2 Fiscal Policy and Mobilization of Savings for Growth

Mario Bléjer, and Ke-young Chu
Published Date:
June 1989
  • ShareShare
Show Summary Details

Mario I. Blejer and Adrienne Cheasty*

I. Introduction

The mobilization of domestic financial resources may be analyzed on two levels: first, one needs to determine the total volume of savings and how these may be channeled into the financial system in order to make them available for investment; and second, once resources are being absorbed into the financial market, one needs to determine how the system may function most efficiently in directing these resources to where they are most needed in the economy. Thus, the government may attempt to mobilize resources by affecting the supply side (savings) or the demand side (domestic investment) of the economy. On the savings side, the government may be concerned, first, with increasing total savings, and then with ensuring that a significant portion of savings is directed toward the financial system. Given this level of “loanable funds savings,” 1 the government may wish (i) to overcome imperfections (distortions) in the capital market, which prevent investors from using these funds to finance projects which yield the highest social rates of return; and (ii) to raise the potential rates of return attainable by savers by itself engaging in capital expenditures (mainly on infrastructure and public goods) which generate no immediate pecuniary profit but which increase the effectiveness of private projects and thus the demand for loanable funds to finance them. These last two roles of government may be characterized as entailing the removal of negative externalities, which distort the social return to capital, and as enhancing positive externalities, which increase the social return to capital.

Underlying this description of the role of government is the assumption that private capital markets are not perfect. If they were, there would be no distortion in the supply of, or the demand for, loanable funds. On the supply side, the socially optimal level of savings and its allocation between markets would already exist, and, therefore, attempts by the government to change either aggregate savings or its composition would lower long-run economic welfare. On the demand side, rates of return to capital would provide the correct signals to investors, and market arbitrage (as described by Coase (1960)) would internalize the external benefits or costs to investment. While the extent and gravity of imperfections in capital markets continue to be a matter of debate, there are several reasons why existing imperfections are exacerbated in developing countries. In consequence, if the government can play any role in mobilizing domestic financial resources, that role will be more crucial in developing countries, where the private sector is less able to generate and allocate loanable funds productively.

This paper will discuss the mobilization of domestic financial resources in three sections. Section II will examine the characteristics of developing countries which hinder the performance of their capital markets. Section III will discuss the traditional tax policies governments have used to try to compensate for poorly developed financial systems. Because there are evident problems (both microeconomic and macroeconomic) associated with such policies, Section IV will discuss an alternative solution open to a government which wishes to generate loanable funds savings. The government itself may become a net saver by generating a budgetary surplus and thus freeing financial resources for use by the private sector. Governments can be compensated for the fiscal discipline this strategy requires by sizable benefits at the macroeconomic level.

II. Characteristics of Developing Country Financial Markets

Several characteristics of developing country capital markets make it likely that savings will be low and that they will include a large share of nonloanable funds—that is, resources not available for the purchase or creation of new capital. Wealth holders in developing countries frequently make intertemporal transfers of wealth by holding consumer durables, such as jewelry, works of art, or livestock, which are not defined as savings; by buying land or housing, forms of saving/investment which only indirectly generate output; or by holding financial assets outside the financial system, in informal (“curb”) markets, which cannot be measured as savings or offered openly to investors. This non-optimal level and allocation of savings may arise for the following reasons:

(1) The capital market is undiversified and fragmented. The size of the capital market in developing countries is small, and the scope for diversification of financial institutions and financial instruments or assets is limited. Typically, the banking system is confined to one central city, supplemented perhaps by a few peripatetic administrators of development bank credit. The equity market in such countries is rarely well developed, which means that investments in physical capital tend to be few and large relative to the total size of the financial capital market, with consequent wide differentials in the rates of interest offered on different loans. These differentials are exacerbated by poor communications, which prevent the flow of information and advertising of the quantities and prices of funds available. Thus, it is difficult to pool funds and to arbitrage rates. (The largest information distortion, of course, is the dissimulation of all transactions which take place in informal markets, where the government, for example, cannot bid for funds.) Furthermore, the government may often have very clearly defined sectoral priorities which it expresses through a conscious policy of making different quantities of funds available, at different interest rates, to different demanders. The result of these shortcomings is the market’s inability to generate a single clearing rate of return to financial capital. Wedges driven between borrowing and lending rates may lead to such anomalies as an individual saver finding, on the one hand, that banking his savings is not worthwhile, but, on the other hand, that borrowing to support a planned enterprise is too expensive. This leads to increased use of retained earnings for investment in developing countries, compared with the countries where financial intermediation is more developed, reducing both the pool of loanable funds and the likelihood that the rate of return to financial saving will equal the cost of capital.

(2) Financial returns to savings and/or investments are insufficient. The discrepancies between borrowing and lending rates described above may make returns in the financial system unattractive. Returns may be inadequate for two other reasons. First, controls and intervention are widespread in many developing countries. Fixed interest rate policies quickly lead to financial repression as foreign interest rates change, exchange rates move, or the inflation rate rises. In this case, even if the capital market were characterized by complete information and open access, savings would fall short of investment and loanable funds would have to be rationed, since their price would be prevented from functioning as an effective allocative mechanism. Second, in less urbanized areas of some developing countries, nonfinancial assets (such as housing, jewelry, or livestock) may serve as currency substitutes, and this may raise their liquidity premium above what bankers would perceive it to be. The rate of return to nonfinancial assets might also perhaps be higher outside cities if these assets were valued for nonpecuniary benefits attributed to them. (An example would be the accumulation of unprofitable land to consolidate an estate.) In such cases, it becomes difficult for the financial sector to measure the opportunity cost of alternative assets, and thus to offer a competitive rate of return. In a large capital market, equilibrium prices which take into account all available alternatives may emerge quickly, but in the markets of developing countries, the number of transactions that take place may be insufficient to permit the market to clear and thus provide equilibrium valuations of nonpecuniary benefits.

(3) Financial assets bear uncompensated risks. The distortions in returns described above would occur even in a riskless economy. In fact, however, potential savers may perceive risks to entering the financial system which they do not consider are compensated for by an adequate risk premium on the rate of return because of the various rigidities in the market which prevent the emergence of a clearing rate. The liquidity of financial savings may not be guaranteed in many developing countries, where frequent changes of government policy (or, indeed, of government regime) close banks, change interest rates, retire bonds, or alter the purchasing power of the currency, often with little or no advance notice. These abrupt changes in policy are often the result of unavoidable exogenous shocks, and even if such shocks do not affect the stability of the financial system or the political system, they may cause the costs of imports and exports to fluctuate widely, making financial assets denominated in domestic currency a bad risk unless interest rates are free to vary with the expected rate of exchange rate depreciation.2

Savers may perceive a further risk in the danger that financial assets will sooner or later be taxed. Typically, the tax base in developing countries is narrow, and governments are constantly searching for ways to widen it. Income and wealth already denominated in financial terms are by far the easiest to tax, so that, while a government may be aware of all the caveats attached to the taxation of savings, it may find the revenue-raising potential of a wealth tax or a capital-gains tax, on financial assets, impossible to resist. Nonfinancial assets, which are intrinsically harder to value and which are much further removed from the notice of the taxing authorities, may be seen by investors as yielding a far more certain stream of future returns.

III. Traditional Role of Government in Developing Countries’ Domestic Financial Markets

In theory, all of the distortions described could be removed by adjusting the incentives to holding financial assets appropriately, either through taxes or credit policy. As is evident from the description of these distortions, however, the most intransigent problem may be one of information. The government may not have complete information on the rate of return that would compensate people for the costs of storing savings in a faraway city, for forgone nonpecuniary benefits, and for perceived risks. In a larger capital market, sufficient numbers of transactions would take place for the clearing rate of return to manifest itself, but in developing countries the whole problem reduces to the fact that, because the capital market is so small, if the government wants incentives to be appropriate, it has to set the level of these incentives itself without receiving signals from the market. In a second-best world, although the government was unable to locate the market-clearing price precisely, it might at least try to encourage savings and investment by increasing returns through tax policy (providing either broad-based or selective incentives) or by favoring certain projects through credit policy.

Tax policies to mobilize saving are intended to work by increasing the return to future consumption—either to savings in general or to savings held in a specific form. Under most tax systems, savings are taxed twice, first when total income is taxed, whether it is consumed or saved, and then again when savings generate interest. It has been argued 3 that if this discrimination were to be lifted or lessened, total savings would rise. The scope for increasing saving in this way extends as far as the extreme case where only expenditure is taxed.4 Likewise, if the penalty for saving were lighter on some forms of assets than others, people would tend to shift into those assets.

The policy recommendations which follow from this argument range from broad-based proposals to convert present income taxes into personal taxes on consumption expenditure or into a value-added tax on the basis of consumption, to highly specific tax incentives, such as altering the tax treatment of social security contributions.5 On the one hand, it is notable that broad-based incentives have not been widely used: of 96 countries surveyed for the 1983 edition of Coopers and Lybrand’s International Tax Summaries, only Vanuatu, the Turks and Caicos Islands, and The Bahamas had well-developed sales taxes in lieu of income taxes. The fact that no country employing a value-added tax used it to replace an income tax suggests that revenue-raising characteristics (and perhaps distributional characteristics) are deemed to be far more important than any gains in aggregate savings.

On the other hand, many countries exempt interest income derived from certain assets or institutions; some countries exempt all dividend income or certain types of dividend income; and some countries’ tax policies favor retained earnings. For instance, interest on various types of bank deposits is exempt from taxation in Argentina, Guatemala, Iran, the Republic of Korea, Malta, Panama, and Paraguay; and financial company and/or investment company income is tax exempt in Argentina, Cyprus, the Dominican Republic, the Netherlands Antilles, and Panama. Housing is favored by the exemption of building society income (in Dominica, South Africa, Swaziland, and Venezuela) and mortgage bank income (in the Dominican Republic). Saving in illiquid, low-risk, low-return forms is encouraged by exemptions or deductions on life insurance or provident fund contributions (in India and Malaysia), the income of insurance companies (in the Netherlands Antilles), and income from pension funds (in the Federal Republic of Germany and Nigeria). Dividends are favored in different ways (which entail different distortions): by exempting from tax the dividends of new companies in France, intercorporational dividends in India, dividends paid to foreigners in the Republic of Korea, and new equity derived from retained earnings in Indonesia. Governments may also try to increase nonbank government financing by exempting interest and gains from government securities (as in Argentina, Panama, South Africa, Swaziland, and Venezuela). Other specific savings incentives are more unusual: for example, Brazil requires employers to put 8 percent of employees’ salaries (above social security contributions) into untaxed bank accounts in the names of the employees. In Brazil also (as in the Netherlands Antilles, Norway, Portugal, and Sweden), certain companies’ reserves (usually designated formally for future purchases of capital) are not taxed. Brazil further exempts the income of corporations (up to 25 percent of the corporations’ total tax liability) if it is invested in the Fundo de Investimentos da Amazônia (FINAM)—the financial organization responsible for developing the Amazon region. More generally, if people are assumed to derive utility from making bequests, it might also be argued that countries which do not levy an inheritance tax are favoring saving compared with countries which impose such taxes.

1. Microeconomic Consequences of Tax Incentives

a. Income and Substitution Effects

The reluctance to apply broad-based savings incentives perhaps stems from the recognition that “even if a change in tax policy would increase welfare by reducing discrimination against capital accumulation, it does not follow that its adoption would result in more saving.” 6 This is so because a reduction in the return to saving through a capital income tax has both an income effect and a substitution effect. The increase in the relative cost of saving makes individuals prefer present consumption to saving (and, therefore, to future consumption). However, the tax-induced drop in the rate of return is equivalent to a drop in the lifetime income the individual can realize. If the individual desires to spread his consumption evenly over his lifetime, he will have to maintain a higher level of saving, in light of the increased cost of saving, than he would otherwise have chosen; indeed, he may find he has to save more than he did before, in order to maintain a stable (though reduced) consumption pattern. Summers (1981) notes a further effect arising from the imposition of a tax if it reduces the rate of interest. As the rate of interest falls, the present value of future earnings rises (because it becomes more costly to substitute savings out of present income for future income in order to sustain future consumption). Thus, lifetime income—that is, present and future earnings—rises somewhat to offset the reduction caused by the income effect.

Given these effects, a reduction in the distortion caused by the tax will have an ambiguous result. Since we cannot, a priori, tell whether capital income taxes will decrease or increase saving, we cannot predict whether the removal of that distortion will move people into saving because it becomes relatively more profitable, or out of saving because, given the higher return to saving, they need to save less than before to maintain the level of consumption they had planned for the rest of their lives.7

b. Interest Elasticity of Saving

From the discussion above, it is apparent that there is no strong theoretical reason why savings incentives should increase aggregate savings. The question becomes an empirical one—whether, in reality, the savings-increasing relative price change (the substitution effect) is bigger than the savings-decreasing income change. This combined effect is measured by the uncompensated interest elasticity of saving. There exists a sizable literature in which attempts are made to estimate this elasticity. Unfortunately, most of these estimates have been made for industrial countries. It may be argued, however, that the better-functioning capital markets in these countries would allow a more responsive flow of funds following an interest rate change, compared with countries where the capital market suffers from the defects described earlier. Thus, interest elasticity estimates for industrial countries could serve as upper-bound indicators of interest elasticity estimates in developing countries. Boadway ((1984), p. 313) cites several studies which found saving to be more or less insensitive to changes in interest rates. This finding is borne out by repeated time-series results on the proportionality of the long-run consumption function. Boskin (1978) calculated an interest elasticity of saving of 0.4 for the United States, but this estimate has been considered a higher limit and subsequent studies have strongly questioned his calculations. For instance, Howrey and Hymans (1978) found that no definition of saving (loanable funds, flow of funds, or national income accounts) showed a significant interest rate effect. While their results have, in turn, been criticized, subsequent studies for specific middle- and low-income countries continued to show either an insignificant relationship between the interest rate and saving or a very small positive elasticity.8

It must also be remembered that, even if there is a positive interest elasticity of saving, it will not be of significance for policy if the change in the interest rate necessary to effect significant changes in saving, welfare, and output is too large. The small size of the estimated interest elasticities suggests that the required interest rate changes would be costly. Even Boskin’s estimate of a 0.4 elasticity would imply that a 2.5 percent increase in interest rates would be necessary to engender a 1 percent increase in total savings. In low- and middle-income countries (including China, India, and oil exporters) an optimistic estimate of the ratio of gross domestic savings to gross domestic product (GDP) in 1982 (according to the World Bank’s World Development Report, 1984) would be 20 percent. Thus, to raise S/GDP by 1 percentage point, it would be necessary to change the real after-tax interest rate by (1/0.2)·(1/0.4) percent—that is, by 12.5 percent. This strong result is indirectly supported by the results of Mackenzie’s (1982) simulation model, which was calibrated for U.S. data. It estimated the changes in output that could be expected after various combined wage and saving tax changes under different labor supply elasticities and interest elasticity of saving scenarios. Mackenzie found that if the tax rate on saving were reduced by 50 percent at the same time labor income taxes were reduced by 20 percent, then, in an economy with an interest elasticity of saving of 0.4 and a unitary labor supply elasticity, output would increase by only an average of 0.12 percent per annum during the first ten years after the policy change. If the labor income tax had not been reduced, then presumably the effect on output of a change in interest income tax would have been weaker.

In sum, given the small and uncertain magnitude of interest rate effects on savings and output (as indicated in economic theory, because income and substitution effects at least partially offset each other, and as indicated in empirical work by the lack of evidence of a strong interest elasticity of savings), the other potential uses of interest rates as policy instruments, and the international constraints imposed on interest rate determination in open economies, it is not surprising that governments have not often attempted to influence aggregate savings directly through changes in the rate of interest.

Studies for industrial countries suggest, however, that the composition of savings can be affected significantly when incentives to hold wealth in various assets differ.9 Unfortunately, these studies are not applicable to developing countries, where the desired change in assets is not between different financial assets but between financial assets as a whole, on the one hand, and “nonloanable funds” assets, on the other. As described in the previous section, the premiums placed on nonfinancial assets because of their liquidity value or their nonpecuniary benefits may not be eliminated by marginal changes in the returns to financial assets. Thus, there exists little statistical evidence of success in overcoming these seemingly qualitative differences between financial and nonfinancial assets in developing countries.

Hence, while the specialized incentives listed above, which are undertaken in a sizable number of countries, may succeed in changing the form of the financial assets in which savings are held, they do not properly address the problem of attracting a larger supply of loanable funds from nonfinancial stores of wealth.

c. Distortions Created by Tax Incentives

Even if tax incentives to save did generate a large response in the quantity of savings, it would not necessarily follow that the government should use them as policy instruments to increase aggregate financial savings, because the distortions inherent in them might not be less detrimental to economic welfare than heavier taxes; this would be true whether or not savings were affected. Any tax on savings distorts the choice between present and future consumption. As discussed, for example, in Becker and Fullerton (1980), the distortions are worsened when different types of savings are taxed at different rates, because if after-tax rates of return are equalized across different types of savings, then, given non-uniform tax rates, the pretax return to each type of asset must be different. This distorts the flow of funds into each type of asset. This is why incentives to change the composition of savings are effective, but if the government wishes only to alter the level of savings, a system of specific tax incentives, implying many different tax rates, may lower welfare more than a heavier, but uniform tax system. The problem is exacerbated in developing countries, where misperceptions and lack of information about the real rates of return to various assets (owing, for example, to their nonpecuniary benefits) make it very difficult to set tax incentives to achieve desired directions and magnitudes of portfolio adjustment.

Some specific examples of distorting tax incentives would be the removal of interest receipts from the personal tax base—which would encourage firms to use debt financing and distort the debt-equity choice unless dividend income were also exempted—and vice versa. The exclusion of capital gains from taxation creates an economic distortion, in that it mainly raises the rate of return to corporate income and thus may change the institutional structure of an economy. Besides, the administrative costs involved in a highly differentiated tax structure are much higher than when the tax rate is uniform and exemptions minimal.

A further distortionary effect is present even in broad-based incentives to save: incentives lead to a less progressive tax structure if the rich have a greater propensity to save than the poor.

One argument often made in favor of tax incentives is that they alleviate some of the distortionary effects of inflation on the tax system. Becker and Fullerton (1980) show that, in inflationary circumstances, effective tax rates will be much higher than statutory rates; and, the lower the tax rate, the smaller the differential between the two rates. Thus, McLure (1980) is able to demonstrate that the greatest welfare gains from saving incentives come through offsetting the inflation-induced boost in effective tax rates. However, this argument ignores two common characteristics of developing countries which would vitiate the effects of saving incentives. First, inflation, particularly at the high rates often observed in developing countries, is difficult to predict. As Sanchez-Ugarte ((1985), p.16) concludes (in a discussion primarily directed at investment incentives, but with wider applicability), “Tax incentives have to be announced in advance, which gives a large margin for error favoring either investors or the government. Correcting the income tax for inflation, rather than using tax incentives, results in less distortion and is less costly.”

Second, financial repression is very common in developing countries,10 so that their real interest rates may be very low or negative and loanable funds may consequently be in such short supply that investment is constrained by a rationing of finance.11 Because interest rates are not free to rise with inflation, the effective tax rate is not boosted by inflation, as it would be in a free-market economy, and, therefore, there is no increased distortion from the effect of the inflation component of the interest rate that a reduction in tax rates on the nominal interest rate could offset. Thus, the “inflation” argument in favor of tax incentives for saving is not valid in a financially repressed economy. Indeed, Ebrill ((1984), p. 13) makes the point that “the distortion to savings decisions implied by the existence of financial repression may be far larger than that associated with the fact that interest income is subject to income tax.” Given an interest rate ceiling, financial repression increases as inflation increases. Therefore, a more appropriate goal for a government which wishes to increase saving would be to use tax policy to prevent inflation, instead of trying to offset inflation indirectly by using incentives which have the effect of (very imprecisely) indexing some selected returns to inflation.

To summarize, neither the theory of savings behavior nor the empirical evidence on the response of saving to changes in its rate of return supports the use of tax incentives as an instrument to increase financial saving. Besides, tax incentives create economic distortions which may leave the economy worse off than it would be under the distortions of a uniform tax system.

2. Macroeconomic Consequences of Tax Incentives

In order to obtain a comprehensive evaluation of the results of using incentive policies to promote aggregate savings and investment, it is not enough to assess the direct effects the specific tax incentives may have, and the distortions they may impose, on particular sectors of the economy; it is also crucial to take into consideration their macroeconomic consequences. Tax incentives involve, in most cases, the loss of actual and/or potential fiscal revenues and, therefore, they are bound to have direct budgetary implications. Such a loss of revenue will require either an increase in some other source of income or a commensurate cut in government expenditures. Without either action, a budgetary gap, which would require additional financing resources, will remain. In other words, the implementation of a policy based on the granting of tax incentives may result in either the need to increase other taxes and/or reduce government expenditures, or in a budget deficit.

It is thus very difficult, and to some extent misleading, to assess the outcome of tax incentives when they are considered in isolation. It is necessary to have information on the additional measures the government will adopt in order to carry out the incentive policy. If existing taxes are increased or new taxes are imposed, it is necessary to evaluate their impact on the overall economy, and their consequences for savings and investment, before a complete picture of the tax incentives’ effects can be obtained. The same applies to cuts in expenditures. If, on the one hand, in order to enable the government to establish a tax-incentive system, it becomes necessary to cut government expenditures which have been complementary to private sector investment, the overall positive effect of the tax incentives may be extremely small. If, on the other hand, the expenditure cuts are in areas related mainly to government consumption, the effectiveness of tax incentives may actually be enhanced.12

When, however, no measures are actually taken to compensate for the loss of revenue, a program of tax incentives will definitely result in budget imbalances, the effects of which can only be assessed if information about the financing of such imbalances can be obtained. In general, it is possible to say that fiscal deficits tend to put pressure on aggregate demand, which, in turn, may result in higher inflation, balance of payments disequilibrium, and—if deficits are debt financed—in higher real interest rates and the crowding out of the private sector. It is evident that all these results will have detrimental effects on the investment and saving process.

If tax incentives ultimately lead to an acceleration in the inflationary process, this outcome may cancel, to a large extent, the benefits to the private sector which are intended to arise from the tax-incentive program. There are a number of channels through which high and accelerating inflation may discourage saving and investment. In the first place, inflationary situations tend to involve higher perceived risk and uncertainty. Such an environment is generally not conducive to dynamic and sustained investment and saving.

High rates of inflation are often accompanied by falls in economic activity or, at least, in the growth rate, and these tend to reduce the incentive to invest.13 High rates of inflation also tend to be accompanied by a progressive overvaluation of the exchange rate. This eventually leads to a scarcity of foreign exchange which subsequently reduces imports of basic commodities. Imports are initially encouraged by the overvaluation of the exchange rate but are subsequently discouraged by the reduced availability of foreign exchange, which inevitably brings about import controls. Sooner or later, domestic firms find that they cannot function at the previous level because of a lack of basic materials which they used to import. This leads to unemployment, falling profits, and, consequently, lower investment.

In addition, a high rate of inflation is always accompanied by an increase in the variance of relative prices. As the rate of inflation becomes progressively higher, the variance of relative prices increases. This increases uncertainty about sectoral allocation and discourages investment or, at least, leads it into less efficient channels. It is widely held that inflation leads to the predominance of financial decisions over technological or cost-reducing ones, which results in a reduction in savings available for productive capital formation.

Another effect of inflation is the attempt by the government to suppress it by controlling the growth of some prices, particularly exchange and interest rates. Intentional overvaluation of the exchange rate tends to depress the availability of savings for domestic uses, since individuals come to expect large devaluations at some future dates and will thus be attracted by the alternative of holding foreign securities and foreign exchange. This is so-called currency substitution, or the substitution, in people’s portfolios, of foreign-issued liabilities for domestically denominated ones, thus diverting domestic resources from financing domestic investments. With respect to interest rate controls, in many cases they lead to the emergence of negative real rates, depressing savings and worsening the currency-substitution problem.

IV. An Alternative Strategy: The Government as Saver

1. Generation of a Surplus

In view of the microeconomic and macroeconomic costs of tax incentives, it seems reasonable to search instead for savings-mobilization policies which do not lead to fiscal imbalances. As has been discussed in the first section, the most serious barrier to the attainment of an appropriate savings level in developing countries seems to be rooted in the lack of a properly working capital market including well-developed financial intermediaries. As an alternative to specific tax incentives, the government can alleviate some of the shortcomings associated with developing country capital markets by constructing some sort of proxy for the financial market. This would provide an argument for a planned budgetary surplus. In other words, the government should aim to set its total tax revenues and its total expenditures (both current and capital) at levels that would yield an overall surplus, which could then be made available, on a competitive and nonconcessionary basis, to the private sector as well as to public enterprises. This would provide the government with a powerful and flexible tool that would facilitate, to a considerable extent, the efficient allocation of private investment. That such an approach may succeed is shown by the Japanese experiences in the decade after the Meiji Restoration (1868) and in the period following World War II. Under such an approach, the availability of credit financed by surpluses in the government budget is seen as an integral part of fiscal policy. But, unlike incentives and controls, it puts in the hands of the government an instrument that is exceedingly flexible, through which resources can be directed toward productive investments carried out by foreign and domestic private entrepreneurs and toward investments with high social significance. In this way, the government, in addition to increasing savings, would stimulate entrepreneurship, attract foreign resources, and thus affect not just the quantity but also the quality of investment. The generation of a government surplus, because of the level of governmental economic commitment and discipline it demonstrates, tends to enhance the confidence of the private sector. This confidence, in turn, stabilizes or boosts aggregate demand and thus maintains the size of the tax base and legitimizes the tax collection that caused the surplus in the first place. A tax system which is uniform and predictable, and which is associated with prudent macroeconomic management, may make higher rates more acceptable than they would be in a tax system with many exemptions that is associated with a fiscal position perceived to be unsustainable in the longer run.

Several surveys carried out in many developing countries have shown beyond any doubt that lack of foreign investment and the inability of many enterprises and individuals to obtain credit are serious obstacles to development. It is therefore probable that credit incentives would be extremely useful, especially for the creation of new enterprises. Although tax incentives may also play a role, they are likely to be less effective in the overall economy. On the one hand, tax incentives, by increasing the liquidity of established enterprises, may facilitate their development, but they normally do not provide very much help to start-up enterprises which do not have access to funds needed for an initial fixed investment. Credit incentives, on the other hand, have an impact in both areas.

As stressed by Tanzi (1976), credit incentives financed by surpluses in the government budget should be a crucial component of fiscal policy in developing economies—a component which has not received the attention it deserves. Moreover, such a course of action puts in the hands of the government an instrument to promote savings and investments which does not suffer from the main shortcomings of tax incentives.

2. Role of Government Investment

The recommendation that the government run an overall surplus and place the funds generated by this surplus in the private capital market does not mean that the government itself should cease to act as an investor. The level of aggregate savings and their channeling into productive uses will tend to be positively related to the expected rate of return on the available investments. Given this, it is possible to postulate that the government could increase domestic savings by undertaking actions which increase the perceived rate of return on private sector investments. One way of doing this would be to invest directly in projects which would result in positive externalities to the private sector.

It is a well-accepted proposition that in developing countries private and public investment are related, although there is considerable uncertainty about whether, on balance, public sector investment raises or lowers private investment. In broad terms, public sector investment can cause crowding out if it utilizes scarce physical and financial resources that would otherwise be available to the private sector, or if it produces marketable output that competes with private output. Yet public investment that is related to infrastructure and the provision of public goods can also clearly be complementary to private investment. Public investment of this type can enhance the possibilities for increasing the expected rate of return on private investment by raising the productivity of capital, increasing the demand for private output by increasing demand for inputs and ancillary services, and augmenting overall resource availability by expanding aggregate output and savings.

The overall effect of public investment on private investment will, therefore, depend on the relative strength of these various effects, and there is no a priori reason to believe that they are necessarily substitutes or complements. In a recent study,14 a private investment function was derived which took into account the effects of government policies, particularly on financial credit and government capital formation. The study attempted to make an empirical distinction between public investment related to the development of infrastructure and other types of public investment which may, in fact, substitute for private capital formation. Empirical results indicate that (i) if the flow of domestic credit available to the private sector were reduced for whatever reason, including greater absorption of credit by the public sector to finance budget deficits, then private investment, and consequently economic growth prospects, would tend to decline; and (ii) an increase in the infrastructural component of government direct investment would raise private investment (probably by increasing its potential profitability), but similar increases in other types of public investment (e.g., in sectors producing marketable output) would appear to crowd out, and therefore reduce, private sector investment.

Therefore, saying that the government should run an overall surplus is not equivalent to precluding a role for the government as investor. However, its investment should be undertaken primarily to mobilize further domestic resources through the more effective use of existing resources.

V. Conclusion

The limitations which characterize developing country capital markets make traditional government policies, and tax incentives in particular, unsuitable for mobilizing domestic savings. Not only does their efficacy remain unproven—at the theoretical, as well as the empirical, level—but they also create distortions and budgetary and other macroeconomic problems which may leave the economy worse off than it would be under a uniform and stable tax system with higher rates. Prudent macroeconomic management and the maintenance of a public sector surplus will not only avoid these distortions but also provide a pool of loanable-funds savings to private sector investors; this will compensate for limitations in the capital market which the private sector could not overcome on its own. Thus, the generation of public sector savings is a fitting role for government in a developing country.


    Becker, C., and D.Fullerton, “Income Tax Incentives to Promote Saving,”National Tax Journal (Columbus, Ohio), Vol. 33 (September1980), pp. 331-51.

    • Search Google Scholar
    • Export Citation

    Blejer, MarioI., and MohsinS. Khan, “Government Policy and Private Investment in Developing Countries,”Staff Papers,International Monetary Fund (Washington), Vol. 31 (June1984), pp. 379-403.

    • Search Google Scholar
    • Export Citation

    Boadway, R.W., and D.E.Wildasin,Public Sector Economics, 2nd ed. (Boston: Little, Brown, 1984).

    Boskin, M.J., “Taxation, Saving, and the Rate of Interest,”Journal of Political Economy (Chicago), Vol. 86 (April1978), pp. S3-S27.

    Coase, R., “The Problem of Social Cost,”Journal of Law and Economics (Chicago), Vol. 3 (October1960), pp. 1-44.

    Coopers and Lybrand,International Tax Summaries, 1983 (New York: John Wiley and Sons, 1983).

    Ebrill, LiamP., “The Effects of Taxation on Labor Supply, Savings, and Investment in Developing Countries: A Survey of the Empirical Literature” (unpublished, International Monetary Fund, April3, 1984).

    • Search Google Scholar
    • Export Citation

    Hills, J., Savings and Fiscal Privilege (London: Institute for Fiscal Studies, 1984).

    Howrey, E.P., and S.H.Hymans,“The Measurement and Determination of Loanable Funds Saving,”Brookings Papers on Economic Activity: 3 (1978), The Brookings Institution (Washington), pp. 655-705.

    • Search Google Scholar
    • Export Citation

    International Bank for Reconstruction and Development, World Development Report, 1984 (New York: Oxford University Press, 1984).

    International Monetary Fund, Interest Rate Policies in Developing Countries: A Study by the Research Department of the International Monetary Fund, IMF Occasional Papers, No. 22 (Washington, 1983).

    • Search Google Scholar
    • Export Citation

    Mackenzie, GeorgeA., “The Macroeconomic Supply Effects of Tax Rate Changes” (unpublished, International Monetary Fund, December30, 1982).

    • Search Google Scholar
    • Export Citation

    McLure, CharlesE., Jr., “Taxes, Saving and Welfare: Theory and Evidence,”National Tax Journal (Columbus, Ohio), Vol. 33 (September1980), pp. 311-20.

    • Search Google Scholar
    • Export Citation

    Molho, LazarosE., “Interest Rates, Saving and Investment in Developing Countries: A Re-examination of the McKinnon-Shaw Hypotheses” (unpublished, International Monetary Fund, October7, 1985).

    • Search Google Scholar
    • Export Citation

    Sanchez-Ugarte, Fernando, “A Supply-Side Look at Tax Incentives: Definition, Design, and Selection Criteria of Efficient Tax Incentives” (unpublished, International Monetary Fund, June5, 1985).

    • Search Google Scholar
    • Export Citation

    Saunders, P., and F.Klau, “The Role of the Public Sector: Causes and Consequences of the Growth of Government,”OECD Economic Studies, No. 4 (Spring1985).

    • Search Google Scholar
    • Export Citation

    Summers, L.H., “Capital Taxation and Accumulation in a Life Cycle Growth Model,”American Economic Review (Nashville, Tennessee), Vol. 71 (September1981), pp. 533-44.

    • Search Google Scholar
    • Export Citation

    Tanzi, Vito, “Fiscal Policy, Keynesian Economics and the Mobilization of Savings in Developing Countries,”World Development (Oxford, England), Vol. 4 (October-November1976), pp. 907-17.

    • Search Google Scholar
    • Export Citation

    Tanzi, Vito, and MarioI. Blejer,“Fiscal Policy in an Economy with High Inflation and High Debt,”Anales de la Asociación Argentina de Economía Política: XX Reunión Anual (Buenos Aires), Vol. 4 (1985), pp. 1423-51.

    • Search Google Scholar
    • Export Citation

This paper was first published in The Role of the Public Sector in the Mobilization of Domestic Financial Resources in Developing Countries (New York: United Nations, 1986).


This phrase has been borrowed from Howrey and Hymans (1978).


The relevant exchange rate here may be the relative price of a particular commodity, rather than the value of the currency, in an economy where asset markets are fragmented because, in such a case, for example, the only funds available to a potential importer may come from his own previous export trade.


For example, as discussed in McLure (1980).


Since it is possible to subsidize investment through making the cost of capital negative, it might also be possible for the government to subsidize saving through a direct subsidy on interest income. However, discussion of the use of tax policy to encourage saving usually envisages a movement to a pure consumption tax but not beyond it.


The effect of this policy has been simulated by Becker and Fullerton (1980).


This argument has abstracted from business saving, which is, of course, a large part of total saving. However, life-cycle utility maximizing individuals will buy and sell equity in firms in order to maintain their desired consumption patterns. Thus, the ability and desire of the firm to save will ultimately be determined by the same income and substitution effects which dictate individual saving responses.


See, for example, Hills (1984) and Saunders and Klau (1985), pp. 177ff.


This is the McKinnon-Shaw hypothesis, as described by Ebrill ((1984), p. 10), and tested by Molho ((1985), especially p. 22).


See Section IV on this issue.


For a discussion, see Tanzi and Blejer (1985).


See Blejer and Khan (1984).

    Other Resources Citing This Publication