No. EC 31 This report is not to be published nor may it be quoted as representing the Bank's views. INTERNATIONAL BANK FOR RECONSTRUCTION AND DEVELOPMENT PRIVATE SAVINGS FLOWS AND ECONOMIC GROWTH March 29, 1954 Prepared by: Mervyn L. Weiner Economic Staff PRIVATE SAVINGS FLOWS AND ECONOMIC GROWTH SUMMARY The importance of domestic savings needs little emphasis. In the long run, the rate at which a country saves is-probably the basic determi7a- ant of the rate at which its capital stock and thus its income may be expee- ted to increase. Measures to increase the volume and improve the utilization of a country's local savings are thus of crucial importance for its develop" ment programming, The purpose of this paper is to illuminate some of the principal objectives to which the private savings part of domestic savings programs should be addressed. The rates at which different countries save reflect, of course, public as well as private savings decisions: savings are generated by public authorities as.well as by individuals and by businesses, and government savings decisions affect the volume and composition f private savings. The problems of encouraging and mobilizing private savings remain sufficiently distinct, however, to warrant separate discussion. From the point 6f view of r1ising the rate of private savings the individual and business components of the private savings stream have to be distinguished. Individual savings are probably unlikely to be overly responsive to measures to raise their rate. Business savings, by contrast, can probably be increased substantially by measures which raise profit expec- tations and thus inducements to invest. Raising rates of business savings is not an unqualified gpal in itself., however., As business savings are usually used for investme4t Within the firm which does the saving and therefore claim rather than release resources, stimulating business savings does not make savings available to finance investments elsewhere in the economy; it may even make investing elsewhere more difficult, Measures to improve the utiliz- ation of savings by inducing savers to save in forma which permit the transfer of resources to investors other than the savers may thus be just as important as measures to increase the volume of savings. In order to elarify the problems of improving the utilization of private savings by inducing savers to change the forms in which they save, private savings are usefully classified as "retained" and "transferred". "Retained" savings are defined to include savings, the control and final investment of which is retained by the saver4 "Transferred"-avings include savings whose control and ultimate investment are transferred to others As "retained" savings represent resources which are already committed to particular uses determined by the saver himself while "transferred" savings represent savings which can be pooled to finance investments in quantities, varieties, and qualities beyond the wealth and entrepreneurial ability of any single saver, the-advantages for any country of enlarging the volume of Savings made in transferable forms are obvious, "Retained" savings may be important. Reinvested business earnings and self-financed improvements on ii farms are only two of the most obvious categories of "retained" savings that have always been and will undoubtedly continue to be major sources of capital formation in all countries. The range of investment possibilities that can be financed from any given volume of savings is increased immeasur- ably, however, once the pooling of separate savings streams is encouraged. The value of enlarging the "transferred" savings stream for the purpose of improving the utilization of resources in underdeveloped countries is thus apparent, quite independently of the rate at which these savings may be accumulated. For the most part, enlarging the "transferred" savings stream consists of correcting for those negative factors which cause savers to choose to retain control over the use of their savings. The most important of these are: (1) lack of convenient savings outlets, (2) lack of convenient savings instruments, (3) lack of trust in others, (4) fear of illiquidity, and (5) fear of a decline in the real value of money. A variety of devices can be suggested to remove these barriers to the growth of transferred savings, but the effectiveness of these devices is ultimately dependent on the appropriate- ness of the environment in which they are introduced. As the economic aspects of this environment are to such a large extent a reflection of past and antic- ipated monetary and fiscal policies of governments and central banks, these policies are as intimately responsible for the success or failure of measures to influence the rate and utilization of private savings as any savings "gadgets" can possibly be. CONTENTS Page I. INTRODUCTION .............................. 1 II. TOTAL DOMESTIC SAVINGS ......................... 2 A. Gross National Savings Rates ............ 2 B. Three Savings Streams ................... 3 III. PRIVATE SAVINGS ................. ......... 6 A. The Rate of Savings ..................... 6 B. The Form of Savings ..................... 8 IV. CONCLUSIONS . . .. ............................. 15 I. INTRODUCTION The importance of local savings for financing economic development needs little emphasis. In the long run, increases in income are directly related to increases in the stock of tools, machines, houses, factory buildings, schools, irrigation systems, roads, power plants, etc., that a country has at its dis- posal. Gross domestic savings must therefore be sufficient to finance invest- ments both to maintain existing productive plant intact and to provide for additions to this stock of capital. Even where a country's income rises sub- stantially because of improvements in its terms of trade or because of increases in the degree to which its existing productive capacities are utilized, the permanence of these increases in income is in large measure related to the rate at which the community chooses to save and invest rather than consume out of this income. Of course, a community may not choose to save, preferring higher (or unreduced) present consumption to the prospect of higher consumption in the future. It may also export its savings rather than invest them at home. All things considered, however, the higher a country's domestic savings, the better will its income prospects be. The fundamental importance of domestic savings is in no way altered by the fact that foreign capital inflows may make possible a more rapid increase in a country's capital stock than it could finance from its own savings. As international loans are unlikely to be made to borrowers who have little pros- pect of increasing their future savings by amounts at least sufficient to service these loans, savings are as necessary a condition for capital inflows as for domestic investment. The rate at which a community saves isthe end result of three separate sets of savings decisions. Savings may be generated by public authorities in the form of current budget surpluses (surpluses of gross revenues over current expenditures) and retained earnings of government corporations; savings.may be made by individuals; and savings may be made by businesses in the form of retained earnings. Each of these sets of savings decisions differs both as to motives and the influences to which it responds. Measures which are intended to influence the rate and composition of domestic savings must therefore be designed in terms of their effect on each of these components of the total savings stream. As the purpose of this paper is to illuminate some of the principal objectives with which the private part of a domestic vavings program should be concerned, the following discussion deals primarily with the individual and business components of total savings. The volume and composition of private savings is, of course, not independent of goveriment d-cisicns as to how much the community should save compulsorily and how tne public revenues should be collected. Even if a government savings program takes this relationship into account, however, the problem of encouraging private savings and influencing its composition still remains. It is to this aspect of domestic savings that this paper is addressed. - 2 - The paper is divided into two main parts. To give some perspective to a discussion of only the private component of total savings, the first part deals with domestic savings as a whole, their magnitudes, and the place of private savings in this total. The second and principal part deals with private savings alone-the forms in which they are.made, the factors which affect their volume, and measures that may be taken to influence their compo- sition and volume. Throughout, the emphasis is on the long-run development of domestic savings. The encouragement of savings and the development of new institutional savings mechanisms is a long-run problem, even though unex- pected changes in income may cause savings to fluctuate widely in any country during short periods. II. TOTAL DOMESTIC SAVINGS A. Gross National Savings Rates A comparison of the rates at which a number of different countries have been saving in recent years reveals wide differences between national savings efforts. Savings refer here to realized gross domestic savings, which are equal to gross domestic investment plus (or minus) the balance of payments surplus (or deficit) on current account. Some countries have been saving in one form or another as much as 25% of their current income; other countries have been saving slightly less; and still others, even less than 10%. As aggregate statistical savings figures are not strictly comparable, one must beware of pressing comparisons too far. The figures represent gross savings rates and thus include, in addition to new capital accumulation, vary- ing provision for the replacement of worn out capital stock. They contain estimates which are not all equally reliable. The ratios are not all based on the same definition of capital formation; they refer to the short postwar period and reflect in varying degrees the peculiar pressures of this period. Even after allowing for these qualifications, however, the spread between gross national savings rates appears to have been rather large, of the same order as the range of savings rates that is observable in earlier periods in those countries for which long-term records of capital formation are available. -3- Gross Savis Per Capita Rate Income Group 25% or more Australia II Finland IV Iceland III Norway III 20 - 25% Canada I Denmark II Germany (F.R.) IV Italy IV New Zealand I 15 - 20% Argentina IV Austria IV Brazil V Burma V Ceylon V France III Honduras V Peru V Sweden II United States I 10 - 15% Belgium III Chile V Dominican Republic V Israel IV Mexico V Southern Rhodesia V United Kingdom II Under 10% Ireland III Philippines V 1/ Sources: UN, OEEC,.ECXA, and IBRD data.. 2J UN estimates for 1949, grouped as-follows: I - over $800; II - $600-800; III - $400-$600; IV - $2004400;: V - under $200. B. Three Savings Streams There are probably many ways-of explaining these differences in rates of savings. None is obvious, however;. One might expect, for example, to find some relationship between per capita income and.gross national savings rates. Certainly, the higher a country's per capita product, the easier*it should be for that country to save at any given rate.. Even after allowing for conceptual differences in the statistics, however, no firm.relationship of this kind -4- appears to exist. In recent postwar years, as wide a variety of countries as Germany, Italy, Norway, Iceland, Finland, Australia, Canada, New Zealand Argentina and Brazil saved at rates equal to or higher than the United States. Most probably, because national savings are made up of the savings of govern- ments, businesses and individuals, and because substantial differences un- doubtedly exist between countries as far as the relative importance of each of these three savings streams is concerned, 1/ national savings rates are ex- plainable only in terms of the variety of factors which relate-to each of these components of the total savings stream. The problem of analyzing national savings rates for the purpose of better understanding how they may be increased is, of course, simplified if only one of these components of domestic savings is important. -Thus, in countries in which business and government savings are automatically limited because the economic role of government is circumscribed and because businesses are under pressure to distribute a large proportion of their earnings, national savings rates might be explained in terms of individual savings alone. 7 For underdeveloped countries, however, such simplification is probably not warranted, Government and business savings are undoubtedly at least as important as indi- vidual savings in most of these countries. More important, they are probably the savings streams that can be most substantially increased. The reasons are the following. A country's savings can increase only at the expense of consumption or out of additions to income. If incomes are not increasing and new savings have to be made at the expense of consumption, these savings will almost certainly have to be extracted by the government through its fiscal system, except in the very unlikely event that the whole community can be persuaded to reduce its levels of consumption. If incomes are increasing, there is a greater probability that new savings can be induced: consumption need only rise less rapidly than income, or at the limit not increase at all; it need not fall by some absolute amount. The average rate of total savings will increase as long as new savings out of income-increments are made at a higher rate than average savings were made in the past; i.e., as long as the marginal savings ratiois higher than the average. Even here, however, increases in individual savings are likely to be limited. The pressures of absolutely low incomes, compounded by the knowledge of higher consumption standards prevailing elsewhere, all make for additional individual incomes going into consumption. New savings can probably accrue in significant amounts only if the stimulus of rising incomes induces the business community to save and expand its activities and if the government saves more by increasing its public reveiues while keeping its own current expenditures from rising as rapidly. 2/ Unfortunately, statistical comparisons cannot readily be made. / S. Kuznets, "Proportion of Capital Formation to National Product", American Economic Review, Vol. 42, No. 2, May 1952, pp. 507-526. -5- Some increase in aggregate individual savings may be brought about by redistributing incomes from the low income consumption-oriented groups in the population to the "automatic savers" 1/ in the higher income groups. Even if this were acceptable on other grounds, however, the extent of the increase in individual savings that.can result from redistributing incomes cannot be expected to be large. First, the numbers of "automatic savers" in underde- veloped countries are small. And second, a redistribution of income, even if it only results in a relative rather than an absolute decline of incomes accruing to the lower income groups, may well increase the numbers of dis- savers in the community and thus partly offset the initial effects of the income transfer. If the thesis that individual savings are determined by relative incomes as well as by absolute incomes has any applicability to underdeveloped countries, the effects of any redistribution of income to the higher income groups may well be to redistribute the ownership of property rather than increase savings. Even if the prospects for individual savings are relatively limited, however, they are still important, for at best the process of raising national savings rates is a slow process. If the entrepreneurial calibre of the business community is low, business savings may also be slow to grow, even in a propitious environment. Even government savings, in principle the most manageable of the three components of savings, cannot increase rapidly over- night. In view of the reliance that is frequently placed on fiscal savings to increase the rate of total savings, VJ the probable rates at which fiscal savings can grow are illuminating in this regard. At their maximum, govern- ment savings can perhaps approach 12% of the national product. This limit is derived from the observation that only in a very few countries are the gross revenues of the public authorities likely to approach 30% of the national product, or are the consumption expenditures of the public authorities likely to account for less than 60o of their gross receipts. In Asian countries, representative ratios of government revenue to national product are very low, about 10%, and government consumption ratios are very high. India and the Philippines, for example, use about 90% of their revenues for current expendi- tures, and so does Ceylon, even though Ceylon's revenue as a proportion of national income is more than twice that of the other two countries. Even allowing for differences between countries, Asian government savings rates are probably much closer to 2% than to their probable limit of 12%. Govern- ment savings rates may be somewhat higher in other countries, but even where high, they fall well below their probable limit. In Turkey and Mexico, two countries in which incomes have been rising and in which the scope for govern- ment savings has probably been more favorable than in many other countries in recent years, gross public savings have averaged less than 5% of their 1/ "Automatic savers" refer to persons whose consumption is unlikely to increase from its present high levels, however much their current incomes may increase. g/ For example, "Some Financial Aspects of Development Programmes in Asian Countries", Economic Bulletin for Asia and the Far East, Vol. III, No. 1-2, Jan.-June, 1952, pp. 1-12. - 6- respective gross national products. Of course, the lower the present rates of government savings, the better the prospect for raising these savings in the future. If, for example, an appropriately flexible fiscal mechanism could siphon, say, 30% of every addition to income to the government, and if a sufficiently well disciplined financial administration could limit the rise in current govern- ment expenditures to, say, 50% of all additional government receipts, the incremental gross public savings ratio might approximate 15%. With this high marginal rate of public savings, government savings could rise by 2 percentage points or more each decade. Over time, this rate of increase would decline but if the conditions of this example were fulfilled, the average rate of government savings would still continue to grow; it would exceed 10% of the national product well before the end of the fifth decade, even in those .countries which started off with very low rates of government savings. How the Averape Rate of Government Savings Would Increase, With Initial Average Rates of 2.3.4,5%, if the Marginal Rate of Government Savings were 15% Period 2% _9AL _% After 10 years 4.3% 5.2% 6.0% 6.8% " 20 * 6.3 6.3 6.9 7.6 " 30 " 7.8 8.4 8.9 9.5 n 40 " 9.1 9.6 10.0 10.5 " 50 n 10.2 10.5 10.9 11.3 If, however, the pressures f6r higher current expenditures could be only partially restrained, even having small positive savings, let alone any increment of savings, would represent a considerable achievement for a govern- ment. For reasons of quantity alone, there is therefore every good reason to encourage private savings, however low they may be to begin with. There are probably more compelling reasons to encourage private savings for their qualitative importance in a developing economy, but this requires little elaboration here. III. PRIVATE SAVINGS A. The Rate of Savings As has already been noted, from the point of view of raising the rate of private savings the individual and the business components of the private savings stream have to be distinguished., Individual savings are probably the least responsive to measures to raise their rate, being liable to change for the most part only with changes in income and with changes in the demographic composition of the community over time. There is good reason to assume that, in the aggregate, they are largely indeDendent of and irresponsive to the rate of return on capital: any increase in the rate of return, while it may induce some people to save - 7- more, may lead others who save for a given income from capital, to save less. Mechanical devices such as payroll savings schemes may possibly result in some net increase in the rate of individual savings. By and large, however, the volume of individual savings out of any given level of income can probably be increased by only very limited amounts. Business savings, by contrast, can be increased substantially under appropriate conditions. As business earnings are retained primarily for re-- investment, their rate of accrual is directly related to the inducement to invest. 'Jhere profit expectations are sufficiently optimistic, the marginal rate of business savings-business savings include the retained earnings of independent farmers as well as the savings of commercial and industrial enterprises--may accordingly be very-high. "The ploughing-back of entre- preneurial profits was historically the major source of capital accumulation in Western economic growth... If the pattern can be repeated elsewhere, it would be an effective and almost automatic way of maximizing a countryts marginal saving ratio". 1/ Now the inducement for businesses to save can be influenced in many ways. Taxes may be used to induce all businesses, or particular types of business, to retain rather than distribute earnings. Accelerated depreciation allowances may make the prospect of internal expansion more appealing. By reducing the expenditures that have to be incurred by an individual firm when expanding, the provision by public authorities of social overhead facilities such as roads, railways, port facilities, sources of power, training schools, etc. may also increase the private rate of return that the business investor expects from any given volume of business investment. Where entrepreneurial endeavor ,is deficient in particular directions, public enterprise may-prepare the way by initiating enterprises which, once established, can be taken over by private owners for subsequent self-financed expansions. In each of these ways, to name only a few, the marginal rate of business savings may raise the average rate of private savings beyond that point which might be reached with individual savings alone. Perhaps the most tempting inducements to increasing business savings are associated with inflation. As profits generally increase more rapidly than other forms of income in the early stages of inflation, the rate at which businesses save for internal expansion may rise considerably. The process is certainly easy to initiate. As long as the lag between increases in profits and increases in other forms of income can be maintained,business savings should most probably rise, in amounts more than sufficient to offset the reductions in individual savings that inflations generally induce. Rising rates of p7ivate savings associated with inflation can only be tempor- ary, however; they depend upon the acquiescence of individuals, particularly those in the lower income groups upon whom the real burdens of inflation fall most heavily, to a continued forced transfer of income to the business communi- ty. Sooner or later, usually sooner, individual income earners will find ways 1/ R. Nurkse, Problems of Capital-Formationin Underdeveloped Countries, Oxford, 1953, p. 155. to prevent further forced reductions in their real incomes. From this point on, the only result of inflation can be rising prices,and a growing maldis- tribution of investments without any associated increases in the rate of savings. Even if businesses could be induced to save more and without social or political complications,,however, raising rates of business savings is not an unqualified goal in itself. As business savings are usually used for investment within the firm which does the saving and therefore claim rather than release resources, stimulating business savings does not make savings available to finance investments elsewhere in the economy. If, for example, important capital-intensive public utility type investments cannot be financed by the retained earnings of utility authorities or corporations or by government savings, private savings will have to be obtained in one, way or another. But business savings will not make savings available-to finance these investments. In fact, where resources are scarce, stimulating business.savings might even increase the difficulties of making particular investments. Measures to increase the extent to which savings are made in forms which permit the transfer of resources to investors other than the savers may thus be as important an aspect of a domestic savings program as measures to raise the rate of savings. It is to these aspects of the savings problem that we will now turn. B. The Form of_Savings In order to clarify the problems of improving the utilization of private savings by changing the forms in which they are made, two separate components of the private savings stream will be identified. For convenience, we shall call these "retained" and "transferred" savings. There are many ways of classifying savings; the purpose of this particular classification is to em- phasize differences in control.over the use of savings. We define "retained" savings to include all savings, the control of which is retained by the saver; e.g., savings invested directly in the saverts own business, farm, or home, or invested directly in inventories or real estate, or held as hoards of currency, or precious metals, or savings made in kind. "Transferredn savings include all savings whose control and ultimate investment are transferred to othersi e.go,rsavings deposited in savings accounts, or savings invested in marketable financial instruments such as equities, bonds, mortgages, or in non-marketable financial instruments such as pension annuities or insurance. When classified in this way, a distinction between savings in "under- developed" and "advanced" countries immediately stands out. "Retained" savings account for the great bulk of all private savings in underdeveloped countries. They are important absolutely in all countries, but relatively much less so in more advanced countries where the "transferred" savings stream is generally very much larger. In the following paragraphs we shall try to point up the principal factors which might explain this high proportion of "retained" savings in underdeveloped countries. As "retained" savings represent resources which are - 9 - already committed to particular uses ddtermined by the saver himself while "transferred" savings are savings which can be pooled to finance investments in quantities, varieties, and qualities beyond the wealth and entrepreneurial ability of any single saver, the possible limiting effects of a high propor- tion of "retained" savings on a country's-rate of growth are obvious. "Retained" savings may be important. Reinvested business earnings and self- financed improvements on farms are only two of the most obvious categories of "retained" savings that have always been and will undoubtedly continue to be major sources of capital formation in all countries. Real capital formation of the community project variety or savings in kind can be equally important. The range of investment possibilities that can be financed from any given volume of savings is increased immeasurably, however, once the pooling of separate savings streams is possible. The value of enlarging the "transferred" savings stream for the purpose of improving the utilization of resources in underdeveloped countries is thus apparent, quite independently of the rate at which these savings may be accumulated. Why Savings are "Retained" In addition to savings which are retained because the profits expected from internal expansion of enterprises appear higherto savers than prospective earnings elsewhere,control over the use of savings may be retained by savers for a variety of negative reasons. The most important of these are: (1) lack of convenient savings outlets, (2) lack of convenient savings instruments, (3) lack of trust in others, (4) fear of illiquidity, (5) fear of a decline in the real value of money. 1. Lack of convenient savings outlets Savings may be retained because savings outlets are not conveniently accessible to the saver. Rural cash savings, for example, may be put into boxes or gold instead of deposit accounts because the nearest deposit facilities may be one-half a day's journey away. Where agricultural incomes account for a large proportion of total incomes, relatively large amounts of savings may be available for transfer if only savings outlets were brought closer to the savers and the savers were taught how to use them. Savings out of given rural incomes may well be even larger than urban savings out of equivalent incomes because the upward pressures on consumption are generally less intense in rural than in urban areas. Probably the most convenient, least expensive, and widely used device for making savings outlets as accessible as possible is post office savings facilities. Administrative and physical facilities are generally already at hand and post offices are known to savers. If organizational and staffing problems can be mastered and people taught how to use deposit accounts, physical inaccessibility can become the least of the barriers to "transferred" savings..- - 10 - Mobile savings banks are one novel device that has been used in the Belgian Congo and in Malaya to bring deposit facilities directly to the saver. They may be expensive, however, particularly where distances are large and terrain is difficult. They may also have a poor psychological appeal to the saver. In Malaya,,for example, recent experience with mobile savings busses revealed that many avers preferred to travel to the nearest post office rather than use mobile deposit facilities. They undoubtedly felt that their savings would be more secure in the custody.of an institu- tion with which they were already familiar., 2. Lack of convenient savings instruments Savers may also retain control over their savings because available savings instruments are either inconvenient or not known to them. This barrier to "transferredu savings is,in principle, also easily corrected. Where lack of knowledge is the barrier, savers need only be made aware of the availability of local deposit facilities and how they can be used, for deposit accounts generally satisfy all the criteria which a saver ordin- arily demands of a savings instrument. They are safe, they yield income, and they are liquid; they are simple to use, and the savinrs can be withdrawn in amounts which need bear no relation to the size of the initial deposit. Systems of government guarantees or deposit insurance can be devised to make the saver feel entirely secure in committing his savings to deposits in particular institutions. Where other savings instruments are used, however, their denominations, rates, maturities and conditions of redemptions have to be appropriate to the needs of particular savers. As small savers generally cannot afford to purchase securities of large denomination and large savers generally cannot be bothered to buy certificates of small denomination, the range of denominations in which savings instruments are issued should be broad enough to appeal to the widest.possible group of savers. Issuing large numbers of small certificates may be costly and inconvenient, More expensive certificates can still be sold to small savers, however. Savings stamps, for example, might be sold for a convenient fraction of the purchase price of a certificate. After stamps to the full value of a certificate have been accumulated, they could be exchanged for the certificate, or for a savings bank deposit, or any other specified instru- ment that may be chosen by the saver.. The nominal rate of interest at which certificates are issued will,.of course, vary with the maturities of the certificate and with the special privileges that may be attached to them. Rates must compare with those ob- tainable from other current alternative uses for savings, however. If they do not, the security will be taken up only at that discount which will bring its yield into line with alternatives open to savers. Certificates also need to be made available in a sufficiently wide range of maturities to attract the savings of savers willing to transfer their-savings for both short and long periods of time. - 11 - 3. Fear of illiquidity A large part of the private savings stream in any country, not just underdeveloped countries, seeks liquidity. Individual savers may desire liquid savings for emergencies. Retained business earnings and depreciation allowances may be invested temporarily in'liquid assets if the accrual of these funds does not coincide in time with investment plans. The degree of transferability of these savings depends on the ease with which the assets in which they are invested can be converted back into money. Now at the extreme, money hoards (or various real goods in periods of heavy inflation) will be the only satis- factorily liquid assets, though securities may be relatively liquid too if their market is sufficiently broad. Where securities markets are narrow, however, the possibility of capital losses at the moment of sale is increased. A saver's estimate of the probable net return he may earn on any investment of his savings in securities is thus reduced substantially once allowance is made for possible capital losses; it may even become negative. In such circum- stances, liquid savings, to remain a store of value, are likely to be retained as hoards of cash or gold rather than be invested in securities. As this barrier to "transferred" savings is a direct reflection of the narrowness of securities markets, broadening the market is the only basis on which it can be eliminated. Now markets cannot be broadened by fiat; they .can only grow if there are lenders and borrowers in sufficient numbers to encourage dealers to devote their attentions to a market to make it effective. Measures can be devised to impart the liquidity which narrow markets cannot initially provide, however. These measures can thus help to stimulate the demand for securities and to broaden their market. Special redemption facilities are the principal device that can be used to impart liquidity to securities while markets remain narrow. They may con- sist of a commitment by the borrower to repurchase at or around face value or purchase value, together with accrued interest, securities which are submitted for redemption under certain conditions; e.g., after specified notice, or on the occurrence of specified events such as death in the family. Redemption may also be authorized at sight. Making particular certificates acceptable as collateral for loans may also add to their appeal. Redemption facilities cannot be offered "in vacuo", however. The borrower, whether a government or a financial intermediary, can safely offer such facilities only with the knowledge that a lender of last resort, generally the central bank, will be prepared to rediscount notes presented for redemption should the need arise. At some point, therefore, the question of the price at which these facilities shall be made available must be faced. Do measures for imparting liquidity to securities have to include supporting the market? Market intervention by the public authorities for the purpose of evening out unusual temporary fluctuations in securities prices may be useful and necessary in the early stages of development of a market; it must be limited, however, if its ultimate effects are not to be the reverse of its intent. Unconditional support of, a securities market, because it is equivalent to monetizing debt, may open the way to an uncontrolled money expansion and only inhibit rather than help stimulate the growth of "transferred" savings. - 12 - 4. Lack of trust in others Another reason for savers retaining control over their savings is lack of trust in others. Where the rule of "caveat emptor" dominates commercial and even financial relationships, it is understandable that savings will only be transferred to others who are personally known to and trusted by the saver. Without trust, the saver's estimate of the risk involved in transferring his savings to strangers will far exceed any possible return he may get on his savings. Financial mismanagement in any type of public or private institution can accordingly seriously inhibit the growth of all transferable savings; it is certain to prevent savers from transferring their savings to the particular institutions that have been or are likely to be mismanaged. Rural savers in some countries may thus not be willing to entrust their savings to branches of distant or unknown urban institutions. And where little or no provision for the information and protection of stockholders exists, even urban savers may not be willing to buy equities, or even obligations, of private companies unless some personal connection with the company already exists. Similarly, and of even more importance, where there is little confidence in the integrity or administrative efficiency of public officials, or where there is a history of default on internal debts, the prospect of savings being voluntarily trans- ferred to governments is also small. At the extreme, where savers fear the loss of their savinrs because of political instability or public policy, they may even send their savings abroad. %vings which are "retained" because of lack of trust are generally potentially transferable domestically under appropriate conditions, though it may take time to bring these conditions about. In.the case of corporations issuing securities to the public, provision for proper reporting and for the legal protection of stockholders can always be developed. Provision for inde- pendent auditing may require prior training of accountants; it certainly re- quires that the business community accept and recognize their professional independence and integrity. Once achieved, however, there is no reason why savers who would be prepared to assume the ordinary risks involved should not invest their savings in corporate securities. Restoring confidence in the public administration may be more difficult, even assuming that conditions do warrant a revival of this trust. If independent public agencies such as central banks or development banks of various types are regarded by savers as somewhat apart from the government itself, however, they can provide a conven- ient vehicle through which private savings can be collected for financing public investments. As savers may be as unwilling to transfer their savings to incompetent as to dishonest administrators, another approach to this barrier of distrust is also called for. It consists of remedying the lack of readily available managerial skills. Management contracts are a possible'solution to this problem. Manage- ment firms are commonly available in particular types of enterprise in ad- vanced countries. Their services can be imported until local managerial skills can be developed. Ultimately, the significance of the development of this particular type of skill can far transcend its influence on the relative share of "transferred" savings in the private savings stream. - 13 - As long as incomes remain unchanged, the elimination of distrust as a barrier to the transferability of savings will, like the provision of new savings outlets and convenient and liquid financial instruments, most probably affect directly only the composition of savings, not its volume& Savings newly invested in financial instruments of corporations, public agencies, or govern- ments will generally only represent either reduced hoards,.or shifts from generally accepted forms of "transferred" savings to new forms which are now also considered safe. They are no less desirable on this account, however. Savings which are institutionalized are far more readily available for financing productive investments than when they are "retained", say as hoards. It has at times been said that savings which take the form of additions to cash hoards can be mobilized even more easily than has thus far been suggested. As additions to cash hoards sterilize money, governments, it is argued, can "mobilize" savings made in this form by creating equivalent amounts of new money to finance desired public expenditures. As long as the new money created does not exceed the addition to hoards, it can have only neutral, not inflationary, effects. Proposals to "mobilize" additions to cash hoards in this manner must assume, however, that government finances have not already overcompensated for cash hoarding and that the amounts involved are known; they must also assume that, hoards will continue to be held, or that should dishoarding occur, action will be taken to contract the money supply. "Mobil- izing" additions to hoards by equivalent expansion of the money supply may thus have inflationary rather than neutral effects. If it does, it will aggravate one of the most pervasive barriers to "transferred" savings--fear of a decline in the real value of money-and thereby make more rather than less difficult the task of increasing the transferability of savings. 5. Fear of a decline in the real value of mone This fifth obstacle to the growth of "transferred" savings, like lack of trust and fear of illiquidity, so affects the saver's estimate of the real rate of return he may expect to receive on various types of assets that he chooses to save in non-transferable forms. Unlike the previous obstacles, however, for which specific institutional remedies are indicated, fear of a decline in the real value of money can be removed only by far-reaching funda- mental changes in public policy. Measures to enlarge the "transferred" savings stream ultimately depend on the willingness of savers to save in money claims such as postal or savings bank deposits, insurance policies, government bonds, government savings certificates, etc. Where there is inflation or fear of substantial increases in the general level of prices, however, assets which take this form are the most undesirable way in which to save. There is accordingly a clearrelation- ship between savers' expectation of changes in the real value of their savings and the size of the "transferred" savings stream. If the decline in real values is expected to be slow, a rate of interest at least large enough to compensate for the expected rise in prices may still permit savings in "transferred" forms. Generally, however, even "high" rates of return are un- likely to compensate for expected real losses in countries where the fear of a - 14- depreciating currency has firm basis in the past; savings will therefore move into assets such as gold or buildings whose real values are more likely to be maintained. The experiences of advanced as well as underdeveloped countries provide ample support for this observation. The savings of the small and large savers may not be invested in the same real goods. Small savers may hoard gold or jewels, or in periods of violent inflation, even commodities,. while large savers are more likely to invest in land and urban building, and businessmen, in large inventories; savers may also hold their savings abroad in various forms. The net effect of all these reactions is the same, however. The compo- sition of investment is oriented in particular directions which are generally uneconomic for the community as a whole, and the "transferred" savings stream dries up. Because of the importance of "transferred" savings in financing public investment programs, various proposals for the issue by governments of certifi- cates with a real value equivalent have been made to overcome this barrier to "transferred" savings. One of the most interesting of these actually put into effect is the French loan of 1952. To counteract the demoralizing effect on savers of a long period of currency depreciation, the French government issued in 1952 a security with a gold guarantee which consisted of defining the face value of the security in terms of the value of a gold piece of a specified weight and name. As the security cannot be redeemed at less than its par value at issue, the holder is assured of having a constant amount of capital in terms of gold and thus a genuine store of value as long as the real purchasing power of gold does not decline. As an extra attraction these particular securities can be used, at the their redeemable value, to pay certain taxes, and annual interest Is free of tax. Real value guarantees of this kind which may be attached to securities can probably restore a borrower's credit more rapidly than might otherwise be possible. They can be successful in restoring public credit, however, only if more fundamental anti-inflationary fiscal and monetary policies are initiated at the same tine. In the absence of such policies, the guarantee itself will sooner or "Uter become just another source of pressure on the expenditure side of the budget, thereby exaggerating the difficulties it was originally intended to overcome. Unlike money claims, equities may retain their real value during infla- tion. Their relatively high rates of return coupled with appreciating money values may be sufficient to offset the risks normally associated with investment in equities with the result that they may attract "transferred" savings which might otherwi8e switch into other "retained" forms. This particular influence towards maintaining the size of the "transferred" savings stream has little merit by itself, however. The great advantage of a large "transferred" savings stream from the point of view of economic growth is that it may permit a better composition of investment than "retained" savings alone might produce. However desirable equity investments may ordinarily be in this regard, the investment of savings in equities, if induced only by a desire to hedge against inflation, may compound the maldistribution-of investment which inflation first induces. - 15 - 6. Other barriers There are perhaps still other barriers to transferring savings such as inertia and high costs of transfer. Social customs may attach high prestige to saving in particular non-transferable forms. If savers desire the satis- faction of saving in tangible form, gold coins may appeal to them whereas a gold certificate may not. Special devices may be necessary to appeal to other motives. Thus, savers may be more willing to pool their s-vings for the use of local borrowers rather than for distant unknown borrowers, or for local improvements which can be seen and whose benefits may be enjoyed rather than for distant investments whose benefits may be enjoyed only indirectly, if at all--this is the basis for the successful development of credit cooperatives in their many and interesting varieties. Because of a traditional familiarity with lotteries of one form or another, some special attraction such as a lottery feature or special premium attached to a bond or a savings account- there are many varieties of devices of this kind--may be required to induce people to save in transferable forms. A long list could be made of these characteristics of savers in all their shadings, and the many "gadgets" that have been thought of to induce savers to transfer their savings by appeal- ing to each of them in turn. They are probably of secondary and rather local interest, however, compared to the more general obstacles that have been enumerated above. Whatever ideal appeals to savings incentives may be made, the "transferred" component of the private savings stream will remain small in the absence of .convenient savings outlets and instruments, in the absence of trust, and in the absence of a stable monetary unit. IV. CONCLUSIONS The preceding discussion has canvassed some of the principal barriers to "transferred" savings flows in underdeveloped countries, and some of the most important measures that can be taken to eliminate these barriers where they exist. The problem of directing these "transferred" savings to partidular investments has not been considered explicitly for this second problem is, in a sense, subsidiary to the first. Measures to rechannel "transferred" savings in particular directions, which consist generally of the development of appropri- ate investment intermediaries, can only make a minor contribution to eliminat- ing the barriers which have been referred to above; their success depends from the beginning on the prior existence of freely-flowing "transferred" savings. One can think of the most elaborate varieties of financial inter- mediaries--institutions which are equipped to receive funds from both public and private sources, which choose between borrowers on the basis of economic priorities alone, and which mobilize savings by insisting that borrowers con- tribute resources of their own to projects financed by them. Even with all these attributes, however, the absence of "transferred" savings forces these institutions to become only devices for the disbursement of public fund-4. For confirmation, one need only consult the experience of countries which are by all standards rich in specialized investment intermediaries but in which the real functions performed by these institutions are only a pale shadow of the purposes for which they were set up. Lfhere the volume of private savings is very small, whatever their form, even the most vigorous measures to enlarge - 16 - the "transferred" savings stream will be subject to obvious limitations. These measures still remain, however, a necessary condition for the successful introduction of new financial institutions in any country. Like financial intermediaries, securities markets must also remain limited in development and function where the size of the "transferred" savings stream is small. Once savings are transferred freely, however, securities markets assume a special importance. Their unique contribution is that they impart liquidity to particular forms of savings which could not otherwise be liquid and therefore make possible the pooling of savings for purposes which might not otherwise be realized. As has repeatedly been noted, the great value of "transferred" savings is to permit this kind of pooling so that scarce savings may be combined to finance the best possible combinations of resources in the economy. By enlarging the range of financial investment possibilities that become acceptable to savers,and financial intermediaries, securities markets facilitate the rechanneling of savings and thus make financial institutions even more useful than they could otherwise be. Of course, savings can be rechanneled without securities markets. In many countries governments which cannot induce private savers to buy their bonds force severs to take them up, collecting private savings for public use in this way. Public securities could, however, undoubtedly be-sold freely to private or institutional investors if their price conformed to the credit standing of the government and to the prices at which these funds could be used in alternative ways. Forced loans thus compel sayers to suffer losses equal to the difference between the price at which they are compelled to buy securities and the price at which they would have been willing to take them up. By reducing the freely available supply of savings and thereby per- petuating the narrowness of securities markets and fears of illiquidity, they also restrict the extent to which private savings can become fully transfer- able. These are disadvantages which must be set against whatever advantages directing savings by decree to governments or other borrowers may have in specific circumstances. After all the merits and demerits of the variety of devices that can be used to influence the volume and utilization of private savings are sorted out, one general observation always remains to be made. The effectiveness of savings devices is almost entirely dependent on the appropriateness of the environment in which they are introduced. As the economic aspects of this environment are to such a large extent a reflection of past and anticipated monetary and fiscal policies of governments and central banks, these policies are as intimately responsible for the success or failure of measures to in- fluence the rate and composition of private savings as any savings "gadgets" can possibly be. To neglect these considerations can serve only to obscure rather than clarify the most important factors affecting private savings flows and the extent to which they can contribute to a country's economic growth.