Chapter

1 Interest Rates and Tax Treatment of Interest Income and Expense: Fiscal Affairs Department

Editor(s):
Vito Tanzi
Published Date:
June 1984
Share
  • ShareShare
Show Summary Details

In recent years the Executive Directors of the International Monetary Fund have expressed concern over the effects on the level and variability of interest rates arising from differential tax treatment of interest income and expense across countries. They have also expressed concern with respect to the effects of these and other related tax provisions on the effectiveness of policies of monetary restraint and on international capital movements. The Executive Directors therefore asked the Fund staff to assess the role of tax factors in the high levels of interest rates prevailing in 1981–82 in the United States and other major industrial countries.1

The situation has changed somewhat in the past year. Nominal interest rates have fallen considerably in the United States and elsewhere, but real interest rates have remained high. Some industrial countries have started taking a closer look at tax provisions relating to savings and capital formation. Attention has focused on the influence of taxation on the after-tax costs of investments and returns on savings, especially in an inflationary situation. Furthermore, the tax deductibility of interest payments unrelated to income-earning activities (viewed by some observers as a “tax expenditure”) has come under close scrutiny in the United States and other industrial countries.

Nevertheless, tax reforms have been slow. Different tax policies, frequently dictated by tradition and by the national goals of individual countries, continue to be the rule. Therefore, insofar as tax factors do affect significantly the level and volatility of interest rates in a closed (domestic) economy, and differential tax regimes affect exchange rates and international capital movements in an open (international) economy, the concern expressed by the Executive Directors remains valid and significant.

I. BACKGROUND

As tax factors have generally been given little importance in the literature on monetary theory and international finance, the professional body of knowledge in this area is limited. Theories of interest rate determination and of demand for money are often discussed in terms of pretax variables, and they omit reference to taxation of interest income or to deductibility of interest payments from taxable income. Similarly, the literature on theorems of interest rate parity and purchasing power parity has largely ignored the effects of tax factors. On the other hand, public finance literature has focused on the microeconomic (allocative) and equity (redistributive) effects of taxes and has ignored their macroeconomic effects.2

Limitations of the paper

As is often the case, the research effort involved in preparing the present study, while answering many questions, has raised many more and has made the Fund staff more aware of the many problems in this area.

  • (1) There is little unanimity among economists as to the determinants of interest rates and international capital movements. As a matter of fact, no single theory of interest rate determination that has been advanced is readily accepted by the majority of economists or can explain the unusually high levels of real interest rates in recent years.

  • (2) The interrelationship between nominal or market interest rates and their determinants, particularly expected inflation, has been unstable over time, so that no specific conclusions can be reached about the precise quantitative effect of inflation on interest rates.

  • (3) The available empirical evidence on the interrelationships between taxation, inflation, and interest rates is sketchy. For the United States and one or two other industrial countries, the evidence that does exist is often conflicting or ambiguous.

  • (4) The tax provisions bearing on interest rates and international capital movements of individual countries tend to be highly complicated and thus subject to a variety of qualifications and interpretations. (Some of these problems are discussed in Appendix I.) Furthermore, recent ad hoc legislative efforts aimed at adjusting the tax systems for inflation have created uncertainties of their own.

  • (5) Finally, little statistical information is available on the taxes actually paid by groups of savers and investors in individual countries. In addition, the existence of many possibilities for tax avoidance (e.g., tax allowances for savers and tax deferrals for investors) and tax evasion (resulting from shortcomings in national tax administration and the existence of tax havens abroad) makes the available legal information of doubtful usefulness (see Appendix I).

The research on the subject, carried out in the Fund’s Fiscal Affairs Department and elsewhere, can therefore claim at best, to have reached, tentative conclusions subject to the above-mentioned limitations.

Outline of the paper

The paper focuses on the following questions:

  • (1) Does the tax treatment of interest income and payments differ markedly among major industrial countries? Do other tax provisions affecting interest rates and international capital movements also differ markedly?

  • (2) How does the tax treatment of interest income and payments influence the level and volatility of interest rates, especially in an inflationary environment? If there is an influence, what is the direction of the effects of these and other relevant tax factors?

  • (3) How do potential international movements of capital and related tax provisions alter the results of changes in tax treatment regarding interest income and expense initiated by a single country?

  • (4) What implications does the tax treatment of interest income and expense and related tax policies have for the impact of changes in the degree of monetary restraint?

  • (5) What are the consequences of tax policy for the volatility of interest rates and what are the procedures for changing tax policy?

Sections II, III, and IV attempt to answer these questions and to present the available empirical evidence. Section V brings together the major conclusions of the paper and states their implications for Fund activities.

Summary of major conclusions

Section II, supported by three tables in Appendix III, shows that, even though the taxation of interest income is nearly universal and that such income is generally taxed at ordinary income tax rates, the tax regimes of major industrial countries differ in major respects. The marginal rates of the income tax differ among industrial countries, and the tax allowances offered to savers by individual governments also differ markedly. Furthermore, the changes that have been made in income tax structures in response to inflation in most industrial countries have been ad hoc ones and have been incomplete. As a result, the effective tax rates for real interest income differ among industrial countries, although no estimates of these rates are available.

The differences in the tax treatment of interest payments are even more marked. Of the 16 industrial countries surveyed,3 7 offer deductibility from taxable income of all interest payments, including those for income-earning activities, mortgage payments, and consumer loans; 8 allow tax deductibility of interest payments for incomeearning activities and for mortgages but none for consumer loans; and only one country does not allow any deduction for either mortgages or consumer loans.

The tax treatment of foreign exchange gains and losses also differs among industrial countries, although the legal provisions are more complex. In some countries, gains are treated as capital receipts and are thus subjected to capital gains taxes, which are generally levied at lower rates; in other countries, they are treated as current receipts and are subjected to the regular income tax.

Section III suggests that, even in the absence of income taxes, the one-to-one relationship between expected inflation and nominal interest rates (the basic Fisherian hypothesis) may not be valid for a variety of reasons. The existence of expected inflation may, in fact, lower the rate of return on real investment and, consequently, the real rate of interest, thereby limiting the increase in nominal interest rates.

The taxation of nominal interest income and the tax deductibility of nominal interest payments, which typically exist in many industrial countries, can theoretically cause a more-than-proportional increase in nominal interest rates as a result of a given expected rate of inflation. But this increase is also limited by the existence of many other tax factors, including the tax treatment of other capital income and capital gains that limit the investor’s capacity to avoid taxes, the tax depreciation allowances and inventory procedures that increase taxation on companies and thereby reduce their ability to pay higher interest rates, the existence of tax-exempt lenders and borrowers, tax evasion, tax havens, etc. Because of these many tax factors, the positive impact of the tax treatment of interest income and interest payments is likely to be somewhat reduced. The few empirical investigations covering this area tentatively suggest that, when all factors are taken into account, the nominal interest rate, on average, does not increase more than proportionately with expected inflation (e.g., see Section III, Table 1). However, this empirical result does not rule out the possible existence of tax effects.

Table 1.Eight Industrial Countries: Impact of Taxation on Adjustment of Interest Rates to Inflation, 1971–811
Country11τ1θ21τβ (of πe)2β(of π)3
Canada1.1760.9060.9910.879
France1.4931.1201.0220.973
Germany, Fed. Rep. of1.5150.9700.9911.050
Italy1.4291.0721.0300.505
Japan1.0750.8601.1620.818
Netherlands1.2990.9840.283
United Kingdom1.3160.9210.9970.061
United States1.2990.9090.9930.968
Source: Background paper by Menachem Katz, “Inflation, Taxation, and the Rate of Interest in Eight Industrial Countries,” in Part II of this volume (Chapter 7), Table 4.

The average income tax rate applicable is denoted by τ and the capital gains tax rate is denoted by θ.

The tax rates used here are given in Table 15 of the Katz paper (see Source). The difficulties of calculating the effective tax rates are well known; hence, it was necessary to adopt a variety of approaches for estimating these rates. The method used, as well as its limitations and the biases it creates, is explained in the Katz paper (see Source).

The coefficients β (of πe) and β (of π) represent, respectively, the Fisher effect for expected and actual inflation rates.

Source: Background paper by Menachem Katz, “Inflation, Taxation, and the Rate of Interest in Eight Industrial Countries,” in Part II of this volume (Chapter 7), Table 4.

The average income tax rate applicable is denoted by τ and the capital gains tax rate is denoted by θ.

The tax rates used here are given in Table 15 of the Katz paper (see Source). The difficulties of calculating the effective tax rates are well known; hence, it was necessary to adopt a variety of approaches for estimating these rates. The method used, as well as its limitations and the biases it creates, is explained in the Katz paper (see Source).

The coefficients β (of πe) and β (of π) represent, respectively, the Fisher effect for expected and actual inflation rates.

Many proposals have been made to restrain the positive effect of the “typical” tax treatment of interest income and payments on nominal interest rates. In particular, these have included (1) a complete inflation adjustment of interest income and expense for taxation purposes, (2) the elimination of taxes on interest income simultaneously with the elimination of the tax deductibility of interest payments, and, finally, (3) limitations on the tax deductibility of interest payments. The discussion in Section III suggests that the first two proposals will have the greatest effect on interest rates but are unlikely to be adopted by many countries (and, because of capital flows, it may not be in the interest of large single countries to do so). The last proposal—limiting the tax deductibility of interest payments for nonbusiness purposes—seems, in reality, to hold the most promise; it is administratively feasible, it would reduce the fiscal deficits, and it would have beneficial effects on capital formation. Its effect on interest rates would, however, not be very significant.

Section IV considers some extensions and qualifications of the discussion in Section III. Differential taxation across countries is shown to affect the levels of interest rates and to induce international capital flows. In addition, tax rates on foreign exchange gains (and losses) that are lower than those on interest income would lead to an increase in the pretax nominal interest rate differential in order for an international equilibrium of credit markets to exist. A change in the tax treatment of interest income and expense through inflation adjustment would lower the effective tax rates on interest income for any single country taking such steps. As the nominal interest rates would also fall, some capital outflows would result. A new equilibrium nominal interest differential would thus emerge, based on the relationship between the new effective tax rate on nominal interest and the tax rates applied to foreign exchange gains and losses.

Economic theory and empirical evidence suggest that expected inflation causes a decline in after-tax real interest rates. When this relationship is considered in the context of policies of monetary restraint, it implies that real interest rates may stay high for some time after the initiation of monetary restraint. Initially, this may be due to negative liquidity effects when money growth is first slowed; then it becomes due to a rise in the real interest rate resulting from a combination of the wealth effect and tax policy.

Section IV also examines the causes of the higher volatility of interest rates since 1979 (see Section IV, Table 2). The tax treatment of interest income and expense, coupled with “bracket creep,” has probably enhanced the impact of changing inflationary expectations on the volatility of interest rates in recent years. As interest rates are determined in forward-looking markets by investors and savers interested in future, after-tax, real returns, any event that tends to broaden the range of possible future outcomes for inflation, fiscal deficits, and the tax code itself also increases interest rate volatility.

Table 2.Level and Volatility of U.S. Interest Rates(Monthly data)
Treasury BillMedium-Term

Government Bond
Long-Term

Government Bond
Sample

Period
MeanStandard

deviation
MeanStandard

deviation
MeanStandard

deviation
1964–694.791.095.201.065.000.77
1970–745.951.686.691.026.830.79
1975–796.672.017.791.258.310.66
Jan. 1980–

Jan. 1983
12.022.7312.882.0112.621.48
Source: United States, Federal Reserve System, Federal Reserve Bulletin (Washington), various issues.
Source: United States, Federal Reserve System, Federal Reserve Bulletin (Washington), various issues.

Section V brings together the major conclusions of the paper and re fleets on their implications for Fund activities. It argues for the use of opportunities offered by consultations under Article IV of the Fund’s Articles of Agreement to review, especially in inflationary circumstances, important tax reforms in areas such as the tax treatment of interest income and expense and taxation of foreign exchange gains and losses. Control of inflation should be an important objective of all governments. If inflation cannot be controlled (and subject to budgetary and other constraints), the country should explore the possibilities for adjustment of income taxation for inflation, including the tax treatment of interest income and payments. Such an exercise should also simultaneously review the scope for (1) curtailing discretionary, and often distortionary, tax incentives given to (the recipients of capital income, and (2) improving the efficiency of collection of income tax from such income so as to reduce the opportunity for tax avoidance and tax evasion. The possibilities for narrowing the differences in the tax treatment of foreign exchange gains and of interest income should also be explored.

Finally, interest is only one form of capital income; dividends, capital gains, business profits, and properly rentals are other important forms. Ceteris paribus, the tax treatment of interest income and payments relative to the tax treatment of other capital income tends to have important allocative effects in an economy (Ben-Zion (1983)), and their differential tax treatments across countries can affect the form of international capital movements and the sectors to which the international capital will flow. This paper does not deal with these important and complex allocative questions, nor does it deal with the effects of taxes in developed and developing countries where most interest rates are regulated.

II. TAX TREATMENT OF INTEREST INCOME AND EXPENSE

The tax treatment of interest income and expense varies among major industrial countries.4 The differences are due partly to historical reasons and partly to the different weights that policymakers attach to the objectives of tax policy—namely, revenue, equity, and efficiency. Prima facie, the differences among industrial countries in the taxation of income—and particularly in the tax treatment of interest income and payments—appear to be so large5 as to defy generalization. However, on closer look, and subject to the problems discussed in Appendix I, some basic similarities are discernible.

Basic similarities

Almost all of the industrial countries have the following similarities:

  • (1) They treat nominal interest income, including that received from abroad, as any other source of income and thus tax it at the progressive global income tax rate.6

  • (2) They exempt interest income earned by certain institutional recipients, such as pension and retirement funds, selected financial institutions, certain government bodies, and most charitable and nonprofit institutions (see Appendix III, Table 3).

  • (3) They exempt, with or without limits, interest on certain debt instruments, such as specified government securities, deposits with selected savings institutions, and bonds of certain public enterprises (see Appendix III, Table 3).

  • (4) They often exempt specified amounts of interest and/or dividend income for administrative reasons or to promote savings (see Appendix III, Table 3).

  • (5) They allow full deductibility of interest payments on all borrowing for income-generating activities (see Appendix III, Table 4).

  • (6) They permit deductibility, with or without limit, of mortgage interest payments on at least one owner-occupied house (see Appendix III, Table 4).

  • (7) They generally tax interest paid to nonresidents at final withholding tax rates (frequently well below the typical marginal tax rates for individuals and corporations); nonresident taxpayers can generally rely on a foreign tax credit to avoid double taxation in their country of residence (see Appendix III, Table 5).

Major differences

Despite the similarities, there are major differences:

  • (1) While most countries tax nominal interest income at the normal income tax rates, Belgium, Japan. Italy, and France permit withholding taxes on them to become final taxes. Some of these rules are pragmatic measures to ensure compliance. The United Kingdom, on the other hand, has a supplementary tax of 13 percent, over and above the income tax, on all investment income, including interest income. In Belgium, too, a supplementary tax. ranging from 20 percent to 47 percent, is levied on interest and dividend income exceeding certain limits.

  • (2) While most countries exempt interest income earned on selected debt instruments within reasonably low limits, and sometimes subject to a ceiling on the taxpayer’s total income (see Table 1; see also Byrne (1976)), the exemption of interest income is relatively generous in Japan (see Appendix III, Table 3). An exemption of interest on savings up to a cumulative nominal amount of US$56,0007 is stat-utorily permitted to each household for bank deposits, postal savings deposits, certain government bonds, and savings for formation of employee’s assets. In addition, in Japan there is no ceiling on the taxpayer’s income to qualify nor is a taxpayer limited to having only one savings account. In the United States, too, there is no restriction on a taxpayer’s holdings of state and local government bonds. The selected debt instruments (government bonds and savings deposits) on which interest tends to be exempt from taxation generally carry lower interest rates, when adjusted for risk, than taxable debt instruments of equal maturity.8

  • (3) While all countries allow tax deductibility of business-related interest payments, subject to few restrictions, only the United States, the Netherlands, Switzerland, and the Scandinavian countries permit the tax deductibility of interest payments on consumer loans and without any limit (Appendix III, Table 4). The degree to which consumer loans are readily available and the degree to which taxpayers habitually use them vary among these countries. The question of the imposition of some limits on this “tax expenditure,” and even a complete elimination of it, has been raised at one time or another in the United States, the Netherlands, and the Scandinavian countries.

  • (4) While many countries allow tax deductibility of mortgage interest on owner-occupied housing (Appendix III, Table 4), they either tax the imputed incomes from housing or limit the amount of mortgage interest that can be deducted, or do both. The United States, the Netherlands, Switzerland, and the Scandinavian countries do not limit the tax deductibility of mortgage interest. Furthermore, Canada, France, Japan, Switzerland, the United Kingdom, the United States, and a few other industrial countries do not tax the imputed income from such owner-occupied housing.

  • (5) While most countries have withholding taxes on interest income,9 Australia, Canada, the Netherlands, and the Scandinavian countries have no such taxes (Appendix III, Table 3). The United States also does not have a withholding tax on interest income and has recently rejected its introduction.

  • (6) While most countries tax long-term capital gains of individuals, either under a separate tax (e.g., the United Kingdom) or under the regular income tax after exempting a certain proportion of capital gains (e.g., Canada, Sweden, and the United States), most industrial countries apply lower tax rates on long-term capital gains than on ordinary income. The Federal Republic of Germany10 and Japan, on the other hand, treat long-term capital gains in much the same manner as they treat ordinary income. The tax treatment of long-term capital gains realized by companies also differs among countries.

  • (7) While most countries tax interest earned by nonresidents, some countries provide for selective exemptions. For example, Belgium, France, Ireland, and the United Kingdom exempt interest earned by nonresidents on certain government securities, and the United States exempts interest earned by nonresidents from banks and other savings institutions, as also do Belgium and Denmark (see Appendix III, Table 5). If withholding taxes are imposed on interest payments to nonresidents, they are typically much lower than normal income tax rates, but they apply to gross payments rather than to net income and are often final tax payments in the source country. Capital gains of nonresidents are usually taxable, especially if they are related to real property or business, but France, the Federal Republic of Germany, and the Untied Kingdom exempt capital gains of nonresidents on all but a few assets. Such provisions can encourage foreign capital inflows for selected purposes.

  • (8) While most countries tax only the realized foreign exchange gains of their residents, legal provisions frequently lend to be complex and subject to different interpretations. Some industrial countries treat these gains as capital receipts and subject them to the tax rates for capital gains; others treat them as current receipts and apply the income tax rate. In most countries, the tax treatment becomes a subject for court decisions, and the case becomes law. The final outcome, though consistent with prevailing accounting practices in individual countries for tax purposes, tends to affect capital flows between countries.

State of inflation adjustment

To the extent that the bases of the income tax, corporation tax, and capital gains tax of industrial countries are not adequately adjusted for inflation, the effective tax rates on real amounts of interest income, corporate profits, and capital gains tends to rise. At the same time, the effective benefits from the tax deductibility of nominal interest payments and various tax incentives for savings and investment also rise. Although certain industrial countries have in recent years introduced some adjustment schemes, none of these countries seems to have fully adjusted its personal income tax and its corporate income tax for inflation.

The most comprehensive adjustment schemes for the personal income tax are found in Canada and the Netherlands, but even these schemes are limited to the adjustment of tax brackets and do not extend to the tax bases. In Canada, income tax brackets and personal allowances are automatically changed annually in accordance with changes in the consumer price index. In the Netherlands, income tax brackets are normally adjusted annually by a certain percentage of the increase in the consumer price index of the preceding year. Several other industrial countries (for example, Denmark, France, Sweden, and the United Kingdom) have also introduced adjustments for inflation. In France, such adjustments are discretionary and limited to years when the inflation rate exceeds 5 percent. The U.K. adjustments are also discretionary and have been limited to personal allowances and deductions. Sweden also has an automatic indexation of exemptions and bracket limits, but recent legislation has restricted application of the system. Legislation in the United States provides for the indexation of income tax brackets and personal exemptions to inflation beginning in 1985. Italy has also experimented with occasional adjustments of the tax bracket.

Selective information collected on changes in the income tax systems of eight industrial countries (Austria, Canada, France, the Federal Republic of Germany, Italy, Switzerland, the United Kingdom, and the United States) suggests that, since 1973, personal allowances (or tax credits) and income deductions have generally been adjusted by less than the rates of inflation in these countries.

With respect to taxation of capital gains, Ireland and the United Kingdom have made legal provision for adjustment of the cost of acquiring assets with reference to the consumer price index over the holding period of the asset. Sweden and France also have such a provision, but it is applicable only to certain categories of assets.

Tax provisions for a comprehensive adjustment of business and corporate profits for inflation do not exist in any industrial country, although a few industrial countries (e.g, Belgium, Japan, the Netherlands, and the United States) do allow use of the last-in, first-out method of inventory accounting. In all industrial countries except Denmark, depreciation of assets is still allowed on the basis of historical cost rather than replacement cost. Most countries, however, have liberalized their investment incentives (accelerated depreciation allowances, investment allowances, income tax credits, tax-free reserves, giants, free write-offs, etc.) in recent years,11 often allowing more than 100 percent of the original cost of acquisition of assets, or have adopted other means of reducing the growth of the real tax burden on corporate entities. The value of such tax subsidies to capital varies from country to country; these subsidies seem to be more generous in Italy, the United Kingdom, and the United States and less generous in the Federal Republic of Germany and Japan (Kopits (1981)).

In conclusion, the tax regimes of major industrial countries are seen to be different, as a result of both different traditions and the mix of domestic objectives pursued by policymakers. Yet the taxation of interest income is nearly universal, and practically everywhere interest income is taxed on nominal rather than real magnitudes. The changes made in response to inflation in the structure of income taxes (e.g., the rates, income brackets, personal allowances, and one or more personal deductions) have been both ad hoc and incomplete. The effective tax rates on interest income, therefore, have tended to differ among industrial countries as a consequence of differences in legal provisions, in degrees of tax compliance, in rates of inflation, and in degrees of exemption accorded to interest income.

The tax treatment of interest expense differs even more markedly among industrial countries. Some countries offer deductibility from taxable income of all interest payments (for business, home ownership, and consumer credit), while others allow it only for income-earning activities and still others allow it for business purposes as well as home ownership (the latter generally in a restricted fashion). But nearly everywhere the deductions for interest payments reflect nominal amounts, making the tax benefits of available deductions differ even more significantly from country to country, depending on the rate of inflation.

Interest payments to nonresident taxpayers are subject to withholding taxes in most industrial countries, but foreign exchange gains and losses are not treated uniformly. In some countries, they are treated as capital receipts and are subjected to (lower) capital gains taxes while in other countries they are treated as current receipts and are subjected to (relatively higher) regular income taxes.

The possible implications of these and other differences in tax regimes on (1) the level and variability of interest rates and the effectiveness of monetary restraint in a given economy and on (2) interest rate differentials between countries and international capital movements are discussed in Sections III and IV.

III. EFFECTS OF TAXES AND INFLATION ON INTEREST RATES

In an inflationary environment, the typical tax treatment of interest income and expense—taxing nominal interest received and allowing a deduction for nominal interest paid—can have a significant impact on the level of interest rates and can also affect their volatility. As inflation distorts the base for the tax on interest income and increases the effective tax rates on income in general, this impact is likely to be magnified over time as long as the rate of inflation continues to be high. The tax treatment of interest income and expense may also alter the redistributive impact of monetary restraint and may affect capital flows across countries.

The survey of taxation of interest income in Section II indicates that the industrial countries generally (1) tax nominal interest income without allowing any adjustment for its inflationary component. (2) allow a deduction for interest expense, again without any adjustment for the inflationary component, (3) permit the issuance of tax-free financial assets by public bodies, and (4) do not tax the income received by certain institutions (such as charitable, educational, and religious ones).

However, these generalizations hide important differences, which are set forth in Section II and in Appendix III. For example, (1) France, Italy, and Japan collect withholding taxes, levied at lower rates than marginal income tax rates, that become final taxes (for taxpayers with high taxable incomes, this is an advantage that may induce them to supply loanable funds at lower rates than they would otherwise do); (2) Japan provides more generous exemptions for interest income than other countries do, again potentially reducing the offer price (the interest rate demanded) for funds; and (3) the Netherlands, the Scandinavian countries, Switzerland, and the United States permit tax deductibility of interest payments on consumer loans and also do not limit the tax deductibility of mortgage payments.12 Thus, it would be expected that the rate of interest would be higher in the latter countries and that a potentially larger share of investment would go toward these tax-advantaged investments.

In recent years, the potential effect of taxation on the level of interest rates in an inflationary situation has attracted the attention of some economists. Two major conclusions have come out of the resultant literature.

In a situation where there is inflation and the income tax is imposed on nominal interest income while nominal interest payments are deductible expenses, taxation should have a positive effect on the nominal rate of interest. To the extent that part of the interest received is not taxed or part of the interest paid is not tax deductible, this tax effect would naturally be reduced.

As the rate of inflation is likely to be negatively correlated with the real rate of interest (for reasons given below), the positive effect on interest rates associated with the existence of taxes may not be apparent in simple quantitative analysis.

Effects of inflation on interest rates in the absence of income taxes

For convenience, the discussion here is organized around the Fisher equation, which simply states that the nominal (or market) rate of interest, i, equals the sum of the expected real rate, r. and the expected rate of inflation, π. That is,

In the absence of expected inflation, the real rate of interest and the nominal rate of interest will be the same. As expected inflation acquires a positive value, the Fisherian hypothesis asserts that, if the expected real rate is constant and therefore independent of expected inflation, each percentage point rise in the expected rate of inflation results in a percentage point rise in the nominal rate of interest. This hypothesis is usually expressed as

It must be emphasized, however, that equation (2) represents a quite rigid or extreme view of how inflation is likely to affect interest rates.13 In reality, there are several reasons why β would not be equal to unity. (It should be stressed that the existence of taxes is still being ignored.)

Real balance effect

The real balance effect, associated with Robert Mundell and James Tobin, postulates a negative relationship between the real rate of interest (r) and the expected rate of inflation (π). In Tobin’s formulation, a rise in expected inflation causes a shift out of money balances and into real capital, thereby depressing the marginal product of capital and the equilibrium real rate of interest. In Mundell’s formulation, a rise in the expected rate of inflation reduces the real cash balances of individuals, making them feel poorer. They react by raising the steady-state level of saving, thus pushing down the real rate of interest.

Liquidity effects

As additional money is injected into the economy, individuals may for a time experience excess liquidity, particularly if the increase in the money supply is not fully anticipated. Thus, before prices and inflationary expectations fully adjust upward, the impact of excess money may, as Keynes argued long ago, lead to a lowering of the rate of interest. However, when the money increase is fully anticipated, as it would be when the rate of inflation has stabilized, this effect disappears. In an economy where inflation has been rampant for some time, this liquidity effect is not likely to be significant. By the same token, the real rate of interest may increase if there is a drastic but not fully anticipated decrease in the growth of the money supply.

Economic activity effect

Various nominal rates of interest can be associated with the same inflationary expectation, provided that the level of economic activity varies. The demand for loanable funds is likely to be lower during recessions or during periods of low economic activity, when many investments are postponed; on the other hand, it is higher during booms, when optimism is prevalent and investment is high. Thus, a slowdown in economic activity is likely to pull the rate of interest below the level that, ceteris paribus, would exist if economic activity remained at a “normal” level. This effect might be reflected in insufficient adjustment of the nominal rate for the expected rate of inflation.

Institutional constraints

There is probably no country where all interest rates are completely free to adjust to the level determined by the market. To varying degrees, the movement of interest rates is constrained by legal or institutional limitations, so that the observed rates may be lower than the rates that would prevail in the absence of any limitations.

Money illusion

Although economists have become increasingly skeptical about the existence of money illusion, there must be at least some individuals who, especially when the rate of inflation is low or when inflation is a new phenomenon, confuse nominal interest rates with real interest rates. As long as some hold this illusion, the nominal rates may tend to increase by less than expected inflation.

Fiscal deficits

Fiscal deficits can influence the rate of interest in various ways. If they change the level of economic activity in the country, they will affect the rate of interest in the same way as described above. If they change the supply of money in the economy, they will affect the expected rate of inflation, π, and may also influence the rate of interest through the liquidity effects described above.14 However, a more direct effect is through the demand for loanable funds. As the government sells bonds to finance the deficit, the supply of bonds, ceteris paribus, increases. The price of bonds falls, and the rate of interest rises. If these fiscal deficits occur during a recession, when private sector borrowing is depressed, the effect of the deficit on the rate of interest may not be obvious. If the fiscal deficit continues into a boom, its effect on interest rates may become more evident, as public borrowing will be added to the higher level of private borrowing, thus pushing upward the total demand for loanable funds.

Uncertainty

Empirical evidence indicates that higher inflation tends to be associated with a greater variance in relative prices. As investments are essentially commitments to a given set of future realistic prices, it is implied that the risk factor associated with investment rises. This rise in the risk factor induces a negative shift in the borrowing schedule which, per se, implies a lower real rate of interest. On the other hand, similar considerations also reduce the willingness of lenders to lend, thus bringing about a negative shift in the lending schedule. It is thus an empirical question whether, on balance, uncertainty reduces or increases the real rate of interest.

To summarize, the most basic theory of the behavior of interest rates in an inflationary situation is that the nominal rate of interest increases pari passu with the expected rate of inflation; that is, the real rate of interest does not change. However, recent amendments to that theory indicate that (even in the absence of income tax) the Fisherian hypothesis of a close correspondence between expected inflation and nominal interest rates may not be valid. For several reasons, when the expected rate of inflation is increasing, nominal interest rates are likely to adjust by less than the expected rate of inflation.15 It is therefore implied that, in terms of equation (2), the coefficient of π (β) will be less than unity. In other words, a rise in expected inflation is likely to reduce the real rate of interest. Only when a substantial fiscal deficit coexists with a strong boom docs the nominal rate of interest increase by more than the expected rate of inflation. Or, alternatively, only when the rate of inflation decelerates significantly does the real rate of interest increase.

In a period of accelerating inflation and in the absence of income taxes, borrowers, ceteris paribus, face lower real costs of borrowing (and lenders face lower real rates of return) than during a period of price stability. But, as inflation is likely to affect the efficiency of the economy, the rate of return on real investment is also lower.

Effects of inflation on interest rates in the presence of income taxes

Within the framework outlined above, it is now assumed that nominal (rather than real) interest income is fully taxed at a marginal rate equal to τ and that nominal (rather than real) interest expense is fully deductible from the taxpayer’s income before the tax is assessed on his taxable income. This is in conformity with the tax laws of most of the countries surveyed in Section II under which there is no distinction between real and nominal values of interest income and expense.

It becomes necessary to make a distinction between a before-tax real rate of interest, r, and an after-tax real rate of interest, r*. For simplicity, it is assumed that the tax rate, τ, is the same for all taxpayers—that is, the income tax is a proportional tax.16 If, given the tax rate, τ, the net-of-tax expected real rate of interest, r*, is to remain unchanged in the face of a rise in the expected rate of inflation, π, the nominal rate of interest must rise by more than π. More specifically,17 the Fisher equation must be modified and rewritten as follows:

In this equation, the effect of π on i is magnified by the existence of taxes. The higher τ is. the greater is the impact of π on i; τ can range between 0 and 1. In the United States, in recent years, the tax rate on interest income has been close to 40 percent. Adjusting for exemptions and tax evasion, the effective tax rate may be closer to 25 percent. It is therefore implied that, ceteris paribus, a 1 percentage point increase in the expected rate of inflation would result in a 1.33 percentage point increase in the nominal rate of interest. And it also is implied that a 1 percentage point fall in the rate of inflation would result in a 1.33 percentage point fall in the nominal rate of interest. This example shows how taxes can potentially increase the volatility of the nominal rate of interest in situations where inflationary expectations are rapidly changing.

In considering the combined impact of expected inflation and income taxation, it is assumed that (1) lenders and borrowers agree on a 4 percent real interest rate, r, in the absence of inflation; (2) the effective income tax rate is 25 percent18 and the real rate of interest is independent of expected inflation; and (3) expected inflation rises from 0 percent to 6 percent. equation (3) implies that the nominal rate of interest would have to rise to 12 percent in order to maintain the purchasing power of a 4 percent interest rate without expected inflation. To be more specific, the lender is paid 12 percent, of which he pays one fourth, or 3 percent, in income taxes and “loses” another 6 percent to inflation. He is thus left with an expected after-tax real interest rate of 3 percent, which is the same as he would have received in the absence of inflation but in the presence of income tax. The same after-tax real interest rate will also result for a borrower expecting the same inflation rate and paying the same tax rate. Obviously, a lower income tax rate would lead to a lower increase in the nominal rate of interest, while a higher inflation rate would lead to a higher nominal rate.

The theoretical example above gives an exaggerated view of the expected change in the nominal rate of interest when income taxes are present and when the expected rate of inflation is positive. The various factors that make the nominal rate of interest less responsive to changes in expected inflation than would be expected from equation (2) were discussed earlier in this section. Some tax-related factors reduce further some of the magnification effect implied by π/1 – τ in equation (3). These are described below.

Taxes on other assets

Equation (3) implies that alternative and untaxed channels of investment are available to the lender, so that he is willing to lend the same amount as before only if he does not suffer any reduction in his net-of-tax real interest income. But suppose alternative uses of his funds, i.e., real or other financial investments, are also taxed, even though at lower rates. (Some of these other taxes may simply be the capital gains taxes on realized gains from real property or from bond holdings.) In this case, the lender’s options are limited, and he may be willing to accept a lower rate of return on all his financial investment. It would therefore be implied that the effect of π on i would be less than π/1 – τ. If the average tax rate on all other operations is indicated by θ, then the effect of π on i is π(1θ)1τ.

if θ is equal to zero, the situation would revert to that described by equation (3). If θ = τ, then there is no tax effect. In general, θ would be lower than τ but higher than zero.

Taxes on borrowers

The theoretical result of equation (3) must also be qualified to take into account the fact that the borrowers themselves may experience tax increases associated with expected inflation that reduce their willingness to pay the nominal rate shown by equation (3). For example, if depreciation is estimated on the basis of historical cost, and if inventories are evaluated on the basis of first-in, first-out accounting methods, as is the case in most industrial countries, corporations will find their tax burden increased during an inflationary period. Therefore, they will not be willing to pay the nominal rate implied by equation (3). Thus, again, the nominal rate of interest is likely to be lower.

Tax-exempt lenders and borrowers

As all financial markets frequently include many lenders who do not pay taxes on interest received and many financial instruments that, because of the nature of the issuers, pay tax-free interest (see Appendix III, Table 3), the impact of taxes on interest rates is reduced. It is unlikely that tax-free institutions are marginal lenders; nevertheless, this factor again reduces the role of taxation in interest rate determination. In the United States in 1976, the latest year for which this information is available, the proportion of tax-exempt income to total interest income was 15.

Tax evasion

The tax systems of all countries require interest income to be taxed. In reality, however, owing to the absence of withholding provisions in the tax systems of many countries (including the United Slates), the existence of bearer’s shares, and the inability of the tax authorities to ascertain all interest income paid, some interest income is not reported to the tax authorities. Thus, the potential effect of taxation on interest rates may be reduced by tax evasion. Here, again, it must be realized that it is the tax treatment of the marginal lender (and borrower) that is significant. But, to the extent that tax evasion brings about a rightward movement in the supply schedule of funds, it reduces the nominal rate of interest.

Capital inflows

If taxes magnify the effect of inflation on nominal interest rates in a given country, these rates could become attractive to foreigners, especially if the latter can avoid being taxed in their own countries and are not taxed in the country in which they invest their money. Furthermore, as capital flows in from “tax-haven” countries, it exerts a downward pressure on the interest rate, which helps to make the rate diverge from the theoretical results of equation (3).

To sum up, the presence of income taxes—that is, the taxation of interest income and the tax deductibility of interest expense—tends to magnify the effect of inflation on interest rates. Many economists have therefore come to expect that β > 1. In reality, many tax-related and nontax-related factors (listed in this section) tend to dampen the value of β. Consequently, the fact that an increase in the expected rate of inflation does not always increase the nominal rate of interest more than proportionately cannot be taken to mean that tax factors do not matter; it could simply be that the positive impact of the tax treatment of interest income and expense may be partially or fully neutralized by the other factors mentioned above.

Empirical investigations

Empirical investigations of interest rates have improved the level of understanding about their determinants but have not yet fully explained the behavior of real rates of interest. In addition, economists attempting empirical investigations of interest rate behavior have struggled with the very difficult problems of accurately measuring expected inflation and expected real interest rates in terms of observable variables. Their efforts have progressed through three stages.

The first stage of concerted empirical investigation of interest rate behavior in the period after World War II was centered on the simple Fisher equation (equation (2) above). The approach was to regress nominal interest rates on various measures of expected inflation. Thus, most investigators were testing a joint hypothesis of market efficiency, whereby a percentage point rise in expected inflation would result in a percentage point rise in nominal interest, conditional on the hypothesis that the expected real rate was independent of expected inflation. As with all empirical investigations of interest rate behavior, it was also necessary to assume that behavior of expected inflation was being accurately measured. Many of these investigations conducted during the 1960s and early 1970s found a less-than-proportional impact of expected inflation on nominal interest; that is, the estimated values of β were found to be persistently below unity.19

The second phase of empirical investigation of interest rates, begun during the late 1970s, incorporated the taxation of interest income and deductibility of interest expense into a modified Fisher equation (equation (3) above) and hypothesized a greater-than-proportional rise of interest rates to changes in expected inflation. Given the persistence of a less-than-proportional response of interest rates to expected inflation, the gap between theory and reality was widened even further.20 This forced investigators to re-examine the hypotheses of constancy of real interest rates and their independence from expected inflation.

The third phase of empirical investigation of interest rates derived expressions for nominal interest rates from more comprehensive models. This approach enabled investigators to incorporate into interest rate equations those variables, other than expected inflation, that theoretically should help to determine the behavior of real rates.21 A number of these variables are discussed above. This broader approach also indicated more clearly the precise nature of the relationship between interest rates, expected inflation, and variables such as taxes, real balance effects, economic activity, and uncertainly, all of which determine the real rates of interest.

The comprehensive approach to modeling interest rates has also helped to resolve the paradox arising from consideration of tax treatment of interest income and expense only in the context of the simple Fisher equation. The incorporation of tax treatment of interest income and expense into the analysis of the relationship between nominal interest rates and expected inflation suggested a magnified impact of the latter on the former (that is, β > 1). Failure to discover such a magnified impact led some analysts to conclude that the effective tax rates on interest income and interest expense must be very low, but this interpretation was questionable for at least three reasons: (1) empirical results usually find a coefficient of 0.7 to 0.9 on anticipated inflation and, even if effective tax rates were only 10 percent, the coefficient should be about 1.1; (2) information on the holdings of tax-exempt financial assets or of tax-exempt interest income suggests that, in the United States, such tax-exempt securities constitute only 10–15 percent of total holdings; (3) and most important, a comprehensive framework of interest rate determination has also yielded a value for the coefficient on anticipated inflation, β, in the 0.7 to 1.0 range.

A recent paper by the Fiscal Affairs Department has investigated the relationship between interest rates and inflation (simple Fisher equation) and the effects of taxation on this relationship (modified Fisher equation) for a sample of eight industrial countries.22 Subject to various limitations that are set forth, the paper investigates the extent to which interest rates have responded differently to changes in expected and actual itflation rates in countries with different legal tax treatments of interest income and payments.23

The eight countries were divided into three groups according to the degree of taxability of interest income and the deductibility of interest payments. The first group consisted of Canada, the Netherlands, and the United States, all of which treat interest income and payment literally in a way that would imply a relatively high value of β (greater than unity). The second group consisted of France, the Federal Republic of Germany, Italy, and the United Kingdom, all of which have a more moderate tax treatment of interest income and payment. The third group consisted of Japan, which has a more generous tax treatment of interest income and an implied low value of β (smaller than unity).

The estimation results for 1971–81 presented in Table 1 indicate that the response coefficients of nominal short-term interest rates to actual inflation, adjusted for changes in real interest rates, were not significantly different from unity for Canada, France, the Federal Republic of Germany, and the United States. The estimated response coefficients were found to be more moderate for Italy and Japan, low for the Netherlands, and insignificant for the United Kingdom.

For nominal short-term interest rates, the response coefficients to expected inflation were significantly greater than unity for Japan but not significantly different from unity for the other seven countries.24

For long-term interest rates, the response coefficients to expected inflation were about unity for the United Slates, moderate for France, and low for all other sample countries (see Table 1). The response coefficient of long-term interest rates to actual inflation was found to be well below unity for Canada, France, Italy, the Netherlands, and the United Slates, and insignificant for the Federal Republic of Germany, Japan, and the United Kingdom.

In general, while some variation occurs across countries in the value of β, none of the coefficients is significantly above unity. This situation suggests that the positive impact of typical tax treatment of interest income and interest expense is perhaps frequently neutralized by the impact of many other factors, as mentioned above at the beginning of this section. This inference can be treated as do more than a tentative hypothesis at this stage; further empirical work will be required to establish its validity.

Policy implications

In the preceding discussion, it is argued that high effective taxes on interest income and liberal tax deductibility for interest payments can, at least theoretically, have a significant impact on the level and variability of interest rates. To reduce this potential tax effect, three alternative policies have been suggested at one time or another: (1) Tax only real interest income and allow a deduction only for real interest expense. This implies the removal of the inflation component from both interest income and expense; (2) For individuals, at least, eliminate from income taxation all interest income and, at the same time, do not allow any deduction for interest expense;25 (3) Reduce the range of deductibility of interest expense at least for consumer loans and possibly for mortgages on owner-occupied houses. Some of the implications of these policies are discussed below. On the basis of existing technology, it is not possible to give robust or reliable estimates of the effects of the above changes on the levels of interest rates. However, on the basis of a few plausible assumptions, some guesses can be made. Some of these are derived from a simple theoretical loanable fund model, outlined in Appendix II.

Inflation adjustment of interest income and expense

The distortions created by the typical tax treatment of interest income and interest expense in an inflationary environment are due primarily to the taxation of nominal, rather than real, magnitudes. Full inflation adjustment would require that the inflation rate be subtracted from the nominal interest rate received by lenders before calculating the taxes due. It would also require that the inflation rate be subtracted from the interest rate paid by borrowers be fore a deduction for interest expense could be claimed.

Such a correction would shift the demand and supply schedules of loanable funds in ways that would result in lowering the equilibrium nominal interest rate associated with a given rate of inflation. The supply of funds schedule would shift to the right as lending became more attractive. The demand for funds schedule would shift to the left is borrowing became less attractive. In the numerical example discussed earlier in connection with equation (3), a nominal interest rate of 12 percent was required to yield an after-tax real rate of 3 percent, given a tax rate of 25 percent and an expected inflation rate of 6 percent. If the expected rate of inflation is equal to the actual rate of inflation and if the tax applies only to the real rate of interest, a 10 percent interest rate will now yield the same after-tax real rate of 3 percent. Thus, in this extreme example, the nominal rate of interest could fall by 2 percentage points.

Realism would require that (1) the negative impact of expected inflation on interest rates be recognized, and (2) the qualifications discussed in the preceding subsection on the effects of inflation on interest rates in the presence of income taxes be taken into account. When this is done, perhaps the reduction in the nominal rate of interest would be less than 2 percentage points. However, even if the fall in the nominal rate was only 1 percentage point (a figure that can be considered conservative), it would still have important effects. it would, for example, by reducing interest cost on the public debt, reduce the U.S. fiscal deficit by $8 billion, and it would reduce the cost of borrowing for developing countries by considerable amounts. Furthermore, the reduction would benefit disproportionately more those borrowers (including losing enterprises, low-income taxpayers, and developing countries) that had not been in a situation whereby they could deduct their interest expense from taxable income.

Not all the effects from this policy change would be positive. A few potentially negative effects are mentioned below.

  • (1) Inflation adjustment of taxable interest incomes would reduce tax receipts on interest earnings; however, this impact would be partly or totally offset by the gain in receipts, owing to the deductibility of (lower) real interest payments rather than (higher) nominal interest payments from taxable income. The net effect would depend on the average tax rates of lenders versus borrowers, with the net revenue impact being positive if the average tax rate of borrowers exceeds that of lenders.26

  • (2) The results would also be affected by enactment or nonenactment of inflation adjustment in other developed countries. If only one country—even one as large as the United States—adopted such a change, the downward pressure on interest rates would be significantly reduced by capital outflow to markets where the absence of full adjustment of interest taxation continued to keep the interest rates higher.

  • (3) Inflation adjustment of the tax treatment of interest income and payments that lowered equilibrium nominal interest rates would confer windfall gains on lenders holding loans contracted at the fixed nominal interest rates that were required in the preindexing environment and would invoke windfall losses on the borrowers issuing such contracts.

  • (4) Even after indexation of interest income and payments for tax purposes is carried out, some distortions will still remain. The unindexed tax treatment of inventory and depreciation allowances, for example, will continue to affect the after-tax profitability of investment, inducing negative shifts in the investment schedule and, as a consequence, reducing the equilibrium rates of interest. Furthermore, indexation to reduce distortions in the after-tax costs and benefits of financing by means of debt will also have considerable effects on domestic equity markets as well as on the flow of debt and equity capital abroad.

  • (5) Finally, and most important, the indexation of interest income and expense for tax purposes would create nightmares for income tax administration. No country has adopted or even attempted complete indexation of interest income and payments, despite the theoretical attractiveness of the proposal. Thus, it would be Utopian to expect that countries would agree to such a change just for its effect on interest rates.

Eliminating taxes on interest income and tax deductibility of interest payments

Total elimination of taxation of interest income, together with the total elimination of deductibility of interest payments, is the easiest solution administratively. It would be a desirable policy from the standpoint of lowering nominal interest rates. In fact, the reduction in nominal interest rates associated with this policy would most likely exceed that associated with the taxation of real interest rates, as discussed above. Such a policy would produce a zero net impact on tax revenue in countries where the level of domestic borrowing and lending and the tax rates applicable to borrowers and lenders are equal. If it was carried out by all countries, it would also equate effective real interest rates in different countries, assuming the equality of expected inflation rates.

Yet such a policy is unlikely to be adopted by many countries. It will convert the existing global income tax system, under which all sources of income are treated equally, into a schedular income tax system. With the exemption from taxation of interest income, pressures would be created for the exemption of other capital income as well, on the grounds that such a policy distorts the flows in the financial and capital markets. Such a policy might be seen as unjust and inequitable between earners of labor income and capital income, especially at present when the real rates of interest and thus the income associated with financial assets are very high. In any case, the principal justification for this policy—that the rate of inflation is about equal to the nominal rate of interest—is clearly not valid at this time for many of the large industrial countries. Thus, the policy could be justified in certain countries but not in others.

Limiting the tax deductibility of interest expense

A realistic proposal would be to eliminate, or at least limit, the deductibility of interest payments for particular purposes, namely, interest on consumer credit and mortgage interest payments by households. To what extent this would result in a drop in the equilibrium nominal interest rate in a given country would depend on the elasticity of the supply of funds and the demand for funds. Because the elimination of the deductibility of household interest payments would cause a downward shift in the demand for funds, the downward pressure on interest rates would be correspondingly higher as the ratio of supply elasticity to demand elasticity is lower. No estimates of these elasticities are available; however, on certain hypothetical assumptions, it appears that the nominal interest rates would not fall by a large amount.27

If deductibility of interest payments from home mortgages was preserved and only the deductibility of interest on consumer credit was eliminated, the impact would be even smaller. In the United States, mortgage borrowing, while volatile, usually constitutes well over half of total household borrowing. A policy that eliminated only nonbusiness and nonmortgage interest payments from deductibility would hardly reduce the equilibrium interest rate.

The elimination of nonbusiness and nonmortgage interest deductibility, despite its relatively small impact on the equilibrium level of nominal interest rates, would still produce significant effects on resource allocation, since more resources would be directed toward capital formation. It would also have some revenue effects that would produce a moderate reduction in fiscal deficits. The effect on revenue enhancement of such a policy change for the United States is estimated to be about $8 billion a year. In addition, a 0.5 percent drop in interest rates could also lower the annual U.S. debt service by about $4 billion annually. An overall reduction in the fiscal deficit for the United Stales of up to $12 billion annually would thus be possible.

In conclusion, inflation adjustment of interest rates for tax purposes or the nontaxation of interest income would have the largest effect on interest rates. The other, more modest, policy discussed above would still have some effect, but it would be somewhat more moderate. Implementation of such policies by any single economy, while helpful, would result in capital outflows. Even if prevention of such flows was considered desirable, it would be difficult to achieve in view of existing statutes in most industrial countries and in view of extensive arbitrage opportunities offered in the Eurocurrency markets.

The elimination of deductibility of interest payments for nonbusiness and nonmortgage (consumption) household purposes seems, in reality, to hold the most promise. It would produce beneficial effects in the form of enhanced capital formation and some moderate reduction in fiscal deficits, but it probably would not have a significant impact on interest rates in any country.

IV. EXTENSIONS AND QUALIFICATIONS OF THE ANALYSIS

This section considers some extensions and qualifications of the analysis in Section III. First, consideration is given to the implications for open economies of tax policies pursued by single countries and to the effects of introducing explicitly the tax treatment of foreign exchange gains and losses. Second, the implications of tax policies for the impact of monetary restraint are discussed. Finally, consideration is given to implications of tax policy on volatility of interest rates and to the related uncertainty regarding expected after-tax real interest rates.

Role of international differences in tax regimes

Insofar as capital is able to move between countries, both domestic and foreign tax policies with respect to interest income and expense will affect the level of interest rates in any given country. Further, as foreign exchange transactions are involved in arbitrage among many international financial assets, taxation of foreign exchange gains and losses will also play an integral role in determining the level of interest rates. To the extent that a country’s capital market is isolated, by controls or other means, from world capital markets, cross-country differences in tax policies are obviously less relevant.28

Perhaps the best method of considering the international implications of the differential tax treatment of interest income and expense is to examine the impact of different tax policies on the after-tax interest parity condition, which states that, in equilibrium, the difference between after-tax interest rates must be equal to the after-tax gain (loss) from expected appreciation (depreciation) of domestic against foreign currency. In the absence of tax considerations, and if exchange rates are determined mainly by purchasing power parity, interest parity can be approximated by an interest differential equal to an expected inflation differential.

Suppose initially that interest parity holds under conditions where countries experience the same rate of inflation and, therefore, the expected change in the exchange rate is set at zero, effective tax rates are equal, and both domestic and foreign governments tax interest receipts and allow full deductibility of interest payments. If, for example, the domestic economy now exempts from taxation all or part of the interest income of households and no longer allows a deduction for all or part of the interest expense, the direct result is to lower the equilibrium nominal interest rates in the domestic economy.29 As a consequence of this fall in interest rates in the domestic economy, capital begins flowing abroad, thus mitigating the initial drop in interest rates and inducing a reduction in the interest rate of the foreign country.30

The flow of capital from domestic to foreign markets will continue until the after-tax interest rates are again equalized, although at some lower level. These effects would be accentuated in an inflationary environment if, For example, the taxpayers in the foreign country experienced bracket creep while the taxpayers in the domestic country avoided it through indexation. This configuration would produce a steady upward drift in borrowing and lending schedules. Equilibrium would require an ever-increasing flow of capital from domestic into foreign markets.

So far, expected changes in the exchange rate have been set at zero by assuming equal inflation across countries or, alternatively, a system of fixed parities. When this condition is relaxed, and expected changes in exchange rates are allowed, tax policy regarding foreign exchange gains and losses becomes relevant in determining the levels of equilibrium interest rates and exchange rates. Assume that tax rates applied to such gains and losses are below the tax rates applied to interest income.31 Such a tax policy would allow interest differentials to exceed expected appreciation or depreciation of currencies. Suppose, for example, that domestic currency is expected to depreciate against foreign currency at a 3 percent annual rate and the domestic interest rate is also 3 percent higher than the foreign rate. If the tax rate applicable to domestic residents on their expected foreign exchange gain is below that applicable to their interest income (regardless of its source), the tax liability to domestic investors on their foreign exchange position would be lower than the income tax liability incurred on their positive interest position. It would, therefore, induce capital outflows. As a result, the equilibrium after-tax interest differential that would not induce further capital movements would have to be above the pretax level. If, say, the capital gains tax rate was 20 percent while the income tax rate was 50 percent, a 3 percent pretax interest differential would have to be matched by a 4.8 percent after-tax interest differential (a factor of 1.6, derived as (1 – 0.2)/(1 – 0.5) = 1.6).32

An alternative to achievement of an after-tax equilibrium by means of interest rate changes would be a reduction in expected depreciation of the domestic currency or a rise in the domestic spot price of foreign currency. That is, if the spot currency rate rose by enough to cut expected depreciation to 1.875 percent, the initial 3 percent pretax interest differential would not be altered by the differential tax treatment, because 3 = 1.6 (1.875). It should be noted, however, that a reduction in the expected rate of devaluation without changes in the underlying interest differentials would imply a deviation from purchasing power parity that had been caused by differential tax treatment of interest income and foreign exchange gains.

Actual equilibrium would probably result from a combination of a change in the interest differential and a change in expected depreciation of currency, if both were accompanied by some capital movements. Whatever the mode, the basic conclusion is that, like tax policy on interest income and expense, tax policy on foreign exchange gains and losses can significantly affect the equilibrium levels of interest rates and exchange rates.

This effect on equilibrium interest and exchange rates of tax policy on foreign exchange gains and losses could be removed by taxing foreign exchange gains and losses and interest income at equal rates. Given the widespread tendency to treat interest as ordinary income, as indicated in Section II, this tax policy would require treating foreign exchange gains and losses as ordinary income as well.33

The realization that neutrality of tax policy calls for a tax treatment of exchange gains and losses similar to that of ordinary income suggests an important asymmetry regarding after-tax interest parity. Consider the example given above and designate as Country A the case in which a 3 percent interest differential satisfies only pretax equilibrium and a wider interest differential (or a smaller currency depreciation) is required when lower (20 percent) tax rates are applied to foreign exchange gains and losses. If Country B applies, from its perspective, equal tax rates (50 percent) to both interest income and foreign exchange gains and losses, the 3 percent differential constitutes an after-tax as well as a pretax equilibrium. The equilibrium interest differential that will emerge in this circumstance would probably result in simultaneous movements of capital in opposite directions between Countries A and B. At anything above a 3 percent interest differential or its exchange equivalent, capital will flow out of Country B into Country A. At anything below a 4.8 percent interest differential or its exchange equivalent, capital would flow out of Country A into Country B. Any equilibrium between 3 and 4.8 percent would result in two-way capital flows. If the situation persisted long enough, equilibrium could eventually be achieved by means of an adjustment of relative real interest rates through changes in the stock of real capital. The burden of adjustment would fall more heavily on the smaller economy.

Operation of the phenomenon just described may be manifested in the unexpected strength of Country A’s currency against that of Country B. Given Country A’s tax policies, equilibrium would be analogous to the 4.8 percent interest differential, while Country B’s tax policies would imply an equilibrium analogous to the 3 percent interest differential. The result would be a flow of capital from Country B to Country A. Even a drop in Country A’s rates would not necessarily stem the flow (and therefore the pressure on the currency of Country B to depreciate against the currency of Country A) unless an analog of the 3 percent equilibrium differential for Country B was reached. The asymmetry could be eliminated by alignment of tax policies whereby both countries would treat foreign exchange gains and losses as ordinary income.

To sum up, differential tax regimes between countries can result in either an increase or a decrease in the pretax equilibrium interest rate differential. Ceteris paribus, countries introducing policies that imply a lower taxation of interest income and a lower deduction for interest expense, in isolation, would have lower interest rates relative to countries that have higher tax rates. However, the negative impact of such policies on interest rates would be mitigated by capital outflows. For small countries acting singly, this impact could be eliminated completely. Finally, countries levying taxes on foreign exchange gains lower than the taxes levied on interest income are likely to experience a greater after-tax differential in interest rates vis-à-vis other countries. Clearly, different combinations of tax treatment result in different outcomes, but a central issue is that real interest rates may also be affected by differential tax policies through their effects on capital flows.

Implications of tax policy for the effectiveness and international impact of monetary restraint

Abrupt application of monetary restraint operates for a time to depress aggregate demand for domestic output through the positive impact on real interest rates as outlined in Section III and through the impact on the real exchange rate. In the absence of taxes applied to nominal interest rates, the result of abrupt and unanticipated monetary restraint would be to raise the market interest rate by the amount of the increase in the real interest rates as long as the initial impact of abrupt monetary restraint does not affect inflationary expectations. Since the full increase in the nominal interest rate would typically be taxed, attempts at preservation of the higher expected real, after-tax rate of return on the part of lenders would require market interest rates to rise by more than the increase in the real rate. The increase may be more easily accepted by borrowers, provided that they can fully deduct interest payments from taxable income. Such a magnification effect is analogous to that displayed in response to a change in expected inflation in the tax-adjusted Fisher equation (equation (3)) discussed in Section III.

However, owing to a somewhat interest-elastic demand for funds by borrowers and the depressing effect of monetary restraint on overall economic activity, the increase in nominal interest rates may be somewhat dampened. Still, the taxation of interest income and the deduction of interest expense would produce a larger positive impact of monetary restraint on interest rates than would otherwise have occurred.

Viewed in this way, the effect of a typical tax policy on interest income and expense is not to reduce the effectiveness of monetary restraint per se as much as to require that a large rise of market interest rates would result from a given monetary restraint if aggregate demand is to be reduced. Tax policy does have implications for the effectiveness of monetary restraint across income tax brackets, however. Households, enterprises, and nations facing effective tax rates below the average tax rate that has been incorporated in the level of interest rates would experience a sharp rise in their after-tax real interest rates relative to borrowers facing above-average tax rates. In sum, the effectiveness of monetary restraint is enhanced by tax policy for individuals and enterprises facing below-average tax rates and is reduced by tax policy for those facing above-average tax rates.

Tax policy regarding interest income and expense in a country applying monetary restraint would also have international implications, particularly if the country concerned is large. If a large country, which has relatively high tax rates for financial market participants—that is, which fully taxes interest income and allows liberal deductions of interest payments—applies monetary restraint, there will be a relative increase in after-tax real interest rates. Countries where investors (including the public sector) have access to international capital markets and where interest incomes are domestically taxed at lower rates or interest expenses are less liberally deductible would suffer as well. This result would be signaled by capital flows into the country applying the monetary restraint. Such inflows are a normal response to monetary restraint, but they are enhanced by differential tax policies on interest income and expense.

As noted above, the discussion up to this point has focused on the initial impact of monetary restraint, prior to any reduction of expected inflation. Once expected inflation begins to fall, the effect of tax treatment of interest income and expense is, symmetrically, to magnify the negative impact on interest rates. As the drop in interest rates reflects the average tax rates, after-tax real interest rates faced by those in above-average tax brackets will rise relative to after-tax real interest rates faced by those in below-average tax brackets. Likewise, countries that experience a relative increase in after-tax real interest rates as a result of monetary restraint will experience a relative reduction in after-tax real interest rates, once lower inflationary expectations begin to reduce interest rates.

The overall implication is that the fluctuations in after-tax real interest rates resulting from changes in monetary policy tend to be exacerbated for domestic and foreign borrowers and for lenders in lower tax brackets, given the differences in tax policy regarding interest income and expense.

The corollary for international capital flows is an enhanced outflow and inflow pattern for countries whose capital markets are closely tied to those of the country executing a policy of monetary restraint. Depending on the degree of intervention, exchange rates may be expected to be somewhat more volatile as well, in view of the tax policies under discussion here. Overall, differential tax policies regarding interest income and expense may increase the volatility of interest rates, exchange rates, and international flows of capital in the wake of application of monetary restraint in a large economy.

The sudden easing of monetary restraint would produce exactly opposite results compared with those caused by the sudden imposition of monetary restraint. Given the taxation of interest income and the deduction of interest expense, a greater fall in interest rates would result from monetary ease if aggregate demand is to increase. The initial liquidity effects of monetary ease would lower the after-tax real interest rates for those who are subject to below-average tax rates relative to those who are subject to above-average tax rates. Subsequent expectations effects would cause a relative increase in after-tax real interest rates for those in lower tax brackets, with the net result that those in below-average tax brackets (including borrowers in developing countries) are likely to experience more volatility in after-tax real interest rates, given either the easing of monetary restraint or its imposition.

A qualification to the analysis of the impact of monetary restraint or ease is called For in the light of possible effects of tax policies beyond those concerning interest income and expense and the empirical evidence of a negative relationship between expected inflation and real rates. The effect of tax policies taken as a whole and the associated phenomena may be to mitigate the negative pressure on nominal interest rates arising from a drop in inflationary expectations. A large body of empirical evidence suggests that the expected after-tax real interest rate may be negatively related to the level of expected inflation. This result implies that the initial rise in real interest rates attendant upon the effect of constrained liquidity may be prolonged at a later time when monetary restraint begins to result in lower inflationary expectations. Conversely, the initial drop in real interest rates resulting from monetary ease may be prolonged if the appearance of higher inflationary expectations is coupled with a drop in real interest rates. The exact timing and stability of such relationships is not understood, but the view is consistent (although crudely) with the recent U.S. experience with the imposition and relaxation of monetary restraint.

Volatility of interest rates

Discussion by the Fund’s Executive Board has often included expressions of concern regarding the particularly high degree of volatility in interest rates and exchange rates since the late 1970s. An effort is made here to explore the possible role that tax policy and its administration may have played in the increased volatility of interest rates, with the implied increase in exchange rate volatility being taken for granted.

Interest rates are determined in forward-looking markets for assets that represent a claim on a nominal stream of interest payments and return of principal over some future period of time. Interest rate volatility at some mean level during a period of time (the variance of interest rates) reflects the volatility in the outlook for inflation and for the expected after-tax real rate of interest. A high volatility of expected inflation or of the expected after-tax real interest rate means that a wide range of possible future outcomes for these variables is contemplated. This is a formal characterization of uncertainty. An increase in such uncertainty entails risks for savers and investors for which they must be compensated. Viewed in this way. increased uncertainty attendant upon increased volatility of interest rates represents an additional cost of capital formation that lowers productivity growth and growth of real output and, in turn, enhances inflationary pressures. In view of such costs and their relationship to interest rate volatility, if is useful lo consider the sources of increased volatility of interest rates.

It is clear from Table 2 that, in addition to reaching historically high levels since 1979, U.S. interest rates have been more volatile. This phenomenon has appeared in varying degrees in most industrial countries. The increased volatility has a number of sources. The two proximate determinants of interest rates—namely, expected after-tax real rates and expected inflation—have been unusually volatile, while the former have probably dominated the volatility of the latter in its impact on interest rates. Volatility of real interest rates in the United States may have been further increased by the volatility in the unanticipated portion of money growth associated with the change in the conduct of monetary policy in October 1979, the imposition and removal of credit controls during 1980, and the implementation of a new policy mix under the new administration during 1981. Another factor that may have exacerbated the volatility of real interest rates is increased uncertainty about fiscal deficits and a wide range of tax provisions associated with debate over the passage and the modifications of the Economic Recovery Tax Act of 1981 and the Tax Equity and Fiscal Responsibility Act of 1982.

Increased volatility of interest rates, like the level of interest rates, can be partially attributed to tax treatment of interest income and expense. The simplest approach, which parallels the discussion in Section III of the effects of these policies on the level of interest rates, is to derive an expression for the variance of nominal interest rates based on the lax-adjusted Fisher equation. Such an expression, which is found in Appendix II as equation (8), describes the variance of nominal interest as a weighted sum of the variances of after-tax real interest rates and expected inflation plus a co variance term. The weights rise with the effective tax rates faced by borrowers and lenders. In effect, if two countries experience identical levels of volatility of after-tax real interest rates and expected inflation, the country with the higher tax rates will experience more volatility of nominal interest rates. If bracket creep raises the effective tax rates in a given country, that change alone, ceteris paribus, will increase the volatility of interest rates.

A more comprehensive approach to interest rate determination—one that takes into account open-economy effects (after-tax interest parity) and a wide range of tax policies that result in inflation-induced changes in real incentives to work, save, and invest—would reveal a more pervasive basis for tax policy to affect interest rate volatility. A given array of effective tax rates on all forms of income (including interest, labor income, profits, capital gains and losses, and foreign exchange gains and losses) is what determines the impact on nominal interest rates of changes in expected inflation and other variables. An environment of high or volatile rates of inflation, given the taxation of nominal values, results in numerous and frequently unpredictable changes in effective tax rates, which actually contribute to the volatility of after-tax real interest rates as well as nominal interest rates. In addition, ad hoc efforts to address tax-policy-induced distortions that are magnified in an inflationary environment increase the uncertainty attached to the future path of interest rates, owing to the volatile political process to which such proposals must be subjected.

V. CONCLUDING REMARKS

This section draws some conclusions from the study and discusses their implications for Fund activities.

Conclusions of the study

While tax regimes of major industrial countries vary widely with respect to the specific rules governing treatment of interest income and expense, some subsidy to consumption that is implicit in the liberal deductibility of household interest payments is prevalent—policies of the United States being among the most liberal and those of Japan being among the least liberal. These and related tax policies may have significant effects, not only on the level and variability of interest rates and the effectiveness of monetary restraint in a given country but also on exchange rates and international capital movements. Although there has been some empirical research on these subjects, a large amount of work remains to be done. This work should help to provide a fuller understanding of the behavior of real and nominal interest rates since 1979 and should allow a better quantification of a full range of effects arising from changes in tax policy. The simulation methodology required for a comprehensive analysis of the effects of changes in tax policies is just being developed.

Despite these qualifications, it is useful to set out two basic conclusions that emerge from this study.

  • (1) Full inflation adjustment, by developed countries acting as a group, of interest income to be taxed and of interest payments to be deducted from taxable income probably constitutes the single most effective tax policy measure for lowering nominal interest rates in developed countries and for lowering effective real interest rates in developing countries. However, inflation adjustment of interest income and expense is administratively very difficult for any country. It is thus unlikely that a concerted action of this type will take place.

  • (2) Elimination of deductibility of interest payments for some or all household consumption borrowing by developed countries would lower equilibrium interest rates and the gap between effective real interest rates in developed and developing countries, but by much less than full inflation adjustment. It would, however, result in removal of a consumption subsidy, with effects on resource allocation akin to those of increases in taxes on consumption. Some enhancement of revenue and some reduction in fiscal deficits would result, depending on the share of household borrowing considered ineligible for tax deductibility.

The corollary to the first proposition is that all of its desirable aspects can also be achieved by eliminating or at least reducing inflation and with it the distortionary effects on after-tax incentives to save and invest in both developed and developing countries. Viewed in this way, the reduction in inflation and inflationary expectations in the United States from the 11–12 percent level prevalent at the end of 1980 to the 3–5 percent level prevalent during the first half of 1983 has helped a great deal to reduce distortions arising from tax treatment based on nominal instead of real interest rates. Still, it is necessary to recall that, since interest rates are set in forward-looking markets, it is the outlook for future inflation rates and related variables, coupled with the knowledge that potential distortionary effects of taxing and deducting nominal interest rates remain in most tax systems, that produces continued uncertainty regarding the outlook for effective after-tax real interest rates. In sum, it might still be desirable in the current environment to contemplate changes in tax policy aimed at reducing the potential of tax systems to magnify the impact on interest rates of further possible changes in the level of inflation. But it must be recognized that tax policy is pursued for other objectives. The effect of tax policy on interest rates has been ignored in the past, and it is not likely that it will be given much attention in the future.

Implications for Fund activities

While tax policies of most countries are determined by domestic considerations, the present study suggests that they may have important international implications as well, especially for exchange rates and international capital flows. In addition, they may also have implications for real debt service burdens of developing countries. The fund may, therefore, have a role to play in making national authorities more aware of the implications of their domestic tax policies for the international economy. The Fund’s surveillance activity and Article IV consultations provide valuable opportunities to review important tax provisions in areas such as the tax treatment of interest income and expense and the taxation of foreign exchange gains and losses

In particular, if the aims in an inflationary situation are (1) to dampen the effects of taxation on the level and volatility of interest rates and (2) to alter the redistributive impact of monetary restraint, adjustments may have to be made in the current tax treatment of interest income and payments. which typically ignores the existence of inflation. Any adjustments that may be carried out will need to ensure a “reasonable” after-tax real interest income to the lenders (i.e., will need to remove “excessive” taxation) and a “reasonable” after-tax real cost of interest payments to the borrowers (i.e., will need to remove “excessive” subsidy), without incurring severe budgetary consequences. This will call for removing many of the gaps in the tax base that exist in the form of exclusions, exemptions, and other tax incentives provided to interest income and payments and that are frequently justified on the premise that the tax system is not duly adjusted for inflation. A policy on inflation adjustment of taxable interest income and payments should go hand in hand with the elimination of special tax allowances for interest earners and interest payers. This general principle is valid not only for interest income and interest payments but also for the income and payments of all participants in the capital market and other taxpayers as well. In the final analysis, government budgets should rely more on the “fiscal dividend” of the rate of economic growth than on the rate of inflation in the economy. It is evident that the best solution is for governments 10 lower and stabilize inflation rates so that the need for the inflation adjustment of the tax system is minimized.

If the aim is to neutralize the effects of tax factors on international capital movements and exchange rates, indirect as they may be, adjustment of the tax treatment of foreign exchange gains and losses relative to the tax treatment of ordinary interest income also becomes relevant. Differences between the tax treatment of the former relative to the latter, wherever they exist, should be discouraged. Simultaneously, neutralization of these effects will also call for adoption of measures and tightening of administrative machinery in individual countries to improve the tax compliance of individuals and enterprises having income earned on international operations.

Finally, the tax treatment of interest income and payments relative to the tax treatment of other forms of capital income and payments tends to have important allocative effects. This paper has dealt with these important questions only in a very limited way and has not attempted to analyze the effects of taxes in countries (many of which are developing countries) where interest rates are regulated.

Appendix I Tax Treatment of Interest Income and Interest Payments: An International Survey

Current analyses of money market developments and the use of interest rates as instruments of economic policy tend to include one or more aspects of taxation. One example is the tendency (shown by Makin and Tanzi in their paper in Part II, Chapter 4, of this volume) of a typical tax regime for interest income to exacerbate the interest rate fluctuations induced by inflation. Another example, of a more institutional kind, is how withholding taxes imposed by some countries on interest paid to recipients abroad may affect the Eurocurrency market and similar markets. A third example, and seemingly one of growing interest, is how the interest rate affects the behavior of savers and borrowers.

Taxes may also influence interest rate structures inasmuch as certain types of interest may be exempt, wholly or partly, from tax, certain groups of investors may be exempt from tax, and (on the opposite side) certain types of interest expenditure may be deductible within certain limits or without limits.

Any analysis of these issues, based on available information on the tax systems, necessarily suffers from three serious shortcomings. One is the interpretation problem often aggravated by language difficulties, as well as the institutional differences and deviations between the law as practiced and the statutes as published. Another is the quantification problem; while some few countries, the United States among them, have detailed statistics based on representative samples of tax returns, other countries have no statistics at all or have only vague estimates of the revenue importance of particular features of their tax laws. Finally, there is the question of how national differences in tax compliance may affect the effectiveness of the provisions of the law. To take just one example: if individual interest income in a country is notoriously underreported, to a point where few if any taxpayers care to comply, a mandatory withholding tax, even if it is imposed at a relatively low rate and is final, may imply a higher effective taxation on interest income than before. This may occur even though an analysis of the legal provisions may lead the innocent reader to believe that the change meant relief.

Taxation of interest

Exemptions for certain types of income

Interest income, as a general feature, forms part of the income tax base. This is self-evident in tax systems based on a global income concept. Systems developed from schedular systems may occasionally retain a final withholding tax on certain interest income as a remnant from the schedular system.

One type of exemption is mainly a de minimis rule. In countries applying pay-as-you-earn types of income tax. it is seen as desirable to avoid deviations from the withholding tax, particularly if the system is based on the premise that the pay-as-you-earn tax is final and no tax return is necessary in the majority of cases. Similarly, interest on postal savings bank accounts and accounts in similar institutions may be exempted, usually up to some maximum level; the background may well be that the interest rate on these accounts is low and that the tax exemption somehow compensates for this while at the same time the rule simplifies the tax system. The same motive can play a role for exempting interest on savings bonds issued lo small savers.

There is, however, no clear distinction between this type of de minimis exemption, and those exemptions offered as measures to promote savings. Often the latter take the same legal form—that is, an exemption for a given amount of interest earnings (legally defined as gross earnings or net earnings after interest deduction, depending on other features of the system) or, alternatively, an exemption for interest on particular types of investment.

If it is sometimes difficult to distinguish between de minimis provisions and savings promotion schemes, the picture is further blurred by exemptions based on constitutional grounds, such as the exemption in the United States from federal income tax for interest on state and local government bonds. The constitutional issue might not be the only consideration; the exemption for interest on these bonds is more often than not seen as a means (of disputable effectiveness) of subsidizing the local governments or the activities the bonds are issued to finance.

Exemptions for certain recipients of income

More often than not, the institutional investors playing the major part in the bond market are social security institutions, pension funds, and the like, working under tax rules exempting them from tax on interest income. The exemption may be construed as a deferral rule: pension funds are allowed to be tax exempt, but beneficiaries will be taxed once they receive their pension payments. Or the institutional investor may be regarded as part of the government, exempt from tax under the somewhat dubious assumption that a government should not pay tax to itself. Other institutional investors may be insurance companies; their interest income may be subject to a lower tax than the usual corporate income tax rate under the assumption that the interest is accruing in the interest of policyholders and should be taxed at their representative average income tax rate or perhaps at a preferential rate aimed at stimulating insurance savings.

Whereas the institutional investors may benefit from preferential tax rules of this kind, they might well have to pay a price in the form of compulsory placement rules. There are many countries where the bond market, or at least the market for government bonds, is limited de facto to institutional investors that are forced by law to invest their funds in gilt-edged securities. If such rules are used to keep the market bond rate down at an artificial level, the tax exemption may be an insufficient compensation for the loss of potential investment income caused by the compulsory placement rules.

Other tax-exempt recipients of interest income are charities. Whereas tax laws often provide for income tax to be paid on business profits arrived at in competition with fully taxable subjects, there is a general tendency to allow charitable trusts and foundations to enjoy interest income without paying tax. Once again, the existence of important investors in this category tends to dilute the effects that a more general tax might have on the security market.

Finally, there are vast discrepancies in the treatment of interest paid to recipients abroad. Some countries abstain from any claim to such tax, presumably under the assumption that the tax withheld on foreign interest payments would be shifted to the domestic borrowers. Other countries impose withholding taxes on some or all interest payments, presumably under the assumption that the interest income has its source in the country and hence legitimately forms part of its tax base; there might also be a secondary argument that an exemption would make foreign borrowing too cheap and affect the relative prices of capital and labor in a direction that is undesirable, at least in countries with substantial unemployment. Attitudes toward these withholding taxes are affected by the existence of banking centers in tax havens and elsewhere that can be used to circumvent the interest withholding taxes of other countries; for competition’s sake, countries are often unwilling to introduce withholding taxes, even if it is argued that in their absence the field is open for international investors to evade tax altogether. A certain moderation in withholding taxes on interest is dictated also by another reason: it is argued that a with-holding tax levied on gross interest income may easily exceed a reasonable level if it is related to the actual net income of an investor who himself has borrowed some or all of the capital lent.

Discrimination

Interest income may also be given less lenient treatment than other income. One traditional method of doing this is by applying earned-income relief to other types of income, such as salaries and wages, but denying the same relief to interest income and other unearned income. The U.K. tradition in this respect is old. The United States formerly applied a maximum marginal tax rate of 50 percent to earned income only, while allowing the top marginal rate to go higher for unearned income. The basic effect is the same.

Other countries achieve the same purpose by net wealth taxes. In a country subjecting interest income to preferential treatment, the interest-bearing bank accounts may at the same time form part of the taxable net wealth. The former provision may be aimed at promoting savings, the latter at redistributing net wealth. The effectiveness of net wealth taxes in this respect is not great, however. Carried too far, net wealth taxes may channel savings into unproductive outlets of a type difficult to tax, such as art, jewelry, and stamps, or they may give an incentive to spend the money rather than to keep it.

Death duties also form part of redistributive taxation. For both these duties and the net wealth taxes (and the two taxes are often seen as substituting for one another), there is a conflict between the redistributive purpose of the law and the potential undesired consequences in the form of dissaving or escape into tax-exempt assets or assets difficult to tax. It is conceivable, though difficult to quantify, that the propensity to save in interest-bearing instruments may be affected by these taxes.

Deductibility of interest payments

Some concepts of taxable income include interest payments as an element of negative income, deductible from the positive income in order to arrive at the proper tax base. The argument may be one of pure theory, referring to the periodical nature of the interest payments; more often, it seems as if this type of deduction is granted with respect to the reduction of the taxpayer’s disposable income—and hence the ability to pay. Some may also have argued that, lacking tax relief for negative net wealth, debt-ridden taxpayers should at least have the benefit of a deduction for interest.

Other concepts of taxable income put interest payments on a level with other expenses. In other words, a deduction for interest is granted when the debt has been incurred to acquire or maintain an incomegenerating investment and is refused when the debt has been incurred to cover costs of living or investment in assets not generating any income. The implications of this approach vary considerably with what is understood by “income-generating” investment. Typically, the most important deduction for individual taxpayers is mortgage interest. If the imputed income of the taxpayer’s own residence is assessed as an income item, the residential investment is obviously income generating and the mortgage interest should be deductible, according to the basic principle. However, even in many countries assessing imputed income of this kind, it is felt that the assessment value is not equal to the market value. Hence, a limit might apply to the mortgage interest deductible, either at a level preventing the home investment from rendering a deductible net deficit or at a level representing a maximum for what is seen as normal. It is noteworthy that a country such as the United Slates, while not taxing imputed income of residences, offers un limited deduction for mortgage interest, whereas several of the European countries that impose income tax on imputed income nevertheless apply limitations to the deduction for mortgage interest.

A common problem in applying this type of rule is the proper allocation of loan interest to the corresponding investment. A rule attaching the debt to the collateral and refusing a deduction for home mortgage interest would prevent a homeowner from deducting interest paid on a mortgage taken up, say, to finance his business. On the other hand, requirement of specific information about the original purpose of a loan, while often possible, might lead to incorrect conclusions if the loan was originally taken to finance an income-generating investment but repayment has been postponed in order to use the money for other purposes. Finally, it is often argued that the nondeductibility of interest on consumption loans or mortgages discriminates against those of limited means, who must borrow, and favors those who can draw upon their income-generating assets in order to finance their consumption or their home investment. The former get no tax relief; the latter enjoy a reduction in income tax.

When the debtor is a corporation, the deduction for interest payments creates a considerable difference in treatment of the yield of an investment between equity and debt capital. Dividends are very rarely a deductible item for corporation tax; on the contrary, there is often a double discrimination against dividends, inasmuch as there is not only no deduction for the dividends but they are often also subject to a higher withholding tax, when paid to foreign recipients, than interest payments. With regard to this preferential treatment of debt, many countries have therefore been compelled to set up rules against “thin capitalization,” treating loans from shareholders or from creditors affiliated with them as disguised equity, the interest on which should be treated as dividends.

Interest income and capital gains

Most countries offer either tax exemption or a more favorable tax treatment for capital gains than for ordinary income. The definition of what is a capital gain as distinct from ordinary income varies, however, with a tendency (or the capital gains concept to be wider in countries taxing capital gains while the concept of current income tends to be wider in countries exempting capital gains.

One such definitional problem concerns interest. Inasmuch as bonds are issued at a discount, the effective interest may exceed the nominal interest. In some countries, issuers of bonds have used the capital gains tax treatment for bond discounts to obtain a favorable tax treatment. Legislative measures have been taken to stop this type of abuse by redefining interest to include bond discounts. There is still a problem, however, insofar as bonds are marketable and can be sold to some exempt entity at a gain representing the discount; if the gain on the sale is defined as a capital gain, the transaction may be a successful avoidance scheme. As is usual in this area, legislation reflects taxpayers’ inventiveness only after some delay.

Interest income and foreign exchange gains

If a loan transaction affects two countries with different currencies, there is the possibility that repayment is offered at a different exchange rate than that prevailing at the time the credit was made. Depending on whether the creditor or the debtor has given or taken up the credit in a foreign currency, be may stand to gain or lose, respectively, depending on the foreign exchange rate.

In a great number of countries, the tax rules governing foreign exchange gains and losses are in a state of flux. First of all, there may be a difference between credits taken up or given in the course of a taxpayer’s business and nonbusiness investment transactions. Second, even within The business sector, some countries recognize a capital gain and loss sector, and there have been cases in which a court has found a taxpayer taxable on his foreign exchange gain on lending in a foreign currency that has appreciated, while at the same time be has been deprived of a compensating loss deduction for corresponding borrowing in the same currency! Even if this type of incongruity is rare, it is still conceivable that borrowing for investment purposes is regarded as a capital transaction, the foreign exchange loss being a capital loss, with the restrictions applied to its deductibility. In this respect, a distinction is often made between different lines of business.

Even in those cases where the business character of foreign exchange gains and losses is recognized, timing may be a crucial factor. The traditional accounting principles may prescribe a recognition of losses while at the same time prohibiting the immediate accounting for unrealized gains; this kind of incongruity has, however, caught the attention of the authorities in several countries, and efforts are under way to establish accounting rules satisfactory to both the accounting profession and the tax authorities. At present, it is difficult to say what the prevailing rules are, even for one particular country. An international comparison of the tax treatment of foreign exchange gains and losses is a forbiddingly complicated subject.

Appendix II Financial Sector Framework for Determination of Interest Rates

Expression for the level of interest rates

This simplified framework equates borrowing and lending, each determined by expected after-tax real interest rates and solves for a nominal interest rate, i, in terms of an expected after-tax real rate, r*, and expected inflation, π.

The negative impact of expected inflation on the borrowing schedule (indicated below by the negative λ1 term) reflects a depressing effect of expected inflation on the level of investment—an effect that arises in turn from a number of sources, some of which are tied to tax policy. First, “excess” taxes result from the use of historical cost depreciation and first-in, first-out inventory valuation in an inflationary environment. With such methods in use, as they are in the United States, a rise in expected inflation results in a foreseeable reduction in after-tax profits, thereby depressing investment. Second, higher expected inflation has been found in the United States to be associated with elevated uncertainty about relative prices. In turn, relative price uncertainty results in reduced investment and a negative shift in the borrowing schedule since most capital is not adaptable to a multitude of uses. Investment really represents an increased commitment to a given set of relative prices and is therefore made more risky by increased uncertainty about relative prices. Finally, if a rise in expected inflation depresses the equilibrium stock of desired money balances, a negative wealth effect requires a lower equilibrium level of the after-tax real interest rate. These effects are discussed, along with others, in Section III of the paper.

With these considerations in mind, lending and borrowing schedules are written as

where β0 > α0 and

Setting equation (4) equal to equation (5) and solving for it gives

The simple, tax-adjusted Fisher equation can be derived from equation (6) by setting τL = τB = τ and

With these conditions satisfied and where (α1 + β1) = 1.0

Expression for the volatility of interest rates

Equation (7) implies an expression for the variance of i (given λ1 = τ):

As (0 < τ < 1) implies [1/(1 – τ)]2 > 1.0, the effect of taxation of interest income and deduction of interest expense is to magnify the impact on the variance of nominal interest, σi2, of changes in the variance of r*,σr*2. The variance of it is likely to be reduced by the negative covariance between r* andπ,ρr*π<0, which arises in connection with wealth, tax, and uncertainty effects.

Appendix III
Table 3.Selected Industrial Countries: Aspects of Tax Treatment of Interest Income
Tax Exemptions of Interest IncomeWithholding Tax on Interest Income of Residents1
CountryTax-exempt

debt instruments1
Tax-exempt

recipients1
General tax

exemp-

tions
Whether

interest

subject

to with-

holding
With-

holding

tax rates

(in percent)
Debt instruments

whose interest

earnings are not

subject to withholding
AustraliaCertain government securities issued prior to 1968 Certain securities issued abroadMunicipal corporations; certain “trustee” savings banks; superannuation funds; life insurance company superannuation business; various categories of associations that are tax-exempt on all income; credit unions (on income derived from their members); nonresidents (in part)2NoneNoNilAll debt instruments
AustriaCertain government bonds (in part)3 Savings deposits (in part)3State railways and monopolies; the national bank; postal savings bank; retirement and pension funds; nonresidents (in part)2Yes20Corporate bonds; government bonds; ordinary loans; mortgage loans;4 bank deposits
BelgiumCertain government loans Certain government-guaranteed debt issues5 Saving’s deposits5 Investments in cooperatives5 Interest income in general (final withholding tax)5Some social bodies;6 nonresidents (in part)2Yes25Mortgage loans4
CanadaMunicipal and other bodies; public enterprises;7 trusts and corporations for

(a) profit-sharing plans,

(b) pensions and retirement savings,

(c) home-ownership savings,

(d) education plans
Up to Can$1,000 (US$815)NoNilAll debt instruments
DenmarkCompulsory savings instituted by lawCooperatives; building societies; certain banks; public utilities; certain superannuation funds;8 mutual insurance societies; nonresidentsUp to DKr 3,500 (US$600) to persons over the age of 679NoNilAll debt instruments
FranceCertain government bonds Savings bank deposits (in part)10 Certain bank deposits in foreign currency Special deposits10 Long-term savings schemes10Agricultural and mutual credit funds; cooperative building societies; investment companies; pension and superannuation funds (in part);11 nonresidents (in part)2F 3,000 (US$610) on French fixed-interest bondsYes10–2512State government bonds; ordinary loans; mortgage loans;4 bank deposits
Germany, Fed. Rep.Certain federal government bonds13 Home-ownership savings13 Employee savings schemes13Federal railways; postal service; certain banks and financial institutions; pension funds; nonresidents (in part)2DM 800 (US$340)14Yes25Corporate bonds; government bonds; ordinary loans; mortgage loans; bank deposits
IrelandPost office savings bank deposits (in part)15 Commercial bank deposits (in part)15 National savings bonds16 Public sector bonds (govt., local authority and some public enterprise)16Approved superannuation funds; savings banks; National Insurance Funds (health and unemployment); nonresidents (in part)2Yes35Bank deposits; profit-sharing bonds
ItalyAll government and nationalized enterprises’ bonds17 Certain post office bonds Some companies’ debt instruments (usually according to particular times of issue)18Corporations (including credit institutions) for holdings of public debt (in part); certain small cooperatives; Mezzogiorno Fund (in part); credit institutions for some loans to local authorities; nonresidents (in part)2Only on nonexempt bonds, deposits, and some other debt instruments18,1912.5, 15, 25Government bonds; selected other instruments
JapanBank and savings deposits20 Certain bonds and debentures20 Certain central or local government bonds Savings for formation of employees’ assets20Local governments; fully government-owned corporations (railways, etc.); qualified nonprofit corporationsYes1920, 35None
NetherlandsNonePublic utilities undertakings; all pension funds; unemployment and sickness insurance funds; investment organizations; nonresidents (in part)2f. 700 (US$270)21NoNilAll debt instruments except those whose income is regarded as dividends, such as debt claims participating in profits
NorwayDeposits with domestic banks22 Deposits with certain savings institutions22 Government savings bonds22 Bonds issued by debtors22Cooperatives; mutual insurance companies; building societies; savings banks (but subject to municipal tax); nonresidentsNoNilAll debt instruments
Sweden23Local authorities; pension funds; benefit (death, unemployment compensation, etc.) societies; nonresidentsMarried couple: SKr 1,600 (US$275)23 Single person: SKr 800 (US$138)23NoNilAll debt instruments
SwitzerlandSavings deposits24Social security; compensation and staff welfare or provident funds; transportation enterprises; nonresidents (in part)24Yes35Ordinary loans; mortgage loans
United KingdomNational savings certificates National savings bank deposits (in part)25 Certain stocks and loan issues26 Contractual savings, SAYE26Local authorities; savings banks; issuing departments of certain foreign central banks;27 approved superannuation funds; nonresidents (in part)2Yes30Bank deposits; profit-sharing bonds;28
United StatesState and municipal government bonds Savings certificates of certain depository institutions (in part)29Pension plans; farmers’ cooperatives; nonresidents (in part)2US$40030No31NilAll debt instruments31
Sources: United Kingdom, Income Taxes Outside the United Kingdon (London: H.M. Stationery Office), loose-leaf service; Tax Management, Foreign Income Portfolios (Washington), loose-leaf service relating to Austria, France, the Federal Republic of Germany, Italy, the Netherlands, and Switzerland; Price Waterhouse, Information Guide on doing business in the countries surveyed, except Austria (New York), latest editions; for Sweden and Switzerland: Harvard Law School, World Tax Series (Chicago: Commerce Clearing House, 1959 and 1976, respectively); for Japan: Yuji Gomi, Guide to Japanese Taxes (Tokyo: Zaikei, 1981); for Australia: E.F. and J.E. Mannix, Australian Income Tax (Sydney: Butterworths, 1981); International Bureau of Fiscal Documentation, The Taxation of Private Investment Income, Guides to European Taxation, Vol. 3 (Amsterdam), loose-leaf service. The rates of exchange of June 1982 (or the nearest month) published in International Financial Statistics, International Monetary Fund (Washington), are applied to derive the U.S. dollar equivalents in the table and footnotes.

Tax exemption is defined broadly here. It also includes preferential tax treatment, such as partial tax exemption, reduced rate of tax, refund of tax, and tax credit. The “tax-exempt” recipients tabulated here are some of the major categories of recipients listed in the income tax codes of these countries and are (in addition to the charitable, social, and religious bodies) institutions of scientific research, cooperative societies, agrarian collectives, organizations for low-cost housing, etc., which, too, are generally exempt in most of these countries. Neither the list of tax-exempt debt instruments nor that of tax-exempt recipients is intended to be exhaustive. Withholding tax on interest is subject to tax credit in all countries except Italy and Japan (see footnote 19).

The categories of interest income of nonresidents to which the exemption relates are shown in Table 5.

Tax exemption applies to bond issues (including mortgage bonds) and local government loans of up to S 100,000 ($6,050) and to interest on savings deposits of up to S 7,000 ($425) per taxpayer in each case.

But not mortgage bonds, which are subject to withholding.

Guaranteed debt relates to some exceptional loans by public enterprises (telegraph and telephone, railways). Savings deposit interest of up to BF 50,000 ($1,100) and interest received from cooperatives up to BF 5,000 ($110) are exempt. Interest of up to BF 10,000 ($220) is exempt if the total taxable income does not exceed BF 350,000 ($7,700). Interest income in general is subject to a final withholding tax insofar as certain limits are not exceeded (limit for interest income: BF 316,000 ($6,952); limit for interest and/or dividend income: BF 1,110,000 ($24,420)). Above these limits, a supplementary progressive tax is levied, ranging from 20 percent to a maximum of 47 percent when interest and/or dividend income exceeds BF 3 million ($66,000).

Exemption is in the form of a reduced rate of tax.

Minimum required government ownership is 90 percent.

In 1984, a tax is payable on real incomes of certain pension funds (in excess of 3.5 percent return) at the rate of 40.5 percent.

Thirty percent of interest income of taxpayers with incomes below DKr 50,500 ($8,620), maximum available deduction being DKr 3,500 ($600) per annum. For taxpayers with higher incomes, tax-exempt interest is reduced by DKr 50 ($8.50) for each DKr 100 ($17) in excess of the specified limit (DKr 50,500 or $8,620), which is adjustable annually in line with the price index.

Limited to a maximum savings deposit of F 32,500 ($5,320); only one third of interest on special deposit with the nonagricultural mutual credit institutions is subject to income tax; limit in regard to the amount of long-term savings is up to F 20,000 ($3,310) or one fourth of the taxpayer’s income, whichever is higher.

At a reduced tax rate of 10 percent.

Lower rate applies to issues after January 1, 1965.

Mainly issues after World War II (1952–55); includes mortgage bonds and municipal government debentures. Preferential treatment in respect of home ownership and employee savings schemes is extended through the payment of limited amounts of tax-free bonuses by government to certain categories of taxpayers participating in these schemes.

This exemption from tax applies only on income (joint return) from movable capital.

Interest on commercial bank deposits up to £lr 50 ($63) and from savings banks up to £lr 120 ($150), with a total ceiling of £lr 120 ($150).

Exemption usually extends to bonds held by persons not ordinarily resident in Ireland.

As expressly provided in the laws sanctioning their issue.

Under various laws for encouraging sectoral or regional investment in agriculture, mining, industry, and low-cost housing.

Withholding tax on interest is final in Italy and Japan—in Italy, on all debt instruments except ordinary and mortgage loans and in Japan whenever the taxpayer has opted for a final withholding tax at the rate of 35 percent.

Interest exemption applies to interest received on bank deposits of up to ¥ 3 million ($11,950), on postal savings deposits of up to ¥3 million ($11,950), on national or local government bonds up to ¥3 million ($11,950), and on employee savings of up to ¥ 5 million ($19,920) deposited with banks or security dealers under a contract, the total tax-exempt savings being ¥ 14 million ($55,770).

Available only on excess of interest receipts over interest payments, not on gross receipts.

Combined interest income from only these sources is exempt up to NKr 4,000 ($660) on joint return and half this amount on single return.

Certain savings deposits, up to a maximum yearly savings of SKr 7,200 ($1,240) and maximum total savings of SKr 30,000 ($5,165), are exempt. Most government bonds bought by private investors in Sweden are premium (non-interest-bearing) bonds, winnings on which attract a flat-rate lottery tax of 20 percent.

Exemption is from the income tax imposed by cantons (not the federal government) as provided in their respective legislation. Nonresidents are not subject to federal income tax except by way of withholding tax on bonds, debentures, and interest on bank deposits, which is generally final (not refundable).

Up to £ 70($105) a year.

The exempt interest relates only to payments received by persons not ordinarily resident in the United Kingdom. SAYE is a “save as you earn” scheme under which employees participate in a scheme to acquire shares at 70 percent of their value by contributing £20 ($36) per month to the scheme; indexed part of interest income is tax free.

India and Pakistan.

Interest on such instruments is treated as a cash dividend.

Applies to tax-free savings certificates issued by certain depository financial institutions (banks, thrift institutions, and credit unions) between October 1, 1981 and December 31, 1982. Limit on interest earnings on a joint return is $2,000.

This exclusion of dividends and interest income on joint returns for 1981 is now reduced to $200 and relates only to dividends. Starting in 1985, taxpayers will be able to exclude 15 percent of the net interest income up to $450, net income being net of nonbusiness, nonhome mortgage interest payments.

Effective July 1, 1983 all interest earnings other than those specifically exempted from withholding, e.g., minimal interest payments (below $150), certain payments by qualified cooperatives, and payments to exempt individuals and institutions, are subject to a 10 percent withholding tax.

Sources: United Kingdom, Income Taxes Outside the United Kingdon (London: H.M. Stationery Office), loose-leaf service; Tax Management, Foreign Income Portfolios (Washington), loose-leaf service relating to Austria, France, the Federal Republic of Germany, Italy, the Netherlands, and Switzerland; Price Waterhouse, Information Guide on doing business in the countries surveyed, except Austria (New York), latest editions; for Sweden and Switzerland: Harvard Law School, World Tax Series (Chicago: Commerce Clearing House, 1959 and 1976, respectively); for Japan: Yuji Gomi, Guide to Japanese Taxes (Tokyo: Zaikei, 1981); for Australia: E.F. and J.E. Mannix, Australian Income Tax (Sydney: Butterworths, 1981); International Bureau of Fiscal Documentation, The Taxation of Private Investment Income, Guides to European Taxation, Vol. 3 (Amsterdam), loose-leaf service. The rates of exchange of June 1982 (or the nearest month) published in International Financial Statistics, International Monetary Fund (Washington), are applied to derive the U.S. dollar equivalents in the table and footnotes.

Tax exemption is defined broadly here. It also includes preferential tax treatment, such as partial tax exemption, reduced rate of tax, refund of tax, and tax credit. The “tax-exempt” recipients tabulated here are some of the major categories of recipients listed in the income tax codes of these countries and are (in addition to the charitable, social, and religious bodies) institutions of scientific research, cooperative societies, agrarian collectives, organizations for low-cost housing, etc., which, too, are generally exempt in most of these countries. Neither the list of tax-exempt debt instruments nor that of tax-exempt recipients is intended to be exhaustive. Withholding tax on interest is subject to tax credit in all countries except Italy and Japan (see footnote 19).

The categories of interest income of nonresidents to which the exemption relates are shown in Table 5.

Tax exemption applies to bond issues (including mortgage bonds) and local government loans of up to S 100,000 ($6,050) and to interest on savings deposits of up to S 7,000 ($425) per taxpayer in each case.

But not mortgage bonds, which are subject to withholding.

Guaranteed debt relates to some exceptional loans by public enterprises (telegraph and telephone, railways). Savings deposit interest of up to BF 50,000 ($1,100) and interest received from cooperatives up to BF 5,000 ($110) are exempt. Interest of up to BF 10,000 ($220) is exempt if the total taxable income does not exceed BF 350,000 ($7,700). Interest income in general is subject to a final withholding tax insofar as certain limits are not exceeded (limit for interest income: BF 316,000 ($6,952); limit for interest and/or dividend income: BF 1,110,000 ($24,420)). Above these limits, a supplementary progressive tax is levied, ranging from 20 percent to a maximum of 47 percent when interest and/or dividend income exceeds BF 3 million ($66,000).

Exemption is in the form of a reduced rate of tax.

Minimum required government ownership is 90 percent.

In 1984, a tax is payable on real incomes of certain pension funds (in excess of 3.5 percent return) at the rate of 40.5 percent.

Thirty percent of interest income of taxpayers with incomes below DKr 50,500 ($8,620), maximum available deduction being DKr 3,500 ($600) per annum. For taxpayers with higher incomes, tax-exempt interest is reduced by DKr 50 ($8.50) for each DKr 100 ($17) in excess of the specified limit (DKr 50,500 or $8,620), which is adjustable annually in line with the price index.

Limited to a maximum savings deposit of F 32,500 ($5,320); only one third of interest on special deposit with the nonagricultural mutual credit institutions is subject to income tax; limit in regard to the amount of long-term savings is up to F 20,000 ($3,310) or one fourth of the taxpayer’s income, whichever is higher.

At a reduced tax rate of 10 percent.

Lower rate applies to issues after January 1, 1965.

Mainly issues after World War II (1952–55); includes mortgage bonds and municipal government debentures. Preferential treatment in respect of home ownership and employee savings schemes is extended through the payment of limited amounts of tax-free bonuses by government to certain categories of taxpayers participating in these schemes.

This exemption from tax applies only on income (joint return) from movable capital.

Interest on commercial bank deposits up to £lr 50 ($63) and from savings banks up to £lr 120 ($150), with a total ceiling of £lr 120 ($150).

Exemption usually extends to bonds held by persons not ordinarily resident in Ireland.

As expressly provided in the laws sanctioning their issue.

Under various laws for encouraging sectoral or regional investment in agriculture, mining, industry, and low-cost housing.

Withholding tax on interest is final in Italy and Japan—in Italy, on all debt instruments except ordinary and mortgage loans and in Japan whenever the taxpayer has opted for a final withholding tax at the rate of 35 percent.

Interest exemption applies to interest received on bank deposits of up to ¥ 3 million ($11,950), on postal savings deposits of up to ¥3 million ($11,950), on national or local government bonds up to ¥3 million ($11,950), and on employee savings of up to ¥ 5 million ($19,920) deposited with banks or security dealers under a contract, the total tax-exempt savings being ¥ 14 million ($55,770).

Available only on excess of interest receipts over interest payments, not on gross receipts.

Combined interest income from only these sources is exempt up to NKr 4,000 ($660) on joint return and half this amount on single return.

Certain savings deposits, up to a maximum yearly savings of SKr 7,200 ($1,240) and maximum total savings of SKr 30,000 ($5,165), are exempt. Most government bonds bought by private investors in Sweden are premium (non-interest-bearing) bonds, winnings on which attract a flat-rate lottery tax of 20 percent.

Exemption is from the income tax imposed by cantons (not the federal government) as provided in their respective legislation. Nonresidents are not subject to federal income tax except by way of withholding tax on bonds, debentures, and interest on bank deposits, which is generally final (not refundable).

Up to £ 70($105) a year.

The exempt interest relates only to payments received by persons not ordinarily resident in the United Kingdom. SAYE is a “save as you earn” scheme under which employees participate in a scheme to acquire shares at 70 percent of their value by contributing £20 ($36) per month to the scheme; indexed part of interest income is tax free.

India and Pakistan.

Interest on such instruments is treated as a cash dividend.

Applies to tax-free savings certificates issued by certain depository financial institutions (banks, thrift institutions, and credit unions) between October 1, 1981 and December 31, 1982. Limit on interest earnings on a joint return is $2,000.

This exclusion of dividends and interest income on joint returns for 1981 is now reduced to $200 and relates only to dividends. Starting in 1985, taxpayers will be able to exclude 15 percent of the net interest income up to $450, net income being net of nonbusiness, nonhome mortgage interest payments.

Effective July 1, 1983 all interest earnings other than those specifically exempted from withholding, e.g., minimal interest payments (below $150), certain payments by qualified cooperatives, and payments to exempt individuals and institutions, are subject to a 10 percent withholding tax.

Table 4.Selected Industrial Countries: Tax Deductibility of Interest Expense of Individuals
Owner-Occupied HousingIncome-Generating

Investment
General Consumer

Credit
CountryWhether imputed

income taxable
Whether deductibility relates to

interest payments, principal (or

equivalent savings), or both
Whether

deductibility

limited
Whether interest

payments deductible
Whether interest

payments deductible
AustraliaNoNone1YesNo
AustriaNoPrincipal and interest2YesYesNo
BelgiumYes3Principal and interest3Yes3Yes3No
CanadaNoPrincipal and interest4Yes4Yes4No
DenmarkYesInterestNoYesYes
FranceNoInterest5Yes5Yes6No
Germany, Fed. Rep.YesSavings or interest7YesYesNo
IrelandNoInterest8Yes8Yes8No8
ItalyYes9InterestYes9Yes10No
JapanNoPrincipal and interest11Yes11YesNo
NetherlandsYesInterestNoYesYes
NorwayYesInterestNoYesYes
SwedenYesInterest12Yes12Yes12Yes12
SwitzerlandYesInterestNoYesYes
United KingdomNoInterestYes13Yes14No
United StatesNoInterestNoYes15Yes
Sources: United Kingdom, Income Taxes Outside the United Kingdom (London: H.M. Stationery Office), loose-leaf service; Price Waterhouse, Information Guide on doing business in various countries under review (New York), latest editions; Tax Management, Foreign Income Portfolios (Washington), loose-leaf service; Yuji Gomi, Guide to Japanese Taxes (Tokyo: Zaikei, 1981). Exchange rates for June 1982 (or the nearest month) published in International Financial Statistics, International Monetary Fund (Washington), are applied to derive the U.S. dollar equivalents in the footnotes.

Mortgage interest deduction ceased to apply to interest accruing after November 1, 1978.

Since 1980, the contribution includes interest on borrowing connected with the building of a house up to a maximum of S 20,000 ($1,215) for taxpayer and spouse and S 5,000 ($303) for each child. Taxpayers are further entitled to receive an allowance for “extraordinary burden” (S 15,000, or $910, on joint return) arising from home acquisition in the first year.

The occupation of his own property by the owner is treated as taxable income from real estate, but an exemption is given up to BF 120,000 ($2,840), to which BF 10,000 ($220) is added for each dependent. Further, if the exemption does not exhaust the taxable amount, this amount is reduced to half when the total taxable income does not exceed BF 950,000 ($20,900). When the total taxable income exceeds BF 950,000 ($20,900), the imputed income may not be raised by more than half of the difference between BF 950,000 ($20,900) and the total taxable income. Mortgage interest payments are tax deductible to the extent that the deduction does not exceed the total income derived from immovable property and from interest and/or dividends. The principal is only tax deductible when the loan relates to a social property or to private property having a market value below BF 2,500,000 ($55,000) (purchase) or BF 2,700,000 ($59,400) (building), to which 5 percent may be added for each dependent or 25 percent when the property is partly for professional use. The deductibility of the principal is subject to different conditions and limitations. The amount of the deductible principal is subject to different conditions and limitations. The amount of the deductible principal is also limited depending on the taxable professional income and may not exceed a limit of BF 45,000 ($990), a limit to which deductible insurance premiums must also be imputed, but in addition a tax deduction of BF 11,250 ($245) or BF 13,500 ($297) is possible. Supplementary incentives are granted, such as supplementary interest deductibility up to BF 200,000 ($4,400) spread over three years for new construction, exemption of taxation on imputed income from owner-occupied property in 1984–95 for new construction, and limited deductibility of expense for renovation of buildings.

Deductibility relates to annual contributions of up to Can$1,000 ($815) for a lifetime maximum of Can$10,000 ($8,150) paid to a Registered Home Ownership Savings Plan (RHOSP) for only one residential property, to be limited to 20 years. Interest on income-generating investment is tax deductible only to the extent of income generated. However, such restriction is not applicable to investment in equity of a private company and to borrowing incurred prior to the 1982 budget for acquiring rental property.

On interest for the purchase, construction, and major repairs of a principal residence at the rate of F 7,000 ($1,145) plus F 1,000 ($165) per dependent, for a maximum of 10 years.

The deduction against real estate income is on the entire amount of mortgage interest. On income from movable property the deduction relates to the purchase of certain domestic financial assets—for example, shares in investment companies, officially quoted domestic securities (mainly company shares), or interest in a private company.

Housing is usually acquired by means of savings with building associations first and mortgage loans from it thereafter. Home ownership savings contributions are tax deductible, provided taxpayer does not claim housing bonus thereon. Such contributions are, however, a part of “special expenditures,” which are tax deductible only up to maximum amounts specified in the tax law. Interest on subsequent borrowing from the loan associations is tax deductible but only up to the amount of imputed income of the house to the owner, the latter being in practice considerably lower than the former.

Except in respect of interest connected with rental income and interest on loans to pay death duties, deductibility of interest payments for all other purposes (currently under review) is restricted to £lr 2,400 ($3,550) per annum on single returns and double that amount on joint returns. However, effective March 25, 1982, the relief will apply only to interest on mortgage loans for the purchase, repair, or improvement of sole or main residence, and on an interim basis (through 1984/85) on loans then in existence. For new loans thereafter, interest on general consumer credit will be deductible only on loans within specified limits of £lr 25,000 ($37,000) and £lr 5,000 ($7,400) for couples. Deductibility of interest on income-generating loans is now available only for trading and rental income and for investment in companies engaged in such activities and in professional partnerships.

Notional income for each cadastral unit represents imputed income. Interest of up to a maximum of Lit 4 million ($3,080) paid on mortgage loans.

The deductibility is limited to the total amount of interest payments multiplied by the ratio of taxable to total income (the latter inclusive of 90 percent of the tax-exempt interest income).

Two schemes are in existence: (1) an employee contributing to a savings scheme for residential housing for at least three years gets a tax credit of up to ¥50,000 ($205) a year; and (2) an annual tax credit equal to 7 percent of the repayment of borrowing with a limit of ¥ 17,000 ($70) to ¥ 30,000 ($125) (higher amount is available where loan is raised by a formal loan agreement) for three years, based on an income mean test (annual income of ¥ 8 million, or $33,000, and below). Scheme (1) is being abolished under 1982 Tax Reform, with reliefs for houses acquired prior to April 1982.

Interest deductions, available on all borrowings, are restricted so as not to reduce the tax by more than 50 percent even if the marginal tax rate is higher.

Interest deductions are limited to a mortgage of £25,000 ($44,750) for acquiring a dwelling and for improvements thereto.

Subject to one of four conditions: (1) on commercial property, it must be on rent for at least six months in a year; (2) taxpayer applying proceeds of borrowing for acquiring a share in an enterprise must either have a material interest in it or act personally in the conduct of trade; (3) proceeds of borrowing are being applied to pay transfer tax or purchase a life annuity; and (4) borrowing is for purchasing or improving land.

Interest on borrowings that are applied to the acquisition of tax-exempt investments (e.g., state and local government bonds) are not tax deductible. The deductibility of interest on all other borrowings is limited to the investment income generated. To the extent that such interest exceeds $10,000 (raised to $15,000 if investor is seeking to increase his minority stockholding in an enterprise to a majority), the excess may be carried forward and may be deducted in the subsequent year subject to the same limitations as in the initial year.

Sources: United Kingdom, Income Taxes Outside the United Kingdom (London: H.M. Stationery Office), loose-leaf service; Price Waterhouse, Information Guide on doing business in various countries under review (New York), latest editions; Tax Management, Foreign Income Portfolios (Washington), loose-leaf service; Yuji Gomi, Guide to Japanese Taxes (Tokyo: Zaikei, 1981). Exchange rates for June 1982 (or the nearest month) published in International Financial Statistics, International Monetary Fund (Washington), are applied to derive the U.S. dollar equivalents in the footnotes.

Mortgage interest deduction ceased to apply to interest accruing after November 1, 1978.

Since 1980, the contribution includes interest on borrowing connected with the building of a house up to a maximum of S 20,000 ($1,215) for taxpayer and spouse and S 5,000 ($303) for each child. Taxpayers are further entitled to receive an allowance for “extraordinary burden” (S 15,000, or $910, on joint return) arising from home acquisition in the first year.

The occupation of his own property by the owner is treated as taxable income from real estate, but an exemption is given up to BF 120,000 ($2,840), to which BF 10,000 ($220) is added for each dependent. Further, if the exemption does not exhaust the taxable amount, this amount is reduced to half when the total taxable income does not exceed BF 950,000 ($20,900). When the total taxable income exceeds BF 950,000 ($20,900), the imputed income may not be raised by more than half of the difference between BF 950,000 ($20,900) and the total taxable income. Mortgage interest payments are tax deductible to the extent that the deduction does not exceed the total income derived from immovable property and from interest and/or dividends. The principal is only tax deductible when the loan relates to a social property or to private property having a market value below BF 2,500,000 ($55,000) (purchase) or BF 2,700,000 ($59,400) (building), to which 5 percent may be added for each dependent or 25 percent when the property is partly for professional use. The deductibility of the principal is subject to different conditions and limitations. The amount of the deductible principal is subject to different conditions and limitations. The amount of the deductible principal is also limited depending on the taxable professional income and may not exceed a limit of BF 45,000 ($990), a limit to which deductible insurance premiums must also be imputed, but in addition a tax deduction of BF 11,250 ($245) or BF 13,500 ($297) is possible. Supplementary incentives are granted, such as supplementary interest deductibility up to BF 200,000 ($4,400) spread over three years for new construction, exemption of taxation on imputed income from owner-occupied property in 1984–95 for new construction, and limited deductibility of expense for renovation of buildings.

Deductibility relates to annual contributions of up to Can$1,000 ($815) for a lifetime maximum of Can$10,000 ($8,150) paid to a Registered Home Ownership Savings Plan (RHOSP) for only one residential property, to be limited to 20 years. Interest on income-generating investment is tax deductible only to the extent of income generated. However, such restriction is not applicable to investment in equity of a private company and to borrowing incurred prior to the 1982 budget for acquiring rental property.

On interest for the purchase, construction, and major repairs of a principal residence at the rate of F 7,000 ($1,145) plus F 1,000 ($165) per dependent, for a maximum of 10 years.

The deduction against real estate income is on the entire amount of mortgage interest. On income from movable property the deduction relates to the purchase of certain domestic financial assets—for example, shares in investment companies, officially quoted domestic securities (mainly company shares), or interest in a private company.

Housing is usually acquired by means of savings with building associations first and mortgage loans from it thereafter. Home ownership savings contributions are tax deductible, provided taxpayer does not claim housing bonus thereon. Such contributions are, however, a part of “special expenditures,” which are tax deductible only up to maximum amounts specified in the tax law. Interest on subsequent borrowing from the loan associations is tax deductible but only up to the amount of imputed income of the house to the owner, the latter being in practice considerably lower than the former.

Except in respect of interest connected with rental income and interest on loans to pay death duties, deductibility of interest payments for all other purposes (currently under review) is restricted to £lr 2,400 ($3,550) per annum on single returns and double that amount on joint returns. However, effective March 25, 1982, the relief will apply only to interest on mortgage loans for the purchase, repair, or improvement of sole or main residence, and on an interim basis (through 1984/85) on loans then in existence. For new loans thereafter, interest on general consumer credit will be deductible only on loans within specified limits of £lr 25,000 ($37,000) and £lr 5,000 ($7,400) for couples. Deductibility of interest on income-generating loans is now available only for trading and rental income and for investment in companies engaged in such activities and in professional partnerships.

Notional income for each cadastral unit represents imputed income. Interest of up to a maximum of Lit 4 million ($3,080) paid on mortgage loans.

The deductibility is limited to the total amount of interest payments multiplied by the ratio of taxable to total income (the latter inclusive of 90 percent of the tax-exempt interest income).

Two schemes are in existence: (1) an employee contributing to a savings scheme for residential housing for at least three years gets a tax credit of up to ¥50,000 ($205) a year; and (2) an annual tax credit equal to 7 percent of the repayment of borrowing with a limit of ¥ 17,000 ($70) to ¥ 30,000 ($125) (higher amount is available where loan is raised by a formal loan agreement) for three years, based on an income mean test (annual income of ¥ 8 million, or $33,000, and below). Scheme (1) is being abolished under 1982 Tax Reform, with reliefs for houses acquired prior to April 1982.

Interest deductions, available on all borrowings, are restricted so as not to reduce the tax by more than 50 percent even if the marginal tax rate is higher.

Interest deductions are limited to a mortgage of £25,000 ($44,750) for acquiring a dwelling and for improvements thereto.

Subject to one of four conditions: (1) on commercial property, it must be on rent for at least six months in a year; (2) taxpayer applying proceeds of borrowing for acquiring a share in an enterprise must either have a material interest in it or act personally in the conduct of trade; (3) proceeds of borrowing are being applied to pay transfer tax or purchase a life annuity; and (4) borrowing is for purchasing or improving land.

Interest on borrowings that are applied to the acquisition of tax-exempt investments (e.g., state and local government bonds) are not tax deductible. The deductibility of interest on all other borrowings is limited to the investment income generated. To the extent that such interest exceeds $10,000 (raised to $15,000 if investor is seeking to increase his minority stockholding in an enterprise to a majority), the excess may be carried forward and may be deducted in the subsequent year subject to the same limitations as in the initial year.

Table 5.Selected Industrial Countries: Aspects of Tax Treatment of International Flows of Interest Income
Foreign Interest Income of Resident IndividualsInterest Payments to Nonresident Individuals
CountryTaxable interest gross or net of tax paid in source countryType of double taxation relief availableWithholding taxes

applicable1
Tax exempt

interest,

if any
AustraliaGross2Foreign tax credit10On certain bearer bonds
AustriaNetForeign tax credit (in part) and ex emption (in part)30, 154On convertible bonds; on profit-sharing bonds; on borrowings on local real estate mortgage
BelgiumNetForeign tax credit or deduction0, 10, 15, 16, 25On bank deposits; on registered bonds of banks; on registered government bonds and debt; on registered government bonds and debt instruments
CanadaGrossForeign tax credit50, 15None
DenmarkGrossForeign tax creditNilAll
FranceGross6Foreign tax deduction or credit70, 157On certain government bonds
Germany, Fed. Rep.Gross5Exemption (in part);8 foreign tax credit (in part);8 deduction (in part)90, 10, 15, 2510On bank deposits and loans that are not mortgaged by local immovable property
IrelandNetNone110, 10, 15, 35On certain government securities
ItalyGrossForeign tax credit80, 10, 12.5 157On public loans; on certain qualifying institutions’ bonds; on certain bonds issued abroad; on certain public enterprises’ bonds
JapanGrossForeign tax credit or deduction10, 15None
NetherlandsGross12Foreign tax creditNilAll interest other than from loan mortgaged by local immovable property and from substantial interest company
NorwayNetNone13NilAll
SwedenGrossForeign tax creditNilAll
SwitzerlandNetNone110, 5, 10, 15All interest except from bonds and certain registered loans
United KingdomGrossForeign tax credit140, 10, 15, 30On certain government securities
United StatesGrossForeign tax credit or deduction0, 5, 10, 15, 30On bank deposits
Sources: Same as in Tables 3 and 4. Also, International Bureau of Fiscal Documentation, Corporate Taxation in Europe (Amsterdam), loose-leaf service; and Coopers and Lybrand International Tax Network, International Tax Summaries (New York: John Wiley, 1982).

The rates given in the table apply only to interest flows from countries with which a given country has a double taxation treaty agreement. Almost all countries covered here have double taxation treaties with each other, and the rates vary in each case depending on provisions of the treaty with specific countries.

Under treaty provisions. For nontreaty countries, interest income that is already taxed in source country is no longer taxable in taxpayer’s country of residence.

Exemption of interest on mortgage loans from some specific countries.

Withholding tax applies only to interest on profit-sharing bonds.

Foreign tax credit is available only to the extent that interest earnings are subject to domestic income tax.

With treaty countries; for other countries, on a net basis and no tax credit is granted.

Foreign tax relief is usually by means of deduction of foreign tax paid, but foreign tax credit often applies under tax treaties. Withholding tax applies to interest on loans.

Under some tax treaties, foreign income is exempt from taxation. Credit against domestic tax is prorated to the ratio of domestic tax due to the total taxable income (both foreign and domestic).

At taxpayer’s option, foreign tax paid may be deducted from the computation of tax liability in the country of taxpayer’s residence.

Only on interest on bonds and mortgage loans.

However, if taxpayer had spent less than ten years in the foreign country from which interest income was derived, he may be allowed some foreign tax credit.

With treaty countries and nontreaty developing countries; for other countries, on a net basis and no tax credit is granted.

Exceptions: (1) interest on mortgage loans, on foreign sites, on immovable property on which tax reduction may be granted on a pro rata basis as stated in footnote 8; and (2) all of specified interest from some specific countries.

Extends even to the applicable highest marginal tax rate.

Sources: Same as in Tables 3 and 4. Also, International Bureau of Fiscal Documentation, Corporate Taxation in Europe (Amsterdam), loose-leaf service; and Coopers and Lybrand International Tax Network, International Tax Summaries (New York: John Wiley, 1982).

The rates given in the table apply only to interest flows from countries with which a given country has a double taxation treaty agreement. Almost all countries covered here have double taxation treaties with each other, and the rates vary in each case depending on provisions of the treaty with specific countries.

Under treaty provisions. For nontreaty countries, interest income that is already taxed in source country is no longer taxable in taxpayer’s country of residence.

Exemption of interest on mortgage loans from some specific countries.

Withholding tax applies only to interest on profit-sharing bonds.

Foreign tax credit is available only to the extent that interest earnings are subject to domestic income tax.

With treaty countries; for other countries, on a net basis and no tax credit is granted.

Foreign tax relief is usually by means of deduction of foreign tax paid, but foreign tax credit often applies under tax treaties. Withholding tax applies to interest on loans.

Under some tax treaties, foreign income is exempt from taxation. Credit against domestic tax is prorated to the ratio of domestic tax due to the total taxable income (both foreign and domestic).

At taxpayer’s option, foreign tax paid may be deducted from the computation of tax liability in the country of taxpayer’s residence.

Only on interest on bonds and mortgage loans.

However, if taxpayer had spent less than ten years in the foreign country from which interest income was derived, he may be allowed some foreign tax credit.

With treaty countries and nontreaty developing countries; for other countries, on a net basis and no tax credit is granted.

Exceptions: (1) interest on mortgage loans, on foreign sites, on immovable property on which tax reduction may be granted on a pro rata basis as stated in footnote 8; and (2) all of specified interest from some specific countries.

Extends even to the applicable highest marginal tax rate.

References*

    Ben-ZionUri“Impact of Inflation and Taxation on the Level Allocation and Financing of Investment: A Survey of Recent Literature” (unpublishedInternational Monetary FundFebruary241983).

    Ben-ZionUri and J.Weinblatt“Purchasing Power Interest Rate Parity and the Modified Fisher Effect in the Presence of Tax Agreements” (unpublishedAugust1982).

    ByrneWilliam J.“Fiscal Incentives for Household Saving,”Staff PapersInternational Monetary Fund (Washington) Vol. 23 (July1976) pp. 45589.

    DarbyMichael R.“The Financial and Tax Effects of Monetary Policy on Interest Rates,”Economic Inquiry (Long BeachCalifornia) Vol. 13 (June1975) pp. 26674.

    FeldsteinMartin S.“Inflation, Income Taxes, and the Rate of Interest: A Theoretical Analysis,”American Economic Review (Nashville, Tennessee) Vol. 66 (December1976) pp. 80920.

    GandolfiArthur E.“Inflation, Taxation and the Interest Rates,”Journal of Finance (New York) Vol. 37 (June1982) pp. 797807.

    HartmanDavid G.“Taxation and the Effects of Inflation on the Real Capital Stock in an Open Economy,”International Economic Review (OsakaJapan) Vol. 20 (June1979) pp. 41725.

    HowardDavid H. and Karen H.Johnson“Interest Rates, Inflation, and Taxes: The Foreign Connection,”Economics Letters (Amsterdam)Vol. 9No. 2 (1982) pp. 18184.

    KopitsGeorge F.“Taxation and Multinational Firm Behavior: A Critical Survey,”Staff PapersInternational Monetary Fund (Washington) Vol. 23 (November1976) pp. 62473.

    KopitsGeorge F.“Fiscal Incentives for Investment in Industrial Countries,”Bulletin for International Fiscal Documentation (Amsterdam)Vol. 35 (July1981) pp. 29194.

    LahiriKajal“Inflationary Expectations: Their Formation and Interest Rate Effects,”American Economic Review: (Nashville, Tennessee) Vol. 66 (March1976) pp. 12431.

    LeviMaurice D.“Taxation and ‘Abnormal’ International Capital Flows,”Journal of Political Economy (Chicago) Vol. 85 (June1977) pp. 63546.

    LeviMaurice D. and John H.Makin“Anticipated Inflation and Interest Rates: Further Interpretation of Findings on the Fisher Equation,”American Economic Review (Nashville, Tennessee) Vol. 68 (December1978) pp. 80112.

    ModiJitendra R.“Survey of Tax Treatment of Investment Income and Payments in Selected Industrial Countries” (unpublishedFiscal Affairs Department, International Monetary FundMay271983).

    MundellRobert“Inflation and Real Interest,”Journal of Political Economy (Chicago) Vol. 71 (June1963) pp. 28083.

    NielsenNiels Christian“Inflation and Taxation: Nominal and Real Rates of Return,”Journal of Monetary Economics (Amsterdam)Vol. 7 (March1981) pp. 26170.

    NowotnyEwald“Inflation and Taxation: Reviewing the Macroeconomic Issues,”Journal of Economic Literature (Nashville, Tennessee) Vol. 18 (September1980) pp. 102549.

    SargentThomas J.“Interest Rates and Expected Inflation: A Selective Summary of Recent Research,”Explorations in Economic Research (New York) Vol. 3 (Summer1976) pp. 30325.

    SteuerleEugene“Is Income from Capital Subject to Individual Income Taxation?”Public Finance Quarterly (Beverly HillsCalifornia) Vol. 10 (July1982) pp. 283303.

    SummersLawrence H.“The Non-Adjustment of Nominal Interest Rates: A Study of the Fisher Effect” National Bureau of Economic Research (New York: Columbia University Press) Research Paper No. 836 (January1982).

    TanziVito“Real Fisherian and Required Rates of Interest Under Inflationary Conditions: Theoretical and Empirical Results” (unpublishedInternational Monetary Fund (Washington) December171975).

    TanziVito“Inflation, Indexation and Interest Income Taxation,”Quarterly ReviewBanca Nazionale del Lavoro (Rome) No. 116 (March1976) pp. 6476.

    TanziVito“Inflation and the Incidence of Income Taxes on Interest Income: Some Results for the United States, 1972-74,”Staff PapersInternational Monetary Fund (Washington) Vol. 24 (July1977) pp. 50013.

    TanziVito“Inflationary Expectations, Economic Activity, Taxes, and Interest Rates,”American Economic Review (Nashville, Tennessee) Vol. 70 (March1980) pp. 1221.

    TanziVitoInflation and Personal Income Tax (Cambridge: Cambridge University Press1980).

    TanziVito“Inflationary Expectations, Economic Activity, Taxes and Interest Rates: Reply,”American Economic Review (Nashville, Tennessee) Vol. 72 (September1982) pp. 86063.

    TanziVito and Mario I.Blejer“Inflation, Interest Rate Policy, and Currency Substitution in Developing Economies: A Discussion of Some Major Issues,”World Development (OxfordEngland) Vol. 10 (September1982) pp. 78189.

    TobinJames“Money and Economic Growth,”Econometrica (Evanston, Illinois) Vol. 33 (October1965) pp. 67184.

Although the paper was a collective effort by Fiscal Affairs Department staff members, assisted by a consultant—Professor John H. Makin of the University of Washington and the National Bureau of Economic Research—specific individuals were responsible for certain parts. Section II was prepared mainly by Ved P. Gandhi, who also coordinated the preparation of the entire paper. The factual information on which that section is based was researched by Jitendra R. Modi, who was also responsible for the preparation of the three tables in Appendix III. Section III was written by John H. Makin and Vito Tanzi, with contributions by Ved P. Gandhi, Menachem Katz, and Mario I. Blcjer. Section IV was mainly the work of John H. Makin, who also prepared Appendix II. The introductory and concluding sections were written by Ved P Gandhi, John H. Makin, and Vito Tanzi, with Mr. Gandhi carrying a greater share of the burden. Leif Mutén wrote Appendix I.

Surveys of recent works on the effects of taxation on interest rates and international capital movements have been carried out in the fiscal Affairs Department; see the background papers by Ben-Zion in Part II of this volume. Other relevant papers prepared by the Department are included in the References. The present paper relies heavily on findings and conclusions contained in those papers.

Australia. Austria, Belgium, Canada. Denmark, France, the Federal Republic of Germany, Ireland, Italy, Japan, the Netherlands, Norway, Sweden, Switzerland, the United Kingdom, and the United States.

Section II is based on a survey of the tax systems of 16 industrial countries: Australia, Austria, Belgium. Canada, Denmark, France, the Federal Republic of Germany, Ireland, Italy, japan, the Netherlands, Norway, Sweden, Switzerland, the United Kingdom, and the United States. See the tables in Appendix III for more detail

The tables in Appendix III show only major differences and are not comprehensive.

In reality, withholding of income tax at the source for selected interest income can introduce a scheduler element.

Based on a rate of exchange of ¥ 250 = US$ 1.

In the United States, for example, the risk-adjusted difference in the interest rates on taxable and nontaxable bonds has historically been of the same magnitude as the average tax rate on interest income.

The existence of withholding taxes often reduces the opportunity for tax evasion.

Only when long-term capital gains arise from disposal of business assets.

The conventional forms of accelerated depreciation are, however, inadequate substitutes for depreciation based on replacement cost; the larger the anticipated rate of inflation and the longer the life of the asset, the wider is the gap. Investment allowances and income tax credits, on the other hand, favor short-lived assets, because every time such assets are replaced the investor can take advantage of these incentives in addition to using full depreciation.

For fiscal year 1984, the tax revenue loss from deductibility of interest on consumer loans in the United States is estimated at about $8 billion. The revenue loss from deductibility of mortgage interest on owner-occupied houses is estimated at about $28 billion.

Fisher himself is reported to have had doubts about this extreme version of his theory.

If people expect that the fiscal deficits will be monetized in the future even though they are not monetized in the short run, the fiscal deficit may keep long-run rates high, which in turn, through arbitrage, may also keep the short-run rate high.

And, by the same token, when the expected rate of inflation is falling, the nominal rate is expected to fall by less than the rate of inflation.

See Tanzi (1977) for a discussion of the implications of progressive taxes.

See. in particular, Tanzi (1976).

Thus, the net-of-tax real interest rate is 3 percent.

It must be recalled that this was generally a period of rising inflation.

this led some researchers to argue that there may have been monetary as well as fiscal illusion at work during the period.

For an example of such an approach, see the background paper by Makin and Tarzi in part II of this volume (Chapter 4).

See Chapter 7. background paper by Katz in part II of this volume. The eight industrial countries are Canada, France, the Federal Republic of Germany, Italy, Japan, the Netherlands, the United Kingdom, and the United States.

These differences refer to the legal treatment alone, since differential possibilities of tax evasion could not be assessed. Furthermore, the capital gains taxes that determine the value of θ are assumed to apply to realized gains on a yearly basis.

This is puzzling because the tax effect should have been lowest in Japan.

This policy has recently been introduced in Iceland. The basic justification for this policy change seems to be the fact that interest rates in Iceland have approximated the inflation rate for most years, implying that real interest income and expense have been close to zero.

An empirical study for the United Stales has shown that gains and losses would approximately balance out. See Tanzi (1977).

Assuming that (1) households account for roughly 40 percent of total borrowing, (2) the elasticities of the supply and demand schedules vary between 0.2 and 0.8. respectively, (3) the expected real rate of interest is 3 percent, and (4) the expected rate of inflation is 6 percent, it can be estimated by the model in Appendix II that such a policy change would reduce the nominal rate by less, and in most cases by much less, than 1 percentage point.

For a discussion of the effects of financial market taxation on international capital flows, see the background paper by Blejer in Part II of this volume (Chapter 8).

As argued in Section III, this tax-induced fall in interest rates may not be very large.

The degree of reduction in the interest rates in the domestic country and the increase in the rate in the foreign country is determined by the relative size of both countries.

This assumption approximates the conditions in Canada, the United Kingdom, and the United States, where realized foreign exchange gains and losses are treated as capital gains and losses and the returns on assets held longer than a statutory minimum period are taxed at lower rates than interest earnings, which are taxed as ordinary income.

The size of the actual gap between tax rates is exaggerated here for purposes of illustrating the impact. In practice, a very small expected return on an after-tax basis can generate large flows of capital.

This practice is followed in the Federal Republic of Germany, France, Japan, and some other industrial countries.

In addition to the background papers in Pan II of this volume, the references listed here were important sources in the preparation of this overview.

    Other Resources Citing This Publication