Journal Issue

Saving Through Taxation—Reality or Mirage?

International Monetary Fund. External Relations Dept.
Published Date:
March 1967
  • ShareShare
Show Summary Details

Stanley Please

IF THE WORLD were morally, politically, and economically more enlightened, the flow of external assistance to the developing countries would be much larger. Developing countries must accept this unenlightened environment as one of the obstacles to their development. In today’s world it is mainly through their own efforts to raise their rates of saving that developing countries can add to their stock of capital. These domestic efforts are not only extremely important in themselves; by them the world judges the strength of a developing country’s claim to financial aid.

Lending agencies, including the World Bank, look closely at a country’s own efforts to overcome the internal obstacles to its development before they do business with it. In part this stems from the moral judgment that only those who help themselves deserve to be helped. It has, however, also come to be recognized that a country’s ability to service external debt depends to a large extent upon two things: first, its ability to invest in such a way as to obtain a sustained increase in its real national product and, second, its ability to save a growing proportion of this increase. If a country can do these two things, then the growth of its external debt gives less cause for alarm than is sometimes supposed.

For a variety of reasons, most economists have argued that domestic savings for national development can be increased only by governments compulsorily reducing (by taxes but also in other ways) the personal consumption of the citizens. These economists can point to the difficulties of raising the level of voluntary savings—the low incomes of a large proportion of the population, the attempt to attain the levels of consumption of people in the highly developed countries, the fear of price increases undermining the real value of savings, political misgivings, and so on. Budget surpluses played a crucial role in the early stages of Japan’s economic development, and have provided the favorite example to be paraded before other developing countries wishing to emulate its success.

This article is based on a paper prepared for the 1965 Cambridge University Overseas Studies Committee Conference on “Overcoming Obstacles to Development.”

Even if voluntary savings alone could do the job they are not favored by a number of economists, who argue that while the investment of tax revenue in development creates no claims on future increases in output, private savers will hold such claims in the form of bonds and other savings media. The servicing of this internal debt, it is argued, will cramp the government’s revenue and expenditure policy in the future.

Public Revenue Rises…

Most economists, then, have advocated a development strategy aimed at raising the level of domestic savings through budgetary policy—i.e., reducing personal consumption by increasing taxation while ensuring that this increased revenue is not used to increase government consumption or defense, civil services, etc. And their advocacy has in one respect been very successful. Over the past decade and a half there has been an immense national and international effort to raise the levels of taxation in the developing countries. Major internal inquiries such as that by the Indian Taxation Enquiry Commission have been undertaken, independent reports by tax consultants have been solicited, and, in appraising the economic performance of their member countries, the International Monetary Fund and the World Bank attach great weight to this aspect of a country’s economic policy. So significant is it in the eyes of many people and institutions interested in economic development that the ratio of tax revenue to gross national product (GNP) is treated as one of the most important criteria for measuring and judging a country’s economic performance.

Nor has this great taxation effort lacked results. In many countries, governments have shown a high degree of political self-discipline in introducing new tax measures and in increasing revenue yields. In India, for instance, Union Government tax revenue has increased at a rate of 10 per cent per annum since the beginning of the country’s program of planned development in 1951 and by 20 per cent per annum during the Third Plan period which began in 1961. As a result, tax revenue (Union and States combined) now represents about 12 per cent of GNP, compared with about 7 per cent in 1951. Likewise, in Ceylon, tax revenue now represents 23 per cent of GNP, compared with 20 per cent only five years ago. Many Latin American countries, too, have raised their tax revenue. In Peru, for instance, current public revenues now account for 20 per cent of GNP, compared with 15.5 per cent in 1960, and in Brazil the percentage has risen from 20 per cent in 1956 to 27.5 per cent in 1964. There are, of course, exceptions, but the general impression is that governments have rarely lacked willingness or ability to control an increasing proportion of their national economic resources.

… But Does Not Flow Into Development

To the extent, however, that the motive of these countries was to channel more of their domestic resources into economic development, the record is depressing. In India, the Union Government has increased its income by Rs 9,000 million since 1960-61, yet, out of this increase, less than 5 per cent (Rs 440 million) has taken the form of investible resources. While the Rs 9,000 million represents about 5 per cent of GNP, the increase in the flow of savings has been a mere 0.6 per cent of GNP. In Ceylon, developments have been even more depressing, the increase in the tax effort since 1950 having been matched by an almost persistent decline in government savings. In 1963-64 these were insufficient to cover the losses of government enterprises, thus resulting in net dissaving by the government sector as a whole. In Peru, the increased tax effort has been matched by an 18 per cent decline in public saving—during a period in which GNP went up by 60 per cent. Thus, public saving fell from 4 per cent to 2 per cent of GNP.

There is nothing novel in this pattern. The disappointing behavior of public savings in the less developed countries in spite of impressive records of tax performance is due to the growth of government current expenditure, and this is a phenomenon which is widely recognized. In 1960, for instance, an unpublished study by the World Bank of the financing of public investment in 19 less developed countries warned that “the proportion of public investment financed by public savings tended to diminish during the 1950’s”; the main reason was a decline in the rate of public savings.

Still More Taxation?

Yet most economists argue for still more taxes. It is still possible to read an article by the best known of all consultant economists in the tax field—Nicholas Kaldor—which is inappropriately entitled “Will Underdeveloped Countries Learn to Tax?” Richard Goode of the IMF in his recent paper prepared for the Rehovoth Conference on Fiscal and Monetary Problems in Developing States argues that “An underdeveloped country that is determined to avoid both stagnation and inflation will have to find ways of raising large and growing amounts of tax revenue.” Likewise, at the most general level, both Alan R. Prest and R. N. Tripathy, in their identically titled books Public Finance and Underdeveloped Countries, give overwhelming emphasis to the need for an increasing tax effort in the developing countries. The World Bank itself, in appraising the development plans of its member countries, attaches immense importance to the increases in tax revenue which can be expected from them.

This attitude and advice are relevant to the development problem, provided that the government can prevent its expenditure from rising as fast as its income. Then, of course, any increase in revenue reduces public dissaving or increases public saving. Because this assumption is so crucial to the judgment that must be made upon the whole strategy of development, which relies heavily on compulsory saving through taxation, its realism needs to be examined.

Judging What a Government Can Afford

Public expenditure is governed by politics, and its general size and shape is not the business of economists—that is the conventional view accepted by most economists. To the extent that it implies that public expenditure is determined aside from any economic considerations, it is, of course, patently not true. Political decisions must, however vaguely, take into account the ability of the country to spend more on defense, education, and so on. But there are two different ways of judging this ability. It can be judged either in relation to the total resources of the country and to the flow of income from these resources, or, alternatively, in relation to a narrower concept of that proportion of these resources over which the government is able and willing to gain control. Governments may say: “Given a national production of so much, what can we do?” Or they can say “Given that we get X per cent of national production, what can we do with our X per cent?” It is the first of these two criteria of ability which economists apparently have in mind when they talk about government expenditure being “politically determined.”

While it may be desirable on occasion to think in terms of the whole economy, the narrower concept is probably more influential in many instances. However complicated budget accounts may be, they are easier for the non-economist policymaker than even the simplest form of national accounts, the understanding of which calls for an understanding of the basic nature of income flows. Money flowing into the exchequer has a reality and significance to the noneconomist which is unlikely to be matched by the flows of money against goods and services in the economy as a whole. However, if this is true, then the independence of government expenditure and government revenue will be undermined; with the danger which this implies to a development strategy that aims to raise the total flow of domestic savings through increased taxation and other revenue-producing policies.

However, in addition to these considerations, all the writers referred to above, and many others, seem to accept the much more dangerous argument that the rise in public expenditure must be almost unconditionally accepted and the whole burden of increasing public savings thrown onto taxation, despite the evidence that this is, in many instances, abortive. It is one of the tragedies of the emotionally packed environment in which economists discuss their problems that one has misgivings at suggesting or even hinting that this might not be altogether true—that possibly some part, and in some countries some large part, of public expenditure and the annual increment thereto might be of lower priority than other calls on a country’s resources.

Of course, a great deal of public expenditure must be accorded high priority, and some built-in rate of growth of public expenditure must be accepted as part of the development process. One obvious example of this is the increased expenditure on teachers’ salaries resulting from a growing number of teaching institutions, but the same factor is at work in the field of health, roads, and other largely nonrevenue producing sectors. However, while it is possible to accept the operation of these built-in factors as “inevitable,” there is a dangerous tendency to extrapolate these arguments into an explanation of and justification for all increases in government current expenditure.

In fact, such “inevitable” factors are less important in relation to our problem than other increases in expenditure which are altogether unrelated to development but are the outcome of autonomous policy decisions by national governments. Expenditure on defense is the most obvious, but many others are familiar—even notorious. They include expenditure on consumer subsidies, the deficits of publicly operated industries, unduly high salaries of public officials, as well as the more obvious forms of wasteful prestige expenditure.

The more fundamental question, however, is whether increased public revenue intended for development may not in fact weaken the administrative and, above all, the political self-discipline in restricting government current expenditure. Is it possible to resist political pressure or popular clamor for increased defense expenditure, salary increases for public officials, and so on, when it is known that the money is in the kitty? The ability to understand even simple economic analysis is not so widespread that the justification for increased tax revenue to finance a budget surplus for developmental purposes can be readily understood, let alone accepted. It is one thing for the World Bank and other institutions and people interested in development to argue for an increased tax effort for this specific purpose; it is quite another matter for governments to be able to act in such a single-minded manner. The gloomy record of inadequate budget surpluses, despite increased tax performance over the years, suggests that those who argued for a development strategy based on increased compulsory savings underestimated and, more frequently, ignored the effect that the increase in taxation might have on public consumption. In these circumstances there is a danger that those who recommend increased taxation in the interests of economic growth may be looking at a mirage.

Policy Implications

What, in this context, does governmental discipline of expenditure actually mean? Which expenditures can be controlled and to what extent can an outside lender involve himself in such questions? It is obviously a most delicate matter to suggest to a government that it should cut defense expenditure. Even comment on such highly sensitive issues can be dangerous. Yet external lenders cannot avoid concerning themselves with this issue; the funds that they lend for a particular project might enable a country to divert into defense the funds it had itself earmarked for that project. External lenders must, therefore, if they are seeking to promote economic development rather than to underwrite a defense program, attempt to obtain an answer to such hypothetical questions as: “How much defense expenditure would a given borrower undertake and how will that expenditure grow if we make a loan for the new railway that they must have anyway?”

As regards expenditure on education, health, housing, and so on, there is, first, the notorious statistical problem that, although most (though not all) expenditure under these heads is classified as current expenditure in government budgets, some should rank as capital expenditure. This would embrace those funds that in fact increase the ability of the economy to produce more goods and services in the future. The financing of this expenditure, therefore, should rank as a contribution to the aggregate flow of domestic savings and, to this extent, public savings are usually understated, as they are, for instance, in the examples from different countries given above. While this is correct, there are two comments which need to be made. First, the acceptance of the principle must not give carte blanche to the acceptance of all social service expenditure as investment expenditure. Secondly, even if the investment element as an item of social welfare expenditure is accepted, its priority in terms of the economic return to the expenditure must still be judged in relation to the return on alternative investments.

Linked to these two points, and pervading the whole field of discussion of social expenditure, is the almost universal acceptance of the fact that this expenditure, judged in vacuo, is “good.” Everyone believes in the desirability of an educated rather than an ignorant population, healthy rather than sick people, and so on. It becomes particularly difficult, therefore, to argue that expenditure on such desirable objectives is itself too high or rising too fast. Yet in the light of the aims of the development decade, this may well be the case.

Can We Change the Situation?

Ultimately the only certain answer to this problem is that governments should be both more rational and more self-disciplined in determining public expenditure policy. Greater rationality requires the development of budgetary processes which clearly set forth the choices of ends open to the government and the implications of each choice in terms of alternatives forgone. This in its turn requires adequate accounting practices, expenditure controls including forecasting analysis, financial reporting techniques, etc. The importance attaching to these developments cannot be exaggerated and is reflected in the establishment in 1964 of a technical advisory service in the IMF, the organization and responsibilities of which were described by Jakob Saper and Timothy Sweeney in another article in Finance and Development (Vol. II, No. 4, December 1965).

Rationality of choice can, however, only be judged in relation to the pattern of ends which a government wishes to satisfy, and rationality does not imply wisdom in selecting this pattern. This pattern, as already noted, comprises many elements, of which a higher rate of economic growth is but one. Other elements are often more appealing, particularly if they offer benefits which are more immediately apparent than those from policies designed to achieve longer-run economic growth. This is especially true where the government is dependent on popular support for its policies at periodic elections and where, as a consequence, considerations of short-run political expediency are likely to be at a premium compared with the known long-run interests of the country. In such circumstances the development of procedures by which budgetary policy can be determined more rationally is likely to be a necessary but inadequate response to the problem raised by this article. In addition, it is desirable that governments should, if possible, be made subject to institutional pressures which would result in more emphasis being given to the longer-term interests of the society. Economic development represents the longer-term interest which the Bank was established to encourage, and it is also one of the primary objectives of the Fund, and thus any institutional device which helps governments to achieve this end must a priori be regarded with favor.

One device which appears to meet this need is the assigning of public funds for the financing of development expenditure in general, or the earmarking of particular revenues for particular developmental uses. Economists have widely argued that the earmarking of public revenues is an undesirable procedure; governments should be free to determine the use to which all revenues are put. Earmarking prevents this by relating given categories or items of expenditure to given categories or items of revenue. It requires an immense effort of will for any theoretical economist to convince himself that it might be desirable in certain circumstances that obstacles should be established to the selection of the most appropriate allocation of a given supply of resources of any kind. Nevertheless, if the amount of non-priority expenditure is a function of the availability of funds, as suggested above, the prevention of the use of significant and significantly growing sources of revenue for such purposes must impose greater restraint upon a government.

At the most general level, such earmarking can take the form of assigning certain revenues to a development fund—as happens already, for instance, with a proportion of oil revenues in some of the oil producing countries. Alternatively, the earmarking can be more specific. For instance, the surpluses generated by public enterprises or by particular public enterprises can be authorized to be retained by the enterprises and used to increase their productive capacity, rather than be siphoned off into general budgetary revenue. Likewise the revenue from user charges, levied for example on road users, can be assigned specifically to further development in the relevant sector. Somewhat different from the earmarking of public enterprise surpluses and user charges is the assignment of revenue from a particular tax to an unrelated sector of development expenditure. For instance, the revenue from liquor duties might be assigned to the financing of technical education.

The danger of a misallocation of financial resources arising out of earmarking which takes any of these forms must be readily acknowledged. Some services might have a surfeit of finance and others a dearth. This danger can be reduced if in the first place earmarking is undertaken in a generalized form, such as the assigning of revenues to a development fund or to wide sectors such as communications and power. It also can be reduced to the extent that anticipated increases in expenditure required under given heads can be matched against the expected growth of revenue from particular taxes or other sources. Certainly nothing which is said here is intended to countenance the proliferation of detailed earmarking which has taken place in some countries and which is unrelated to the development needs of the country. At the same time the danger of a less than optimum allocation of public development funds resulting from earmarking must be set against the fact that there would be the likelihood of an increased flow of such funds. Providing the distortions are not too great, therefore, a net gain can be expected.

Apart from administrative devices of this nature, which are designed to make it more likely or less unlikely that any increased revenues will be made available for meeting developmental objectives, the time has, perhaps, arrived to question for certain countries the strategy of domestic financial mobilization which is based almost exclusively on increased taxes. This emphasis on public saving has stemmed from the strength attributed to factors which undermine the willingness of the private sector in less developed countries to make an ever-increasing contribution to the flow of domestically available savings—the urge to increase consumption (stemming in particular from the desire to attain the standard of living found in developed countries), political uncertainty, fear of inflation, etc. There is undoubted validity in this attitude though, for certain countries, the readiness to expand public expenditure appears to be such that total savings would probably have been increased if taxation had been reduced and private disposable income thereby increased. However, even if the strength of these factors which weigh against private saving is fully admitted, it does not follow that the only solution is to rely upon publicly generated saving. A more effective solution, or at least a solution which could be given far more emphasis in developmental strategy than it is generally accorded, is to provide the necessary encouragements to savers. These encouragements would take the form of an appropriate pattern of terms offered to savers in securities, insurance policies, savings deposits, and other financial assets, aimed at offsetting the factors which weigh against savings and in favor of consumption. There is, however, no guarantee, except by coincidence, that this pattern of terms will correspond to the pattern which can be offered by private borrowers—private firms, etc.—seeking to finance their activities. In these circumstances, it would be possible for the terms of the securities which constitute the national debt to be adapted in such a manner that the supply of private saving was stimulated. In effect, the government would act as the bridge between the terms under which the ultimate users of borrowed funds need or wish to borrow and the terms under which lenders are willing to lend. Particularly in less developed countries, where capital markets are relatively underdeveloped and this bridge is not always supplied by specialist financial institutions, the need for government action in this respect is vital.

The simplest and most obvious change in the terms savers receive would be a straightforward increase in the structure of interest rates. This, however, is an unselective measure. If it is apparent to the government that specific factors are operating to reduce private saving or domestically available private saving (i.e., allowing for capital exports), it is likely that specific terms in a financial asset to take care of the adverse effect on savings of these factors would be preferable. In particular, for instance, it is possible to give savers guarantees against rising prices undermining the real value of their savings, by the issue of bonds or other savings media, the value of which and/or the return on which is linked to some measure of the change in the internal value of money.

This paper is not the place in which an extended discussion of such devices can be undertaken. However, experience with the use of the value-linking device by governments is already available—notably in Israel, Finland, and France. These experiences are instructive, both regarding the positive advantages of value-linking in encouraging savings and regarding the problems to which such linking gives rise.

It might be argued that a change in the orientation of policy in this way, so that more emphasis came to be given to the encouragement of private saving through enlarging the role of the national debt, and less emphasis to the generation of compulsory saving through budgetary surpluses, would still channel funds into the hands of the government which it could use to finance an increasing volume of current expenditure. This, of course, is true; but it merely reflects the fact that if a government intends, come what may, to undertake a given level of expenditure, then no administrative or institutional device will prevent it from doing so, even though it unduly augments the money supply and forces up prices. Except in emergencies, however, governments tend to follow a set pattern of budgetary policies, reacting conventionally to prospective surpluses and deficits. This being so, a change in the balance between revenue receipts and capital receipts could significantly affect governmental attitudes toward financing increased current expenditure. By increasing the government’s deficit “on current account”—or (in terms of the less strict British conventions) the deficit “above the line”—such a change could generate more resistance administratively and politically to increases in public current expenditure. This resistance could be particularly significant in those countries which attempt to follow the conventions of sound budgetary policy, even if this policy is not developmentally orientated.


The nineteenth century saw the dominance in policymaking of the idea of economic man as the decision taker—economic man, forever visualized as weighing gains against costs in all his decisions. We now seem to be dominated in our thinking by the idea of the economic government, and this domination is by no means the monopoly of people and institutions which explicitly accept a socialist ideology. The acceptance of the idea of the all-rational government was, in part, based on the notion that the government is a timeless institution, and can thus bring to bear a telescopic faculty on the decisions of society, deferring consumption today in order to increase consumption in the future. The government would be more farsighted than the individual, with his natural preference for present consumption. The evidence is that this counterforce, even to the extent that it has operated at all, has been weak; that the pressures on governments to spend have been irresistible, even if often quite understandable. In particular, the pressures to spend are made less resistible when it is apparent to everyone that the funds are immediately available to the government in the exchequer. If this thesis is accepted, I have tried to indicate some devices which might possibly alleviate the problem. These cannot be foolproof for the very obvious reason that governments, in the end, are sovereign within their borders and subject only to the ultimate constraint of the basic and universal economic laws which stem from the scarcity of resources. However, even sovereign states can be assisted in keeping to the path of virtue if temptations to take easier and politically more popular decisions are removed.

Other Resources Citing This Publication