Journal Issue

The Management of Public Debt in Developing Countries

International Monetary Fund. External Relations Dept.
Published Date:
June 1970
  • ShareShare
Show Summary Details

Jonathan Levin

LIKE TAXATION, government borrowing is undertaken to finance government expenditure. Whether the expenditure is for consumption or investment, for military needs or economic development, every borrowing government must face the same set of questions: first, how to borrow the money—from whom, for how long, and at what interest rate—and second, what effects each borrowing choice will have upon the economy as a whole and upon particular sectors within it.

Like the decision to tax, every government decision to borrow represents a political determination that additional resources are to be diverted to the public sector and allocated to the fulfillment of particular goals. Where these additional resources are to come from will depend upon the kind of borrowing that a government chooses. It may borrow outside the country, which will bring in additional resources from abroad. Or, it may borrow domestically, which—in the absence of unemployed resources that may be readily mobilized—will in one way or another reduce the use of resources in the rest of the economy and allocate them to government. There are several ways by which the “incidence” of domestic borrowing—the source from which resources are to be diverted—may be fixed. When a government borrows from individuals or institutions through compulsory loans it diverts resources from those to whom the funds would otherwise have gone—potential borrowers from the institutions, alternate saving recipients, and perhaps the individuals themselves, thus reducing their own outlays. When a government borrows in the market, it diverts resources from other borrowers who will be unable to pay the higher interest rates that result from the competition for funds. In the process of borrowing in the market, redistributive side effects will also occur, directing increased interest payments from all borrowers to lenders and, perhaps, increasing the flow of savings from potential lenders as well.

Unlike taxation, government borrowing does not always succeed in reducing the use of resources in the rest of the economy at the time the proceeds are obtained. Sometimes the diversion of resources will come only later on, when the loan proceeds are expended. When a government borrows funds that would not otherwise be used—as from the central bank or from hoards or idle balances—its expenditure of the loan proceeds is expansionary, bringing a net addition to demand for resources. In the absence of unemployed domestic resources that may be quickly mobilized, or of available foreign exchange, this diverts resources not from the lenders, who were in any event not going to use their funds to purchase resources, but instead from those who pay the “inflation tax” when prices go up and they can buy less. By whichever domestic means the government chooses to borrow, therefore, it is the diversion of resources from other users—not the interest to be paid—which is the cost of spending the loan proceeds. The interest paid represents only the cost of persuading the lenders to part with their claims on the use of resources in exchange for the government’s promise to repay, voluntarily for market borrowing and at a politically acceptable price for compulsory lending.

When the time comes for the government to repay the loan, the cost or burden consists of diverting the resources from use by the government—or its taxpayers if taxes are to be increased—to those who hold the government’s debt. As one would expect, this cost is quite different in character for a loan from abroad, which brings in additional resources to begin with and which must be repaid through a parallel transfer of resources abroad. In a sense, foreign debt is like a personal debt which will have to be repaid to others out of the country’s future income or borrowing. A domestic debt, on the other hand, is more like a country’s debt to itself, on which interest and eventually capital is to be paid by the taxpaying citizens to the debt-holding citizens. With domestic debt, the strain of interest payments on the budget, the frictions of raising taxes to repay the debt, and the redistribution of domestic income resulting could be important if the debt is large. However, there need not be the reduction in total income and the strain on the balance of payments that repayment of a foreign debt could entail. The “burden” caused by both foreign and domestic debt, however, will depend on how fruitfully the borrowed funds have been used. It is usually assumed that, while taxes are paid by reducing consumption, lending to the government is accomplished by reducing other investments. If the government’s borrowing diverts funds from productive investment to current consumption, or to projects bringing no increase in the nation’s income, the country—like anyone who consumes his capital—will be poorer for it in the future. In this sense a public debt may be a “burden” both upon the present and future generations. This is a “burden,” however, to be weighed in the same scales as the large endowment of capital facilities to which each generation falls heir when it comes upon the scene.

The Choice of Borrowing

Almost all the governments in developing countries have had to choose some forms of borrowing to supplement their tax revenues.1 Borrowing abroad has been their most frequent choice and foreign debt forms the major portion of the public debt in most developing countries. The sources of this foreign debt over the past two decades have been chiefly international agencies and official lending institutions, a marked contrast to the sale of foreign bonds to private bondholders which flourished until the early 1930’s. Governments have also drawn upon short-term and medium-term external credits from suppliers and banks. The contracting of most foreign debt, therefore, is now dependent mainly upon a country’s ability to prepare and carry out investment projects and on the maintenance of its capacity to repay. Sound management of a government’s foreign debt, however, requires (1) that the particular project generate a sufficient return—either in itself or to the economy as a whole—to cover repayment of the loan; (2) that the volume and maturity structure of the debt be such as to hold the over-all burden of foreign debt repayment, public and private, to within a reasonable proportion—often taken as being 20 per cent—of foreign exchange earnings; and (3) that the burden of debt repayment remain sustainable in the light of the economy’s growth and level of income. While these principles, other than the foreign exchange limitations, apply to domestic debt as well, it is debt placement that is the primary concern of domestic debt management—from whom to borrow and with what effect on the economy—so it is debt placement that mainly concerns us here.

Selected Countries: Per Capita Gross National Product and Domestic Public Debt as a Per Cent of Gross Domestic Product
Debt as a

Per cent

of GDP


GNP in US$
Debt as a

Per cent

of GDP


GNP in US$
United Kingdom78.81,620Mexico9.5470
Ceylon52.0150Trinidad and Tobago9.0630
New Zealand49.51,930
United States41.23,520Colombia8.3280
South Africa38.0550Uganda Zambia7.8 7.8100 180
United Arab Republic31.1160
Ghana27.2230Japan Peru7.1 7.1860 320
Costa Rica19.9400Cyprus6.4690
Dominican Republic17.2250Honduras4.9220
Sweden17.12,270China, Rep. of4.6230
Germany, Fed. Rep. of15.31,700Haiti4.370
Portugal13.5380El Salvador2.5270
Iran10.5250Central African Republic0.8110
Austria9.91,150Malagasy Republic0.190
Sources: United Nations, Statistical Yearbook; International Monetary Fund, International Financial Statistics; and World Bank, Atlas, 1969.
Sources: United Nations, Statistical Yearbook; International Monetary Fund, International Financial Statistics; and World Bank, Atlas, 1969.

The difficulties of placing government debt within a developing country spring mainly from the insufficiency of savings and the limited development of financial institutions. Many developing countries have only recently advanced from substantial involvement in barter economy to a wider use of money and bank deposits. Contractual savings institutions, such as social security systems and insurance companies, have begun to mobilize savings in some countries but others still lack the wide base of industrial employment and commerce upon which such institutions can be built. The evolution of markets for claims and equities, moreover, is dependent not only upon the generation of savings but also upon the emergence of salable and reasonably secure promises to pay or share. Against this background of limited financial development, governments have tried a variety of means to place their debt within the country. The cumulative results of these efforts may be gauged from a comparison of total domestic public debt as a per cent of gross domestic product (GDP) in some 73 countries for which data are available (see Table). A surprisingly wide range of ratios is apparent, with some developing countries showing a higher ratio than many developed countries. While an examination of particular country ratios is not attempted here, the higher ratios of debt to GDP generally appear to exist in those countries where no inflation has intervened to shrink previous debt accumulations and where financial institutions have evolved. This seems to be particularly the case in countries of British or Commonwealth background.

Government efforts for the placement of domestic debt may be divided into two sorts, those outside any market and dependent upon compulsion and those operating within the realm of free personal or institutional discretion.


Because the government is not a private borrower it has a certain degree of power to compel persons and institutions to lend it money. In a completely socialized economy this may extend to central collection and allocation of virtually all savings and investments. In an economy with some dependence upon private enterprise the power to compel private lending to the state is exercised to a more limited extent.

The Public Sector

The first source of funds from which the central government may borrow without going through a market lies within the public sector itself. The unused cash balances of other public sector entities outside the budget are readily turned to central government uses by concentrating them in a single bank, though in some countries they remain in separate or earmarked accounts while the government seeks to borrow elsewhere. Where the financial or commercial institutions within the public sector are large these current balances can represent a significant source of potential funds to the central government. Thus, in most French-speaking countries the Treasury (Trésor) acts as cashier and banker for all official entities—sometimes receiving private deposits as well—and as part of its normal functions makes these funds available as overdrafts and longer-term loans to the central government.

In addition, governments often find longer-term funds collected by public sector thrift or insurance institutions available for the placement of longer-term government debt. Thus, in many countries the post office system serves to mobilize small savings which are invested in turn with the government. Probably the major sources of longer-term funds for governments, however, are the assets of governmental contractual savings institutions, such as social security systems, provident funds, or retirement funds. Where such systems exist they are usually required by legislation to invest all or some proportion of their accumulating funds in approved investments, quite often government securities alone, and the management of such funds may be in the hands of debt commissioners or trustees within the central government. A recent study by Professor Franco Reviglio found social security institutions in Burma, Ceylon, India, Israel, Malaysia, and Nigeria investing all their assets in public debt. Many others in developing countries were placing over 50 per cent in such debt, while in very few instances was less than 20 per cent of social security assets in public debt.2 A basic requirement for the creation of a social security system of the pension type, Professor Reviglio found, is a reasonable degree of monetary stability, avoiding the inflationary erosion of a pension’s real value. Only rarely have sizable social security reserves been generated and maintained in the face of inflation through the issuance of indexed government bonds, as in Israel and for a time Brazil.3

Selected Countries: Per Capita Gross National Product and Domestic Public Debt as a Per Cent of Gross Domestic Product

On the fringe of the public sector lies the most important and controversial of all sources of funds for government—the central bank. Most central banks are prohibited by their basic statutes from lending more than a limited amount directly to the government. This amount is usually some proportion of the previous year’s total government revenues and is designed to cover seasonal variations between the pattern of revenues and expenditures within the fiscal year. The actual independence of the central bank varies widely from country to country, however, so that such prohibitions do not always withstand the pressures of particularly difficult situations. Whether a particular central bank is to be an “engine of inflation” or the “watchdog of financial conservatism” has been determined not by laws alone but by the whole constellation of influences which lead a government to pursue either expansionary or more conservative policies.

As is generally recognized, the central bank is the most expansionary source of government borrowing, first because it can draw upon funds that need not otherwise be used, and second because the money lent to the government by the central bank can serve to increase the reserves of the commercial banks and provide a base for a multiple expansion in their credit. An expansionary impact may be desirable in some circumstances, to be sure, depending upon over-all developments in the economy. In fact, determining the appropriate amount of central bank credit to the government constitutes one of the most important elements of government economic policy. The dependence of governments in many developing countries upon central bank credit may be gauged from Professor Raymond Goldsmith’s findings.4 These indicated that while central banks in the developed countries studied held an average of 15 per cent of their government’s public debt, this proportion averaged about 50 per cent in the developing countries reviewed. In the absence or inadequacy of other sources of domestic financing, the central bank becomes the lender of last resort not only for the banks but for the government as well.

The Private Sector

Within the private sector, the compulsory placement of government securities has focused upon several targets, most of them within the circle of financial institutions. Most commonly it is the commercial banks which are compelled to purchase government securities, by reserve requirements or minimum liquidity ratios which must be met in whole (as in India) or in part by holding government securities equal to specified portions of their deposits. Even when government securities form only an optional part of such liquidity ratios or reserve requirements, the interest they pay, however low, makes them so much more attractive than the alternative of vault cash or noninterest bearing deposits at the central bank, that the option to hold securities is likely to be exercised.

In some instances, holdings of government securities have been required as reserves against only a particular kind of bank liability, for example, foreign exchange deposits in the Philippines. In many countries also banks that wish to become depositories for the operating funds of the central or local governments or of other official entities are obliged to hold instruments of government debt as collateral equal to a given percentage of such deposits. Elsewhere, as in Mexico, selective credit controls may operate through a complex series of reserve requirements to direct bank credit into a number of favored fields, among them securities of the government or its autonomous entities.

The extent of government borrowing from commercial banks differs widely among developing countries, however, and Professor Goldsmith finds that commercial banks did not appear to hold a distinctly higher percentage of the total public debt in developing countries than in developed countries. Their holdings in both groups of countries ranged from very little up to about 30 per cent of the total public debt.

In its expansionary effects, government borrowing from commercial banks may be second only to borrowing from the central bank. When commercial banks are in fact permitted to hold government securities in place of required cash reserves, the expansionary impact may be equal to lending by the central bank, since it adds to the base of required reserves upon which a multiple credit expansion may occur. Just how expansionary government borrowing from commercial banks is to be, however, will depend upon whether commercial bank lending to other borrowers is curtailed to accommodate the government, or whether, on the other hand, government borrowing is added to other credit so that total credit is expanded, through the use of unused reserves or of funds made available by the central bank.

A second source of compulsory private sector lending to the government lies in the nongovernmental contractual savings institutions—insurance companies, pension funds, and, where they exist, mutual funds and investment companies. While a few countries have brought life insurance into the government sector with the creation of a national insurance company (India and Costa Rica), in most countries insurance companies remain in the private sector and are required to hold at least a given portion of their reserves in government securities. Because most contractual savings institutions will have to make parallel reductions in other investments, their lending to the government should generally have no expansionary impact. When insurance companies are closely tied to firms abroad, their compulsory purchase of government securities may sometimes be an alternative to investment abroad.

Compulsory lending to the government by those outside the circle of financial institutions—private persons and nonfinancial corporations—is a quite different and less frequent practice. Resort to such schemes may be preferred to an outright increase in taxation primarily for psychological reasons, as when a government may feel that the present limits of taxation have already been reached. While seldom sustained over long periods, such compulsory savings schemes have been attempted in a number of countries, levied usually on the basis of income, according to Professor A. R. Prest,5 and utilizing five to ten-year nontransferable bonds, paying 3 or 4 per cent in tax-exempt interest.

Quite another form of compulsory lending to the government requires no formal legislation and bears no relationship to income or any other equitable base. It arises from a government’s failure to pay for purchases made or services received. The resulting “floating debt” may work a considerable hardship upon unpaid employees and sometimes upon suppliers caught unawares. A government with a bad payment record, however, will pay an implicit—and often quite high—interest cost for such floating debt, as contractors and suppliers will tend to pad the prices quoted for future government business to compensate for anticipated delays.


A different set of conditions surrounds government borrowing from voluntary lenders. To sell securities in the market a government must be ready to compete, and this means, above all, payment of competitive interest rates. In some countries exemption from income tax has added to the competitiveness of government securities, but this affects the relative return on government paper only to the extent that a significant tax is actually collected upon other income. The question of offering competitive interest rates on government securities is often only a small part of more general government policy toward interest rates.6 While some may consider that higher interest rates discourage needed business investment and redistribute income toward those who own the funds, others will look upon the interest rate as the most efficient means of allocating loanable funds among competing users and of stimulating greater savings. Some governments may be unwilling to let market interest rates rise because of the decline this would bring in the capital value of a large volume of outstanding government securities carrying lower interest rates. Others, in the throes of inflationary price increases, may consider it inappropriate for the government to “recognize” the going market rates by borrowing at them. Whatever the merits of these positions in each set of national circumstances, and whatever their wider implications for government economic policy, a government cannot expect to place its debt among voluntary lenders without paying a competitive rate.

One means of paying a more competitive rate that has been proposed, and occasionally resorted to, is to tie the value of the government bond to be repaid, and at times the interest payment as well, to some index of prices in the economy. As a selling tool, such indexing has the advantages of effective government borrowing: it can reduce purchasing power in the hands of the public and particularly any speculative expenditure for real assets or other hedges against inflation. Where the debt instruments widely held by small savers are redeemable, rather than marketable, or carry relatively short maturities, indexing can help discourage their redemption and thus restrain such additions to demand. By providing small savers with protection against inflation an indexed bond may serve the objectives of equity, at least more validly than the sale of indexed securities to banks, for example. Whether all taxpayers should be forced to protect any single group against inflation, however, may also be questioned on the grounds of equity.7

The need for government securities to compete in the market does not mean that they must always carry as high an interest return as other financial paper. Government securities may enjoy a premium over other securities as a result of their lower risk and, at times, their liquidity. It may be these qualities, indeed, that contribute most significantly to the establishment of a market for government debt instrument. The risklessness which can characterize government securities—in contrast to the securities issued by private firms or institutions that may become insolvent—is earned through the scrupulous repayment of maturing obligations.

While all government securities in the market may be riskless as regards the payment of interest and redemption at maturity, only shorter-term issues are likely to be liquid, that is, salable before maturity without delay and without much loss of capital value. This is demonstrable by simple arithmetic, since any variation in interest rate will bring a greater change in the price of long-term securities than of short-term securities (long-term securities capitalize a longer chain of interest payments). Liquidity is also the result, however, of the number of buyers in the market, and these are likely to be more numerous for shorter maturities in most circumstances. In developing a market for their securities, governments may sometimes seek to ensure a greater degree of liquidity through the market intervention of the central bank or of some other agency acting on the government’s behalf. Somewhat the same result may be sought through the central bank’s rediscounting of government securities presented to it by commercial banks. There is a danger in market support operations, however, in that they may be carried too far, sacrificing the objectives of monetary policy to facilitate government borrowing.

Sliorter Maturities

A market for the more liquid, shorter-term government securities is the first to emerge in many developing countries. This is so because banks may welcome the interest yield of a riskless, liquid instrument other than cash in which to invest the secondary reserves they may keep as a means of absorbing fluctuations in loans and deposits. Companies too may be interested in such a security as a repository for temporarily idle working capital or for funds being accumulated to meet periodic liabilities, such as taxes. In the absence of such debt instruments, the short-term funds in some developing countries may even be invested abroad in more developed money markets. Not all developing countries possess such a supply of short-term funds, however, as some suffer from a shortage of working capital. Businesses may depend upon banks to finance any upward fluctuations in their working capital needs, while banks, in turn, may meet their occasional or seasonal needs through rediscounts of business paper at the central bank rather than from any secondary reserves. Where unused balances of working capital or secondary reserves do exist, moreover, government sales of highly liquid short-term securities that can be held instead of money may serve to reduce the lenders’ holdings of money but not the use of resources—as long as the balances would otherwise have remained idle and not been loaned to others. Government expenditures of these funds would increase the velocity of money, adding to total demand, and would therefore be expansionary.

A disadvantage to the placement of short-term debt lies in the need to return to the market at frequent intervals as the debt matures. This difficulty may be eased by establishment of a regular pattern of short-term borrowing so that lenders may anticipate the availability of new government securities and maturing issues may be rolled over with minimum disruption to the market. In the refinancing of all maturing debt, however, a government must weigh carefully the characteristics of both the holders of maturing issues and of potential new lenders if the desired effects of expansion, contraction, or neutrality are to be achieved for the economy.

The instruments of short-term government debt have most often been treasury bills. Carrying maturities which may vary from three months to a year, these bills may be sold at regular auctions—perhaps weekly—on a discount basis, that is with the interest deducted in advance, and in either bearer or registered form. Some bills—tax anticipation bills—are sold to mature a few days after a tax date. Taxpayers choosing this means to accumulate funds for their tax liabilities can use these bills to pay their taxes and receive as a premium the interest for the extra few days still to elapse before the formal maturity date.

Longer Maturities

The market sale of medium-term and longer-term government securities averts the need for frequent re-borrowing and more effectively diverts resources from other domestic uses. The prospects for such sales in some countries, however, may be adversely affected by the limited development of the financial institutions likely to hold such securities. A survey of potential holders—thrift institutions, fire, accident, and life insurance companies, and pension funds—of their annual flow of receipts and investments and of the maturity structure of their liabilities, and hence asset preferences, can offer a valuable guide to government sales of such securities and prevent the offering of issues beyond a market’s absorptive capacity. The instruments of medium-term and long-term debt usually pay an annual or semiannual interest, rather than being sold at a discount. Denominated notes, certificates, bonds, or stocks are generally not offered at auction but at a fixed issuing price during a subscription period of several days. When the offering is oversubscribed, securities are allocated on a percentage basis, sometimes with the smaller subscriptions allocated in full. In some countries, bonds have been made available “on tap” for an extended period by the central bank or by other fiscal agents. At times, short term instruments may also be sold on tap.

Apart from placement of securities in the market, some governments have taken particular efforts to sell specially designed bonds to small savers. Typically such savings bonds would be available in small denominations for maturities of under ten years, and might be purchased on a systematic basis through payroll deductions. They may be the subject of a broad publicity campaign appealing to motives of patriotism, defense, or perhaps economic development. To make savings bonds liquid, in case of an individual’s unexpected need for cash before the bonds mature, and yet to protect the holder from market fluctuations which might lower their capital value, such issues are usually not marketable but are redeemable by the government, usually at a price representing a higher interest return the longer the bond is held.

No World of Easy Choices

A government that chooses borrowing rather than taxation to pay for a part of its expenditures does not escape to a world of easy choices. Foreign borrowing must meet the rigorous standards set by most lending institutions for project rentability and balance of payment considerations. Domestic borrowing, whatever one’s opinion of the “burden” upon the resources and income distribution of future generations, presents unavoidable costs in the present as embodied in the resources other users must forego if the government is to use them. These costs must be of political and economic concern to most borrowing governments along with the more general effects upon aggregate demand, prices, and interest rates. With all good intentions, however, domestic government borrowing can go no further than development of the country’s financial institutions permits, and in the earliest stages of development this may mean getting no further than the central bank, the most expansionary of sources. Whatever the source of funds with which a government must content itself, it must weigh the consequences of this borrowing in determining its over-all economic policies, while it sets about encouraging the growth of financial institutions to permit a wider choice in the future.

THE INTERNATIONAL MONETARY FUND, 1945-1965 Twenty Years of International Monetary Cooperation

Vol. I. Chronicle, 663 pp.

Vol. II. Analysis, 621 pp.

Vol. III. Documents, 549 pp.

Price: $12.50 the set ($5 a volume if sold separately), or the equivalent in the currency of any member of the Fund.

The volumes are available in English only. Orders should be addressed to

The Secretary International Monetary Fund

19th and H Streets, N.W.

Washington, D.C., 20431


1While this is the usual reason for borrowing, governments may on occasion issue debt instruments to facilitate development of a market for their securities.
2Franco Reviglio, “Social Security: A Means of Savings Mobilization for Economic Development,” International Monetary Fund, Staff Papers, Vol. XIV (1967), p. 354.
3A discussion of indexing bonds appears in Sanjaya Lall’s “Countering Inflation: The Role of Value Linking” in the June 1969 issue of Finance and Development.
4Raymond Goldsmith, Financial Structure and Development (New Haven, Yale University Press, 1969), p. 161.
5A. R. Prest, “Compulsory Lending Schemes,” Staff Papers, March 1969, pp. 27-52.
6For a broad discussion of government interest rate policies see the article by A. G. Chandavarkar in the March 1970 issue of Finance and Development.
7Sanjaya Lall, op. cit.

Other Resources Citing This Publication