6 Lending Programs, Guarantees, and Quasi-Fiscal Operations
- A. Premchand
- Published Date:
- March 1993
“The time has come,” the Walrus said,
“To talk of many things:
Of shoes—and ships—and sealing-wax—
Of cabbages—and kings—
And why the sea is boiling hot—
And whether pigs have wings.”
LEWIS CARROLL, Through the Looking Glass
The financial activities of a government in the current setting may be considered flagrant violations of Polonius’s advice.1 Governments have become both lenders and borrowers—and as borrowers, they have done so on an unprecedented scale. This chapter and the one that follows discuss government lending programs and associated matters and public debt budgeting.
A number of considerations underline the importance of lending, guarantees, and the quasi-fiscal operations of governments.2 All three have become important elements of a government’s financial management with a substantial impact on the status of government finances and what they imply in terms of taxpayers’ likely future burdens. For example, in the United States (which is by no means unique), the Government’s exposure from its credit and insurance programs has risen from approximately $438 billion in 1965 to about $6.5 trillion in 1991. The exposure is reported to have doubled during the second half of the 1980s. In addition to the risk of future loss and, consequently, the likely burden on taxpayers, what is more worrisome, as the U.S. General Accounting Office (1991) has reported, is that the Government does not presently have the financial management capability to evaluate all the liabilities.3 This lack, which is even more conspicuous in countries with less sophisticated financial management systems, is also indicative of the tasks that lie ahead.
Government lending has frequently been used, along with tax expenditures (or incentives), as a substitute for direct government expenditures. Such an approach gathered currency because of the general perception that these loans are repayable and are therefore not part of the durable programs of expenditures included in the budget. Sometimes, because the grant of such a loan did not require legislative approval, it may have been considered an easy and liberating influence on the policymakers to resort to lending for expediency. But as vast amounts remain unpaid and the prospect of repayment recedes with time, they end up as part of government expenditures. Countries like Australia and Canada were thereupon prompted to insist on loan programs in conjunction with other expenditures being considered part of the portfolio and envelope budgeting, respectively, that they adopted. In some countries, that legislative approval is not needed may have contributed to the less transparent track that in effect became parallel to the budget itself. Another consideration is that until recently CPEs had no lending programs, but as resources became strained, and to induce financial discipline, some of them (for example, China and Viet Nam) have introduced lending with specified terms of interest and repayment. This action suggests that lending is likely to become even more widespread than before. Thus, the potential risks to the taxpayer, the preference to soft lending as a political expediency, the extended use of this approach in more countries, and the absence of machinery as yet to deal with these aspects make detailed consideration essential. In recent discussions on the subject, however, attention has been paid to the total credit activities of the public sector, which includes for the purpose lending by government as well as by financial institutions owned or sponsored by government and guaranteed loans. The issues here are restricted to government lending.
Nature and Extent of Government Lending
National goals may be pursued through loans, loan guarantees, or other intervention in capital markets rather than through tax-supported direct expenditures. This trend has developed over the past four decades and its origins can be traced in part to budgetary deficits that could be incurred with the assurance that the loans raised for financing them would be used for productive ventures that were expected to be self-financing over time. It was expected that if productive ventures were separately organized the budget would be used as a vehicle to provide loans and that the receiving agencies would generally be free from the traditional financial controls exercised in government. Later, as government-owned enterprise activities grew, the overall borrowing requirements of those agencies were pooled in some countries and were raised by the government in the market. Such central pooling and related mobilization were considered an essential part of the overall framework of controls needed to pursue appropriate macroeconomic policies. Such pooling also proved beneficial to those enterprises that could not borrow on their own from the market.
Government lending now takes, broadly speaking, three forms: direct lending by the administrative agencies whose transactions are included in the public budget; on-lending of funds raised on a pooled basis and transmitted to government-owned and controlled agencies; and lending by government-owned financial institutions organized specifically to further the policy goals of government. Outlays of the first two types are included in the government budget and are shown, for administrative control purposes, in gross terms. Transactions in the third group are not included in the budget, although analytical statements on their operations have in recent years become part of the information provided in budget documents.
Data on government net lending—new lending minus repayment received during the year—are provided in Tables 9-12. These tables show net lending as a percentage of GDP and as a percentage of total expenditure in the OECD countries and in selected developing countries, at central government level for 1980–90. In general, net lending activity is understated in recorded government accounts, reflecting in a number of cases the lag between approval of a loan and its eventual inclusion in government accounts.4 In some cases, owing to government reorganization, some lending could have been debudgetized. The data in the tables do not distinguish between direct loans and on-lent funds. In addition, the bunching of repayments tends to reduce the magnitude of loans provided in some years.
Total expenditure includes net lending.
Total expenditure includes net lending.
Total expenditure includes net lending.
Total expenditure includes net lending.
Experience during the eighties shows that government lending as a share of GDP declined in both the OECD countries and the developing countries. Even in Nordic countries, which had a high rate at the beginning of the decade, it tended to decline gradually toward the end of the decade, although the share of lending continued to be high in Norway. These trends are supported when net lending is considered as a percentage of total expenditure also. In the developing world, many countries also showed a similar trend, although the decline in India is less sharp than in others, and in Sri Lanka lending operations more than doubled.
Although the experience of the eighties is not very different from that of the seventies,5 it is probable that the growing fiscal problems and diversion of some traditional government activities to the private sector as part of the privatization programs may have diminished the importance of government lending in terms of magnitudes. Also, in some of the highly indebted countries such as the Philippines, the budget had little leeway for lending programs to be maintained at their previous levels.
In addition to direct government lending, a large number of government-owned financial institutions are engaged in providing credit. Most of these are in insurance, export guarantees and related export promotion, housing credit and mortgage operations, and a wide variety of other development financing activities. Management of pension funds is also common. These institutions are invariably organized with initial capital provided by the government; some of them also accept deposits and are a recognized part of financial intermediation. The form in which annual capital is provided to those institutions from the government budget differs among countries. Practices are far from uniform and the terms are also differently interpreted. In some countries, a few activities are organized on a revolving fund basis and are therefore expected to be self-sustaining. These revolving funds and others may sometimes receive additional loans from the government for specified purposes. In other cases, substantial subsidies are provided annually to compensate for the pursuit of noncommercial objectives. In some countries, these purposes may also be served through equity participation and through capital transfers that may be used differently by the receiving agencies.
The purposes of government loans and their magnitude are influenced in several ways, including economic and accounting factors. Although the applicability of these factors differs from one country to another and from one phase of the economic cycle to another, it is clear that government net lending is not a matter of managing liquidity to earn a return but is undertaken for public policy purposes. But the social service to be served is slightly more remote than that sought through direct public expenditure, and the financial solvency of the recipient is important in determining the conversion to a loan of what would otherwise be regular expenditure.
The purpose of a loan is often also to supplement or stimulate private lending rather than to substitute for it. Also, after initial successful efforts by governments, specific activities have sometimes been eliminated from public budgets and entirely financed from the commercial banking system and related financial institutions. The practical procedures for providing loans are also often the same as private sector practices, and the rate of delinquency in repayment of public sector loans may also be similar.6 While this could raise the issue of the need for government lending, its distinguishing features are that it is made available to income groups and sectors on far more lenient terms than is private sector lending. Determining the implicit cost of such distributional concerns is significant and needs to be addressed during the budgetary process.
Government lending is also utilized because it does not generally affect the level of taxation (although there is bound to be impact on the credit markets that in due course would affect the budgetary outlays on interest). It is also far more flexible than expenditure in that specific and categorical loans extended to other autonomous organizations and levels of government may be reduced or increased more easily than expenditures. Although loans, like expenditures, will lead in due course to the formulation of clients’ groups and lobbies, their impact is less than with expenditures.
Lending is also considered more important as an instrument of countercyclical policy that can be used more effectively during recession, and, once the normal pace of economic activity is resumed, the magnitudes of promotional loans decline.7 In promoting such lending during recession, however, particular care is needed to ensure that the objectives of lending do not conflict with those of monetary policy.
Lending transactions are part of the overall fiscal stance of governments. Although the element of lending included in public budgets is not substantial, when viewed together with the financial requirements of the budgets and other borrowing undertaken by autonomous organizations, it could convey unintended signals to the financial markets. The contribution of government lending to the overall fiscal strategy should therefore be explicitly recognized in formulating the budget.
Budgeting for loanable funds involves the same considerations as are applicable to direct expenditures. The extent to which the budgetary process is entirely successful in fully considering the various implications of loans depends on the specific attention given to the following factors.
At a policy level, it is necessary to determine the obligation to intervene in the management of the economy, and, more specifically, in a given sector. The need for a pump-priming role has to be specifically determined, followed by more detailed consideration during financial planning of technical issues relating to the quasi-commercial nature of the recipients’ activities and their capability to service loans in the future.
Normatively, the budgetary process should also devote itself to considering the allocative and stabilization aspects of the proposed loan programs. The allocative aspects include the changes in the direction of economic activity as a result of loans; the magnitudes of subsidies and their impact on the allocation of resources in the economy; and the allocation of credit among public and private sectors in the economy and its impact on the allocation of resources. These considerations involve identifying and specifying the noncommercial objective to be fulfilled and its short- and medium-term budgetary and real implications. Subsidies computed for the limited portion of a fiscal year may not show the whole picture and should therefore include the costs involved for the duration of the loan. Allocative considerations, when extended beyond the normal fiscal year, would also have to include the intertemporal aspects of any consequent debt financing. Debt issued today for financing loans implies that taxation for expenditures (if these are not really loans but only in formal and technical terms) is only being delayed and will eventually be paid by a generation that has not benefited from them.
The stabilization aspects include considering the impact of government lending on demand management and on financial markets (to ensure synchronization with monetary policy). Broadly, the impact of lending on the economy is the same as direct expenditure, although some empirical evidence suggests that the multiplier effect of loans is slightly lower than expenditures. Both loans and expenditures add to the deficit and have to be financed by taxation or borrowing. It is, however, difficult in practice to identify and measure the initial incidence and eventual impact of loan transactions as a separate category. Often, an increase in loans could coincide either by intent or otherwise with a reduction in expenditures or an increase in tax revenues. Thus, the expected impact might be fiscally offset by movements in the other ingredients of the budget. Therefore, conventions were developed in recent years to add net lending to expenditures, so that the overall impact of the budget could be assessed.8
The impact of fiscal policy on output is also felt through interest rates and inflation rates. To minimize any adverse impact, an appropriate fiscal stance that seeks a balance between expenditure taxation and borrowing is therefore needed. Thus, to avoid excessive pressure on the monetary variables, the level of public borrowing consistent with the proposed fiscal policy signals should be specified. The dimensions of such public borrowing are much larger than contained in the budget. The total of borrowing computed for the purpose comprises direct budgetary borrowing, borrowing by other organizations and other levels of government, and guaranteed loans.
If these aspects are satisfactorily taken care of in the budgetary process, the choice between taxes, loans, and expenditure to pursue public policy can be relatively easily determined. But in practice they impose different requirements that are often met by different organizations. Taxes and associated revenues are estimated by certain agencies, while the economic impact of the budget on the economy is estimated elsewhere. The impact on the financial markets is analyzed in conjunction with monetary policy, and the budgetary process is usually limited to determining the requirements of the agencies. It was partly to compensate for this compartmentalized approach to budgeting that governments started, notwithstanding difficulties, formulating forward budgets and medium-term financial strategies. Thus, a more systematic convergence of different approaches is achieved through the improved financial planning machinery that is inevitably, but appropriately, larger than the normal budget process itself.
The issues encountered in the management of loans may be divided into two broad categories: presentational and analytical; and those relating to the technical aspects of control.
The first issue from a presentational point of view relates to the coverage of the loans provided by the public sector. These loans, as noted earlier, transcend the boundaries of the traditional budget and include those by other autonomous agencies and guaranteed loans. Although during the early sixties these loans were examined and analyzed in terms of the respective jurisdictions, the much broader basis of financial planning has made this issue somewhat academic nowadays.9 But if the practices of medium-term financial planning are uneven, the issue may not receive the due attention.
The second aspect concerns the presentation of data in gross and net terms. To assess the impact of expenditures and lending on the economy, lending has to be treated in “net” terms. From the point of view of financial markets, however, the gross terms—particularly the magnitudes of the new loans issued in the market—are more important. From an accounting point also, gross treatment is considered essential to facilitate the implementation of the series of checks and balances on flows into and out of government funds. Maintaining these two sets of data is not difficult because the budgetary data are presented on a gross basis while the analytical statements attached to budget documents are in net terms.
The third aspect relates to the form in which data have to be provided on lending activities. The experience of countries reveals three broad approaches: a separate lending or credit budget covering all types of loans by all agencies, including guaranteed loans; a capital budget covering government loans, including borrowing by government; and analytical statements, distinct from the main budget, containing relevant information. The purpose of all these approaches is to provide more information for use by the public or for the necessary action by the legislative bodies. Although no particular approach displays any demonstrated superiority over others, a comprehensive credit budget may imply an exaggerated picture of control, because, in practice the activities of autonomous agencies may not be as amenable to control as government loans. The second approach is less important as capital budgets are being relinquished in a number of industrial countries that previously had them. Analytical statements, although providing much information, are not operational in the same way as the main budget.
The major issue in operational terms is the computation of the element of subsidy involved in government lending programs. Such a computation should take into account: (a) the interest rates charged that are less than market rates; (b) the costs of administering programs; (c) the monetary implications of the social objective pursued; and (d) future losses arising from default in payment. It is frequently pointed out that most loans are so “soft” that they should in fact be treated as grants from the beginning. While this aspect raises the more intractable issue of what is in fact a loan, the operational issue revolves around how the subsidy, however computed, is to be shown in the budget. In theory, the whole interest subsidy could be capitalized and shown in advance in capital budgets, but this approach will result in notionally higher estimates of expenditures. Another approach is to adopt the pay-as-you-go procedure and show the interest subsidy for each year for the duration of the loan. Yet another way is to present these subsidies as memorandum items so that the problem is quickly recognized. Although all calculations will clearly involve some degree of imputation, a more pragmatic approach is to follow consistently over a period a specified method—even if it is not comprehensive—to yield comparable results. In this context, the practices of the U.S. Government are instructive. In the U.S. budget, legislation requires that the subsidy element in its credit programs be shown as a budgetary outlay and that the budgetary authority be specifically approved by the legislature for this purpose.10 In addition, the Government is expected to show estimates of loans written off as well as loans likely to be defaulted. A third measure that the Government is expected to introduce, following the intent of the Credit Reform Act of 1990, is an accrual method of accounting both for credit and insurance programs so that the costs to the Government can be recognized as they arise and even if financed from the budget over a number of years (cash budgets do not provide the full measure of the ultimate costs to the Government).11 These measures, which are comparable to the practices of the private sector for identifying nonrecoverable loans or bad debts, provide institutionalized approaches to consider the full implications and disclose the costs of continuing existing programs unchanged and the implications of any policy changes. Credit repayments and outflows have to be considered in an overall macroeconomic context. This may then lead to a reconsideration of the total framework of on-lending to other levels of government.12
The technical aspects of control are best examined in terms of the agency exercising the control and the point at which it is exercised. For government loans, control broadly follows the pattern of other outlays in that review procedures, release of funds, and monitoring tend to be the same. However, loans to individuals are often demand determined and to that extent tend to be open ended. The amounts lent may therefore be more than budgeted, depending on the economic situation. In such circumstances, the only viable means of control is to seek a change in the relative legislation. More important, any depiction of government loans as the only area in need of control may have an unintended consequence in that the other areas of the budget may acquire a legitimacy that can often prove counterproductive later. That is not to underestimate, however, the importance of recognizing lags in the repayment of loans and associated administrative action.
Loans extended by autonomous organizations are broadly controlled by the respective organizations. The role of government, depending on the framework of formal and informal controls, may vary from being that of an innocent bystander to that of an effective agency that guides and controls every working aspect of these organizations.
Because these organizations are mostly autonomous, the role of the government may be restricted to showing greater vigilance in providing funds annually and leaving the day-to-day operations to the organizations. These aspects raise a host of issues that are best examined as part of the overall relationships between government and its autonomous agencies. It is clear, however, that orderly management of the lending operations requires systematic recognition of their subsidy elements and assessment of the impact of their operations on the financial markets and on the economy, and these budgetary processes should be strengthened to reflect these concerns.
The government also provides guarantees for loans undertaken by agencies, enterprises, and other autonomous agencies under its broad control as well as private sector corporations in some situations. These guarantees may be provided either for domestic loans or for loans obtained from international financial institutions or raised in foreign capital markets. Although the provision of such guarantees has long been in practice, its importance, particularly for the level of fiscal deficits and protracted serious fiscal imbalances, became recognized in the early eighties, mainly as a result of the experience of several Latin American countries. They had provided numerous guarantees (for which there were no comprehensive records), most of which were defaulted by the original borrower, and their servicing and repayment therefore had to be assumed by the government, thereby adding substantial burdens to the budget over a period of years.
A guarantee is a potential or a contingent liability on the consolidated funds of the government. Such guarantees are warranted where the borrower does not have the requisite creditworthiness or where limited creditworthiness is likely to entail high borrowing costs (in interest rates and repayment periods). The government may consider providing such a guarantee as an extension of its industrial, agricultural, or other sectoral or macroeconomic policies. The use of these guarantees has become widespread as most of the government-owned enterprises were undercapitalized to start with and lacked creditworthiness in the markets because of their own lackluster financial performance. Governments considered the provision of guarantees in such situations soft options with no immediate adverse impact on their finances while facilitating the pursuit of their policy objectives. Any soft option carries with it the potential for lax control. As most of the guarantees were neither approved by the legislatures nor conveyed to them as budgetary information, no institutionalized process existed to monitor the progress of their recipients and the potential for default. In some countries, the prevailing cash-oriented budget systems that permitted liabilities to be recognized only when they landed on the guarantor for payment aggravated the situation. Even where accrual systems of budgeting were in effect, the lack of comprehensive records prevented the systems from delivering the information for which they were designed.
From the viewpoint of effective management, three aspects merit consideration. First, an adequate system should exist to force the ramifications of the proposed guarantees to be considered. As a corollary, there should also be a system that permits the subsidy element inherent in such a guarantee to be computed. Second, policy or procedural safeguards to minimize the adverse impact on the public exchequer should be designed, and third, a monitoring system of the progress of the recipient of the guarantee should be available.
Any system designed to review the merits of guarantees should take into account the role of these guarantees in the broader arsenal of policy instruments available. As noted earlier, direct expenditures, loans, guarantees, and tax incentives each offer some scope for pursuing a stated objective. If ceilings apply to the resources directly provided from the budget or associated means, a limit on guarantees could also be prescribed. Without such a ceiling, liberal provision of guarantees could adversely affect a government’s own creditworthiness and, as a consequence, its borrowing costs may be higher in the medium term. Moreover, such ceilings may induce more rigorous scrutiny. Finally, only the strongest may survive the claims of rivals. The subsidy element needs therefore to be computed and explicitly shown in the budget documents.13
The practices of some Nordic countries (and Germany) show that a small fee is levied (as a percentage of the total loan in some cases and as a flat rate in others) payable by the recipient of the guarantee. The advantages of this procedure are twofold. A user charge provides a source for a type of insurance fund that could be utilized in the event of a default. Although it is recognized that the fee or the premium may not be adequate to minimize the impact, its value as a disciplining factor cannot be ignored. Second, the procedure is also helpful in forcing consideration of an issue that might otherwise be relegated to the background.
Associated with the above, a monitoring program, parallel to the budget system, that compels periodic review and anticipation of possible defaults and ways of financing them should exist. As a first step, the publication of data on guarantees as part of budgetary information and the completed accounts of the government may have a salutary effect on the whole process.
These accounts, unlike lending programs and guarantees, have been less visible and for that reason have not been considered extensively either in the literature or in expenditure management practices.14 This benign neglect may have arisen because quasi-fiscal accounts belong to the operations of a country’s central bank and therefore have not received much attention in fiscal literature.
The accounts essentially reflect the activities of a central bank, but they are more like the fiscal operations of the government and are financed from its own revenues. Although comparative data are not available on the extent of these activities, they broadly include selected public debt transactions (illustrated further in Chapter 7), financing the losses of exchange rate stabilization efforts, credit to commodity boards (involving the provision of subsidized credit for the procurement of coffee, rice, wheat, soybean, and other crops), refinancing of agricultural, industrial, and housing activities (at differential interest rates and usually below the government’s borrowing rates), the bailout of ailing banks and other financial institutions, and contributions to development funds. The composition of these activities differs from one country to another. In heavily indebted countries, public debt transactions dominated the quasi-fiscal accounts during the eighties. In some countries, credit to and refinancing of agricultural and industrial activities are prominent. Although the dividing line between monetary transactions and quasi-fiscal operations may in theory be somewhat difficult to draw, in practice, however, the activities represent part of government policies, but if they are pursued outside the government budget, they understate government expenditures and the size of its fiscal deficit. To that extent, the scope of a central bank as a fiscal agency is considerably larger than that traditionally perceived in that role. Two consequences of this have been the loss of transparency (that would have been there if these activities were included in the budget) and the phenomenon of central bank losses.
In origin, quasi-fiscal accounts were more like the soft option exercised by the government to obviate the need to mobilize additional resources. The view may also have been that expenditures transferred to banking institutions take them out of the rough and tumble of annual budget making. Once the general guidelines are set, the bank, following its traditional role, assumes a fact-based judgment in its lending operations. In practice, however, the exclusion of these accounts reduced the boundaries used in fiscal policymaking while creating a sphere of entitlements. More significant, central banks may also have reduced their own stature and autonomy by pursuing, like any public enterprise, some noncommercial objectives at the behest of the government.
From the point of view of effective expenditure management, more is likely to be gained from integrating these accounts into the budget and subjecting them to the same financial discipline as other expenditures. While restoring the credibility of central banks, integration would also enhance transparency and legislative accountability.
Strict observance of the advice to be neither a borrower nor a lender would make the operations of the banking system somewhat questionable.
The usage of the terms varies: for example, in the United States, the term generally used is credit programs, which includes direct loans, loan guarantees, loan insurance, or financial contracts designed to support borrowing. For the purposes of this chapter, these are treated as separate categories and loan insurance or financial contracts designed to support borrowing are not included. Loan insurance is a type of guarantee in which an agency pledges the use of accumulated premiums to secure lenders against default by the borrowers. A financial contract is an agreement made by an agency, whose primary purpose is to make private credit available to a nongovernmental entity on more favorable terms than without the contract by agreeing to assume the risk involved. It is a type of guarantee.
The agency noted that in fiscal year 1990 government agencies wrote off $3.4 billion in loans. The Office of Management and Budget estimated that, at the end of September 1990, of a total amount of $210 billion outstanding in the form of direct loans, over $57 billion (or 27 percent) was expected to result in losses. Some of the loans extended by the Agency for International Development to other countries as part of the country’s foreign aid program are considered grants for purposes of national income accounts, because of the long amortization periods, low rates of interest, and frequent write-offs of interest and principal for political purposes.
Such problems do not arise in industrial countries. For example, in the United Kingdom, the National Loans Fund (the only major fund except for the all-inclusive Consolidated Fund) includes all lending transactions as well as service of the national debt. For an illustrative discussion of these flows, see Mowl and Todd (1990).
For this period, see Premchand (1982).
Both are affected by changes in the economy. Generally, as an economic downturn reduces income and employment, loan delinquencies and defaults occur for both the private and the government sectors.
For an interesting empirical study of this aspect, see Break (1965); the issue of whether as a result of economic downturn the losses stemming from defaults should be borne by the taxpayer is not discussed.
Lending transactions are not included in the national income accounts as they are not considered to be income-generating expenditures.
The narrowing definition of government in financial planning, as in the United Kingdom, may contribute to a loss of information. The U.K. coverage has narrowed over the years from the public sector to general government. From 1990 even local government expenditures are excluded from the planning total. See Heald (1991).
Although there are various methods for calculating the subsidy—the explicit and the implicit parts—concern so far has been with calculating only the explicit subsidy. In the United States, the subsidy element is equivalent to the difference between what the borrower pays and the amount required to reinsure the loan. As the U.S. budget documents state, “cost is defined as the present value, discounted at Treasury rates of comparable maturity, of the expected cash outflows from the Government minus the expected inflows to the Government” (United States, 1992a). See also Wattleworth (1988).
In India, for example, a substantial part of the outstanding public debt reflects on-lent funds to centrally owned public enterprises and state governments.
In the U.S. Government budget, the subsidy element of direct loans as well as guarantees is shown in the budget. A similar practice is found in Germany and the Netherlands. In Canada, guarantees need parliamentary authority and the expenditure estimates include cash payments in the event of default (the latter is common in other countries).