by all life insurance companies has been 8.4 percent net (and 8.7 percent gross). From this it is quite clear that the rate of return on government securities compares quite favourably with that obtained on all assets held by life insurance Lo0nlpanies. Moreover, the low risk on government sectirities would enhance their attractiveness vis-a-vis corporate stock. DISTRIBUTION OF ASSETS OF LIFE INSURANCE COMPANIES (MS million) 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 m I In 0 1. Investments 71.6 73.6 87.1 t 90.9 i 104.3 125.6 146.4 174.1 20S.6 268.8 323.1 34 1, 385.0 cs (%increase) (n.a.) (2.8) (18.3) (4.4) (14.7) (20.4) (16.6 (18.9) (18.1) (30.7) (2(.2) (6.5) (I1.9) 0 (a) Goi'ernment Sec4rities: 48.0 47.3 55.8 42.4 47.2 50.4 51.6 72.2 91.4 I25.9 144.7 h16.9 179.7 g (i) Federal Gov. 19.4 19.4 18.8 19.1 22.1 23.5 25.9 48.4 69.2 104.4 123.3 135.9 151.5 0 (ii) Foreign Gov. 28.6 27.9 35.2 22.1 24.0 25.8 24.6 . .5 21.7 21.2 21.1 '03 20.2 (iii) State & Local Gover.ment . mentt See:uritis 1.8 1'2 1.1 1.1 1.1 V.3 1 0 0 ' 0 > (b) C'orporate ¢ Securities 23.6 26.3 31.3 4 .: 57.1 75 I Ui 14.2 142.9 178.4 7. 3 V. L.oans 23.7 26.2 2 4 53.3 64 8 9.2 93 IP) .y26 Z ' ,increasoe in.a.) 415) 443.1 (11 H4i 2 - 2.32 ; (a) Land and T ; 1 17. . > Loans n.a. 5.2 ; 7 '". K 22.0 24.2 ,- 9.3 bj foh :iann- n. 19.4 245 - l- 44* 4')2 5I n Ca. (cc Other na. 1.6 1 . 0.., 1.3 r1 f 1 l ' t l:.5 I8. . . ..-.. 3. Fi7xed & (rf; Nr e .Pt. 2. 4 .4 47.5 5 6. 5 51.1 63, .4 ,,r.I - :23. ..- (" increae) n.a. 14J (13 A.4) (:*. i. 91 111 1, 9) ' 41 208) (-. 4 20.8) i.i ( 1 ) Totili Assets 12:. I 141.7 172i. 192.4 21.7 2S I - 75/),* 321' - 448.0 (0nincr.,c) 1n.a.J (13.3) (I Ui.3i t13,7) I !1 . 1. (2'.61 03 Ii7., t .-2) .S'nmru - RJ2r:r) ! In-'rj::. - . 4n.atru Ru * r, uarios is%;ii;. rtan- Qer Ic 'rci '-:. b.' . -tri. . uis. TABLE 8 INVESTMENTS AND LOANS OF THE EMPLOY2ES PROVIDENsT FUND (M$ million) 1960 1 1961 1962 1963 11964 1965 1966 11967 11968 11969 11970 11971 1972 11973 11974 1975 In vestments 9 (% increase) 609.2 706.8 814.0 929.5 i10.9 1375.6 1554.6 1669.9 1891.6 (122.1 2432.4 2783.3 65.41352317 400 15 (a) Government SecuriLies: n.a. (16.0) (15.2) (14.2) (14.1) (14.1) (13.6) (13.0) (9.3) (11.3) (12.2 (14.6) (14.4) j3. |(11.3) (13.1.j (i) Federal Government1 I (ii) Foreign Governments 558.0 655.5 762.2 870.6 998.4 1148.4 I 312.1 1494.4 1639.8 1831.3 2062.0 2384.4 2734.6 3116.6 3474.9 3962.7 (iii) State and Local Government 46.2 46.3 46.3 46.4 46.5 46.5 46.6 46.7 46.7 46.8 46.7 34.5 34.5 34.6 34.6 34.6 Securities2 5.0 5.0 5.5 12.5 16.3 16.0 16.9 13.5 13.4 13.5 13.4 13.5 14.2 14.2 14.2 14.2 Loans 10.0 19.0 31.0 45.0 54.0 63.0 7I.0 74.4 94.4 89.1 96.2 99.6 98.4 103.3 1275.5 215.1 (% increase) n.a. (90.0) (63.2) (45.2) (20.0) (16.7) (12.7) (4.8) (26.9) (-5.6) (8.0) (3.5) (-1.2) (5.0) (69.9) (22.6) (a) Building Societies3 10.0 19.0 26.0 40.0 49.0 58.0 66,0 68.4 87.0 81.1 65.4 69.4 75.4 85.4 j125.4 157.4 (b) Other4 - - 5.0 5.0 5.0 5.0 5.0 J 6.0 7.4 8.0 30.8 30.2 23.0 17.9 50.1 57.7 Total Investments & Loans 619.3 725.8 845.0 974.5 1115.2 1273.9 1446.6 j1629.0 1794.3 1980.712218.3 2532.0 12881.7 3273.1 3699.2 14216.6 1. Excludes Treasury Bills; include Federal Government Securities issued outside Malaysia. 2. Between 1960 and 1969, obtained as the difference between Loans and a catch-all item including Loans -F State and Local Government Securities. 3. Loans on Mortgage + Debenture Loans for 1960-1969. 4. Includes investment in Land and Buildings, i.e., fixed assets. Source: (i) Bank Negara Malaysia, Quarterly Bulletin various issues. (ii) Lee, H.L., "Household Savings in West Malaysia", for "Loan" Figures only, between 1960-1969. 68 RESOURCE MOBILIZATION AND ALLOCATION The investment pattern of EPF is dominated by the holdings of government securities. The statutory requirement is that a minimum of 75 percent of total assets should be invested in federal government securities. In practice, this percentage has been much higher, ranging from 90 precent of investments plus loans in 1960 to 94 percent in 197517 (Table 8.2). This preponderance of assets held in the form of government securities contrasts with the investment pattern of the life insurance companies, which held a much more balanced portfolio, and the amount of corporate securities held was approximately the same as holdings of government securities. This difference in portfolio prevails also in other countries, such as India, Turkey and Sri Lanka, and one explanlation may be that life insurance companies have operated, at least initially, from the private sector, whereas provident fund have tended, in these countries specified, to be government-owned and government-managed. Finally, the funds generated by the Social Security Organization have also been invested primarily in long-term federal government securities (Table 9). It can be seen that the contractual savings inistitutions have become a major lender to the government for financing the latter's projects. Out of the total amount of government securities subscribed to each year between 1963 and 1975, these institutions have taken up significantly more than half, on average. Of the total outstanding holdings of government sccurities, these institutions have consistently held more than half of the total. H-loldings of the EPF consistently account for the bulk of these holdings witlhin the contraCtUlal aving institutions (Table 10). A major question which has been raised regarding the above pattern of investment is that the fact that the contractual saving institutionis are such an important subscriber to government securities does not necessarily imply a direct contribution towards development. However, some indication of their contribu- tion can be derived by assuming that if the government maintains a consistent current surplus overtime, or at least, if it maintains a current balance, then any borrowing which it undertakes must be for use in the capital budget. In the Malaysian budget capital expenditure items are all specified as "development" items, including defence expenditure. In fact, Malaysia's current account has been in surplus over the years. 1 8 It is possible, therefore, to derive some indication of the contribution of Malaysian contractual savings institutions towards development. The totalannual development expenditure of the Malaysian Government is calculated, excluding expenditure on defence since its contribution to developmenit is questionable. The annual ciange in holdings of government securities is expressed as a per- centage of the total, and it can be assumed that this percentage is the contribution to development. This percentage has been very hiigh in Malay:;ia, ranging 17. No data are available in fixed assets held by EP3F, making it impossible to arrive at total asset figures. It is for this reasozi that holdiings of dillcreniL assets have been expressed as totalloans and investments. 18. 1966 1967 1968 1969 1970 1971 1972 Current Revenue* 1667.0 1839.6 1891.0 2094.3 2400.3 2417.9 2922.0 Current Ex.penditure* 1619.6 1789.2 1798.8 1933.4 2163.0 2398.0 2912.0 * In million M$ RESOURCE MOBILIZATION AND ALLOCATION 69 between 25 % to 50 % approximately with the EPF contribution comprising the major portion (Table 11). It seems, therefore, that contractual savings institutions have indeed contributed substantially towards the socio-economic development of Malaysia. TABLE 9 SOURCES AND USE OF EMPLOYMENT INJURY INSURANCE FUNDS (1971-72) (Stock Figure: M$ '000) I 1971 17 oa (3 months) 1972 Total Sources of Funds Employer Contribution 73.133 2294,556 Investment Income - 2.574 2370.273 Uses of Funds? Benefit Payments +Fees 2.067 202.483 204.550 Cash at Bank 70.785 167.496 238.281 Cash in Hand .281 2.160 2.443 Investment: (a) 6 % p.a. Federal Securities 30.000 maturing 1977 (b) 7 ", p.a. Federal Securities 1805Q000 maturing 1992 (c) Fixed Deposit with Bank 90.000 Negara 2370.273 * Federally paid administrative expenditure, amounting to $1,053,900 in 1972 are not included here. Source: Annual Report of the Social Security Organisation, 1972. Finally, one may look at the question of whether contractual savings have affected the composition of investment since it is this composition which has clear connotations regarding any contribution towards development channels. Indeed, if one assumes that the alternative pattern of investment that would result in the absence of contractual savings would be similar to the investment pattern of commercial banks (households now saving mainly through the bank- ing system), then there is reason to believe that there is a vast difference in the two patterns. In 1972, the foreign investments of commercial banks in Malaysia amounted to 7.3 %,19 whereas for life insurance companies it was 4.7% and neglible for the EPF. There is no doubt that foreign investments do not contribute to domestic development. As such, not only is it clear that the compo- sition of investment has been affected, but also that it is in a welcome direction. Another index of the change in the composition of investment may lie in the ratio of private vis-a-vis public sector investment. Here again, while com- mercial banks invested 50.5 of their total assets in the private sector in 1972, 20 the EPF invested only about 3.4%/; however, life insurance companies invested 19. Foreign Securities, loans to and balance with banks outside Nfalaysia, and trade bills payable outside Malaysia as a percentage of total assets in 1972. Refer Bank Negara Nalaysia, Quarterlv Economic Bulletin, September 1973. 20. Loans and advances to non-banks in Malaysia plus corporate securities as a percentage of of total assets, TABLE 10 AMA LAkYSIA: HO 1)DI NGS OF FEDERAL GOVERNMENT SECU RIT IES' B3Y CONTRACTUAL SA\'1 -GS INSTI TU1T IONS 1963 1964 1965 1966 j1967 1968 1969 1970 1971 1972~ 1973 I1974 1975 Total Amount Suibscribed 223.2 1711.0 258.3 287.9 58'0.0 404.5 430.0 395.0 635.0 1045.0 11.53.0 950.0 11092.0 of which amount sub- scribed By: C'ontractual Sav-ings I Inistituitions 122.3 123.7 163.7 170.0 1242.0 153.9 199.8 261.3 313.3 548.4 506.2 363.0 497.6 EPF 110.8 1 118.2 157.8 162.6 234.2 146.1 188.3 232.1 298.3 503.1 464:0 328.5 463.4 Other Provdent FundS2 9.5 5.4 I 5.0 7.2 7.3 6.8 10.8 11.5 3.0 11 9.2 10.' 13.8 2.4 Insurance Cornpanies33.4 2.0 0.1 0.9 0.2) 0.5 1.0 0.7 17.7 12.3 36.1 32.1 20.7 31.8 Total Ouitstaniding Holdings 1384.0 1-557.7 1733.4 1932.5 2353.3 2711.1 3091.4 3479.5 4049.4 4835.6 5722.1 6444.21755 of which outstanding holdings By: Contractual Savings II Inrtitiunaii na. nz.a. n a. 1363.1 I 1563.6 1726.0 11941.9 2223.1 25,81.8 29719.8 3397.8 I3781.1 4285.9 EPF 888.04 'Il022.0 4 1169.04 1288.1 1469.8 1615.1 11806,6 i 038.8 2359.1 2 7 09. 1 3093.6 3451,9 3930.1 TPF In.a. n.a. 2.. -8.8 38.0 49.0 61.7 76.2 92.5 99.9 103.8 I105.7 102.3 Other Provident FundS2 nna n.a n.a 27.1 33.7 38.4 47.7 59.7 61.1 66.2 77.1 87.6 102.0 Life InsuranceI Companies 19.4 j19.4 18.8 19.1 22.1 223. 5 25.9 48.4 69.2 104.4 123.3 I135.9 151.5 ('lhange in Outstanding HT-oldings 218.0 173.7 175 7 199.1 420.8 357.8 380.3 388.1 569.9 786.2 886.5 722.1 910.3 of which change in outstanding lhoidin-s by:I Co,uia u nS z.a. n.a. ii.a. n.a. 200.5 162.4 259 812 3.7 9.0 418.0 383. 0. EPF 119.0 134.0 147.0 119.1 181.7 145.3 191.5 232.2 320.3 350.2 383 353 472 TPF n.a. n.a. n.a n.a. 9.2 11.0 12.7 14.5 16.3 7.4 3.9 1.9 -3.4 Other ProvidentI52 Funds2 n.a. In.a. n.a n.a. 6.6 4.7 9.3 12.0 1.3 52 10.9 10.5 14.4 CompIsuanies . . 4 2 2 0 5 8 2 5 Lifompuance n.a. I 0.0 0.6 0. . . 24 2 2. 5 2.. . 1. 26 1 1. Excludes Treasury Bills, which are not held by contractual savings institutions. 2. Other provident and trust funds, and Social Security Organization. 3. Includes non-life insurance companies. 4. Includes state and local g-overnmient securities. Source: Bank Negara Mfalay,sia, Aninual Report, various i-'ies; Bank Negara NMala)sia, Qutarterly Economiic Bulletin; various issues. TABLE 11 t3l CONTRIBUTION OF CONTRACTUAL SAVINGS TO DEVELOPMENT - 1966 1967 1 3968 1969 1970 1971 _ 1972 1973 1 1974 1975 Change in Outstanding holdings of federal government I I i securities by: I Contractual savings institutions (Smn) of which, n.a. 200.5 162.4 215.9 281.2 358.7 398.0 418.0 383.3 504.8 0 EPF 119.1 181.7 145.3 191.5 232.2 320.3 350.2 384.3 358.3 478,2 a TPF i n.a. 9.2 11.0 12.7 14.5 16.3 I 7.4 3.9 1.9 -3.4 Other Provident Funds* n.a. 6.6 4.7 9.3 12.0 1.3 5.2 10.9 10.5 14.4 E Life Insurance Companies 0.3 3.0 1.4 2.4 22.4 20.8 35.2 1 18.9 12.6 15.6 > Government Development Expenditure 477.2 478,1 512. 3 488.1 531.9 531.9 1007.4 1 986.2 1591.0 1615.4 5 Percent contribution of contractual savings to I . l -development expenditure of which, n.a. 41.9 31.7 44.2 52.9 42.2 39.5 42.4 24.1 31.2 Z EPF 1 25.0 38.0 28.4 39.2 43.7 37.7 34.8 39.0 22.5 29.6 > TPF n.a. 1.9 2.1 2.6 2.7 1.9 7 1.4 .1 9 2 Other Provident Funds I n.a. 14 9 .2 2.3 .2 7 .5 1.1 ,7 .9 Life Insurance Companies 0.0 .6 .3 T 5 4.2 2.4 3.5 1.9 .8 1.0 $ Other provident and trust funds, and Social Security Organisation. o t Excludes Defence and Security and General Administration from the government's definition. Includes developmental expenditure in g agriculture and rural development, industrial and mining development, transport, telecormnunication, utilities, education, health and family > planning, housing, social and community services. o Sources: Quarterly Bulletin, Bank Negara Malaysia, several issues. Z 72 RESOURCE MOBILIZATION AND ALLOCATION 52.8 % of theirs. Since the EPF comprises the major portion of contractual savings, we may conclude that c-, tractual savings instit-tions have altered the composition of investment betwe- the private and puolic sector. V. CONCLUSIONS Contractual savings institutions have not affected aggregate savings adversely and, indeed, may have a positive impact on them. If that is true, i.e. the total assets in the economy have increased, and since the additional asset form is primarily available to middle and lower income groups, then the wealth distribution in the economy must have improved, On the allocation side, it is quite clear that the composition of investment has been affected. First, these institutions have purchased a major bulk of new issues and have continued to hold a substantial share of all outstanding govern- ment securities. From our analysis, it seems clear that much of these borrowings by the government went to the Development Budget since the current account has run a siirplus on a trend basis. Since we find, further, that the government's allocation in the development budget conforms more towards socio-economic development than that of the banking institutions, we concluide that contractual savings institutions have contributed more towards development than would have occured in their absence. Since an investment pattern geared toward socio- economic development should lead eventually to better income distribution, we can carry the poirnt further to say that these institutions have contributed to better income distribution. The phrase "pension fund socialism", therefore, is well justified in the Malaysian case. American University, Washington, D.C., and The World Bank, Washington, D.C. Lab-wrndio ot'il, Vol.5, No. 1, January 1980 ,L11 ;d . w. iitxids in Singapore and 1sw I'.arthal.lsarahi Shome}zl. Assoceiate F rotfrssor of 'Feo.anoics. A mericran Un;ivrersity, and Nnsulltant, Wrildrr Bank, 'tWa,sh ilwt nn 02^.'q 141!ULl 5 ri'. s: i.. 'is a basic nit'ii- '. in huiman affairs. The development of socia*l /.t wr'rou ur'li.;u ''; 1>ti'l7t pr;;Nitlenlt funds and pension scheme.s reik.-.s this nmotive in) the7iir !' dulCrit'u! of the financial insecurity associated wvith In-t'. old itcc, and ir;lv&iti;v. i1 fe (mergenceM of such institutions, howtever, *\' n ;i5, i.s as a b1- itrtae the acctumulalrtion of financial resources, since the s'tr;* gtiC- .,ai..l;w Tsrls .ced heir p)avmn ets. In econlomies 'vne;lt^t laboutr force an,d ill&ncoms are rapidily growingt, the surpluses of such irns il^ 'i:ns typAically peiL rmn an1 important role in the inobilisation and ,'idl .v lion of r5e'.our'LtC . I-ht . y:ftr iS prnma rily concerned with the latter effect, n.nm. iv' the f4 lc8''.;A'I tOt5 nit .estrueent pBolicies of social security institutions on Jao..*ur and ocet; JI ';l'Ct ace f man .Tl d elovNA 1 untri es bears witness to the significant conIrI!tut!an whra h cotrctln, tU sehenies such as pension funds and other social Associate ssroft-is crn make, at a crucial devaelopmnt stage. by innovatively lIt c:' :ag the '.plus hindsi. 1 Although in many developing countries social ofe sIr al ,n1 are Ni ii liniit&h in opweraltion and coverage, in sme o ef them tl,: havelt itieid tdevelope into maojr tluasi-iinancial institu tions. We have .t% co.d two such cndnt%Jzs'. S.ini-e:Iotlre and the Phsiuclpinest ith u igniawevenr M1 .1 security ins:til ut ios.^ in h' rmn . of both} the working population covered1 as \%l eis volumes of resoura n zlioijised , ntIn( yet providing in1teresting contrasts in heir FC%)uitU as llion tlicie'c 11uiu lprimnary s.ocial securits schemne in ofiorc iS; _h. C entral Provident ondn (cPU, blegun in 1955, while theos in theiipprinesi are the Goserumnit Seice Insutranc System (GS atSi begun in . - .:1!o I nternational Labour ()rganisation iln: oernati edw Institute tfor Labur Studies) 1980 20 Laboutr atzlh Societv 1936, and the Social Security System (S5S), begun in 1954. Our analysis in this paper ref,crs to these systemis. In Singapore. 100 per cent of the working population was covered by CPF provisions by 1976, atnd savings through CPF comprised an average of 32 per cent of total housenold savings between 1961 and 1976. In the Philippines these figures were 48 per cent (in 1975) and 21 per cent (1951-1974 average) respectively. The impact which these substantial volumes of funds can make on the over-all economnic and social development of the country depends critically oIn how they are invested. Before examiniing the allocation patterns of social security surpluses, certain gvnieral criteria guiding the inivestment need to be specified. This is done in the next section. The actual experiences of Singapore and the Philippines are inalvscd in the last section, followed by concluding remarks. Inestnment criteria for social security finds It is generally true that the obligations of social security institutions are of a l'ilt ermn nature and are, to some extent, actuarily determinable. It may not be n ree ,> iherefore. for these institutionis to maintain large liquid reserves. For cxai.ple, in the case of provident funds, withdrawals are permitted only on death, permanent disability or retirement at a specified age, so that expected payments can be actuarily determined. This makes these funds suited to financing social projects with relatively long gestation lags. At the same time, genuine security and optimum yield on these investments are also iml-pcortant from the point of view of the contributors who have entrusted the social security institutions with the proper care of and return on their savings through some form of social contract. Thus the funds should be invested in assets which are both safe and easily marketable to ensure the continued confidence of contributors and to be able to fuifil contractual oblig,ations speed ily.2 Obtaining an optimal rate of return for policy holders may not be consistent with the government's perception of the investment pattern most conducive to econonmic development. The government may be sensitive to social and political dnemands for uses of what are often termed as "public funds". When social ~ecurity institutiions are subject to constraints on their pattern of investment, then this does curtail their administrative freedom with an adverse effect on cash mnanagement. In most cases, however, there are restrictions on inivestnment policies, and possibly with their own beneficial effects. Such restrictions should be acconmpanied with responsibility, however, as we explain below, utate influence on the funds A suir,ey of the use of social secuLritv fundls in some developing nations clearlv reveals the extent of State influence over these funds.3 Penision and Provident funds. hiistorically having been tinder the government's wing, are Social secuittfunds in Sin pauporw and the Philippines 21 subject to an array of stringent restrictions in terms of their investment policies. In Sin; t.pore. State influence on the allocation pattern of the funds has been very sit"n.itficant for the >tare, through a requirement that the entire CPF fund be investeet in g(g wernnikct stock, has enmbarked on an ambitious programme of mass housing, already cOveliiog 50 per cent of the population. In the Philippines, there is no xuch requirement, with investment in a fairly wide variety of a,ssets hav1ng be en auth(orised. If the State does, however exert this much influence ovqer the investment of surpluses, then it shouilld be obliged to a positive rate of returm, comparable to those from alternative investment possibilities. The State should bear the capital loss, if any, of social investmernts. Tbus State responsibility must accompany State control and regulation of the investment of the accunmulated surpluses. The ih3 rptlve carf,MFct lue5ion1ij Any attemipt towards the appraisal of the portfolio of social security inistitutions has to be looked at in terms of the investment opportunities which are available. This is particularly true when debating the allocation of funds to the public vis-;l-vi the prii .tt -t ei t. In developing countries, it is pertinent to ask wvhether or not; the :'.. siuctor offers viable long-term securities which woulld be appiroprinte pen, i5nn fiund assets. If the private sector offers primarily short-term securilties. tlen it Abviously cannot take advantage of the long-term horizon during which social security funds may be invested. The whole question of the limited absorptivc eapacity of the private securities market in developing economies needs to be assessed. It may, therefore, be unrealistic to criticise those social security schemes which have invested predominantly in government securities without properly assessing alternative investment opportunities. It may be possible, sometimes, to combine social aims with the financial imperatives in allocating the funds. In the long run, failure to adhere to sound financial principles benefits few. As in Singapore, social objectives may be successfully achieved subject to a minimum profitability constraint. In this senise, it is profitable to invest funds in specific projects rather than put them into a general budget sinice the latter's profitability is not so easily reckoned. In another developing country, India for example, the share of government secuirities in the total portfolio of social security funds, to be used for non- specific purpouses. lias been falling on a trend basis and direct, approved, investnments in water supply. electricity and other forms of social infrastructure h ave been rising. Impact on wealti and incoijie distkiliution Social securitv institutidoos can have a decisive impact on wealth and income distribution. Inasniuch as eLmnployers are contributing to social security funds, the m1aniagers of these funds are being entrusted with the function of distributing, -within society, the wealth which these employers, together with their employees. crezite. To quote Harbrecht :4 22/ /h?&'adStit financial i I'. ivii-s i ni ai fiIn&i'. iniS raiIbt'e . iiI' i-"i . prtnision trvPti' stand ready to take the' . ,! i;r I in 'ot w u tl-n !htC'ii sVNtil1 In iil xiu'1t ;aj 5L'.I'. ji>^ w:' ;1/df. ¢W E3^i ti!lw^ ( it; E I, investmcx i a 't, '' 'bto ris capital are TItIW r )tae "avo ra iv, ,f . ' 4 t ris k "u is intrign:rw to v tti,5nt if 1I ''. . . s '' iv ,ii'. , aflfl the inutual i]i '- i t na . -.iltnip us 'apiiallisrii, th . In thie eou ''! . . i *h conclusion mc'' *1nt114 f IT'coIII distritilbtioi , .4 iIs ai if'bio. is we shall see below Peop)l'. i; 'tl bx ' '''- ' 'i t 'i: tt . t r 'on prod1 t iigj wealth, and hi t 'ra pt ' be to trliot'iituc this kind . t . ' Isir's Ii corporatitins, PI *, r ii ,- . 5 h a :iLMfliCt icili'triinltg assets, jtil ti" , 'I ;. i Yt .'In . .. llt ii h 5 shlar e in the itie20fliC-''s4,'iXj0'" * h ;e as r ''r' "' -i>'3 3 3 l t'u''\hi; int 1t;' right dir: t'i' in ' .1 r .' s'1itV !idi'. I1aVT|!le 1v4lool , . *'; 'i ? ;'.,2i'f 4; X atAl4}{r;ti!\ I'ti 1i riOiIS, the 'liilir -M.- The te`tUa' ' 1)1. 1 . i1 1er '.d i ini ;i)ei and the Phlfipr"pine: - 1 : t u' holir aire these investnment p t.? ic 'eli e. . - 'I 'iid! 1k il.'tlar'e'id'iS f thle investment di. '" . ' ,wr 10 '' 1 ix' slineiitioi of the investment poi.o i-e hI . i iWud!'d il the investmer poflv- How ' aw . ii . erl.C'ir the social criteria . if v thie 1 cc .o,o' th e;'' i r, :rlJxt'i cr ft this section. Social securht; l,tliiciir4 4t74cO .Viigizpt're' f; c'IJr. "e :i' (, .,r' r - X , t ; Ct P F ( Pn iip.il t t'mj'eli r of both corporatoe anf'l sc';vm ni_ e , .;, rli. to Contribute: in tact. "any whom all eniplowve' is '' t . litii tlh p1.imeatoll of Wxages to aii igm 1'. i u i: i .! ' '' " 'd 'I 'Ii'^ ' lo 1ye (aI'11np1i'twO t' pay 15C.2 I L . 4 Ull tt .% ix'I. i, I I . o i>uth.;n, th e rate hIs i liig expanded grail iml V.i * . iT 'S CP1F i, admniiai;.tred by a hoard., eno'r' 'i0'' i V 'e' ' . 'F, . a. fi' governmilent, an'.l six other mienmbers,. Of thie ''itAw'' *il .t rI e.'oi Ailc'Ir5,. two repr,elit ernployers. andl two is ipr'r"ril en'pl' "o' I hi'!e dud is entrusted %;itlh policing the actual in;'c-swim f. ( Cl*Fi ' . 'tL .:rovp'rnnent supui,a;nrr. Social securityfunds in Singapore atnd the Philippines 23 A CPF member can withdraw the sum standing to his credit (plus a guaranteed minimum of 21/2 per cent interest) at 55 years of age or upon incapacitation debaring employment. The investment criteria for the interim funds were not clearly laid out prior to 1968, except that almost all of the funds had to go to government securities. In 1968 there was a major shift in policy, when a very ambitious urban renewal was undertaken with the objective of rehousing a large number of Singapore's population, especially those of lower income, in Housing Board flats. The CPF Board decided that low-income CPF members could borrow against their contributions to help finance their housing purchases. These funds could also be used to finance the purchase of a house or flat fiom the Jurong Town Corporation-a major urban development scheme begun in the early 1960s. The CPF Board could authorise the whole or part of the amount standing to the member's credit in the CPF to be transferred to the housing scheme as a deposit or down payment. The Housing Board operates two schemes, and people entitled to purchase a house or flat under either scheme can use their CPF contributions to finance such a purchase. The two schemes are differentiated by the quality of the flat and the income of the purchaser. Those people whose annual income is less than Singapore $1,500 may purchase small flats at a very low cost; those whose annual income is between S$1,500 and S$4,000 may qualify- under the M'viddle Income Scheme and purchase a larger flat or house in a better location, but at a higher cost. Those whose income exceeds S$4,000 are not eligible to purchase Housing Board property.6 The Philippines. Governznent Service Insurance Svstem: Legislation in the Philippines stands as a sharp contrast to Singapore in terms of the diversit of the investment portfolios of their respective institutions. GSIS provides life and property insurance as well as retirement benefits to government employees. through the Retirement Insurance Fund, and is compulsory at the national, provincial and local levels. GSIS was established in 1936, the earliest such institution among developing countries, for the express purpose of consolidating fragmentised pension schemes for government employees. GSIS is supervised by a five-member board of trustees appointed by the president and directed by the general manager, within the Office of Economic Co- ordination.7 There is no mention or practice of employee representation here in contrast to the Singapore case. GSIS engages primarily in insurance, lending and investment activities. It provides both ordinary and endowment life insurance. Originally, no employee was allowed to start a policy larger than his annual salary; this was later increased to two years' salary and, finally in 1966, the salary ceiling was removed. Even though GSIS is government-owned, there is very little government regulation regarding the investment and loan of GSIS funds. Purchases of government securities comprise a much smaller portion of its portfolio, than its counterpart in Singapore. GSIS has extended real estate, salary and policy loans, although the hierarchy in policy-making regarding such decisions is difficult to decipher. 24 Laibozur (itd Societ.y Social Security Syvstem: A government-sponsored scheme to provide old age, death, disability and sickness benefits to workers in the priaute sector, SSS is compulsory for all paid employees betweer. ages 16 and 60. SSS also covers specific contingencies such as invalidity and work injury for public sector employees, since GSIS covers only retirenment and life insurance. Established in 1954 and beginninig to operate in 1957, SSS met witlh ..poradic opposition from both labour and management, and was anmended several times since then.8 Indeed it has been an interesting and important exanple of how, initially, the introduction of a social insurance system may be unpopular not only among employers who are forced to contribute for the protection of their employees, but also among employees who must contribute. The latter being unfamiliar with the systenm, perlhaps perceive their contributions more as a tax on their income than as premia for future benetits. SSS is operated by the Social Security Commission. headed by the Secretary of Labour and consisting of six appointed members, who represent labour, management, and the general public. Basic SSS policies iin(d programmes are determined bv the comnmissior. under the guidelines of the Social Security Act (1954). SSS is authorised to invest in a fairly wide variety of assets, over time it has had a fairly diversified portfolio unlike in Singapore, where the bulk of the funds went into long-term, federal gok-erlnmLnt securities and the remainder went to the central ban k. In the Philippines the authlorised assets of SSS inelLide loans, government securities. private securities and deposits as well as direct loans. The Social Security Act authorises the system to give housing loans to either employee or employer to the extent of 60,00 pesos. In 1967. the ceiling was reduced to P 15,000 in an attempt to assure loans to lower-income groups. Repayment can occur within 25 years. Similarly 40 nmonthly contributions qualifV a contributor for a salary loan. Such detailed categorisation of loan conditions, however, does not reveal the actual extent of low-income loans as we shall see below. Thus while in both the Philippines and Singapore, low-income housing has been an inmportant feature in the invzstment motive of social security institutionis, in the former the attempt, through direct housing loans to contributors, has been truncated especially when compared with the latter, where it has been extremely successful through consolidation and direct participation by government. This is expanded upon next. Experiences in resoiarce allocation Singapore Since its inception in 1955, CPF of Singapore has anmassed assets of almliost US$2 billioti. The rate of growth of assets became rapic from 19 0 when an nuial increases began to exceed 20 per cent. Indeed, from 1973 to 1975, they exceeded 30 per cent annually (table 1). hlost of' these assets have been used for public sector expenditures. Not only has CPF invested more than 98 per cent of its assets in government securities in recent years but, in the process, it has also Socia acc fii inds in Sinigapore, and the Philippines 2 Table 1. Investments of the Central Provident Fund (S$ million) Government stock Advance Fixes Mortgages Accrued Total 2 deposit deposit on Singapore interest Singapore Foreigni with with property government government M AS t bank 1961 145.3 40.8 .-- - 2.3 4. 3 192.7 1962 1 . 2 14.7 29.2 - 1.5 3.6 236.2 1963 234.2 15.9 26.4 0.6 3.8 280.8 1964 304.5 15.8 4.1 - 0.3 6.5 331,1 1965 304.5 15.4 601.9 - 0.1 527 386.6 1966 41j.7 14.8 14.2 - - 6.8 447.5 1967, 488.5 11.5 1. - - 8.4 510.1 1968 5,40.-7 35.3 - - 9.7 .585.7 1969 630.4 - 39.1 - - 9.5 679.0 1970 752.3 -. 60.7 - - 15.3 828.4 19711 963,9 92.6 - 14.8 1 071.3 1972 1 158.6 201.8 - - 27.6 1 388.0 1973 1 634.7 -*171.7 --32.3 1 838.8 1974 2 036.4 381.0 - -59.9 2 477.3 1975 2 865.6 369.5 2.o 65.4 3 303.1 Prior to 1970 and the establishment oif thie Monetary Authority of Singapore. this consisted of advance deposit with the Treasury. 2 These investments comprise between 98 and 99 per cent of the total assets of CPF. The remainder of the assets consists of cash in hand fixed deposits. Source: Central Provident Fuind, Singapore. held more than 70 per cent of all government registered stock and has accounted for more than 60 per cent of government domestic debt in 1976.9 These indices of participation in the governiment sector have, in fact, increased steadily since the establishment of CPF in 1955. As argued earlier, State responsibility has to accompany such overwhelming investments in the public sector. In Singapore this seems to have been the case. CPF has earned a rate of return on its investments of around 5 per cent between 1966 and 1975. Since this is tax free, it compares favorably with the rates earned at commercial banks.10 It does, then, appear that contributors earn a return on their pr-ehiia, comparable to the alternate return they would obtain from the private sector, in addition to the benefits derived ft in their access to funds for housing. In termis of quantifiable results, under both the low- and middle-inicomne housing schemies, over 150.000 Singapore citizens have borrowed a total of US$396 miillion fr-om their CPF savings to purchlase thieir own homnes.1' As a result, more than., half of Sinigapore's 2.3 mlillion population have lived in rented or purchased public housing since 1976 and, of these, the majority had purchased the flats under the home ownership schemes. By 1985, approxima- tely 75 per cent of Singapore's population are expected to be housed in governmenit flats. 26 Labour anid Society Table 2. Investments and loans of the Government Service Insurance System (P million) Stocks and bonds Real estate & Salary Policy Devel- Other Total mortgage loans loans opment Government Other loans bank securities 1954 39.6 16.0 19.7 12.1 40.7 8.7 136.8 1955 71.1 37.3 32.7 16.8 17.6 9.8 185.3 1956 9(.8 75.9 35.8 20.4 2.7 16.0 241.6 1957 30.0 37.0 135.8 40.0) 23.9 16.9 11.2 294.8 1958 28.0 26.6 1 75.4 43.5 27.6 32.2 1.4 334.7 1959 57.0 40.1 1'8.7 57.8 31.8 34.0 1.0 400.4 1960 95.0 37.5 197.4 82.3 36.4 35.8 1.1 485.5 1961 89.0 62.5 223.4 98.8 40.3 37.6 3.7 555.3 1962 101.0 81.1 241.0 100.6 44.5 39.9 5.1 622.2 1963 144.0 74.3 269.3 125.9 50.5 42.5 21.8 728.3 1964 1S0.(0 92.5 359.8 146.9 58.9 44.7 38.8 891.6 1965 153.0 116.1 495.0 164.8 69.9 - 32.1 1 030.9 1966 130.0 148.1 b68.3 203.0 83.6 - 22.5 1195.5 1967 110.0 172.2 707.6 246.1 99.9 21.1 1 356.9 1968 127.0 201.4 808.9 264.1 119.9 22.9 1 544.2 1969 135.0 229.6 932.6 284.4 143.9 - 22.7 1 748.2 1970 94.0 258.3 1 041.1 346.1 167.1 - 23.1 1 929.7 1971 78.0 262.1 1 210.1 380.7 190.2 - 58.1 2 179.5 1972 122.0 220.1 1 430.3 357.1 215.8 - 114.9 2 460.5 1973 1 104.0 309.0 1 839A 382.8 255.3 -- 159.1 3 049.6 1974 169.0 213.2 2 38 1.3 n.a. 280.9 - 526.92 3 5'3.3 1 1973 and 1974. Figures are frorn Insurance Commissioner Annual Reports. 2 Includes salary loans. Sources: GSIS Annual Reports and Insurance Commissioner Annual Reports (1973.74, 1974-'.S). Clearly, CPF, acting as the financing agent for these public housing schemes, has successfully utilised its funds to house Singapore's low-income population. In this way, CPF has played a major role in the development of Singapore's social infrastructure and has made possible a general sharing of capital wealth created in the economy. The Philippines Unlike in Singapore where the investment pattern of CPF has been constrained into purchasing government securities, the portfolio choices of GSIS and SSS in the Philippines has taken a more diverse pattern, with a concentration on direct policy, mortgage, housing and educational loans to pG'icy holders. Recently, however, the investment policies of the Philippines social security institutions have been shrouded with lack of clarity and stand as a sharp contrast to Singapore. GSIS has, in the past, allocated a big share of its funds to real estate and mortgage loans, which reached a high of 67 per cent of total loans and Social securit fuinds in Singapore and the Philippines 27 investment of GSIS in 1974, a marked increase from the 12 per cent of twenty years earlier (table 2). Ml, annual average over the 21-year period was 45 per cent. The steadv increase in this percentage indicated the growth in the importance of GSIS in the provision of funds for real estate, However, to what extent these funds went to hotel construction and to what extent they went to housing cannot be estimated. Te remainder of GSIS loans and investments consisted of salary and police loans (an annual average of 24 per cent over the 21-year period), and stocks and bonds, which declined steadily, however, from 38 per cent in 1955 to l1 per cent in 1974. A recent change in the investment policy of GSIS has altered this allocation, with more investments in public sector bonds of the Development Bank of the Philippines (DBP), as the government has felt increasingly the need for these funds, with lagging tax revenue: GNP ratios for the economy as a whole. The direct participation of GSIS in a wide-ranging loan policy has, thus, been truncated. 12 SSS has also granted substantial real estate loans, averaging 34 per cent of total loans and investments over the last ten years (table 3). In recent years this share has been declining, and SSS has, instead, been channelling a growing percentage of its funds through DBP, as has been the recent policy of GSIS as well. DBP, begun only in 1970, has become increasingly important in the SSS portfolio, DBP bills amrnounting to 38 per cent of total SSS loan- and investments in 1975. Prior to 1970, SSS had, in a minor way, financed low-cost housing; however, this has now been suspended. Time and saving deposits and commercial loans have declined steadily as a component of SSS portfolio, whereas there is some progress towards socially-oriented investments such as community hospital loans, investment incentive loans and salary and education loans, while remaining small in the over-all portfolio. The share of private sector corporate stocks and bonds has remained small throughout the years of operation of SSS. On an average, between 1960 and 1973, GSIS and SSS provided nearly 70 per cent of total real estate loans extended by all Philippines financial institutions.13 Thus the role of the public sector in real estate finance in the Philippines has been extremely significant. Given the importance of real estate loans in the respective portfolios of GSIS and SSS, a closer look at just how much of the housing finance provided by social security institutions have been specifically for low income groups is called for. The nation's President's address on the occasion of the signinig of the Insurance Code stressed the point: 14 I am certain that the insurance companies are involved in the effort at reform.... One of our projects to meet recession is low-cost housing. I would like to invite you to participate in massive scale in this effort.... The public sector is already committed. GSIS and SSS.,. will put in P 5$$ million every year in low-cost housing.... Social security institutions, in conjuniiction with the National Economic and Development Authority and other task force groups such as PARAH, responsible for the resettlement of squatters, took up the task of determining the direction of development in the cities, the possible location of new towns and the use, of land, actions that could be taken on speculative dealings, etc. Table 3. Investments, loans and earnings of the Social Security System (P million) 19651 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 Investznent and loans 428.8 534.2 596.6 780.8 919.7 1 085.5 1 265.8 1 519.2 1 890.5 2 348.8 2 915.3 G (%o increase) (23.0) (24.6) (11.7) (30.9) (17.8) (18.1) (16.6) (20.0) (24.4) (24.2) (24.1) Time/Savings deposits 70.5 151.4 149.1 202.5 254.9 64.5 55.0 40.2 12.7 7.6 7.9 PNB/DBP notes receivable - - - - - 219.4 232.9 319.1 564.1 827.3 1 099.1 Treasury notes/bills 80.0 51.2 9.3 - - 35.0 109.0 300.0 332.3 421.6 617.4 Government bonds 41.2 41.3 43.4 75.2 116.2 114.7 113.7 107.7 110.0 99.6 104.5 Notes discounted - 1.5 1.4 2.4 2.9 - - - - - - Preferred/Common stocks 15.9 21.4 23.0 31.4 27.0 24.3 23.8 21.2 22.8 19.9 19.4 Corporate bonds 12.9 12.6 11.2 10.6 8.5 4.3 3.2 2.7 2.7 52.4 51.7 Real estate loans2 135.5 175.9 212.1 243.7 361.4 449.6 539.3 536.1 595.5 637.0 706.8 Salary/Education loans 19.1 36.2 55.3 68.9 90.6 115.1 130.6 133.8 164.6 209.1 240.5 Investment incentive loans - - - - 0.4 2.3 3.2 3.0 3.9 3.1 1.9 Commercial/Industrial loans 50.9 40.3 89.4 143.9 54.5 49.9 46.5 44.4 69.5 58.1 52.3 Community hospital loans - - - - t.4 4.6 7.0 9.4 11.2 12.1 12.9 National orthopedic hospital 2.8 2.4 2.4 2.2 1.9 1.8 1.6 1.6 1.2 0.9 0.9 Year end assets 471.8 583.5 696.3 820.6 963.2 1 128.2 1 330.6 1 572.3 1 949.2 2388.9 2 997.0 (f5 increase) . (25.2) (23.7) (19.3) (17.9) (17.4) (17.1) (17.9) (18.2) (24.0) (22.6) (25.5) Yearly earnings 19.5 27.4 32.9 40.5 52.0 66.4 79.3 79.3 128.6 173.3 220.1 (% increasei (19.6) (40.5) (20.1) (23.1) (28.4) (27.7) (19.4) (25.4) (29.4) f34.8) (27.0) Per cent return 5.15 5.87 6.00 6.06 6.31 6.85 6.98 7.40 7.84 8.53 8.73 t Break-up of investment and loan figures are not available prior to 1965; however, total figures are: 1957 1958 1959 1960 1961 1962 1963 1964 Investments and loans 6.0 28.2 66.2 108.S 158.6 212.9 275.4 348.9 (% increase) - (370.0) (134.8) (63.9) (46.2) (34.3) (29.4) (26.7) o Year end assets 6.4 31.3 68.2 110.7 165.7 225.2 295.2 376.9 (% increase) - (389.1) (117.91 (62.3) (49.7) (35.9) (31.1) (27.7) 0 Yearly earnIngs - 0.5 2.3 3.6 6.0 8.6 12.3 16.3 (% increase) - - (360.0) (56.5) (66.7) (43.3) (43.0) (32.5) Per "ent retuan 3.82 2.92 5.05 4.18 4.58 4.71 5.16 5.36 C° 2 Note that loans to households comprise only the first two categories. Sources: SSSAnnualReports. NEDA Statistica., Yearbook, 1974. Social security funds in Sinlougre 'and the Philippines 29 GSIS was supposed to be a source of finance for this over-all housing programmine from 1968. Hovever, this direct participation in the provision of hiolusinig was not sustainod, W.th more of the funds being appropriated by the DBP, as explained earliur. The result of this change in policy, however, has not indicated any d ireet or better results. Conclusion The relative merits of contrasting allocative patterns of the social security institutions in Singapore and the Philippines in terms of their impact on the financial, economic and social development of the country can be cited as follows. In terms of financial development, it is clear that, in the case of Singapore, the channelling of huge volumes of social security funds through government securities must have had a developmental influence on the securities market and Singapore has, after all, become the financial capital of the East. In the Philippines. however, the emphasis in the portfolio of social security institutions, in the early years, may have limited this effect, while having its own merits. In terms of economnic and social development, the results are, again, more obvious in the case of Singapore. State responsibility has accompanied State control in Singapoze. While direct loans to participants have their own merit, they may also increase bureauic-acy and show less concrete results in terms of benefits to a wide section of society, as seems to have been the case in the Philippines. In Singapore. the investment policies may have come more under obvious State supervision, but the positive results thereof have been more apparent. In the Philippines, while community projects, hospitals and education of participants' family members have been funded, they have remained quite small in terms of the total assets of social security institutions. The main investment has been in real estate finance. While, in the beginning, GSIS and SSS co-operated with other State institutions in low-cost housing, from published data, it is unclear as to what extent such participation has occurred and, to what extent, they have financed luxury housing and the hotel industry, all of which incorporate real estate finance. It is difficult to ascertain whether the President's coninmitinent to earmark substantial GSIS and SSS funds for low-cost housing has, in fact, been realised. Finally, these decisions have not been especially influenced by the participating public for all members of the board, of GSIS as well as SSS, are appointed by the President even though, two members of the latter are sUJpploseL t'-1 represent labour, and two, the general public. By contrast. thc CPF Authority in Singapore has centralised decision- making channels while incorporating a minimal emnployee representation. In practice, it is probably as constrained in this regard as the Philippine example. However, while the State has exerted extensive control, it has also acted with considerable responsibility. The result has been that CPF has been a major impact on the development of low-cost housing in Singapore. By acting as the 30 Labouir anid SocietY financing agent for public housing schemes, CPF has successfully utilised its funds to house Singapore's low-income population. In the process, a general sharing of the city-state's wealth, generated by corporations, public and private, seems to have taken place. Notes I To quote, "the efficiency of certain rnarkets, in particular those in the United States, the United Kingdom, the Netherlands and in Switzerland seems to stem to a considerable extent from the part played in these markets by intermediaries that collect funds on a contractual basis and invest them principally in long-term forms." OECD Committee for Invisible Transactions: Capital markets study, General Report (Paris 1967), p. 12. For an early review of this role in developing countries, see F. Reviglio: "Social security: A means of savings mobilisation for economic development", in International Monetat Fund Staft Pa pers, July 1967, pp. 324-365. 2 For greater detail, see C. Mouton: Social securitv in Alriica (Gerneva, International Labour Organisation, 1975). 3 In Argentina. Ghana and Togo. surpluses from insurance and pension funds cover large percentages of government current expenditures. Colombia requires insurance companies and mutual funds to invest in securities of the national government and several government organisations. In Brazil, too. the government has attempted to influence production and investment decisiors by channelling savings through intermediaries and controlling their allocation of credit. In Malaysia, 75 per cent of the fund's surpluses are required to be allocated in government securities; in Sri Lanka it is exclusihelh in government securities. See P. Shome: "The role of contractual savings institutions: A preliminary survey", in Studies itt Domestic Finance, No. 29 (Washington D.C., The World Bank, November 1976), and P. Shome and K.A. Saito: "The impact of contractual savings on resource mobilisation and allocation: The experience of Malaysia", in Malavan Economic Review. pp. 54-72. 4 See P. P. Harbrecht: Pension funds and economic power (New York, The Twentieth Century Fund, 1959), pp. 280-281. 5 See CentralProvident Futnd, CPF Board (Singapore, 1978). 6 This information was obtained from the Housing and Development Board, Singapore, while on a World Bank mission in 1978. Actual figures on the numiiber of flats constructed, etc., are not being presented here due to shortage of space. 7 See R.F. Emery: The financial institutions of South East Asia (New York, Praeger, 1970), pp. 463-466, and GSIS AnnualReports, for details. 8 Originally passed through the Republic Act No. 1161 in 1954, it was amended in 1957, 1960, 1963, 1965. 1966, and 1969. due to the fear of both employer and employee regarding possible declines in respective financial positions. See Emery. op. cit., pp. 467-470. and SSSAnnuial Reports. for details. 9 The time series on these are not being presented here to conserve space. 10 Comparable rates of return, at three-year intervats, were: 1969 1'92 IT97 CPF 5.1 5.1 4.8 5.1 12-month fixed deposit 6.0 6.0 5.75 5.79 Savings deposit 3.5 3.5 3.5 3.5 Source: Monc.,rY Auth.mit, {,t %ingap.re. l See Asian Wt'aill Street Journal, 7 December, 1977. 12 This was indicated during a World Bank research mission by the authorities, 13 The sources of total real estate loans, as derived from different financial institutions, have been calculated but do not appear in the text due to space considerations. 14 SeeAtinnalReport ol'the Insurance C170111nissioner. 1973-1974, p. 1. INVESTMENTS Of SOCIAL SECURITY FUNDS IN INDIA AND SRI LANKA: LEGISLATION AND EXPERIENCE PARTHASARATHI SHOME & KATRINE A. SAITO* I. IntroductiolA Social security institutions such as pension and provident funds serve the function of providing funds in the case of specialized contin- gencies, such as death, old age or retirement. These operations are usually accompanied by the accumiulation of large volumes of funds. As contributions exceed benefit payments, the surpluses generated by these sclhemes constitute a major financial instrument with the potential of exerting a suibstantial impact on the allocation of financial resources. In many developing countries social security schemes are still in their infancy, but in India and Sri Lanka, their respective Employee Provident Funds (IEPF for India, SEPF for Sri Lanka) have become major institutions in terms of financial resources mobilized, comprising 22 percent and 43 percent of total household financial savings respectively.' Hence a study of the investment patterns of the surplus funds of these institutions is of major interest.2 In most Asian countries the investment of social secuirity insti- tutions is strictly regulated by the central government, but the extent of government influence varies widely.3 In Sri Lanka, for example, all SEPF funds are appropriated by government in the form of * Parthasaratli Shome is Associate Professor of Economics, American University, Washington. D. C., Consultant, World I ank, Wasitington, D. C. and Katrine A. Saito is Economist, world Bank, Washingt-n, I). C. 1 Refer Shome and Saito (I 97Pb). 2 While the fiiancing mechanisms and savings generation aspects of social security institutions in developing countries have been studied in the past [Rtcviglio (1967a, 1967b)], the investment aspects of s& cial security funds are only beginning to be studied. 8 In Malaysia, 75 percent of the ftunds have to be in goverament securities [Shome and Saitb (1978a)]. In Singal)ore all of tie funds lhave to go into governmcnt securitics, which are then channeled into a massive governmcnt housing project covering more than 50 percent of the population, while iai the Philippines iavestment policies are flexible [Shome and Saito (1980)]. 349 350 INDIAN JOURNAL oF ECONOMICS government securities, In India, on the other hanid, stipulations reg,arding investments in governmcnt securities have varied over time, -and invevstments in alternative channels have also taken place. In Section- II, we summarize the legislation governing such invest- ments. If government does exert control over the way social security are itnvested, it muist be accompanied by government rresponsibility in, at least, two channi-Lels. First, government must see that contri- butors get a reasorlalble ratc of return on their inpayments, com- parable to those that they would have received had they alternately deposited their savings ii a cormmercial hank, bought slhares, etc. Second, the government becomes responsibleh for using the funds in tlhe l.est possible way for the couni-try's dcvelopmenit. For exanmple, xvhetlevr the appropriated funds are uised for development purposes, tliroutg.l either current or capital expendituires, or wlhetlher they are used simply for cotisunmption or defenice puiposes, is of tultimliate importance. Otlhrwise, tlhc issule arises as to wlhy tlhese fuinds shotldl nlot be iiivested according to the vill of the contrihbutors themsel\ es. 'T'hc olnly way to cm'iiIer;rt this a,rguiment is to slhow that govern- mnelt lhas acted responsibly in its invcLt1r stnut dccisions. Section III takes up these issues by sttudyin-g the actual investments of the social security funds of India and Sri Lanka. This is followed by a concllusion. II. Legislation Regarding Coverage and Investment of Provident Funds India The IEPF was formed in 1952 under the general supervision of the Nlitnist.r\ of Labour. The IEPF covers all employees in firms with 10 or more workers, after 240 days of employment. Another major provident ftund is the Coal Mine Provident Fund Scheme (CMPF), ni,der which cvery coal miner, bclow a certaini wage, is coveireCd. Ind..,ed the sticcess of' CMPF, establislhed in 1948, encou- raged tlhe government to enunciate formation of IEPF. For IEIPF, the insuired person contributes 6 25%" of earnings, 8 .', in factorieis with 50 or more workers, with matchlinig funds by employers (who also pay towards an administration cost amounting between 37',} to 2A4% of payroll). Government contributes 1-16%o of payroll, plus cost of administration for survivor benefits. The mininium earning for contribution purposes is Rs. 3 a day. IEPF SHOME AND SAITO: SOCIAL SECuRTY FUNDS 351 benefits are paid at 55 (miners 50), or at retirement, or permanent emigration. Gratuity Fund benefits is a lumpsum equal to 15 days' wages for each year of continuous service, with a maximum equivalent to 20 months' wages. For invalidity, a lumpsum of all contributions plus 7-5 /a interest are paid. Survivors, at the employees' death, get the benefit if nominated the beneficiary. Factories may be exempted from the statutory scheme and allowed to run their own provident futnds provided the total quantum of benefits, including gratuity, is not lets than under the statutory schemes and provided that they iinvest all providend fund surpluses, as prescribed for the statutory funds. The legislation governing the investible resources has undergone changes from time to time, with a declining stress on investment in government securities.4 According to the provisions of the Indian Trusts Act, fund monies are required to be invested in government securities or securities guaranteed by the central or state governments with regard to both payment of interest and repayment of principal. Certain discretionary invest- ments in first mortgages of immovable property and other authorized securities are allowed for certain categories of funds; in general prac- tice, however, trustees have preferred to invest primarily in govern- ment securities and certain other public sector investments, A provident fund formed by a company for its employees (1956 Act) is required to deposit in a post office savings bank or invest in specified securities. The Employees Provident Fund (1952 Act) is required to deposit in the Reserve Bank of India, State Bank of India or in any scheduled bank approved by the Central govern- ment, orin specified securities. They cannot beinvested inimmovable property unlike providend funds of companies. The power of invest- ment is vested in government, or in the governing body of the insti- tution. In general, the funds are invested in Trustee securities. In actual practice, therefore, investments by all provident funds are guided by statutory restrictions or directed by government authorities. Ouir analysis refers to the investmelnts of all provident funds. Sri Lanka Before the establishment of the SEPF in 1958, a few big firms provided provident fund benefits for their employees. The SEPF, on 4 The c,rmplicated rules and regtulations regarding provident ftund investments are not amenable to be presented here in an adeqaute fashion. HoA ever, they are available and the source is the Employees Provident Fund, rombay, India. Here we present con- cise irformation, just sufficient for our purpose. 352 INDLAN JOURNAL OF ECONOMICS the other hand, covered all establishments with three or more emplo- yees, thereby bringing in its folds an additional million workers durin g its first year of operation (1959 to 1960). By an amendment in 1970, all employees were covered. Firms which already had provi- dent fund schemes were given the option of joining the SEPF or to operate separately after obtaining approval of tlhe Commissioner of Labor. The employer contributes 9 X of the gross wage, and the employee (between ages 14 and 55) contributes 6OO Contributors are guaranteed a minimum of 2-1/2 ,. interest. Benefits are paid at retiremenit (55 for males, 50 for feniales) or for permanent invalidity or deathl. Every member receives an annual statement of his aCcomilt, anid rcceives a lumpstuni aimouInt pltis at least 3-1/2 x, com- pounid rate of interest as b;iieflits. The fund is available lor investment in goverlnment loans and other securities approved by the Monetary Board, The SEPF has, in praz -tiCe, inlested, solcly ill g )vernml ent securities. III. Inivestment Patterns of Social Security Fund Surpluses India Indian provident funds, including IEPF, started xith a major share of the:r invlstments in governmilenit securities (Table 1). How- ever, over the years, the. share of government securiti(es has steadily declinie(i fromn a high of 90 'i in 1960-61 to a low of 44 ,, in 1972-73, except for the var-year 1965 to 1966, when tlle ratio actually increa- sed by 6 'o. Indeed, cross-cotuntry data indicate that the purchase of government sce.uritic s has invariably increased during war-years by contracttual sax' ings itstittitio is (social securities and life insurance).5 The rise in the sliare of small savings, on the other hand, has becn phenomenal over the years, stai-tirng fror1 91O in 1960 to 1961 and reaching 47 ,b y 1972 to 1973. It must be remembered, however, that most of these small savings comprise postal ccrtificatcs issued by the governmenit. ThUs putting government securities and small savings togecher, one may conclude tlhat the providenit funds have invested prim.-rily in thle pnbl1 ic sc(.tor. Indeed, on tlle lasis of data uip to 1965 to 1966, a Research Programnmcs Comnmittec of thie l'lailninig CoImmlission (1970) fou(ld :6 51 1is has happetied iin devel')ped 2s well as devcloping countries., Refer Mohsin (Il96). 0 Rcfer Si-hlt (197)), p. 82. SHOME AND SAITO * SOCIAL SECURITY FUNDS 353 Investments in the corporate sector (shares and deben- tures) have been ignored by the Provident Fund autho- rities. Fixed Bank Deposits... increased... perhaps due to the higher rates of interest allowed by banks on such deposits in recent years. From our data (Table 1), it is clear that inve stments in ban} deposits increased dramatically in the ycars following 1965 to 1966; but have begun to fall again in recent years. In fact, bank dcposits lhave fluctuated the most among all the different lines of investment, but have shown remarkably hiiglher rates (approximately 9%,) in recent years (1971 to 1973) compared to very low figures (approxi- mately 25%') in earlier years (1961 to 1964). The above mentioned research project directly inquired provi- dent fund teustees/authorities regarding the factors affecting the choice of investments. The questionnaire mentioned Safety, Income and Liquidity: 77 6;', gav e Safety as the first preference, while only 6', picked Income. However, Iincomc wvas the second prefcrence for 66' ., while Liquidity received the largest proportion of third preferences (46' ,), emphasizing the ability of mobilizing long-term funds by these institutions, Thus, on the average, the fund autho- rities seemed to prefer Safety, Income and Liquidity in descending order. The subscribers' opinion on the same question was somewhat varied: 41"' emplhasized Safety, while 35 "j gave equal importancc to both Safety and Income. Only 6'%0 chose Income as the first choice. No one preferred Liquidity. Income-" ise examination further revea-, led that the emphasis on Safety decreases with increasing income-' groups, with a rise in the importance of the other two factors as well. People earning more then Rs. 12,000/- per annum cared more for Income than lower income groups. Thus it would seem that emplo- yees in the higher income brackets may be persuaded to subscribe to provident funds at higher discretionary rates if they can reasonably expect a satisfactory rate G return on their balances. Total loans and investments of Indian provident funds have comprised an increasing percentage of domestic capital formation, being an average of 2'21°o b)etween 1958-62, 2 46%M between 1963-67, 3 29 ' between 1968-72, and 3 86"',, between 1973-74- It remains to be seen to what extent the purchase of government securities by contractual savings institutions (provident funds and life insurance) 7 Refer Lal (1973) for figures on gross capital formation, 354 INDIAN J OURNAL OF EcoxNomics was chan!elled towards developmenit exp-en1(littli-res.8 Lookinq at data from 1968 to 1973 it is clear that the * unreit account has lbee in surplus (Table 2). This would iuuuply that g.vernuneit breOwinlg must have been for capitail e.xp)endlitutr. Thwse institutions prt)id- ed nearly 25 , tof goe le rnwiuug let eti 1967 and 1 972, and iearly 8', of capital (.X)ciiClitire betxx ee 1 9ti7 and 1973. It may be noted, however, that itt 1970 anI(t 1972, thesc shares dec- lined relatively. B1ut on the whole, they havre Inalitt,LiJz,d a steady share of domestic borrowing as well as capital expenditiiro inlspite of the steady decline of govcrnenvnt S(eCttie iCS itli t1i ) totil portfolio of investiiei lts oftlhese inistittitiouus. On1e inay conclide tlhat tlle p!) uclhase of goveriinionLt SCUVitiCi lias goii" twar(ls capital cxpemtditi tie; fuir- ther, there has been mainitaiiied a steady slare of this typc of expenditure finianiced by these inistitutions. Sri Lanka The SEPF hasi1ltveste(d enitirely in government sectirities jTable 3). It extend(Is no loans to policylholdeRi-4, and neitheir has it in- vested in private sector Secut'ities. Silnce its i1ic 'ptioil thie entire volumre of its iti\stible fuinids has been ciannelled (iel ctlv to the governinclet. Of the go% et'iuiiuent sccillities pnirchlased, most have been Rupee securities, anid only a small shiare have hecli Treasuiry Bills. The rate of return oni SPF investments has closely followed movements in the illte'rest rate on Treasury Bills, l)ut lias generally been slightly higher than the 'Freitsitry B3ill rate. In 1974, for example, the difernenrcc N%elas 1w 2 percentage poinits (5.0 rx comipared to 6 2%).9 The difference probably rflefcts the fact that SEPFportfolio of government stock is heavily weighted by Rupee loans, which have a sliglitly higher yield than Treasury bills. There is little doubt that con1trib)tors to the SEPF could liave received a higher rate of return on thliir prernia liad the SEPF diversified its portfolio, for at.XiIlI)i ill thle patternl of thle IEPF. The gross rate of return on thle invesLitents of private p)rovidelnt funds in Sri Lanka lias been substantially highler thlain eithier SEPF or the Insurance Corporation of Ccylon (ICC), al;thLlo;gh it lias 8 It Seenms n1oIC rtLjional to look at the role of contractoal % iiig7 il1,tititi'olh, ag a whole, in development. Of coursc, from data p)resented in Ttble 2, oun can calculate the contributionns of provirlent funds alorne. It is clear thlat while purcli lse of governmeAt securities bv LIC ihas increased st. adilv, with EPP this slhare has declined. 9 While the series on rates of return on the investmcnts of different financial iastitutions arc avilaable, we are not presenting them here to conserve space. SIO0tIF AND SAIro: SOCIAL SECURITY FUNDS 355 fluctuated much more widely. This clearly raises the question as to what exto it contributors to SEPF have a reasonable return on tlheirz pyinY1 lIts.10 Conlitractual savuigs institutions are the single most important instituitional l110llr of governiment debt, cuirrently holding almost 25'! - of (outstanding public debt (Table 4). This share has grown steadily from 11 in 1959, to 20,) by 1965, to 24%, by 1974. The niajor h1oldr ; are provident and penisioni funds, which hold around 183 :, lih islralice cGm1pallivs hold around 6!/{,. The fact that tllese instictutions reguilarly purobiase sulch a substantial volume of governmenit -'tcritics would ccrtainly help to stabilize the market for gover nient securities, as well as placing at the disposal of the go,ver merit a large volumie of funds. It may also, however, have liad a ii-egative imiipa-(ct on the mnarket for private sector securities in the sense of 'squleezing-out" the private sector from the market. This is likely to be true, given the prominence of these institutions as an investor and thc limited number of other investors. Also the funds have gone to fiiiaic, the gmvernment's net cash deficit rather than towsards investenciit spendinig. Suclh use of the fuinds may be questionerl in terms of its developmental inmpIact, since the funds hiave beecn used for Sii allanka's welFare system of extensive social services."l IV. Conclusion The prinicipal characteristic of the investment patterns of s,.cial secturity iiistittitioi1s is the emphasis on govecrnment secturities. While in Sri Lanka, all of the surplinses of its Employee Provident Fund have been invested in governmcnt securities, in India the shlare of govern ent secuirities has declined sigrnificantly over time, with riore going inito other forms of investment. Investments in government securities have, however, provided a steady sliare of (Iomi1estic capital developmental expenditure in India. TwIced, in termys of contribution towards domestic capital formationi, conitractuial savings institutions as a whole, have shown 10 letween 1971 anid 1974 the average rates of return on .he investments of SEPF, private prov:;clnt fuin(ds, ITC anc1 Treasury l'ill hax e been 5.6",0,, 5-0,', (for 1971 only), and 5H re'iw'c tivc1'. . 11Thiis, (if course, is a trid iioiial conclusion. Contrarily, one increasingly comes awcross the arre,im1etLt that a IlW,ltliN and well-fed population should have a rmuch bigger iinpact on development in the long ruin. Tn that case, Sri Lank-a's use of its social iecurity fuids has been ju'liciou'. 13 356 INDIAN JOURNAL OF ECONOMICS an increasing percentage contribution over quinquenniuins from 1958 to 1972. In Sri Lanka, contractual savings inistitutions are the largest instituitioIIal purlihsatr of government stock, even though such purchases have. l had no dIirect (levclop-nental impact. They have been used to finanice uip to one-third of the government's net cash (India) Table 1 INVESTMENTS OF PROVIDENT FUNDS* (Flows)** (Rupees in Crores) Investments CGovcnn-nmen t S-mall Bank Others Year Tttal Fund Securities Savings Deposits (1) (2) (3) (4) (5) (13) 3960---61 56- 1 50 !) 5-2 - - 1961 - 2 65.5I 52- .5 12 - 2 0 * 2 0.2 1962-63 77-] 58,-9 14- 7 0.1 3- 7 1963- 64 90 - 2 67 -9 20 - 7 0*3 1-3 1964-65 113-8 84-5 27 -9 0- 6 0-8 1965- 66 122 -4 98-3 20-4 4-2 -0.5 1966-67 138 -3 112-9 21-8 2-5 1.1 1967-68 155-9 96-4 12-2 45- 8 1]5 1968-69 175 - 9 100-0 32-5 37- 0 6-4 1969-70 200 8 94-5 87 -7 15*0 3-6 1970-71 252-9 120 -7 82-4 43 -8 6 0 1971-72@w 287- 1 133-2 110 7 26-5 16-7 1972-73 w 336 -7 147 1 156-8 30-1 2-7 * Include tVie provi(dent fund of the employees of,- 1. [Ze. erve P'anlk of Tndia 2. Tnsuiraice Companies 3. Local aitllioriLirs 1. I'actory establishments covered by the Emiployces Provioleut FIundAct. 5, Coal mines 6. Assam Tea Plantations and 7. llanks ** Tflie aw-r-re'aLe investmeit of provi lent fund accumulations stood at Rs 2915-84 crores on 31 st M.irch 1976. Those of x-enipti1 cstalWihnienti being iP s. 1597 14 crorcs. Su rovi:onartol. Souirce: I)c partnient of Stati,.ticq, neserve.RanA; of 1I*dia, B?ombay, Irtndia, SHOME AND SAITO: SOCIAL SECURITY FUNDS 357 deficit and the bulk of the current expenditure is on maintaining the extensive welfare services of Sri Lanka. The funds raised by these inrstituti,w.; thus [nostly flow back to the household sector in the pro..irion of social services, ilihout any direct contribution to financial, industrial or agricultural development. Finally, the impact on the rates of return of a more diversified investment portfolio is clear. The Indian Employee Provident Fund and private provident funds have both invested widelv; in Sri Lanka, howexver, only the private provident funds have a diversified portfolio. The result has been that the rates of (India) Table 2 COmrRACTUAL SAVINGS INSTITUTIONS AND DEVELOPMENTAL EXPENDITURES (Rupees in Crores) 1967 1968 1969 1970 1971 1972 1973 Capital Expen- 1760 1960 1990 2270 2700 3040 3280 diturel Sources: Current Surplus 480 940 1000 1070 1190 1400 1680 Foreign Sources 10 80 30 40 60 20 - Domestic Borro%v- ing 580 530 420 820 1000 1260 n.a. Foreign Borrow- ing 690 410 540 340 450 360 ii.a. Parchase of Gov't, Secu ri ties (a) LIC 27'57 53 03 56'93 62-45 94 53 88.71 115 45 (b) Provideult 112-90 96A40 100 00 94 50 120 70 133 20 147410 Funds Total 150'47 149 43 156 93 156 95215-23221.91 262 55 This as ((, of: (a) Domestic Borrowing 25'94 28 19 37 36 19 14 21'52 ]7 61 - (b) Capital Expenditure 8 55 7 62 7 89 6 91 7.97 7 30 8 00 1 Agriculture, :d ucation, HIcalth,l Transportation, Comrnmunication, Industry and Power, Housing, Loans and Advances, others. Sources: Table l World Tables, The World Bank (1976); LICI Annual Reports, several issues; EI'F Ann1ual Reports, several issues; Lal (1973). 358 INDIAN JOURNAL OF ECONOMICS return on their investments have been much higher than those of its Employee Provident Fund. This indicates that a diversification of the portfolio of public institLutions would, givc the participaints a hiaher returin on their contributions. (Shri Lanka) Table 3 INVESTNMENTS OF TEIi EMPLOYERS PROVIDENT FUND AND THEIR RATE OF RETURN (Rupees in Alilliutt) Total Investmeint Computed (entirely in Rate of government Income -from Return securities)' Investments 1959 9-376 0 108 1.15 1960 62!676 1 212 1 93 1961 121h787 3 562 2 92 1362 183*996 5 983 3!25 1963 245 371 8*758 3*57 1964 311*801 11-611 3 72 1965 382*095 14 771 3'87 1966 459-804 18$172 3 95 1967 542*937 21 778 4 01 1968 641-033 26111 4 07 1969 742'768 31 614 4-26 1970 840-446 37-554 4.47 1971 1001 012 48-211 4*82 1972 1157-600 63 659 5.50 1973 1354 135 80'090 5-91 1974 1554'375 96 995 6-24 1 3oth Rupee Securities and Treasury 3i1is. Sources: EPF Annual Reports, several issues; Centrcl Rank Anntal Reports, several issues. SHOME AND SAITO: SOCIAL SECURITY FUNDS 359 (Sri Lanka) Table 4 THE FINANCE OF THE GOVERN\SENT DEFICIT BY CONTRACTUAL SAVTNGS INSTIIUTIONS (Rups s in Million) Purchase of Government Governricit (1) / (2) Secuirities by Social Net Cash Security Institutions Defcit (1) (2) (3) 1960 73*7 417-5 17 7 1961 105 4 462 5 22 8 1962 94.5 456-1 20 7 1963 115-1 391*7 29A4 1964 124*2 4617 26 9 1965 144-0 430 4 33.5 1966 172*2 566 0 30 4 1967 174 5 606'8 28 8 1968 179 4 715 7 25 1 1969 176-1 787-6 22 4 1970 198-5 935'6 21 2 1971 246 2 1083 3 22 7 1972 321*9 1294 7 24-9 1973 310Q4 991*7 31*3 1974 323 8 1034*8 31*3 Source: EPF Annual Reports, several issues. 360 INDIAN JOURNAL 07 EooNOMIas References [1] Lal, D., Capital Formation in India, Ph. D. Thesis, University of DIelhi, 19 73. [2] Molisini, M., Investment of L-fe Insurance C(orjoration's Fund, Ali- garli Muslimii University Press, India, 1966. [3] NCAER, Contractual Saviig ian Urban J.dia, New Delhi, July 1963. [41 Reviglio, F., "Tlhe Social Security Sector and Its Financing in Developing Counitries," IMIlP'Staff Papers, November 1967 (a). [5] Reviglio, F., "Social Security: A Means ofSavings Mobilisation forEconomic Dt-velopmniti," AIMF Staff Papers, July 1967 (b). [61 Shome, P., "The Role of ConItractual Savings Institutions: A Preliminary Survey," Studies in Domestic Finance No. '9, The World Bank, Washington, DC, November 1976. [7] Shome, P., -Social Security as a Mobiliser of Savings," Inter- national Social Se-curit, Review, 11.0, No, 1, 1978. [8] Shome, P. aiid K. A. Saito, "The Imptct of Contractual Savings on Resouirce Mobilisation and Allocation: The Experience of Malaysia," Malayan Economic Review, April 1978 (a). [9] SlhOmP. P. anid K. A. Saito, "The Impact of Social Security Institutions on Household Savings: The Asian Experience,,' presented at tlle Anniual Amterican Economic Association Conference, Chicago, August 1978 (b). [10] Shame, P. anid K. A. Saito, ,Social Sccurity Funds in Singapore and the Philippines: RainificaLiolis of Investment Policies, Labor and Society, International Institute of Labor Studies, Geneva, January 1980. 11] Singh, D., "Ain Inquiry into the Resource Potential and Invest- ment Policy of Provident Funds," Findings of the Project, Planning Commission, New Delhi, 1970. THE WORLD BANK Headquarters: 1818 H Street, N.W. U Washington, D.C. 20433, U.S.A. European Office: 66, avenue d'Iena 75116 Paris, France Tokyo Office: Kokusai Building, 1-1 Marunouchi 3-chome Chiyoda-ku, Tokyo 100, Japan The full range of World Bank publications, both free and for sale, is described in the World Bank Catalog of Publications, and of the continuing research program of the World Bank, in World Bank Research Program: Abstracts of Cuirrent Studies. The most recent edition of each is available without charge from: PUBLICATIONS UNIT THE WORLD BANK 1818 H STREET, N.W. WASHINGTON, D.C. 20433 U.S.A.