Abstract

John R. King

The Concept of Income

John R. King

Income is not a simple concept. In the theoretical literature, several alternative definitions have been proposed and none of them has achieved universal acceptance as an appropriate definition for all purposes. But practical and reasonably precise measures of the income of an individual or a business in a particular period of time are needed, not only for tax purposes, but also for other purposes—such as financial reporting by a company to its shareholders, and social accounting by governments. In the absence of a single, universally accepted theoretical definition of the concept of income, many aspects of these practical measures remain controversial.

Theoretical Definitions of Income

In what is probably the most influential definition of the personal income of an individual, in a particular period of time, Henry Simons asserted it to be the following:

The algebraic sum of: (1) the market value of rights exercised in consumption; and (2) the change in the value of the store of property rights between the beginning and end of the period in question. In other words, it is merely the result obtained by adding consumption during the period to “wealth” at the end of the period and then subtracting “wealth” at the beginning.1

Similar definitions had been advanced in earlier writings by R.M. Haig (1921) and G. von Schanz (1896), and so this is often referred to as the Schanz-Haig-Simons (SHS) definition. All three writers were concerned to find a measurable concept which would serve as an adequate basis for a personal income tax system. Such taxes had already been in existence in some countries, of course, for many years. In general, however, they had been based on an underlying concept of income as a flow (or yield) of services from specific sources—sepa-rate and distinct from any changes in those sources themselves. Since the SHS definition brings changes in wealth within the income definition, on a par with such services, it is commonly referred to as a definition of “comprehensive income.” A central feature of this definition is that it requires appropriate procedures for valuing “wealth” at the beginning and at the end of the income period.

In subsequent writings, Hicks offered an alternative, broad definition of an individual’s income in a given week as “the maximum value which he can consume during a week, and still expect to be as well off at the end of the week as he was at the beginning.” Elaborating on the concept of “well-offness” in this broad definition, he reformulated the definition as “the maximum amount of money which the individual can spend this week, and still expect to be able to spend the same amount in real terms in each ensuing week.”2

The obvious difference between the SHS definition and that of Hicks is the importance that the latter accords to the subjective expectations of the individual. Hicks was, explicitly, not proposing a definition for use as a practical basis for an income tax. Indeed, in commenting on the difficulty that would be faced by a statistician who might seek to apply his definition in measuring the social income in an economy, Hicks advised that:

The best thing he can do is to follow the practice of the income tax authorities. But it is the business of the theoretical economist to be able to criticize the practice of such authorities; he has no right to be found in their company himself!3

Nevertheless, many particular features of income tax systems are based in practice on something that seems closer to Hicks’s definition than to that of SHS. This definition remains, therefore, an important source of theoretical support for certain departures from the SHS comprehensive tax base.

Personal Income Tax

Theory of Optimal Income Taxation

Howell H. Zee

  • What is the nature of the trade-off between efficiency and equity?

  • What problems does the theory of optimal income taxation address?

  • What are the main economic implications that can be drawn from the literature on optimal income taxation?

The theory of optimal income taxation brings together the efficiency and equity consequences of taxation—the two fundamental concerns of tax policy that hitherto have only been separately dealt with, respectively, in the first two sections of Chapter II. In principle, there is, of course, no reason why equity issues cannot also be integrated into the theory of optimal commodity taxation (see Chapter III).4 Indeed, general optimal commodity tax rules in the multi-person setting have been derived in the literature. These rules, however, have little policy relevance, since it is seldom feasible in practice to impose, on a broad scale, differential tax rates on consumption to achieve equity objectives.5 Hence, the taxation of income remains the most relevant setting for addressing the combined efficiency and equity concerns of the policymaker.

Trade-off Between Efficiency and Equity

At the heart of the theory of optimal income taxation is the trade-off between efficiency and equity. It is, therefore, important to understand the nature of this trade-off before embarking on a formal analysis of the optimal income tax problem. It will be recalled from Chapter II that the only tax which entails no efficiency loss is the lump-sum levy. Hence, if such levies are feasible, the problem would be reduced to one purely of equity, and the issue of the trade-off would not arise. In reality, of course, any attempt at income redistribution has to be carried out through the use of distortionary taxes, such as the income tax. The more they are relied upon to effect an income transfer from the rich to the poor, such as by increasing the progressivity of the income tax structure, the higher the efficiency costs they will impose on society. The optimal trade-off point is reached when the marginal enhancement to equity from an additional unit of income transfer is just worth the marginal efficiency cost of that transfer.

Utility possibility frontier

The efficiency cost associated with an income tax is in its disincentive effect on work effort, particularly at high tax rates.6 In an economy in which individuals differ in their productivities—and therefore in their before-tax incomes, their after-tax incomes will, in general, also be different under any reasonable income tax system.7 Now, if the relative tax burdens between the rich and the poor are altered, subject to a given government revenue requirement, with a view to reducing the inequality in their after-tax incomes (or utilities), at moderate tax burdens, the reduction in the rich’s utility could probably be accompanied by an increase in the poor’s utility. At high tax burdens, however, a point will soon be reached where the tax’s disincentive effects are so pronounced that any further increase in the tax burdens would reduce the utilities of both the rich and the poor. In the limiting case of complete equality, the equality is realized only because the incomes (and utilities) of both have been reduced to zero.8 This is illustrated in Panel (a) of Figure IV.1 for a two-individual world, where the utility levels of individuals 1 and 2 are respectively denoted by U1 and U2 and measured along the horizontal and vertical axes.

Figure IV.1.
Figure IV.1.

Utility Possibility Frontier and Optimal Distributions

For any given level of utility for one individual, there is (are) a maximum attainable level(s) of utility for the other individual. A curve that traces out all such utility combinations is called a utility possibility frontier (UPF).9 Hence, point A represents the maximum U2 when U1 is zero.10 As individual 2s income is being redistributed toward individual 1 through an increase in the progressivity in the income tax system, there will initially be a range of redistributions where the UPF is negatively sloped (segment AR). Once point R is reached, however, any further redistribution through the distortionary tax system can only reduce both individuals’ well-being; the segment of UPF beyond point R (segment OR) is accordingly positively sloped. Indeed, the closer it approaches the 45-degree line representing complete equality, the lower the utility levels of both individuals. Ultimately, complete equality is attained at the origin where U1 = U2 = 0. The most important insight provided by this construct is that the UPF lies entirely to one side of the 45-degree line, thus underscoring the enormous efficiency cost to society in achieving complete equality.11

Optimal distributions

How should the policymaker choose a particular distribution of individual utilities on any given UPF,12 so as to maximize social welfare? The answer clearly lies in the particular notion of distributive justice he subscribes to. In Panel (b) of Figure IV.1, the UPF in Panel (a) has been reproduced, but with the iso-welfare contours of three benchmark theories of distributive justice, developed in Chapter II, superimposed on it. The optimal distributions under the utilitarian, Nash, and Rawlsian theories are given, respectively, by points B, TV, and R. While it should come as no surprise that the Rawlsian (utilitarian) distributive justice produces the most (least) egalitarian outcome among the three, all three optimal points, even in the limiting case of Rawls, provide for some degree of inequality to prevent the complete disappearance of efficiency in society. Hence, complete equality of well-being across individuals can seldom be an outcome where social welfare is maximized,13 even if the policymaker attaches great weight to distributive objectives in taxation.

Optimal Income Taxation

Defining the problem

In view of the nature of the trade-off between efficiency and equity discussed above, what can be said about the optimal degrees of income tax progressivity under different theories of distributive justice? Models of optimal income taxation, which seek to provide an answer to the aforementioned question, have the following important common mixeds: (1) there is a known distribution of (otherwise identical) individuals with differently endowed productivities (skills), thus giving rise to different levels of individual pretax income; (2) the government’s objective is to maximize a given social welfare function,14 with the income tax as its only instrument, subject to a given revenue requirement (which could be zero, in which case, the tax is used only for redistributive purposes); and (3) the income tax produces a disincentive effect (its efficiency cost) through its impact on an individual’s work effort.

The above government maximization problem has been investigated on the basis of a variety of social welfare functions, including the limiting cases of the utilitarian and Rawlsian formulations of distributive justice.

Main results and implications15

A number of important results, some fairly intuitive but a few quite surprising at first blush, have been obtained from solving the optimal tax problem defined above.16

First, in the case of a linear income tax schedule, that is, t=α+β·γ, where τ is tax revenue, γ is pretax income, and α and β are (constant) tax parameters, the optimal solution implies that α < 0 and 1 > β > 0, that is, it is optimal to have a guaranteed minimum income (equals -α) and a positive (but less than unity) marginal tax rate, so that the income tax schedule is basically progressive.17 Moreover, the higher the degree of society’s aversion to inequality (i.e., the lower the value of the parameter ε in the social welfare function),18 the higher the optimal marginal tax rate ß.19 The precise value for the optimal B depends, of course, on a number of factors, including the nature of the distribution of individual productivities, the sensitivity of an individual’s utility to income variations, and the elasticity of his labor supply.20

Note that even if the social welfare function is utilitarian, that is, with ε = 1, some redistribution is still called for. This is primarily because a unit of income has a much higher value to the poor than to the rich. Hence, society as a whole would generally gain from the redistribution even when it has no aversion to inequality per se. The obvious exception to this would be if an individual’s valuation of his income does not vary across income levels (the case of constant marginal utility of income), then under utilitarian distributive justice, the optimal marginal tax rate would be zero.

Second, in the case of a general income tax schedule, that is, t = t(y), so that the marginal tax rate is allowed to vary across income levels, the optimal income tax problem becomes significantly more difficult to solve, since now, a complete solution would involve determining an entire nonlinear schedule. Indeed, even with fairly simple model structures, re-search in this area so far has yielded only limited insight on the optimal shape of the entire tax schedule. Nevertheless, two fundamental results have been obtained: (1) the optimal marginal tax rate is nonnegative (but, of course, less than unity) throughout the entire income range, and (2) the optimal marginal tax rates at both the top and bottom ends of the income range are zero. These results hold irrespective of the underlying formulation of distributive justice.

The first result establishes the important property that there are no income ranges where the optimal tax schedule falls as income rises. The second result provides the biggest intuitive surprise in this literature: it implies that, as one moves from the low to the high ends of the income range, the optimal marginal tax rate first rises and then falls at some point at high income levels—ultimately falling to zero at the top level. To understand this seemingly counterintuitive out-come, particularly in cases where society has some aversion to inequality, it is necessary to realize that a change in the marginal tax rate at any income level only affects individuals situated at that level and above. The smaller the number of individuals with incomes above that level, the more important the efficiency relative to redistributive consequences of the tax will become. Since by definition, no individuals will be situated above the top income level, the optimal marginal tax rate there is thus entirely determined by efficiency considerations.21 This reasoning is, therefore, independent of the particular theory of distributive justice adopted.

As noted earlier, the optimal shape of the tax schedule, except at its end points, and that it should not fall as income rises, are not well understood. Hence, the optimal zero end-point tax rate outcome may well lead to a misguided design of the tax schedule as a whole. In fact, numerical simulations of optimal income tax models have revealed that, under a variety of assumptions concerning the nature of the individual utility function and the distribution of individual productivities, the curvature of the optimal tax schedule is not very pronounced across alternative formulations of the social welfare function; that is, it can be approximated by a tax schedule with only a few linear segments.

The basic insight to be grasped from the theory of optimal income taxation is that the importance of achieving redistributive goals through increasing the progressivity of the income tax system should not be overemphasized, as the efficiency costs of doing so are likely to be extremely high.

The Base of the Personal Income Tax

Janet Stotsky

  • What is the difference between a global and schedular income tax?

  • How do income tax systems arrive at the concept of taxable income?

  • What is the rationale for some typical deductions from income?

  • How do we make a tax system neutral to inflation?

Income tax systems can be designed on either a global or schedular basis, although, in practice, most global income tax systems have schedular features and some schedular income tax systems have global features. A global income tax aggregates all sources of income while a schedular income tax imposes tax on each source of income separately. Global income taxes generally are used in industrialized countries, while schedular income taxes are more commonly found in developing countries. Many developing countries have adopted global income taxes in form though they may be administered similarly to a schedular tax with heavy reliance on withholding and few taxpayers filing final returns or being assessed on global income.

There are several advantages to a schedular income tax. The main advantage is that it may be more easily administered in countries without a sophisticated tax administration. Tax generally is collected by withholding so that the number of taxpayers who must file returns can often be significantly reduced with a schedular system. Moreover, the administrative advantages are greatest if there are few taxpayers with multiple sources of income. An additional advantage is that a schedular tax allows for different treatment of different types of income if a country wants to tax either labor income, capital income, or certain kinds of capital income more lightly. This differential treatment, however, can be obtained under a global income tax with schedules for particular kinds of income, as is typical.

The main advantage of a global income tax is that goals of vertical equity more easily can be achieved since the tax is based on an aggregate measure of income. It may also have administrative advantages if there are many taxpayers with multiple sources of income, because only one return is filed for each taxpayer. Schedular systems often end up with a patchwork of overlapping schedules, multiple exemptions for the same income, high marginal tax rates, and so on.

Under all income tax systems, the base of the personal income tax falls far short of the SHS notion of income. The starting point is “gross income,” defined differently in each tax code. Although termed gross income, this definition of income may be in part a net or gross concept. Under an SHS notion of income, “gross income” would consist of wages and salaries, business income (e.g., partnerships, sole proprietorships, farm income), capital income, rents, royalties, fringe and in-kind benefits, imputed rent from consumer durables, income transfers, pension income, and gifts and bequests.

Some components of income usually are excluded from gross income, meaning that they are not even included in the concept of gross income for the purposes of the income tax. Gifts and bequests, while taxable to the decedent’s estate, are typically excluded from income. Death benefits and disability or sickness benefits are also typically excluded from income on the grounds that they represent compensation for a loss rather than an increase in an ability to consume.

The treatment of fringe benefits varies greatly across income tax systems. Some income tax systems exclude all employer-provided fringe benefits from employee taxable income. Others impute some value for these fringe benefits and tax them at either the employer’s tax rate or at the employee’s tax rate. Still other tax systems deny a deduction at the company level for the cost of the fringe benefits. The imputed value for tax purposes should be market value. If employees value these fringe benefits at less than their market value, then these benefits should be valued at their cash equivalent value. Fringe benefits are rarely valued at market value. Often these fringe benefits are valued at some proportion of the employee’s wages or on some other arbitrary basis. The tax exclusion of fringe benefits generally results in significant erosion of the tax base. It leads to a higher demand for fringe benefits, because fringe benefits are effectively purchased out of pretax income. It also leads to an erosion in the progressivity of the tax code because fringe benefits are disproportionately provided to highly compensated employees.

Imputed rent from consumer durables is in most cases excluded from tax. The largest component of this imputed rent is on owner-occupied homes. This imputed rent would be properly measured as the gross rent minus any costs associated with the home. In part, this exclusion stems from the administrative difficulties in measuring the value of the imputed rent and in part, it stems from the political unpopularity that would be associated with any attempt to tax this nonmonetary form of income. The tax exclusion, however, results in encouraging homeownership, which is generally considered a socially worthwhile goal. Nevertheless, the exclusion of imputed rent creates inefficiencies and inequities.

Income transfers are also typically excluded from tax. Income transfers take a variety of forms and the rationale for taxing them differs. Transfers from public pensions, such as social security, may in part represent payment for previous taxes paid. To the extent that social security taxes are not deducted from taxable income when they were contributed, the benefits are properly excluded from income when paid out. Many income tax systems, however, allow taxpayers to exclude social security taxes from taxable income. In this case, the benefits are properly included in income when paid out. Unemployment compensation and aid to low-income households may also be excluded from tax, although, if income below some threshold is not subject to tax, there is no persuasive rationale for this on efficiency or equity grounds. In-kind transfers to low-income households should also be included in income, but this again poses the problem that their value to the recipient may be lower than their monetary cost. Thus, it would be fair to tax the recipient only on the cash equivalent value.

Many kinds of capital income are frequently not included in the income tax base in developing countries. Dividends and interest on certain forms of savings are often explicitly tax exempt. The rationale for this exclusion is to encourage capital investment. Nevertheless, the favorable treatment afforded capital income may lead to inefficiencies in encouraging only tax-favored forms of investment and inequities in reducing tax burdens on higher-income taxpayers who earn most of the capital income. Capital gains is an area that introduces considerable complexity into an income tax and as a consequence, many developing countries do not include capital gains in income. These issues are discussed later in Chapter IV.

Taxable income is defined as gross income minus tax reliefs. Income tax systems differ in how they reduce gross income to taxable income. Tax relief can take the form of adjustments, deductions, exemptions, allowances, and credits. Adjustments to income are generally tax reliefs that are available to all taxpayers. For instance, alimony paid or pension contributions are typically adjustments in that all taxpayers may reduce their gross income by these amounts.

Under some tax systems, adjustments are grouped with other deductions and these deductions are available to all taxpayers. While under other tax systems, deductions take the form of a standard deduction or itemized deductions, and the taxpayer is left the option of choosing the most beneficial approach. The standard deduction may vary across characteristics of the filing unit or taxpayer. Taking other characteristics of the tax system as unchanged, standard and itemized deductions reduce the tax burden on the taxpayer by the product of the taxpayer’s marginal tax rate and the amount claimed as a deduction. Their value is thus directly proportional to the marginal tax rate of the taxpayer. Itemized deductions typically benefit higher-income taxpayers more than lower-income taxpayers because higher-income taxpayers are more likely to incur expenses that are itemizable. Allowable itemized deductions vary across income tax systems, but may include unreimbursed employee business expenses, charitable contributions, educational expenses, pension contributions, life insurance contributions, personal and mortgage interest payments, medical and dental expenses, taxes paid to other governments and social security tax payments, and casualty or theft losses.

There are various arguments given in favor of allowing these items to be deducted from income. The rationale for allowing a deduction for unreimbursed employee business expenses is that income should be measured net of expenses incurred in earning it. The rationale for allowing charitable contributions, educational expenses, pension contributions, and life insurance contributions to be deducted rests primarily on the argument that the deduction encourages a higher level of these activities and they are socially beneficial activities. The extent to which the deduction encourages a higher level of these activities depends on the price elasticity of demand.

The rationale for allowing a deduction for mortgage interest is to encourage more homeownership. This deduction for interest would be justified if the imputed income from homeownership were taxed. There is little rationale for allowing a deduction for other personal interest, since this interest is generally incurred in purchasing goods that have no particular social merit and thus this deduction favors consumption over savings.

There are several arguments in favor of allowing a deduction for medical and dental expenses, and taxes paid to other governments and social security tax payments. One argument is, as noted above, that the deduction encourages socially beneficial activities (i.e., good health and greater provision of government services). Another argument is that these expenses are to some extent involuntary and reduce the taxpayer’s ability to pay. This rationale has its limitations. Many medical and dental expenses are voluntary and some such expenses are incurred by most taxpayers. It may, therefore, be appropriate only to allow a deduction for extraordinary expenditures on medical or dental care. Taxes are paid to other governments, and for social security tax payments in return for services. It may, therefore, be inappropriate to view them as reducing a taxpayer’s ability to pay. The deduction for casualty and theft losses is also based on the notion that such losses impair ability to pay, although, here again, this applies only to extraordinary losses.

In some cases, these deductions may only be taken in full if they exceed a threshold or may only be taken to the extent that they exceed a threshold. The threshold is typically based on gross or taxable income. Although these thresholds are frequently imposed for revenue reasons, the rationale for such thresholds is that only extraordinary expenses impair ability to pay and should be allowed as a deduction. In some cases, deductions may be limited to a certain proportion of gross or taxable income, again, generally for revenue reasons and to prevent tax evasion. Itemized deductions may be phased out for some higher-income taxpayers, as well, effectively raising their marginal tax rate in the phase-out range.

In addition, many income tax systems allow personal exemptions or family allowances to be deducted. These exemptions are typically based on the number of individuals in the filing unit and may also be related to characteristics of the filing unit. Personal exemptions, like deductions, reduce the tax burden by the product of the marginal tax rate and the amount claimed as an exemption. Their value is thus also directly proportional to the taxpayer’s marginal tax rate. Some income tax systems in developing countries have eliminated personal exemptions because the tradition of extended families makes it difficult to determine the number of people in the household. Some tax codes have converted these personal exemptions into a credit against taxes to enhance the equity of the tax system because a credit reduces the tax burden by the same nominal amount regardless of the taxpayer’s marginal tax rate.

Once taxable income has been defined, the income tax code specifies rates and brackets that apply to this income. There is considerable variation in the rates and number of brackets across income tax systems. Marginal tax rates vary from 1 percent to close to 100 percent, although a typical maximum rate is about 40 percent. Some tax systems have as few as two or three rates and brackets, while others have more than ten. If there is a system of joint filing, different schedules may apply to different filing units.

After computing tax on taxable income, the individual taxpayer may be allowed to claim certain tax credits. These credits vary but are often designed to provide tax relief to low-income households and married couples. Credits are also frequently allowed for taxes paid to other countries on foreign source income that is also subject to domestic tax. These credits are typically nonrefundable, that is, they cannot be claimed by taxpayers with no positive tax liability, but in a few cases, they may be refundable. Refundable credits require an able tax administration to deter tax evasion.

Certain forms of income may be taxed on a schedular basis using the same rate schedule and brackets as the income tax or using an entirely separate schedule. In some countries, pension income, capital gains, partnership income, individual proprietors’ income, farm income, fringe benefits, and other forms of income are treated on a schedular basis, entirely separate from the rest of the income tax.

Inflation interacts in many ways with the individual income tax. Any magnitudes in the income tax that are set in nominal terms, such as brackets and tax reliefs, change in value with inflation. If brackets are left unchanged in nominal terms, then in the face of inflation, which raises nominal incomes, taxpayers are pushed into higher tax brackets. This phenomenon, known as bracket creep, is one means by which the average tax rate rises over time. To prevent bracket creep, the brackets may be indexed to changes in the overall level of prices. Standard deductions, personal exemptions, and credits are also set in nominal terms and lose value in the face of inflation if they are not indexed to changes in the overall level of prices. Some income tax systems index all of these nominal magnitudes to inflation, but the record is that governments often renege on promised inflation adjustments. In some cases, however, these nominal magnitudes are adjusted periodically, giving governments more budget flexibility and the opportunity to score political points with taxpayers by offering them a tax cut when they are merely counteracting the effect of inflation.

The taxation of capital income is another area where inflation interacts with the tax code to change real tax burdens. Capital gains are typically taxed on a nominal basis. Some tax codes index capital gains to inflation, for example, the United Kingdom, but this introduces certain administrative complexity and raises issues as to the proper means of indexing. In addition, nominal interest rather than real interest is generally taxable when received as income and deductible on borrowing, resulting in an increase in the tax burden on lenders and a decrease in the tax burden on borrowers, when inflation leads to high nominal interest rates.

The ideal personal income tax would have a broad base of income and a large basic allowance to eliminate low-income households from the tax rolls and to ensure a degree of progressivity in the tax code. The broader the base of the tax, the lower the rates that are needed to raise any given amount of revenue. Most income tax would be collected through withholding, minimizing the number of taxpayers who would have to file tax returns. In practice, most tax systems fall far short of defining income comprehensively and then fall short again in allowing overly generous deductions, exemptions, and credits. Comprehensive income taxation is an elusive goal.

The Choice of Taxable Unit

Janet Stotsky

  • Why is the choice of taxable unit important in the design of a personal income tax?

  • What do we mean by a “marriage tax”?

  • What are the advantages and disadvantages of joint versus individual filing?

  • What are common practices?

There are several different ways in which the taxable unit may be defined in a personal income tax system. First, each individual may be taxed separately, regardless of marital status. Second, couples may be taxed on their joint income. Third, couples may have the option of being taxed jointly or separately. Finally, families may be taxed on their joint income. Determining the taxable unit is not as obvious as it might appear at the outset. This section will address the considerations appropriate for the selection of the taxable unit and their various ramifications.22

Interaction of the Taxable Unit and Other Features of the Tax System

There are, in principle, four different features of a personal income tax system that determine tax liability: first, the choice of taxable unit; second, sources of income subject to tax; third, tax preferences, such as allowances and credits; and fourth, the tax schedule. Tax systems may combine these factors in many different ways to achieve specific objectives. The choice of taxable unit, along with these other critical features of a tax system, thus has important equity, efficiency, and administrative implications.

It is possible to illustrate the interactions among the different features of a tax system with a simple, hypothetical tax system. Assume the following tax schedule: no tax on income up to $5,000, a 10 percent tax on income from $5,001 to $30,000, and a 20 percent tax on income above $30,000. Assume two hypothetical married couples, A and B and C and D. Table IV.1 gives their taxes under different definitions of the taxable unit. Under individual taxation, column 2 presents their incomes and column 3, their tax liability, under joint taxation, column 4 presents their joint income and column 5, their tax liability.

Table IV.1.

Choice of Taxable Unit—Hypothetical Tax System

(In U.S. dollars)

article image
Source: IMF staff calculations.

As Table IV.1 indicates, A and B pay the same total tax regardless of the choice of taxable unit because A is the only earner of income in the family. C and D, in contrast, under individual taxation, pay a total of $33,000 in tax, and under joint taxation, pay a total of $36,500 in tax. This increase in tax burden under a system of joint taxation with increasing marginal tax rates is often termed a “marriage tax.” As a result, tax systems often include provisions to provide relief from this additional tax. One method, known as income splitting, effectively taxes each spouse as an individual on one-half of the couple’s total income, generally by applying the individual schedule to the couple’s total earnings but doubling any standard deduction and the width of the tax brackets. In the example, with income splitting, A and B would each pay tax on $100,000 of income, reducing their combined tax bill to $33,000. C and D could likewise reduce their combined tax burden to $33,000, thereby eliminating the “marriage tax.” A, in fact, gets a “marriage benefit” because he lowers his tax burden by getting married. In some systems with joint taxation, income splitting is not permitted. Instead, different schedules may apply to couples filing jointly; or, couples may be allowed to file separately for which another schedule applies. In this case, depending on the variation in schedules and other tax preferences as well as the distribution of earnings between spouses, there may be a “marriage tax” for some and a “marriage benefit” for others. In an individual taxation system with increasing marginal tax rates, the nature of the problem is somewhat different. The tax system is neutral with respect to marriage in that when two individuals get married, their tax burden does not change. When considered as a unit, however, married couples with equal incomes will generally pay different taxes. Also, a couple will pay different amounts of tax than an individual with the same total income. For instance, in the case of equal total incomes, if there is only one earner in each couple, then the tax burden is identical for the two couples and for the individual. If both spouses earn income, however, then their tax burden will be less than that paid by a couple with one earner or by an individual.

Under a system of individual taxation, marriage relief can be seen as equalizing the tax burden on two couples who have the same total income but have a different distribution of earnings between the spouses. In this case, this relief does not take the form of income splitting or different schedules but rather tax preferences, such as credits, which may be transferable between spouses. For instance, the husband may get a tax credit which, if he does not wish to claim, can be transferred to the wife as a credit on her return. This tax relief is most important for low-income married households for whom any tax preferences will represent a larger share of their tax burden. With respect to the issue of equity between a couple and an individual with the same income, it is sometimes argued that to the extent that the tax system should correct for ability to pay, it is fair for a couple with the same income as an individual to pay less in tax because the couple has a higher cost of living. Marriage relief is thus a correction for differences in ability to pay.

Equity

The choice of taxable unit thus has important equity implications, particularly with respect to considerations of horizontal equity. The basic notion of horizontal equity is that those with equal income should pay equal taxes. As the discussion illustrates, it is by no means clear to which definition of the taxable unit this criterion for equity should apply, since under every tax system except a very simple one, the criterion will mean something different depending on the choice of taxable unit. Should individuals with equal income pay equal tax or should couples with equal income pay equal tax or should individuals and couples with equal income pay equal tax? The answer is not obvious. If ability to pay (for which income is generally the proxy) depends on household income rather than the income of any one individual in the household, then it might be more appropriate to base the measure of horizontal equity on the household. Marriage is, however, only one, readily identifiable means of creating a household. Households may comprise one individual or many, legally unrelated individuals as well, with varying degrees of sharing of household resources. Defining the household unit as one based on marriage is thus somewhat arbitrary. Thus, the individual may ultimately be a more appropriate measure of ability to pay on which to base judgments of horizontal equity.

Society’s values thus influence the choice of the most equitable taxable unit. To the extent that a married couple has typically been regarded as the standard household unit, this provides justification for taxing the couple as a unit. Today, however, in western societies, a much greater number of individuals are living alone or in a variety of household arrangements, including reconstituted families and unmarried individuals living together, and in more traditional societies, atomistic families are still not the vogue and large extended households—often inclusive of grandparents and cousins—tend to predominate. Under these circumstances, justification for taxing the traditional married couple as a unit may not be as persuasive.

Efficiency

The choice of taxable unit also has important implications for efficiency since it has an effect on the marginal tax rate of the unit and thus has an effect on decisions to work, save, and invest, and on the composition of the household. The importance of these efficiency effects depends both on the degree to which the tax system alters the incentives that individuals face and the responsiveness of their behavior to changes in the tax rate and other features of the tax system. Considerable effort has gone into investigating the effects of changes in taxes on behavior.23 Evidence from both industrialized and developing countries suggests that changes in the marginal tax rate as well as other features of the tax system alter labor supply decisions, and savings and investment behavior.24

To illustrate how the choice of taxable unit affects the marginal tax rate, suppose that under the hypothetical tax system given earlier, A and B decide to get married and file jointly. B goes from the position of facing a marginal tax rate on income of 0 percent to 20 percent (since the marginal tax rate applies to the last dollar earned of the couple, it applies to dollars in excess of A’s income). Economic theory would suggest that the increase in B’s marginal tax rate might discourage her from working, thus causing inefficiency. As another example, suppose that B had $2,000 of capital income. If she got married and filed jointly, she would face an increase in marginal tax rate on this income, which might alter her investment behavior.

Other Issues Affected by the Choice of Taxable Unit

Nonlabor income poses a problem in that there is no ideal way to treat it in a system of individual taxation. The original British practice of attributing all of the nonlabor income to the husband is outdated. Another approach, which attributes all of the income to the higher earner is unappealing in that a taxpayer may be taxed on income over which he or she has no control. Another approach, which allocates the income equally between spouses, suffers from the same problem. A final possibility is to allow couples to allocate the income however they like, but this can lead to tax avoidance through shifting of income to the lower-earning spouse. This treatment of nonlabor income imparts a schedular aspect to the taxation of personal incomes. Similar problems arise in the treatment of income from unincorporated businesses and closely held corporations because it must be allocated between spouses.

Treatment of nonlabor earnings of dependents further complicates this issue. A common tax avoidance scheme has been to transfer nonlabor income to children, to reduce the tax on this income. It might thus be desirable to require that nonlabor income be taxable to the parents, if they file jointly, or to one of the parents, if they file separately, and not allow it to be attributed to dependents. It would probably make sense to tax labor income on the basis of the earner, as this is less subject to tax abuse, and to require dependents to file their own tax return, if their labor earnings are above some threshold.

Itemized deductions that are linked to some measure of income also pose a problem. In some countries, certain expenses, such as medical, are itemizable deductions only on the amount that exceeds a given percentage of income. By combining two taxpayers, the sum of the expense may not, however, change income increases, thus reducing the likelihood that this expense can be itemized.

Choice Between Individual and Joint Filing

Taking all of these arguments into consideration, there are several arguments in favor of requiring individual filing. First, it eliminates the marriage “tax” and “benefit” based on the progressivity of marginal tax rates. Second, it imposes no value judgment on what constitutes the standard filing unit. Thus, marginal tax rates and overall burdens are neutral with respect to marriage and other household forms. On the other hand, there are several arguments in favor of allowing joint filing. First, it gives the tax code another way of adjusting the tax burden to account for differences in ability to pay based on household composition. Second, it eases administrative complications connected with the allocation of nonlabor income and the treatment of dependents. Although, a system of individual filing can also deal with these issues.

Practices in OECD Countries

Within the Organization for Economic Cooperation and Development (OECD) countries, in recent years, the trend has been away from joint taxation and toward individual taxation. Fifteen countries use the individual as the unit of taxation (although four of these countries use different methods for taxing nonlabor income and labor income), four countries use joint taxation, and five countries give married taxpayers a choice between joint and individual taxation. Table IV.2 provides a summary of major characteristics of OECD personal income tax systems.25 The countries are grouped in the table according to the manner of labor income taxation. Column 2 indicates the major changes since 1970 in the treatment of the taxable unit. From 1970 to the present, ten countries have switched from joint to individual taxation, while three others have modified the allowable form of taxable unit.

Table IV.2.

Tax Units Under the Personal Income Tax in OECD Countries—Structural Measures, 1990

article image
article image
Source: The Tax/Benefit Position of Production Workers, Organization for Economic Cooperation and Development, Paris (1991b), pp. 267–68 with slight changes.

Countries are classified according to the taxation of labor income.

Holds only for unused portion of the pensioner’s rebate.

Join t above a ceiling.

Holds only for unused portion of the married couple’s allowance.

At federal level; varies at cantonal level.

Applicableto 1988 income.

Column 3 indicates the nature of relief for marriage. Only four countries have no tax relief for marriage. Seven countries provide relief through credits, five countries through allowances or deductions, and eight countries through income splitting or different tax schedules. As an example of tax relief through a credit, the Australian tax system allows a married taxpayer to claim an additional tax credit if he or she contributes to the maintenance of a dependent spouse. This credit diminishes in value by $1 for every $4 by which the spouse’s income exceeds a fixed amount. Thus, if the dependent spouse’s income is A$100 over the threshold, the credit is reduced by A$25, increasing the marginal tax rate by 25 percent in the phase-out range. As an example of tax relief through a deduction, the Japanese tax system allows a married taxpayer who lives with a spouse to claim an additional deduction. The level of the deduction depends upon the spouse’s income. As an example of tax relief through different tax schedules, the U.S. system allows couples to choose whether to file jointly or separately.26 The schedule for individuals has a smaller standard deduction and lower thresholds for applying higher marginal tax rates than the schedule for married couples filing jointly. Married couples filing separately, however, face a schedule which has a smaller standard deduction (income that is subtracted from gross income to arrive at taxable income for all taxpayers) and lower thresholds than the standard individual schedule. Thus, filing separately is seldom advantageous on tax grounds.

Column 4 indicates the nature of reduction in marriage relief if the spouse is gainfully employed. In eight countries, this relief is reduced if the spouse is employed or if his or her income exceeds a fixed amount, as in the examples of Australia and Japan, cited earlier. Column 5 indicates the degree to which spouses can transfer unused credits or allowances between themselves. In nine countries, unused credits or allowances may be transferred between spouses. For instance, in Iceland, the taxation is based on the individual and each taxpayer is entitled to a basic tax credit. If one spouse does not fully use the tax credit, he or she can transfer up to 80 percent of the unused credit to the spouse.

Column 6 indicates the treatment of nonlabor income. In general, nonlabor income is treated as labor income. Four countries using individual taxation, however, provide a different treatment for nonlabor income. Two countries tax nonlabor income according to the spouse with the highest labor income, and in two other countries, the income is taxed jointly. In one of these countries, the tax is divided between spouses according to their respective share of total nonlabor income. For instance, in the Netherlands, nonlabor income is taxed in the hands of the spouse with the higher labor income.

Informal unions outside of marriage pose another set of problems. Six countries have provisions that apply to informal unions between members of different sexes allowing them to be treated the same as married couples, where joint residence is the basic test for an informal union. Two countries have provisions for informal unions between members of the same sex. No pending-doubt, countries that use the joint method of taxation will face more problems with this issue in the future if these unions become more socially acceptable.

Selected Practices in Developing Countries

The choice of taxable unit varies across developing countries, although the majority of countries use individual taxation. Any schedular income tax system must rely on the individual as the filing unit. In Latin America, Argentina and Mexico use individual filing, while Brazil and Venezuela use joint filing but permit individual filing under certain conditions. In Asia, the same range of practices is found, although most countries rely on individual taxation. China, India, Indonesia, Iran, Japan, Korea, and Pakistan all use individual filing, while Taiwan uses a hybrid individual-joint filing system. Africa, Egypt, Nigeria, and South Africa use individual filing, while Kenya uses joint filing.

Conclusion

The choice of taxable unit is a complicated and difficult one for a personal income tax system. The ultimate decision involves certain judgments with respect to the equity, efficiency, and administration of the tax code. Tax systems based on individual taxation have certain virtues in that they are neutral with respect to marital status and avoid the efficiency problems that result from joint taxation. They do, however, result in different treatment of married couples with the same income, depending on the distribution of earnings between spouses. On the other hand, tax systems that are based on joint taxation while achieving equal treatment of married couples with the same income, regardless of the distribution of earnings between the spouses, create inequalities in taxation between married and un-married couples and individuals with the same income.

The Progressivity of Personal Income Tax Systems

John Norregaard

  • What is progressivity of a personal income tax system and why is it important to be able to measure it?

  • Which factors determine the level of progressivity in a given system?

  • How can progressivity in practice be measured?

It is well known that a poll tax is the least distortive tax available since it does not interfere with the saving and consumption behavior of individuals. Because of its unacceptable distributional characteristics, however, it has not been successfully applied in any country. It is safe to state that in all market economies, there is broad public support for some redistribution via public finances. The main instruments used to achieve distributional objectives are provision of free public goods and income transfers between different groups of the population, combined with the use of progressive personal income taxes.

Taxes other than the personal income tax also have important redistributional implications, particularly payroll taxes, property taxes, and consumption taxes. This section deals with three important aspects of personal income tax progressivity: how progressivity can be defined; what determines the level of progressivity in a given tax system; and how the actual level of progressivity can be empirically measured. The issue of equity and fairness is dealt with in more detail in Chapter II.

The need for measures that reflect the level of progressivity, described in more detail below, follows from the redistributional impact of income taxes being determined by a number of factors of a widely different nature. Thus, one cannot simply ascertain the level of progressivity by looking at the statutory tax schedule: in most of the tax reforms which have been implemented in recent years, nominal marginal rates have been reduced, but concurrently, tax bases have been broadened by bringing untaxed income sources within the tax net. Evaluation of the total impact of such changes on the global level of progressivity requires measures of a nature described below.

What Is Progressivity?

Income tax progressivity is normally associated with the notion of a tax schedule with an increasing rate. Progressivity may be formally defined as follows: let T(Y) represent total income tax liabilities of an individual with income Y, let m(Y) represent the marginal tax rate, and t(Y) represent the average tax rate. An income tax could then be defined as progressive when the elasticity of the tax with respect to income exceeds unity for all income levels, and regressive if the elasticity is below unity. A proportional income tax would have a unitary elasticity. In other words, progressivity implies:

(dT/T)/(dY/Y)>1, or:m(Y)/t(Y)>1<=>m(Y)-t(y)>0

This is equivalent to saying that a tax system is progressive if the marginal tax rate exceeds the average tax rate, or if the average tax rate is an increasing function of income, which is the same thing. Note first that the above relates to progression at a given point in the income scale, and therefore does not provide an unambiguous index of overall progressivity. As described in more detail in a subsequent section, different concepts have been established which measure different aspects of overall progressivity. Second, note that the requirement that the marginal tax rate exceeds the average tax rate may be satisfied in some cases which normally would not be considered to reflect a progressive tax system. As an example, consider a tax system, which is characterized by the following nominal tax rate schedule that applies to—say—income intervals of 10 units: 0 percent, 15 percent, 20 percent, and 30 percent. Above the zero-rated first bracket, the requirement that m(Y) > t(Y) is satisfied for all Y, although this system could not be perceived as a purely progressive one. For this reason, one may want to add the requirement that the second order derivative of the tax function be positive, that is, that the marginal tax rate should also be a positive function of income.27

Determinants of Progressivity

Assuming that a comprehensive (Haig-Simons) measure of gross income can be agreed upon as the relevant income measure against which to measure progressivity,28 the following four basic determinants of progressivity can be identified.

Choice of tax unit: Dependent upon whether individual or family taxation has been chosen, the ratio of marginal to average tax rates may differ for spouses with a given aggregate level of income (see preceding section).

Sources of income subject to tax: It is well known that the composition of personal income changes as one moves up in the income distribution. Generally, the role of personal capital income (i.e., interest income, dividends, and capital gains) increases with increasing levels of income. To some extent, the same applies for retirement income. This has the important implication that the level of overall progressivity (defined in more detail in the following section) depends greatly upon the extent to which different income sources are included in the tax base of a global income tax, or are exempt or subject to flat rate schedular taxation.

It is common to find that at least some sources of capital income are taxed differently from wage and salary income. Interest income in general or some specific types of interest income may not be taxed at all, or are subject to a (relatively low) flat withholding rate. Similarly, capital gains may, wholly or partly, be exempt or subject to fairly modest tax rates at the personal level. Depending on the integration system in function, dividends may be taxed differently from other personal income sources. Although many countries have moved toward more global income tax systems, none has so far introduced truly global systems with identical treatment of all components of personal income. This is true for both developed and developing countries.

To the extent that especially capital income sources are excluded from the personal income tax base, or are taxed at lower flat rates, a seemingly progressive nominal tax schedule may translate into a substantially more modest overall or real level of progressivity, taking into account all personal income sources.

Tax allowances and credits: The provision of tax allowances and thus the way taxable income is defined may substantially affect overall progressivity. By allowing, for example, full deductibility for all interest expenses, which normally increases as a proportion of income when income rises, the level of overall progressivity may, ceteris paribus, be reduced compared to a system with more modest or no interest deductibility (even taking into account that interest deductibility to some extent will be “capitalized” in the interest rate).

A special issue is the choice between tax allowances, which are deducted from gross income to arrive at taxable income, and tax credits, which are deducted from gross tax liability to arrive at final tax liability. Some OECD countries have moved from the use of tax allowances, which in combination with a progressive tax schedule benefits high incomes more than low incomes, to tax credit systems, which can be designed so that the “tax value” of the relief is the same for all taxpayers, independently of the level of income. Tax relief resulting from an allowance of a given size equals the allowance multiplied by the marginal tax rate confronting the taxpayer. Thus, under a progressive tax, the amount of the relief will invariably increase with income. On the other hand, a tax credit may, for example, be calculated as a specific amount per taxpayer, that is, independent of the level of income.

The tax schedule: Whereas the large majority of countries apply tax schedules under the personal income tax with increasing marginal nominal tax rates, it is striking how different are the ways in which this is achieved.29 Some countries have chosen bracket sys-with very elaborate “progressivity” as reflected in a substantial number of brackets and corresponding rates. Other countries have chosen systems with very few brackets and rates, in some cases with fairly wide first brackets, emphasizing administrative simplicity and the need to make to the extent possible tax withheld equal to final tax paid. Developing country tax reforms in the 1980s have generally proceeded in this direction.

An “overlapping” issue between the choice of tax schedule and tax allowances is the question of the size of the tax threshold, which is that level of (gross) income at which tax is first paid. A given level of the tax threshold, which may in itself substantially affect the level of overall progressivity, can be achieved by different means: by a basic tax allowance applying to all taxpayers; by a similar tax credit; or by zero-rating income up to a certain level. Again, it is characteristic that different countries have chosen different solutions.

Measures of Progressivity

The issue of how to measure progressivity in a given tax system has been intensively discussed in the literature during the last decade or so. There is universal agreement that no “correct” measure exists, and that—following Kiefer30—existing measures may broadly be classified into two groups: structural indices and distributional indices. Structural indices are generally based on calculations of tax liability at selected levels of income, for example average income, or fixed multiples of a single reference income. The purpose is to illuminate important aspects of the tax system in question with regard to tax liabilities at different income levels. They differ from distributional indices, which depend upon information about the entire distribution of income.31 Distributional indices may again be classified according to the measure of inequality upon which they are based. Since most distributional indices are based on the well-known concept of the concentration index—or the Gini coefficient—which again is based on the Lorenz or concentration curve, the following brief summary focuses mainly on these measures.

The importance of the inequality measure used follows from the close relationship between inequality and progressivity: if the average tax elasticity exceeds unity, posttax income will be more equally distributed than pretax income. Since the level of progressivity is reflected in this change in income inequality between the before- and after-tax situation, measures of progressivity must, either explicitly or implicitly, be based on some measure of inequality. The Lorenz or concentration curve, on which the Gini coefficient is based, is defined as the relationship between the cumulative proportion of income and the cumulative proportion of income-receiving units. The Gini coefficient may in turn be defined as one minus the ratio of the area under the Lorenz curve to the area under the diagonal or the egalitarian line as shown in Figure II.3.

These measures of progressivity may also be divided between “tax-scale invariant” and “redistribution” indices. This distinction actually captures an important aspect of the measures that has been at the center of the significant controversy over the measurement of progressivity in the last decade. The distinction may perhaps best be explained by exemplifying each of the two different types of measures.

Musgrave and Thin developed a progressivity index M which they called “effective progression,”32 based upon a comparison of the Gini coefficients for income before tax (Gb and after tax (Ga):

M=(1-Ga)/(1-Gb).

This is an indicator of the relative equality of the before- and after-tax distribution: values greater than 1 indicate a progressive tax.

Kakwani33 developed a progressivity index K based on a comparison of Gb and Gb, where Gt is equal to the Gini coefficient of taxes, calculated on the basis of a Lorenz curve showing the cumulative proportion of taxes against the cumulative proportion of income receiving units (using pretax income as the classifier). The measure is defined as:

K=GtGb,

According to this measure, a tax is judged to be progressive if the tax is more unequally distributed among taxpayers than is pretax income, thus resulting in a tax concentration curve which is more concave than the Lorenz curve (i.e., K is greater than zero).

The last decade has seen a proliferation of this kind of Gini-based measures, most of which fall within the two broad groups mentioned above. The basic difference between the two indices M and K may be explained as follows: imagine a tax system which is extremely “progressive” in the sense that all taxes fall on the richest decile of taxpayers, but with a modest total tax burden equal to, say, 1 percent of GDP. This system will generate a large value of K because the Lorenz curve for taxes will be very concave, but because of the low total tax burden, the system will hardly affect the after-tax distribution of income, and will thus simultaneously generate a low value of M. In other words, taxes looked at in isolation are very unevenly distributed among taxpayers, and thus in this sense very “progressive,” but because the total amount of taxes is so modest, they do not “move” the income distribution very much.

Alternatively, imagine a tax system where the tax burden is much more evenly distributed across deciles (although still being clearly “progressive”), but with a “Scandinavian” level of tax burden equal to about 50 percent of GDP. This system will generate a relatively large value of M because pretax and posttax distributions of income will be markedly different (taxes “move a lot of income”), but on the other hand, a relatively low value of K because of the Gini coefficient of taxes is not that different from the Gini coefficient of pretax income.

The above examples show that the measures in question illuminate two different aspects of progressivity: tax-scale invariant measures do not yield different progressivity measures when the tax paid changes with the same multiple at each income level; they depend upon the distribution of tax. Redistribution measures are invariant when the after-tax income changes by the same multiple at each income level and depend upon the distribution of after-tax income. The difference is important because alternative measures do not always give consistent rankings of different tax systems with respect to the level of progressivity. This is an argument for using several different measures, as many empirical analyses actually do.

As shown by Kakwani (1976), however, there is a formal link between the two classes of measures, as reflected in the following equation:

Ga-Gb=tK/(1-t).

where t is the average tax rate. For a given value of K, the change in inequality brought about by the tax system is an increasing function of t. If t is small, a tax system may be judged highly progressive according to the index K, but at the same time be approximately proportional according to the group of redistributive measures. Depending on which aspects of progressivity they consider most important, different authors advocate different measures.

The Gini-based progressivity measures dealt with here have, however, been criticized on the basis that, underlying any summary statistic of the level of progressivity, there is some concept of social welfare, and ideally the analysis should explicitly recognize and focus on the welfare function in question.34 In this vein, indices of inequality, such as the Gini coefficient, can be thought of as assigning weights to income transfers from richer to poorer individuals. Thus, the Gini coefficient has been criticized on the basis of first, its attaching most weight to income transfers among individuals close to the mode of the income distribution, and second, its possessing the same symmetric weighting scheme regardless of how equal or unequal the income distribution in question is. A more satisfactory system would be one in which transfers to the poor probably would be assigned a greater weight, the larger the initial inequality is. Nevertheless, the kind of measures discussed here have to be recognized for their simplicity which perhaps explains their much wider use compared to the more complicated alternatives available.

Corporate Income Tax

The Concepts of Business Income and Taxable Income

John R. King

  • What are the main issues that need to be determined in measuring the income of a business in its accounts? In what areas do the principal problems arise in practice?

  • In the area of business income measurement, how are tax laws related to accounting practice?

Measuring Business Income in Company Accounts

Business accounting systems in market economies are constructed around two principal summary statements of the financial position of a business. The first is a balance sheet which shows its assets and liabilities at a given point in time, usually the end of the accounting period. The second is an income statement (or “profit and loss account”) which shows its revenues and expenditures for a particular period of time, between two balance sheet dates.

The balance sheet may classify assets and liabilities in a variety of ways, but a fundamental distinction that is usually drawn is that between “monetary” assets and liabilities whose value is fixed in nominal terms (such as cash, accounts receivable and payable, and borrowing), and “nonmonetary” assets such as land and buildings, plant and equipment, inventories, and investments in subsidiaries or the shares of other companies. The net worth of the business is the difference between the balance sheet value of its assets and of its liabilities. Double-entry principles ensure that changes in the net worth of the business between two balance sheet dates are equal to the amount of profit earned in that period, as shown in the profit and loss account—except insofar as profits are distributed to shareholders.

In constructing such accounting systems, two fundamental issues that arise concern the timing of recognition of revenues and expenditures in the income statement, and the principles governing the valuation of assets in the balance sheet.

Timing

Except in the case of very small businesses, income is generally measured on an “accruals” basis. Sales revenues are recorded when goods are supplied (or an invoice is issued), rather than at the time the corresponding cash payment is received. Expenses attributed to a particular period consist of assets used up in obtaining the revenues of that period. Capital expenditures are thus spread over the number of periods that each capital asset is used to generate revenues.

Valuation

Accounting systems value assets (and liabilities), for a balance sheet, in different ways. Because of the close relationship between the balance sheet of the business and its income statement, different choices of valuation basis may have important implications for the measurement of income attributable to the owners of the business. For most capital assets, the major practical alternatives are original cost and market value (measured on a disposal or replacement cost basis).

A strong case can be made that, in principle, it would be appropriate to value assets in business accounts on the basis of their opportunity cost or “value to the owner.”35 A profit-maximizing firm would sell a particular asset that it owns if the disposal proceeds were greater than the discounted present value of expected net earnings from using the asset in production. The higher of those two amounts may be referred to as the “economic value” of the asset to the firm. In most cases, however, this will not be the asset’s opportunity cost. If the firm were to be deprived of the asset, and it could be replaced for less than its economic value, then the profit-maximizing firm would replace it; the loss suffered by the firm would thus be limited to the asset’s replacement cost. Hence, “value to the owner” principles imply that assets should be valued at the lower of their replacement cost and economic value—where economic value is defined as the higher of net realizable value on disposal, and the discounted present value of future net earnings from the asset.

In practice, however, accounting systems value most assets on the basis of their original historical cost—after adjusting, where appropriate, for past depreciation (amortization) that has been charged to the profit and loss account. (In Latin America, some accounting systems also adjust the original historical cost in accordance with changes in the general price level since the date the asset was acquired.) The main justification for this choice is that historical cost typically provides a more objective standard than other valuation bases such as replacement cost or value to the owner.

When assets are valued at historical cost, no capital gain or loss will be recorded in the income statement unless and until an asset is disposed of and the gain is “realized” and recognized. With valuation at current market prices, however, such gains or losses need to be brought into the accounts in each period—either by means of an entry in the income statement, or by a change to a “capital reserve” liability in the balance sheet.

Problem Areas in Business Income Measurement

Most of the major practical problems that arise in measuring the income of a business in a particular accounting period concern the capital assets and liabilities of the business. This section summarizes some of these difficulties, which are considered in more detail in later sections.

Depreciation

As noted above, accounting systems allocate the costs of capital assets over time—in principle, according to the extent to which those assets are “used up” in generating revenues. This concept of the cost of a capital item in a particular period cannot usually be measured straightforwardly, however. Accordingly, conventional rules of thumb are applied to derive an estimate of the depreciation of capital assets; for example, the (original) cost of the asset may be allocated evenly over its expected useful life.

Inventories valuation

In the case of inventories, when large numbers of assets of a particular type are held at any point in time but the actual acquisition and disposal dates of particular items cannot be separately identified, valuation at cost requires that assumptions be made about the pattern of acquisition and disposal. The main alternatives adopted by accounting systems as conventional assumptions, in practice, are “first in, first out” (FIFO) and “last in, first out” (LIFO). Valuation at cost, on a FIFO basis, will approximate valuation at current market values. LIFO valuation will result in lower values when inventories’ prices are rising over time; it will also result in lower increases in inventories’ values, at least when the volume of inventories held by the business is constant or increasing over time.

Intangible assets and “goodwill”

Expenditures by a business on advertising or re-search and product development create assets that generate revenues in the future. Accordingly, such expenditures should, in principle, be capitalized and written off (in the same way as fixed capital expenditures) over those future periods, rather than treated as costs in the current period. Estimating an appropriate depreciation charge for intangibles produced by the business itself is, however, particularly uncertain, and accounting systems differ widely in their treatment of these expenditures. Similarly, there are wide differences in the treatment of “goodwill,” which is the difference between what a business pays when it acquires another business as a going concern and the book value of the assets of the business that is acquired.

Exchange rate changes

When a business has assets in a foreign country (such as a foreign branch operation), or an asset or liability which is denominated in a foreign currency, exchange rate changes may have implications both for its balance sheet and for its income statement. The appropriate method of accounting for these changes has, however, been a matter of controversy in several countries; the accounting practices that have been adopted vary. Two main issues arise. The first is whether the particular asset or liability should be valued in domestic currency terms in the balance sheet of the business at its original cost, or at current market value (i.e., using the current exchange rate). If market value is chosen, the second issue is whether gains and losses arising from exchange rate changes should be included along with other revenues or expenditures in the income statement, or taken instead to a capital reserve in the balance sheet.

Domestic price level changes

The most difficult and controversial problems of business income accounting arise under conditions of inflation, which may distort the measurement of income in several areas. Depreciation based on historical cost will understate the true cost of assets to the business in the current period. Increases in the book value of inventories, particularly when measured on FIFO assumptions, will overstate the increased value of these assets that is properly included in the income of the business. Similarly, capital gains recorded in the accounts—on either a realization or an accrual basis—will overstate the true gains of the business. On the other hand, the accounts will not show as income, as they should, the capital gain that accrues to the business as a result of the fall in value of net liabilities that are fixed in nominal terms—that is, its net “monetary” liabilities, primarily debt.

The extent of these distortions depends, of course, on the valuation conventions that are adopted in the accounts. For example, if depreciation is based on asset values measured at current replacement cost, the amount shown in the accounts will be largely unaffected by inflation, except for relative price changes.

Tax Laws and Accounting Practice

Income tax laws do not usually seek to define “in-come” comprehensively, either in general terms or in detail. Most practical issues of business income measurement for tax purposes are left to be determined by “generally accepted accounting practice,” which may be governed by a separate accounting law (or other legislation such as a Companies Act), and which is sometimes codified in national standards of accounting practice. The relationship between tax laws and accounting practice varies. In some countries (such as Germany), the published accounts of enterprises must conform to specific provisions in the tax law. More commonly, specific provisions in the income tax law, or associated regulations, override practices which may be adopted in published accounts, in certain areas, in determining the tax liabilities of a business.

Almost all income tax laws specify the depreciation rates that must be applied in the case of particular assets. They also normally specify certain costs which are not allowable as deductions for tax purposes. Such costs commonly include, among other things, certain “provisions” that may be made in the accounts (such as a general provision for bad debts); payments of certain other taxes (together with fines and penalties for late tax payment); and certain expenditures which the government may wish to discourage (such as business entertainment), or which are not incurred “wholly and exclusively” for business purposes. In addition, income tax laws generally restrict the valuation conventions that may be employed in measuring the income of a business. LIFO valuation of inventories, for example, is not allowed for tax purposes in many countries. In general, the main purpose of these specific provisions in income tax laws is to make the measure of income for tax purposes more objective than it may sometimes be in commercial accounts that are drawn up to serve other purposes.

Finally, income tax laws must provide for specific deductions that are not necessary in the commercial accounts of a business, such as any deduction for losses brought forward from earlier years.

Depreciation Schedules

Dale Chua

  • What are the main types of depreciation schedules?

  • What are pooled asset accounts?

  • Some examples of tax depreciation and economic depreciation rates.

An ideal tax depreciation schedule for an asset is one that is designed to provide as closely as possible a tax deduction profile over time that mimics the profile of the asset’s true economic depreciation. The depreciation expenses or capital cost allowance charged against revenue should represent, if possible, the real decline in value of the depreciating asset. Any departure from this rule will imply that the taxable profits of the firm will also deviate from the ideal tax base. In other words, the taxable profits of a firm will be either over- or understated in real terms when deductible tax depreciation deviates from the true economic depreciation of an asset.

Depreciation charges for an asset account for the loss in value from wear and tear, economic obsolescence and/or the change in monetary value of the asset over time. In general, an asset must have a useful life of more than one year to qualify for a depreciation deduction. This implies that virtually all machinery, vehicles, equipment, plants, and buildings will qualify. Land, on the other hand, is not depreciable because it has an infinite useful life for most purposes.

Given that it is administratively impossible to monitor and impractical to design a depreciation system that will track over time the real economic depreciation charges for each asset group to allow the real yearly decline in each asset value to be written off as revenue, a set of arbitrary rules is generally applied. These rules or depreciation schedules become applicable when an asset is acquired, and are consistently applied to the asset in subsequent years until it is retired or sold. In what follows, we discuss two commonly adopted depreciation rules.

The Straight-Line Method

Under the straight-line method, the historic cost of the depreciating asset is apportioned in equal amounts for deduction over the period of its estimated economic life. The latter is normally specified in the tax code for each broad group of assets. For instance, if a piece of office equipment were to be purchased for $800 and the tax code specifies that this asset group can be written off at a rate of 10 percent, then, assuming a zero salvage value, the yearly depreciation deduction will be $80 for ten years. The example given here will fall under a particular asset class that yields a straight-line depreciation rate of 10 percent. A longerlived asset—say, buildings—will be classified into another class yielding a lower straight-line rate of, for example, 2 percent. In this case, a building costing $300,000 will attract an annual depreciation charge of $6,000 for 50 years.

The Declining-Balance Method

In contrast to the straight-line depreciation method, the declining-balance method permits larger deductions in the earlier years of the useful life of an asset and smaller deductions toward the later years. Under this method, the computed depreciation charge is obtained by multiplying a fixed rate by the asset’s book value. The asset’s book value itself, unlike historical cost, declines from year to year by the amount of depreciation accumulated in earlier years. In practice, this method is also employed in conjunction with asset classes. Like the straight-line method, each group of depreciable assets is assigned to an asset class and written off according to the declining-balance rate applicable to that class.36 The rates applicable to each class are supposed to accord roughly with the useful life of the asset. But, in practice, this may or may not be the case.

There are many instances where the rate of economic depreciation may not correspond to the tax depreciation rate. Again, for the sake of simplicity, if the total number of asset classes is deliberately compressed so that depreciation schedules serve little to discriminate between the treatment of longer-lived investments and shorter-lived ones, then the former will benefit relative to the latter because longer-lived investments are given the same, higher depreciation rate as shorter-lived assets. This is true for both the straight-line and declining-balance methods. All other things equal, whether the declining-balance method would be more advantageous to a firm than the straight-line method, would depend on the given depreciation rates and the applicable discount factor. Since the two methods yield different time streams of depreciation, this comparison must be made on a present-value basis.

Other Depreciation Rules

Less common methods of depreciation include (1) initial allowance, where a more generous portion of the cost of capital assets is recovered in the first year, and the remaining undepreciated capital base is depreciated either by a straight-line or declining-balance method in subsequent years; (2) the sum-of-years-dig its method (a variant of declining-balance method), where capital assets are depreciated each year at a linearly declining rate; (3) immediate expensing, where the total amount of the investment is deducted as cost in the first year of operation; and (4) switching (between declining-balance and straight-line methods), where the capital asset is depreciated according to the declining-balance method for a certain period starting from the purchase date, at a given declining-balance rate. At the end of the period, a switch-over is affected where the balance of the undepreciated capital base is depreciated according to the straight-line method at a straight-line rate.37

Pooled Asset Accounts

To simplify the treatment of depreciation deductions, some countries have pooled fixed assets into several classes for tax depreciation. By placing all depreciable assets, which may or may not include real property, into a small number of pools (or classes) and by allowing a deduction of a specific percentage of the balance in each account every year, the number of asset classes will effectively be reduced.38 The justification for prescribing a correct depreciation schedule for each asset so that, in theory, the ideal tax base of a firm can be derived, is replaced in this instance by the desire for greater simplification under pooling.

Generally speaking, it is possible to design as many asset classes as one desires so as to discriminate better in the treatment of long-lived and short-lived assets. But, for administrative ease, the number of classes of different assets is usually kept to a desirable minimum. It is not uncommon, for example, for the authority to group all vehicles into one asset class, even though a truck may depreciate at a faster rate than an office car; all plant and equipment into another class, even though a lathe may depreciate at a different rate than a furnace; all office equipment into a third class, even though office furniture may depreciate slower than computers, and so on. Compressing the number of asset classes tends to favor longer-lived assets over short-lived ones. Therefore, other than keeping the number to a manageable size, there is no reason to recommend a reduction in the number of asset classes.

In practice, pooled accounts are normally open-ended accounts.39 Such an account is managed by adding annually the total costs of all newly acquired assets less the proceeds from the disposal of any old assets to the balance of the pooled account. This system, which is compatible only with the declining-balance method, is attractive because it is simple to operate. It also solves the tax treatment problem concerning the disposal of assets. The deductible depreciation expense for the firm is the determined statutory depreciation rate multiplied by the balance value of the appropriate type of account every year.

Illustrative tax depreciation rates and evidence on the rates of economic depreciation

A set of tax depreciation rates for buildings and machinery is presented in Table IV.3 for selected OECD countries.40 It is instructive to note the diverse methods and rates employed by the different countries in the table. For example, the straight-line method was employed throughout the 1980s by Italy, Luxembourg, Portugal, and, since 1991, Greece. On the other hand, Japan, Switzerland, and Spain (for the most part) have adopted the declining-balance method. In the 1980s, the United Kingdom provided an initial allowance of as much as 50 percent for buildings and machinery, while Ireland granted immediate expensing to both. The switching method (from declining-balance to straight-line) is used by Belgium (for buildings and machinery), and by France and Germany (for machinery only). Countries that implement another form of switching (from a higher rate to a lower one according to the straight-line method) are Denmark (for buildings and machinery) and Germany (for buildings only). Lastly, the United States adopted in the mid-1980s a complex system of depreciation provisions under the accelerated cost recovery system, where machinery was depreciated along a straight-line method at 8 percent for the first year, 14 percent for the second year, 12 percent for the third year, 10 percent for the next three years, and 9 percent for the next four years.

Table IV.3.

Typical Depreciation Rates on Industrial Buildings and Machinery

article image
Source: Report of the Committee of Independent Experts on Company Taxation, Commission of the European Communities.Key: SL=straight line; DB=declining balance.10KDBx7 then 5%SL means 10 percent declining-balance depreciation for seven years followed by depreciation at 5 percent straight-line until the asset is fully depreciated.

The accelerated cost recovery system (ACRS) in the United States during the mid−1980s involved complex depreciation provisions; typical straight-line depreciation rates for machinery were 8 percent for the first year, 14 percent for the second year, 12 percent for the third year, 10 percent for the next three years, and 9 percent for the next four years. Industrial buildings might typically be depreciated at 6 percent for 10 years and 5 percent thereafter.

By contrast, Table IV.4 provides the results of an empirical study showing the various asset classes estimated in the United States.41 Although it is in no way a definitive statement on economic depreciation rates because of difficult problems underlying such empirical work, Table IV.4 gives examples of how quickly various capital assets would decline in value in a year through normal use. From a group of 22 producer durable equipment and machinery items, and 10 groups of private nonresidential structures, the study found the annual percentage rates of decline to be (1) between 1.9 percent and 5.6 percent for structures; (2) between 6.6 percent and 18.3 percent for equipment (other than automobiles, trucks, and office equipment); and (3) between 25 percent and 33.3 percent for automobiles, trucks, and office equipment. The average rate of depreciation for all producer durable equipment was 13.3 percent and that of structures was 3.7 percent.

Table IV.4.

Asset Classes and Rates of Economic Depreciation

(Annual percentage rates of decline)

article image
Source: Hulten and Wykoff (1981).

Inventory Valuation

Dale Chua

  • What are the conventional approaches to valuing inventory?

  • What are the tax implications for choosing FIFO or LIFO?

As part of doing business, a firm holds inventory stocks such as raw materials, work-in-process, and final goods that will eventually be used to produce revenue for the firm. For a firm, holding a stock of current assets, such as inventory, is associated with three types of costs. First, there is the storage cost. This is deducted against revenue whenever it is incurred. Second, there is the cost associated with financing the inventory stock. Like other capital assets, only the interest cost of financing is a deductible expense, equity financing is not directly deductible against revenue. Third, there is the actual cost of the good held as inventory. The cost of the good becomes a deductible item only when the good in stock is removed from the inventory holding and used in the production process. This is the main issue addressed in this section.

Under standard accounting principles, inventories are generally recorded at the historical cost of acquisition or production,42 although the purchasing price of an inventory stock may have a different current value at some later point in time because of changing market conditions or rising prices. Two basic criteria have been used for the valuation of inventories, with different impact on costs, and hence different tax liability. Because of this, an enterprise must decide on a particular inventory costing principle, and once a method is adopted, most tax laws would not permit an enterprise to switch from one method of inventory valuation to another within an unjustifiably short time.

The first convention is FIFO (first-in-first-out), which assumes that the flow of goods is such that the earliest or first unit acquired is charged to cost first. The second convention is LIFO (last-in-first-out), which assumes that the most recent or last unit acquired is the first that is charged to cost.43 Notice that the significance of adopting a particular convention quickly evaporates when prices are stable because LIFO and FIFO will yield identical costing. During inflationary periods, however, these methods have a different cost implication and tax liability for an enterprise.

Impact of Taxes on Method of Inventory Valuation

During inflationary periods, which cost accounting convention is most likely to yield the greatest benefits to an enterprise? Under LIFO, the higher-cost goods (those acquired later) are charged to cost first before the lower-cost ones (those acquired earlier). Therefore, the costs of stock at the end of the period are calculated for the inventories sold valued at the most recent price. As a result, the nominal profits calculated are less likely to be distorted by inflation; that is, LIFO will lead to lower reported profits and the tax liability of the firm will not be unduly increased with inflation. In other words, the “tax holding cost” of inventories will not be very high under LIFO. For this reason, at least for growing firms, a few experts have recommended that this method be adopted during times when prices rise.44

The opposite is true under FIFO where lower-cost goods acquired earlier are charged to cost before the more recently acquired, higher-cost ones. Therefore, the costs of stock at year end are calculated on the basis of the price of earlier acquisitions, resulting in inventories sold being valued, during inflation periods, at below their replenishment cost. In this way, the “tax holding cost” of inventories will be higher than under LIFO because this method of accounting for the costs of stock at the end of the period results in higher reported nominal profits and higher taxes for a FIFO firm.45

Assuming a single corporate tax rate, the present value of taxes payable is thus higher under FIFO than under LIFO. A LIFO firm, however, will be less able to benefit from the calculus of accounting when its inventory stock is drawn down because whenever a LIFO firm dips into lower cost stock accumulated earlier, it cannot avoid expensing them at a lower cost.46 Hence, a LIFO firm cannot avoid having to pay more taxes on its higher profits that arise from the depletion of its inventory stock. In the extreme, when a LIFO firm completely exhausts its inventory stock, its marginal inventory costs to be expensed against revenue will be exactly identical to those of a FIFO firm. Therefore, the marginal tax liability for a LIFO firm will be the same as that of a FIFO firm when its inventory stock is exhausted. If so, LIFO firms will have a tax incentive not to deplete their inventory stock or, alternatively, LIFO firms will have an incentive to hold larger inventory stock.47

In practice, however, there are nontax factors that will also influence the FIFO/LIFO choice. For example, by following the designated convention, a LIFO firm will tend to have lower reported earnings (if it is required by law to have conformity between book and tax accounts) as well as a lower book value. As a market signal, such characteristics may not be well received. Hence, it may not be in the interest of a firm to adopt LIFO, given that in practice, it has to approach the market for funds. By the same token, a LIFO firm tends also to exhibit greater fluctuation in its reported profits, especially during periods when inventories are being depleted, whereas FIFO firms tend to show more stability in their profit stream. Furthermore, the bookkeeping cost for LIFO may be higher than for FIFO.48 In summary, although tax considerations in times of inflation may suggest that a firm should adopt the LIFO convention if it seeks to maximize its after-tax cash flow, other firm-characteristic considerations may provide a counterweight to its widespread use.

Loss Carryforward and Loss Carryback

Dale Chua

  • What are loss carryforward and loss carryback?

  • Shouldfull loss offset be a feature of a corporate income tax?

  • What is the impact of imperfect loss offsetfor investment?

By the very nature of doing business, firms under-take risks. A successful firm is rewarded with profits, and corporate tax laws worldwide require all profitable firms to pay corporate taxes. Now, what happens if a risk-taking firm incurs losses? Because a well-designed tax system should not discriminate against risk-taking enterprises that may end up making losses, operational losses incurred by a firm should ideally be treated symmetrically with profits. In other words, given that a profitable firm pays taxes, a loss-making firm should receive a tax refund. Although most corporate tax systems recognize the need to provide some tax relief to firms that report tax losses, a perfect loss-offsetting mechanism that treats positive and negative tax liabilities symmetrically remains ideal.

Tax authorities are unwilling to introduce full refundability for corporate taxes because it implies a loss of government revenue. (Full refundability may also be subject to abuse and the cost of monitoring, potentially high.) Although a fear of revenue losses is a legitimate concern in times of budgetary constraints, whether this fear is substantiated over the longer term remains an issue in the loss-offsetting literature. Proponents argue that the provision of loss offsetting to tax-loss firms today will, on average, be accompanied by future tax revenue as these firms regain their competitive position, even though the present value of the future tax dollar may not be known with certainty. From the lack of concrete evidence of outright, full refundability to individual firms, it is obvious that tax authorities are not persuaded by this argument.49 Their preoccupation in an imperfect world is the fall in net tax revenue collected in any given year if perfect loss offsetting is grafted onto the corporate tax law. Furthermore, opponents argue that changing economic conditions that bring new opportunities for investment also imply a necessary foreclosure of less efficient and economically obsolete firms. Full loss offset prolongs the breakup of such firms, thereby tying up capital and scarce resources.

Loss Carryforward and Loss Carryback

In practice, tax laws provide less than full loss offset for tax-loss firms.50 Most tax systems permit tax losses to be carried forward either indefinitely or over a fixed period of time. Such losses, however, are carried forward without interest. Naturally, allowing carryforward with interest could be equivalent, in present value terms, to an immediate refund assuming they are ultimately usable. Since the present value of loss carryfor-ward without interest diminishes with time, the future tax savings derived from such a provision will likewise fall over time.

Some countries, however, permit limited loss carry-backs. Loss carrybacks are usually restricted to those tax-loss firms which have paid some taxes during the years prior to incurring a loss. Provided that the current year tax loss is not greater than the sum of taxes paid in earlier years—usually restricted to three or five years—a firm permitted to carry back its losses would receive a tax refund from the tax authority. This is a limited tax loss provision because it is plausible that a firm may have a particularly large current year tax loss that is greater than potential carrybacks. Or, alternatively, a firm may string up a series of tax losses in consecutive years that will be larger than the potential carrybacks.

Losses that cannot be carried back can usually be carried forward to be offset against future tax liability. Although loss carryback can put strain on the government budget, its main weakness is that it puts newly established firms at a comparative tax disadvantage compared with more mature firms. The reason for this is that new firms will have no carryback and therefore will not be able to benefit from this provision.

Imperfect Loss Offset and Its Effect on Investment

How does a partial loss offset affect the incentive structure of a corporate tax? The two areas most seriously affected by an imperfect loss-offset provision are (1) new investment and (2) the financial structure of a firm. First, imperfect loss offset dampens a firm’s incentive to invest. To see this, recall that for any given firm, the cost of investment is lowered when depreciation allowance can be charged against revenue whenever they are accrued. In the case of a tax-loss firm, accrued depreciation charges may not be realized immediately when there is less than perfect refundability, although it will be absorbed at a later point in time. The present value of a loss deduction carried forward without interest, however, is less than if accrual and realization occur at the same time. The result is that the investment plan for some firms could be changed. For example, if the corporate income tax favors short-lived investment, then compared to a profitable firm, a tax-loss firm will have a greater incentive to invest in long-lived investment. This is because a tax-loss firm gains more from the postponement of tax payment from the asset’s earnings compared with the loss from the postponement of its tax depreciation benefits.51 Hence, profitable firms are more inclined to invest in short-lived investment projects than tax-loss firms.

Second, a less than perfect loss offset impacts on a firm’s financial structure because fully deductible interest cost of financing is worth less to a tax-loss firm if the full interest payment cannot be deducted at cost when it is accrued. A firm in a tax-loss position is therefore less likely to favor debt financing than a profitable firm. But the argument does not stop here. Not only does imperfect loss offset affect a firm’s debt-equity structure, but a chosen financial structure will also influence a firm’s subsequent profitability through its future interest deductions. Hence, imperfect loss offset may be even more costly for a highly levered, tax-loss firm. Turning the argument around, a more generous tax-loss offset provision will induce greater corporate borrowing.

Full Loss-Offsetting Mechanisms

Although full loss offsetting is rarely ever implemented, for completeness, this section discusses two ways to achieve it if necessary.52 The first is for the tax authority to provide a tax credit equal to the value of the tax write-off of losses. Under this method, a refund is paid to the tax-loss firm, just as a tax is paid when the firm is making profit. It has the advantage of injecting cash flow into the tax-loss firm but it could be a very costly option for the government in times of budgetary constraints.

A second way to achieve full loss offsetting is to provide the losses carried forward with interest, at the nominal market rate. This will be equivalent to the immediate refund in present value terms as long as the corporate tax rate remains unchanged over time. In case the enterprise is liquidated, tax losses carried forward with interest should be applied to the revenue received from the sales of assets. A tax refund adjustment could then be made at that time if the revenue from asset sales is too small to be fully written off against loss carried forward. This method is less costly for the tax authority but it provides little immediate help for cash-strapped tax-loss firms.

Inflation Adjustment

Dale Chua

  • Why is inflation adjustment necessary?

  • Whatfactors cause taxable profits to be mismeasured and why?

  • What are the common schemes for inflation adjustment?

Inflation presents a basic problem for all unindexed corporate income taxes. A firm’s corporate tax liability is generally based on a measure of its profits derived from historic cost accounts, and such accounts mis-measure the real profits of the firm during inflation. The consequence is normally an inflated tax liability for the enterprise. Therefore, in the interest of maximizing its after-tax returns to its shareholders, an enterprise may undertake those activities that would generate the most favorable tax treatment so as to minimize its tax liability rather than those that would yield the highest returns to society. Furthermore, the full impact of inflation on an enterprise also depends on its financial structure, the kind of capital employed in its operations, and its need to hold inventory stocks in the course of doing business. Hence, not only does the tax system distort the economic behavior of each firm during times of rising prices, it also creates interfirm distortions that are dependent on the basic characteristics of each firm.

Why Profit Is Mismeasured with Inflation

The reason why profit is mismeasured is because the balance sheet of an enterprise contains items of a nonmonetary nature (such as buildings and machinery) as well as items of a monetary nature (such as accounts payable and receivable, and cash). A comprehensive adjustment system must account for gains and losses on liabilities and assets fixed in money terms and some other items.53

Specifically, there are five areas of concern when inflation is present. First, the real value of depreciation allowances is lowered with inflation. Because depreciation is generally based on historical cost accounting, its value may be far lower than the real cost of depreciation. As a result, the real tax liability of a firm is over-reported. In the absence of any compensation, the unintended side effect will be to discourage firms from undertaking large capital investments in general and those that are long-lived in particular.

Second, similar to depreciation allowances, without inflation adjustment, the value of loss carried forward to a tax-loss firm will be eroded by inflation.54

Third, tax systems based on FIFO conventions effectively treat the increase in the value of stock inventories as a source of income, with no allowance for that part which is due to inflation; a firm holding stock inventories will be obliged to pay taxes on a fictional source of income. In the absence of inflation adjustment, such a tax system will induce a suboptimal level of inventory investment.55

Fourth, without inflation adjustment, enterprises are permitted to deduct as costs the decline in the real value of their debts due to inflation. This is because the full nominal cost of debt finance is a deductible expense even though an ideal measurement of real profit would deduct only the real interest cost. Because nominal interest rates tend to rise with inflation, the real value of the firm’s debt falls when inflation is anticipated. By allowing the firm a full deduction for the nominal interest cost, the tax system allows the firm to deduct part of the repayment of principal, over and above the real interest cost of debt financing. Hence, the true profit of the firm is understated. Therefore, without inflation adjustment of interest, a firm seeking to maximize its after-tax returns will be induced to undertake more debt financing with inflation.56 This unintended tax-induced effect will lead to higher leverage ratios.

Fifth, inflation creates illusionary capital gains on assets and liabilities. To illustrate, suppose, at the risk of oversimplification, that an enterprise holds an asset, A, which earns a return. In addition to the income earned through its operation, the taxable income of the enterprise would also include (1) income from the holding of assets, rA, where r is the rate of return on A, and (2) the accrued capital gain or loss, ΔA, on the asset held over the period. The taxable income base57 for the enterprise will be

Y=X+rA+ΔA,

where X is profits from the operation of the firm (excluding the holding of asset A). Ideally, with inflation, the taxable base should be defined in real terms as:

Y/P=X/P+rA/P+Δ(A/P),

where Δ(A/P) is the change in the real value of assets held over the accounting period. This can be rewritten as:

Y/P=X/P+rA/P+ΔA/P-ΔP/P(A/P).

If this is the base used for taxation, inflation will not present any problem. But, the enterprise’s tax liability is based on nominal terms, and the nominal income tax base consistent

Y=X+rA+ΔA-πA,

where π is the annual inflation rate. Therefore, the real asset value with inflation should be

rA+ΔA-πA,

where πA is the loss in the real asset value due to inflation. Because there are no provisions for the loss in the real asset value in inflationary periods, a corporate income tax, like the personal income tax, imposes a burden on the enterprise through the taxation of illusionary capital gains.

In sum, on the basis of historical cost accounting alone, inflation could either overreport or underreport the real profit of an enterprise. Therefore, an enterprise may be either better off or worse off when adopting a historical cost accounting convention. The magnitude of mismeasurement of taxable profits will be dictated by the strength of opposite forces delineated above, including some firm-specific characteristics such as the debt-equity ratio and the capital investments that are undertaken.

Adjustments Scheme: Balance Sheet Versus Single Account Approach

To tackle the problems associated with the mismeasurement of taxable income in inflationary periods, this section considers two inflation adjustment schemes: (1) an overall approach and (2) a partial account-by-account approach.

Balance Sheet Approach

A full balance sheet adjustment approach, or a comprehensive profit adjustment scheme, is based on the indexation of all assets and liabilities in the balance sheet. Both assets and liabilities are revalued frequently, usually yearly, to correct for an over- or under-statement of income during high inflation periods.58 The principle behind this approach is that a monetary correction is made to the unadjusted profits and income statement to arrive at an adjusted tax base on which the corporate tax is then applied, while in the process, this approach accounts for the impact of inflation on all aspects of the firm’s operations. This approach is made operational through the accounting identity:

Assets - Liabilities + Net Worth.

Since assets and liabilities comprise in each a monetary component and a nonmonetary component, the above identity can be expressed as:

Monetary Liabilities - Monetary Assets =

Nonmonetary Assets - Nonmonetary Liabilities --Net Worth.

From an economic viewpoint, the adjustment for the inflation-generated profits and losses for each period can be made through either the monetary side or the nonmonetary side of the accounting identity. But, an adjustment through the monetary side creates some accounting difficulties. Accordingly, to overcome this, established adjustment mechanisms in South American countries are affected primarily through the nonmonetary side, whereby the inflation-generated profits and losses are given by:

(Nonmonetary Assets - Nonmonetary Liabilities --Net Worth)

multiplied by the monetary erosion factor,

where the monetary erosion factor is defined as the ratio of the annual rate of inflation to one plus the annual rate of inflation.

This method of correction for inflation requires all nonmonetary assets and nonmonetary liabilities to be appropriately indexed. First, the values of all nonmonetary assets such as inventory, depreciation charges (both actual and accumulated), and all other fixed assets are adjusted upward for inflation through various price indices. This results in an equivalent monetary correction gain in the income statement of the enterprise. Second, nonmonetary liabilities such as indexed debt and debt denominated in foreign currencies, among other items, are likewise revalued to account for inflation. The difference between the adjusted (up-ward) values of these nonmonetary liabilities and their nominal accounting values will yield a corresponding monetary correction loss in the income statement of the enterprise. Finally, the initial net worth is also corrected by an appropriate index, and would also give rise to a monetary correction in the income statement.

The issue concerning the appropriate index is important for a full balance sheet approach. For instance, the use of a general price index such as the CPI for all adjustments may be adequate only for some nonmonetary liabilities such as indexed debt, but it may not be sufficient to capture the true cost of depreciation charges on assets. The following questions will continue to present challenging problems for inflation adjustment: (1) Is there a general price index of depreciable assets or even for a specific class of assets? and (2) If one exists, how accurately does it measure the true, real value of depreciation? To minimize this concern, whenever possible, a set of price indexes covering as many different nonmonetary liabilities and assets as possible should be constructed and updated regularly.

Single Account Approach

With the same goal of correcting for inflation-induced distortion, the single account approach provides piecemeal, partial adjustment for an enterprise. Rather than being comprehensive, adjustments are made only to selected individual accounts, for example, the capital depreciation account or the inventory account. Either these adjustments could be ad hoc with only a temporary effect, such as a once-and-for-all adjustment, or they could be periodic, whereby an annual correction is allowed. Like the full balance sheet approach, these adjustments encounter the same difficulties underlying the choice of a proper price index for adjustment.59

Loan Loss Provisioning

Julio Escolano

  • How is bad debt owned by banks treated for tax purposes?

  • What are the advantages and disadvantages of different tax treatments of bad debt?

  • Should bad debt allowances be tax deductible?

An important part of a business’ worth is kept in the form of loans, credit, or financial claims on others. Different forms of financial claims constitute the majority of assets of commercial banks and other financial institutions. To determine business income, for either tax or book purposes, it is essential to assess the change in the value of these assets that takes place during the relevant fiscal or accounting period. When financial assets are frequently transacted, their market price provides an adequate method to ascertain their current value. Market prices, however, are not readily available for many financial claims such as loans, consumer credit, etc. Nevertheless, some of these assets may become wholly or partially worthless before maturity if circumstances show that they are likely to be uncollectible. Hence, specific rules for the tax or financial treatment of bad debt and its relation to taxable or financial income need to be established.

This section focuses on the tax treatment—as opposed to regulatory or book treatment—of bad debt of banks and financial institutions. Loan losses of financial institutions are considered to have a potentially larger destabilizing effect than capital losses in other industries, and consequently, the tax treatment of bad debt of banks and financial institutions, in most countries, follows specific regulations.60 Furthermore, financial institutions are subject to tight financial regulations and the regulatory treatment of loan losses has influenced and, on occasion, determined their treatment by the tax authority. Nevertheless, it should be noted that the principles underlying financial accounting regulations and tax accounting are not necessarily the same. Financial regulations, aiming to preserve the integrity of the balance sheet and the soundness of the financial system, will encourage or force the provision of reserves, thereby diminishing income. Tax accounting, on the other hand, will define taxable income seeking to match economic income as closely as possible to minimize distortions and tax avoidance. This implies that provisions for some reserves, even if they are mandated by financial regulations, do not need to be tax deductible.

Deduction for Bad Debts

Historically, two methods have been used to compute the deduction for bad debts. These are the charge-off and the reserve methods.

Charge-off method

The charge-off method recognizes an expense for bad debts only as they actually become either wholly or partially worthless. At such time as a receivable is determined to be uncollectible in whole or in part, the receivable is reduced by the amount of uncollectibility, and an expense is recognized in an equal amount. Thus, the bad debt deduction claimed for any year must be supported by showing that the debt had some value at the beginning of the year and that some change in the debtor’s condition prompting the loss occurred during the year. If an amount previously charged off as uncollectible is later recovered, the recovery is treated as a separate income item at the time of collection or reintroduction of the loan in the books.

In determining whether a debt is worthless, all pertinent evidence, including continual nonperformance, adequacy of the collateral and financial condition of the debtor may be considered. An inherent difficulty in identifying the year of deduction is that worthlessness often results from a gradual deterioration in the debtor’s financial condition rather than from an easily identifiable event. Simple mechanical rules can be used to establish partial or total uncollectibility for some categories of standardized loans such as consumer credit, real estate loans, etc. Less standardized loan contracts and those with face values above some threshold may require specific evaluation.61

Reserve method

Under the reserve method, receivables are recorded at their face value until they become effectively worthless. A reserve account is set up, however, as an allowance against the eventuality that some of the receivables may prove to be uncollectible.62 For tax purposes, the balance of this reserve cannot exceed certain limits, which depend on the type of reserve. The annual deduction for bad debts is the amount necessary to bring the initial balance of the reserve, adjusted for actual bad debts and recoveries that occurred during the year, to the allowed ending balance.63 Thus, it is the annual addition to the reserve that reduces pretax income while actual loan losses, when they materialize, are charged against the reserve, leaving income unchanged.64 Reserves for bad debts can be classified in two main categories according to whether they are of a general nature or whether they are linked to specific loans with identified loss potential.

General reserves. The allowed level of reserves is computed as a percentage of the face value of the loan portfolio and is not linked to the collectibility of any particular outstanding loan. The reserve-to-loans ratio can be a predetermined percentage for each type of loan, identical for all taxpayers, or it can be computed as a moving average of the proportion of bad debt to total value of the loan portfolio in past years (“experience method”). Under the latter method, the averaging period should be longer than a typical business cycle to maximize the income smoothing effects of reserve formation.65

Specific reserves. Additions to bad debt reserves are allowed only on a loan-by-loan basis whereby individual loans or receivables are evaluated as to their future collectibility. The tax deductible allowance may vary from a low percentage of the face value to a complete write-off of the loan.

The specific reserve method differs from the charge-off method in that under the charge-off method, a debt is written off only when it has been proven wholly or partially worthless. In contrast, under the specific reserve method, loans do not need to be worthless when the provision is made and taxable income is reduced. A tax-deductible provision can be made, provided that the facts and circumstances are such that collection is highly questionable or improbable. Changes in the financial situation of the obligor—such as suspension of payments or bankruptcy—or losses in the value of the collateral may justify provisioning. Actual worthlessness of the loan, against which a specific reserve is made, might actually occur years later.

Comparison Between the Reserve and Charge-Off Methods

A loan is a contract stipulating a stream of payments to be made by the borrower to the lender. This stream of payments includes interest payments and repayment of principal. The value of the contract at any point in time is the value of its expected future cash flows, discounted at the prevailing rate of interest on alternative riskless assets—such as, for example, treasury bills.

Because the lender recognizes the possibility of borrower default, the contract interest rate is generally higher than the risk-free interest rate. The difference between the contract and the riskless interest rate is the risk premium. Amounts received as risk premium compensate the lender for expected loan losses.66

To define the tax base, taxable income derived from a loan portfolio should match as closely as possible the underlying economic category, that is, economic income. Economic income from a loan portfolio equals net cash flow yielded during the fiscal year plus change in the value of the loan portfolio. The net cash flow yielded by a loan portfolio is equal to earned interest including the risk premium, plus repayment of principal minus new additions to the loan pool. The value of a loan portfolio is the present value of its expected flow of future payments discounted at the prevailing rate of interest on riskless assets. It can be shown67 that if the tax base does not match economic income, the value of a portfolio with loan losses will be altered by the presence of the tax and by its rate. Tax-induced changes in the value of risky assets distort their profitability in relation to riskless assets. Thus, the income tax can penalize or encourage the assumption of risk as a side effect of the treatment of bad debt.

Receipts associated with the risk premium are part of interest receipts and take place throughout the life of a loan portfolio in proportion to outstanding loans. In contrast, loan losses may be concentrated in the earlier or in the latter periods of the life of the loans. The disparity between the timing of risk premium receipts, which create a tax liability, and associated loan losses, which give rise to a tax deduction, is the cause of the tax-induced distortion. When this difference in timing results in a deferral in the recognition of income with respect to the recognition of the associated loss, a corresponding deferral in tax liabilities occurs and the portfolio value is favored by the tax. Similarly, the value of a portfolio is impaired by the tax when the recognition of risk premium income occurs earlier than the recognition of the associated loss.

Neither the charge-off nor the reserve method taxes economic income accurately. Because of the mentioned discrepancies in timing of risk premium income and corresponding loan losses, both the charge-off and reserve methods tend to favor loan portfolios with early loan losses and to disadvantage those with late loan losses. It can be shown that the charge-off method is neutral when losses are evenly spread throughout the life of the loan portfolio. Actually, diversification of portfolios and continuity in the flow of amortization and new loan issuances smooth the flow of defaults over time. Correspondingly, the charge-off method could be expected to be roughly neutral in practice.68

The reserve method, however, adds a specific distortion. Since a present deduction is allowed for losses that may occur far into the future, the reserve method results in a treatment of financial risk uniformly more favorable than the charge-off method.69 This was one of the reasons based on which the U.S. Congress repealed the reserve method in 1986.70

Both the reserve and charge-off methods share the difficulty of identifying uncollectible debt without opening loopholes for tax avoidance. It could be argued that the problem is more severe in the specific reserve method since, due to its prudential nature, the deduction needs to be granted before there is conclusive proof of uncollectibility. As a result, intertemporal arbitrage of tax liabilities may result in significant losses of tax revenue. This is one of the reasons behind the present trend toward tighter conditions for the deduction of specific reserves, disallowance of general and hidden reserves, or adoption of the charge-off method.

Generally, tax deductions for losses in receivables of nonfinancial institutions and contingent losses of non-financial assets are allowed only upon realization. In this sense, the reserve method amounts to a tax expenditure in favor of the financial sector.

Nonetheless, tax deductibility of some reserves may be advisable in special circumstances. Risk management techniques, such as portfolio diversification, can be applied only when the risk of default is independent across loans. For some groups of loans, however, the risk of default might be highly correlated. This is the case, for example, of sovereign debt from some particular country (or set of countries) undergoing significant external imbalances. When a large proportion of assets of the financial sector are affected by this type of possible loss, the allowance of tax-deductible specific reserves might be necessary.71 It should be clear, however, that tax deductibility of provisions for loan losses does not alter, by itself, the timing or the size of the losses.

Selected Country Practices

United States

Prior to 1986, banks could deduct from taxable income provisions for loan losses in an amount based on actual loss experience over the last six years (“experience method”) or equal to a set percentage of eligible loans (“percentage of eligible loans method”). The tax reform of 1986 adopted a pure charge-off method. Consequently, tax deductions for loan loss provisioning were disallowed. The only exceptions are banks with assets below $500 million and some specific reserves related to international loans to specific countries and mandated by federal regulators. Notwithstanding their tax treatment, banks routinely allocate after-tax profits to specific and general reserves that are normally included in regulatory capital and are subject to review by regulators and external auditors for consistency and prudence.

Germany

In Germany, as in some other European countries, there is a somewhat tighter link between tax and book accounts. German banks create specific, general, and undisclosed (“hidden”) reserves. Specific loan loss reserves are not included in regulatory capital and are generally tax deductible.

Taking effect on January 1, 1989, the mandatory and tax deductible character of general reserves was repealed. Hidden reserves have to be disclosed to the regulatory authority and they are not tax deductible. These hidden reserves allow German banks to smooth out fluctuations in earnings and report steady profits.

France

French banks are allowed to make specific and general loan loss provisions. Specific provisions are established against individual assets whose recovery is doubtful and tax deductibility is granted on a case-by-case basis. General provisioning is established against other assets and is tax deductible provided that it does not exceed certain limits. The main limits are that additions to reserves cannot exceed 5 percent of pretax income and that the stock of loan loss reserves cannot be more than 0.5 percent of medium- and long-term outstanding loans.

Japan

Tax-deductible specific reserves can be established up to 50 percent of the nominal value of loans that meet at least one of the following two conditions: there is a formal declaration of default against the borrower, or principal and interest have not been paid in the last four years.

Additions to general reserves are tax deductible until general loan loss reserves reach a level equal to 0.3 percent of outstanding loans, or total outstanding loans multiplied by the average loan loss ratio for the three years preceding the reporting year.

Issues Relating to Personal and Corporate Income Taxes

Integration of Personal and Corporate Income Taxes: Advantages and Disadvantages

John R. King

  • What are the arguments of basic principle for and against a classical (or separate entity) corporate tax system?

  • What particular distortions does such a system create compared with an integrated system of taxation for personal and corporate income?

  • What disadvantages could there be in moving from an existing classical corporate tax system to a more integrated system?

An income tax is usually levied not only on individuals but also on legal entities. In many countries, the same income tax law applies to both, but particular provisions (concerning rates of tax, for example) distinguish between the income of individuals on the one hand and that of companies on the other. In other countries, separate tax laws are applied to the incomes of individuals and companies.72 From an economic point of view, the main issue of substance in this area, however, is not the legal form of the tax on the incomes of different entities but rather the extent to which provisions are made under the corporate income tax, the personal income tax, or both, to reduce or eliminate “double taxation” of income which is earned by a corporation but accrues—in one form or another—to the individuals who are its ultimate owners.

Such provisions seek to integrate, to a greater or lesser degree, the income taxes that are levied separately on companies and on the returns they provide to individuals who are their ultimate owners. Particular schemes of integration are described and discussed in the next section; this section focuses on the general case for—and against—some forms of integration.

Arguments of Basic Principle

In its strongest form, the case for integrating individual and corporate income taxes rests on two basic propositions. The first is that taxation should be levied, as a matter of fiscal equity, according to “ability to pay”—as measured by income. The second is that corporate entities do not have an ability to pay taxes, in the relevant sense; they are simply a “conduit” through which income flows to individuals who are their ultimate owners.

Combined, these propositions appear to suggest that corporate income should only be taxed in the hands of the individuals to whom it accrues. They do, however, leave room for a separate corporate income tax to be justified as a withholding tax, which may be a useful means of ensuring that income flowing through the conduit is taxed in a comprehensive and timely manner and that the base of the individual income tax is protected. Many economists, including some who have not advocated full integration, have argued that this withholding function is indeed the main argument for the imposition of a tax on corporate income.

It is less easy to summarize the case in principle against integration—or in favor of what is commonly termed a “classical” corporate tax system, in which income tax is levied separately, both on company income and on dividends received by shareholders. The defense of this system may be based on denying one of the propositions on which the integrationist case rests, or both.73

First, some are content to assert simply that companies are “separate entities,” legally distinct from the individuals who own them, and to rest the case on that issue of legal form. Second, it has been argued that the integrationist case is based on a concept of “ability to pay” developed by utilitarian philosophers such as Bentham and Mill, which now seems narrow and outmoded. The principle of taxation according to ability to pay can be interpreted more broadly, as requiring taxes to be levied on income—and indeed on other tax bases such as consumption and wealth—in such a way as to minimize loss of social welfare.

A third defense of the principle of a classical corporate tax system rests on the “benefit” principle that taxes should be levied according to the benefit provided by the taxing authorities. It has been argued that corporations enjoy benefits in the form of limited liability, and from government services that are provided more directly, and that some form of taxation of those benefits is appropriate. When the U.S. corporate income tax was introduced in 1909, it was seen, for example, as an “excise tax” on the privilege of limited liability. Defenders of a classical system now usually place little weight on this argument, however. The reason is that it is difficult to establish any direct connection between the benefit of limited liability and the income of a company.

More generally, these arguments of basic principle are now rather unfashionable among economists. Empirical considerations have played a much larger part in recent debates on the appropriate relationship between corporate and individual income tax systems.

The Case Against the Classical (Nonintegrated) System

Compared with a fully integrated system, a classical corporation tax which taxes the equity income of companies at a positive rate may distort incentives in four main ways.

First and most obviously, it acts to discourage businesses from incorporating, and hence from taking advantage of benefits which are associated with the corporate form of organization—such as the benefit of limited liability, which reduces the cost to companies of raising outside capital for expansion.

It should be noted, however, that the discouragement to incorporation applies only insofar as the business is financed by equity. A corporate tax on equity income allows interest payments to the company’s creditors to be deducted from the tax base. Hence, when investment is financed at the margin by debt rather than equity, the resulting income bears no tax at the corporate level; the only tax paid on the income is the tax on the lender’s interest income. Effectively, then, a classical corporate tax is “integrated” in respect of income from debt-financed projects, and hence may not discourage incorporation when the firm is free to vary its financial structure.

But this point leads to a second adverse incentive effect of a classical corporate tax—namely, that it encourages companies to finance their projects by using debt rather than equity finance. This distortion increases the risk of bankruptcy. It will, therefore, bias companies toward relatively secure investments. Even so, the number of bankruptcies is likely to increase. The productive assets of companies that go bankrupt need not be lost to the economy; but resources must nevertheless be devoted to redeploying them in new activities.

At the same time, the bias in favor of debt financing gives companies an incentive to disguise the returns they provide to their shareholders, as far as possible, as “interest” payments rather than dividends. Most classical corporation taxes thus require extensive anti-avoidance provisions to limit what may be deducted from the tax base in the form of interest payments.

Third, a classical corporation tax encourages a company to retain its equity earnings rather than distributing them to its shareholders. When dividends are paid, the shareholder is subject to income tax at the appropriate rate. When earnings are retained, the shareholder should benefit, instead, from an increase in the market value of the company. In many countries, that capital gain is not subject to tax; and when there is a tax on capital gains, it is usually levied at a lower effective rate than the income tax on dividends. As a result of this bias in favor of retentions, equity funds may be “trapped” within particular companies rather than allocated between companies in the most efficient manner by financial markets, according to the investment opportunities that the companies face.

Fourth, a classical corporate tax system reduces the incentive to invest, and may therefore inhibit growth.74 This criticism can of course be made of fully integrated income tax systems as well, since any tax on capital income will tend to discourage investment in the relevant activities. The additional tax that is levied on company income under a classical system, however, represents an additional discouragement.

Combined, these four points represent a powerful case against the classical form of corporate income tax. This case has in practice been an influential one; there has been a general—though not entirely universal—tendency over the last two decades for existing classical systems to be replaced by some form of integration of corporate and individual income taxes. Nevertheless, support among economists for a move toward integration is not universal.

The Case for Retaining a Classical System

The first, and most powerful, argument against integration is that it will generally entail a loss of revenue, compared with what was generated by the classical system that is replaced. This revenue loss must be made up in some way; the corporate tax rate might be increased, or some other taxes might be imposed. In either case, there are likely to be economic costs that must be set against the benefits of integration.

Second, doubts have often been expressed about the empirical significance of particular benefits from integration, such as the reduction in bankruptcies, and in the costs of reorganizing the activities of bankrupt firms. In addition, to the extent that equity is trapped within companies by an existing classical system, the burden of the additional tax that is payable on dividends when those earnings are eventually distributed may already be capitalized into share prices. In this case, much of the benefit of a shift to an integrated system could simply accrue as a windfall gain to existing shareholders.

Recent simulations by the U.S. Treasury of the effects of replacing the present classical system in the United States with some alternative integrated systems do suggest substantial benefits, amounting to up to 0.2 percent of consumption;75 but in the present state of the art, such estimates could be quite unreliable.

Finally, some major benefits that may be claimed for a classical system, compared with most integrated systems that have been adopted in practice, are its simplicity and transparency. These features generally make a classical corporate tax system easier to administer than an integrated system. They also avoid most of the severe difficulties that arise in devising an appropriate tax treatment, in an integrated system, of dividends paid or received from abroad. For example, most countries that have adopted “imputation” systems, which provide tax credits to shareholders in respect of corporate tax that has already been paid on the income that they receive in the form of dividends, are reluctant to give such credits for corporate tax that has been paid in a foreign country. As a result, their tax systems may discriminate heavily in favor of domestic, as opposed to foreign, investment. In addition, many countries have adopted integration schemes that allow them to discriminate against foreign shareholders. Thus, integration schemes may discourage both inbound and outbound investment. These biases are much easier to avoid when the corporate tax system has the classical form.

The Mechanics of Integration

Dale Chua and John R. King

  • What is meant by a “fully integrated” system of corporate and personal income taxation?

  • What problems would arise in implementing such a system?

  • What methods are available to achieve integration in respect of the taxation of profits distributed by companies to their shareholders?

  • In what respects do those methods differ?

The Meaning of “Integration”

The term “integration” has been used in different ways.76 Traditionally, “full integration” has been used to denote an arrangement under which the incomes of all entities would be attributed in an appropriate manner to the individuals who are their ultimate owners. The income tax due would then be collected from those owners at the relevant rates, depending on their total incomes.

A scheme of this kind was advocated by the Royal Commission on Taxation in Canada (the Carter Commission) in 1966.77 Many economists have seen “full integration” in this sense as an ideal arrangement in principle. But detailed study has generally led to the conclusion that it would be administratively impracticable. The first reason is that there would be an enormous amount of information reporting required: in many economies, a single company may have a very high number of ultimate owners, many of whom will have held shares for only a part of any tax year. Second, attributing retained earnings to different owners is problematic when there are different classes of corporate security holders, with heterogeneous claims such as ordinary shares, preference shares, warrants, options, rights issues, and convertible notes. Third, many company shares are held by other companies. Hence, tracing the ultimate owners can often be difficult. A fourth general difficulty is that if tax were to be levied on shareholders’ earnings whether they are retained or distributed by the company, severe liquidity problems could result: shareholders could often be liable to pay large amounts of tax without having received cash with which those liabilities could be met.

No country has tried to apply a full integration scheme of this kind to the taxation of all corporate income. Many countries, however, do effectively integrate company and individual income taxation, along these lines, in the case of small companies with a limited number of owners. For example, in the United States, certain companies with no more than 35 shareholders can qualify (as “Subchapter S” companies) to be taxed in a similar way to partnerships, with their income being allocated directly to their shareholders in the appropriate proportions. A similar effect may be achieved indirectly if the tax system allows small companies to pay out all of their taxable income to their owners in the form of tax-deductible directors’ remuneration. This is sometimes referred to as “self-help integration.”

While full integration would tax corporate income at the relevant rates for individual shareholders, irrespective of whether that income is distributed or retained within the company, a more limited form of integration is confined to corporate income that is distributed to shareholders rather than retained. Note that this form of integration, confined to the taxation of corporate income that is paid out as dividends, can eliminate the liquidity problem mentioned above. A wide variety of methods may be used to achieve this limited form of integration. The main methods are illustrated below.

Illustrations of Integration Schemes

This section shows, with the aid of a numerical example, how different integration schemes can reduce the overtaxation of corporate income that is inherent in a classical (or separate entity) corporate tax system.78

Benchmarks: the classical system and full integration

The tax burden under the classical system, adopted by countries such as the United States, the Netherlands, Luxembourg, and Switzerland, can be illustrated with an example in Table IV.5. Assume that the profits of a company are $1,000 and that the corporate tax rate is 30 percent. After-tax profits of $700 are distributed fully as dividends. Suppose that the individual shareholders’ tax rates are 20 percent and 40 percent. The resulting personal income tax liabilities will be $140 and $280, respectively. Hence, the combined corporate and personal tax payment on the same income source is, respectively, $440 or $580. To calculate the tax burden, we express the combined payment as a percentage of the original income; hence, the effective tax rate is 44 percent for the lower-taxed individual and 58 percent for the higher-taxed individual. Comparing these effective tax rates with the relevant income tax rate, the “overtaxation rate”79 is 120 percent for the low-tax shareholder and 45 percent for the high-tax shareholder.

Table IV.5.

Classical System

article image

If the company decides to retain all profits, however, the effective tax rate confronting all shareholders in the short run will be the corporate tax rate of 30 percent. A shareholder facing a 20 percent marginal personal tax rate will then be overtaxed by 50 percent, and one facing 40 percent will be undertaxed by 33.3 percent.80 Between the extremes of full distribution and retention of profits are various possible outcomes; each will result in a specific overtaxation or undertaxation for a given shareholder depending on his marginal tax rate.

Full integration of the corporate and personal income tax is the other extreme in the company/shareholder tax spectrum. Under this method, a corporation is treated like a partnership whereby the company’s incomes, both distributed and retained, are attributed to its shareholders according to their respective holdings: the attribution is then taxed at the shareholder’s marginal tax rate. Working through the various steps, the example in Table IV.6 shows that the effective tax rate is also the personal tax rate. Hence, by definition, over-taxation does not exist.

Table IV.6.

Full Integration

article image

Tax relief is defined as the difference between the classical overtaxation less overtaxation under the present method, divided by the classical overtaxation.

In particular circumstances, full integration could be achieved in principle by several systems besides the partnership method.81 One such system would be to abolish the corporate income tax completely and let shareholders pay taxes under the personal income tax on the dividends received plus net accrued capital gains on shares—that is, on a comprehensive income base.82 Second, full integration could be achieved straightforwardly in the special case where the personal income tax is levied at a single rate, by levying tax on corporate income at the same rate, while exempting dividends and capital gains on company shares from the personal tax.

Partial integration in respect of distributed profits

Between the classical and fully integrated systems are many systems which provide a degree of integration in respect of distributed profits (but not retained earnings). In these systems, relief for the double taxation of dividends can be given either at the corporate or at the shareholder level. The principal methods of providing relief at the corporate level are the dividend-deduction and split-rate systems. The principal methods of providing relief at the shareholder level are the “imputation system” and various schedular methods.83

Corporate level (1) Dividend-deduction system. Under this system, a fraction of company profits distributed as dividends to shareholders can be deducted against the company’s corporate income tax liability.84 The proportion of dividends deductible from the corporate tax base varies across countries. How is this system related to the two benchmarks? In the limit, if the full pretax income of a corporation were distributed and there were full deductibility for dividends (as in the case of Greece, for example), then the corporate tax liability would be zero. All taxes would be paid by shareholders, as in the case of a fully integrated system. On the other hand, this system approaches the classical system when the proportion of dividends allowed as a deduction is lowered. In general, the tax burden on the same source of income will be lower than that under a classical system, but higher than under a fully integrated system.85

To illustrate, we assume that the authorities provide dividend tax relief at a rate of 50 percent. Table IV.7 shows that a dividend deduction is given at the corporate level (in this example, 50 percent) before corporate tax is levied. Again, assuming full distribution of profits, a shareholder then pays taxes on a dividend income of 850 according to his marginal personal tax rate.

Table IV.7.

Dividend-Deduction System

article image

This example is designed to show that relief from double taxation on the dividend is equal to 50 percent of classical overtaxation. Accordingly, the effective tax rate for the low-tax shareholder is 32 percent rather than 20 percent, and that of a high-tax shareholder is 49 percent rather than 40 percent. As is to be expected, the economic burden of overtaxation is lower than under the classical system but higher than under full integration.

(2) Split-rate system. Under this system, profits distributed as dividends are taxed at a lower rate than retained earnings or undistributed profits.86 If the tax rate differential between distributed profits and retained earnings is small, then the split-rate system approaches the classical system. On the other hand, if the differential is large, then it acts essentially like a tax on undistributed profits.87 To show that this system can be designed to be operationally equivalent to the dividend deduction system, we continue to assume that the goal is to provide a 50 percent tax relief measured against overtaxation under the classical system. To achieve the desired level of relief, distributed profits are now taxed at a lower rate of 15 percent. Undistributed or retained profits are taxed at the unchanged rate of 30 percent.

By assuming that all profits are distributed, the contrived example in Table IV.8 shows that the desired level of overtaxation can be easily achieved. In practice, some profits are retained by companies. This complication is ignored in our numerical example, but will add to the overall tax burden in the future as the profits are distributed.

Table IV.8.

Split-Rate System

article image

Shareholder level (1) Imputation system. Perhaps the most common approach to providing relief from double taxation of dividend income is a method generally referred to as the “imputation system.” This method gives shareholders a credit for taxes paid by the company under the corporate income tax; this credit can be used as an offset against their personal income tax liability on dividends.88

The imputation system acknowledges that corporate income paid out as dividends has already been taxed once at the corporate level; therefore, an explicit relief is given against the personal income tax at the shareholder level. The level of tax relief given to shareholders is commonly known as the rate of imputation. Under this system, the personal income tax base for the shareholder is the sum of dividend received plus the tax credit, that is, the grossed-up value of the dividend. The marginal personal tax rate is then applied to this grossed-up value to derive the gross income tax payable. The corporate tax is then credited against the gross tax of the shareholder, and the balance is the net tax payable (or the amount refundable). In general, if the shareholder’s marginal tax rate is higher than the imputation rate, then additional taxes are payable; but he will receive a refund if the imputation rate is higher than his marginal tax rate.

We illustrate the mechanics of this system in Table IV.9. For comparison with the cases shown above, only partial relief (50 percent) is provided; to achieve the stated goal, the net dividend is grossed up by a factor of 3/14.89

Table IV.9.

Imputation System

article image

(2) Schedular methods. Other methods of providing relief from double taxation of dividends, at the shareholder level, are classed collectively as schedular methods. One such approach is the separate tax method, sometimes called the dividend-exemption system. Before distribution, corporate source income is taxed at the corporate rate. Distributed dividends are then partially taxed and partially exempt under a personal income tax which taxes dividend income at a lower preferential but usually flat rate.90

Another system, widely known as the tax credit method,91 is very similar to the imputation system method in that it provides a credit to shareholders against their personal income tax, usually specified as some proportion of dividends received. It differs, however, from the imputation system in that (1) the tax credit is available to the shareholder whether or not the corporate tax is actually levied; (2) the net dividend received is not grossed up; and (3) in general, no refund is given if the tax credit exceeds the gross income tax liability.

Differences Between Integration Methods

As the above numerical illustrations show, the same effective rate of tax on corporate income that is distributed to particular shareholders can be achieved by a wide variety of different schemes. But in many other respects, the implications of those schemes are quite different.

A first, obvious distinction is between schemes which attempt to eliminate the double taxation of dividend income entirely, and those that do so only partially.

A second important distinction is between schemes that levy tax on distributed corporate income at different rates according to the marginal income tax rate of the shareholder, and those that effectively apply a single rate to corporate income.

A third practical difference between integration systems for distributed income concerns the extent to which the benefits of any tax incentives (or “tax preferences”) that are provided to companies, such as accelerated depreciation and investment tax credits, can be passed on to shareholders when the income is distributed. Suppose, for example, that a company earns distributable profits of 100, but preferences reduce its taxable profits to 50. When the profits are fully distributed, some integration systems would effectively credit the shareholders with corporate tax on the full amount of the distribution—that is, on 100; others would limit the credit to the profits of 50 on which corporate tax had actually been paid.

In a similar way, there are important practical differences between integration schemes in respect of their implications for income that is distributed to foreign shareholders and the foreign income of companies that is distributed domestically. For example, dividend-deduction and split-rate systems provide relief for double taxation of dividends, at the corporate level, irrespective of whether the shareholders are resident or non-resident. On the other hand, under imputation systems that provide relief at the shareholder level, that relief can be confined to resident shareholders. This basic difference may be modified in practice by bilateral treaties to avoid double taxation. But it is clear that the choice of an integration system that provides relief at the shareholder level, rather than the company level, will strengthen a government’s bargaining position in negotiating such treaties. International considerations of this kind have often been paramount in determining different countries’ choices between alternative integration methods.

Taxation of Capital Gains

John R. King

  • In what ways do tax systems vary in their treatment of capital gains?

  • What are the arguments for taxing capital gains in a different way from ordinary income?

  • What are the main features of capital gains tax structures in different countries?

  • What are the main difficulties that arise in imposing tax on capital gains?

Capital Gains and Business Income Taxation

The Schanz-Haig-Simons comprehensive measure of income during a given period would include, on a par with other types of income, the change in value of all the capital assets and liabilities of an individual or business during that period. Personal and corporate income tax systems, however, treat many types of capital gains in a different way from other types of income, by applying specific provisions in the case of particular kinds of gain, often exempting them from tax or subjecting them to tax at different rates.

Some of these special provisions may differ according to whether the gains accrue to companies on the one hand, or to individuals on the other. Capital gains taxes at both levels are relevant to the taxation of corporate income; in the larger OECD economies, a major part of individuals’ capital gains consist of gains on company shares, and these gains often rival dividends as a means by which individuals derive returns from their ownership of corporate equity.

In practice, it is much more difficult to compare the treatment of capital gains in different countries’ tax systems than to compare other aspects of corporate taxes, such as depreciation allowances or loss carryforward. Capital gains or losses can arise on many different kinds of assets (or liabilities), owned by different kinds of economic agents; and countries differ in respect of whether particular types of gains would be treated as income, or as “capital gains” that might be subject to special treatment under either the regular income tax or a separate capital gains tax. Three examples of these differences are worth noting.

First, there is often an important distinction between assets that are depreciable for tax purposes and those that are not. In the case of depreciable assets, many income tax systems provide that depreciation allowances should be “recaptured” (or recovered in the form of a “balancing charge”) when the asset is sold for more than its written-down value for tax purposes. With a system of this kind, the gain realized on disposal of a depreciable asset is, thus, treated as income. But separate provisions in the tax law are applied to nondepreciable assets such as land and securities.

Second, what is treated as a capital gain that may be subject to special rules also depends on the valuation conventions that are applied in measuring assets and liabilities, and these often vary in different countries’ tax systems. For example, if liabilities that are denominated in foreign exchange are shown in the balance sheet at current market values, and changes in those values are reflected in the profit and loss account in each accounting period, the exchange gain or loss accruing in each period may automatically be taxed as income. On the other hand, if the liability is carried in the books at historical cost, a gain or loss will be shown as “realized” only when the loan is repaid. The issue then arises whether that realized gain or loss is to be treated as ordinary income or as capital gain.

Third, the same kind of gain can be treated in different ways, according to the nature of the economic agent to whom it accrues. For example, when individuals or businesses trade in property or securities, they are generally subject to income tax on the aggregate gain or loss that accrues in each period. The same gains or losses are often treated differently, however, when they arise in the hands of others, for whom they are in the nature of “incidental” revenues.

In comparing different countries’ tax systems, therefore, it is not sufficient to examine the particular provisions that they contain relating to capital gains. It is also necessary to determine how in practice they draw the line between capital gains and “ordinary” income.

The Case for Differential Taxation of Capital Gains and Other Income

Historically, differential taxation of income and accrued capital gains arose in many European countries from an underlying concept of taxable income as a flow of services from specific capital sources, separate and distinct from any changes to the value of those sources themselves. Under this income concept, capital gains were not subject to tax. Differential tax treatment can, however, also claim some theoretical support from more modern concepts of income such as that of Hicks, which would exclude from the income measure certain gains and losses that are in the nature of unanticipated windfalls.

As the concept of “comprehensive” income gained support as an ideal standard for an income tax base, the practical difficulties of measuring capital gains as they accrue in each income period provided further arguments for differentiation of their tax treatment from that of other forms of income. Valuation of all the assets of each individual or business at the appropriate market prices, at the end of every period, appears generally impracticable. In the absence of such valuations, capital gains could only be subjected to tax when they are “realized” in the disposal of the asset. A supplementary, practical argument in favor of taxation of gains at the time of realization, rather than on an accruals basis, is that the disposal of an asset by way of sale generates cash that can be used to satisfy the tax liability. Taxing gains on an accruals basis, on the other hand, may result in severe liquidity problems for the individual or business concerned.

Postponement of the tax liability on gains to the point at which the gains are realized—which is generally at the discretion of the taxpayer—constitutes differential treatment of accrued capital gains relative to other income that is received in the form of cash. It also gives rise to arguments for further differentiation. In particular, when income in a particular period is subject to progressive rates of tax, it can be argued that it would be inappropriate to tax the gains realized in that period as a “top slice” of income—since those gains may have accrued over many earlier periods and are “bunched” when realized.

A rather different argument that has sometimes been used in favor of differential taxation of (nominal) capital gains is that, under conditions of inflation, they are a distorted reflection of the real gains that would appropriately be included in a comprehensive measure of income. What is measured as a nominal gain may correspond to a much smaller real gain, or even to a loss in real terms.

Capital Gains Tax Structures

In general, when capital gains are subject to tax as they accrue, they are taxed as ordinary income. Special provisions usually apply, however, when gains are taxed at the time of realization.

Among OECD countries, there has been a general—though by no means uniform—trend in recent decades toward bringing gains accruing over a period of several years within the tax net.92 The U.S. provisions date from the introduction of the Federal income tax in 1913, though they have been changed frequently since that time.93 Provisions relating to long-term gains were introduced in Japan in 1946; in Denmark, in 1958; in Sweden, Portugal, and the United Kingdom, in the 1960s; in Canada, France, Ireland, and Spain, in the 1970s; and in Australia, in the 1980s. In many of these cases, gains realized over short periods had been subjected to tax at an earlier date. The main justification for imposing a tax on capital gains has been the equity argument, that such gains are equivalent to other forms of income in their effect on “ability to pay.” In addition, however, an important reason for the spread of capital gains provisions has been the need to protect income tax revenues from erosion through tax avoidance, in the form of transactions that result in the conversion of taxable ordinary income into nontaxable gains.

These special provisions sometimes take the form of a capital gains tax that is separate from the normal income tax (as in Ireland, Italy, and the United Kingdom, among OECD countries). More commonly, however, they are part of the income tax law. This section summarizes the most common special provisions that apply.

Chargeable gains

Capital gains that are taxable under the provisions generally consist of those arising on the disposal of real property and financial assets, such as company shares. In some countries, the provisions apply to only certain assets in these categories. Gains on particular assets, such as owner-occupied residences and government securities, are often exempted.

Occasion for a tax charge

The typical “realization” for tax purposes is the disposal of an asset for consideration. Other events may, however, be treated as “deemed realizations” for tax purposes—for example, transfers by way of gift, or upon death of the owner. On the other hand, some countries allow gains of certain sorts, or in certain circumstances, to be “rolled over” so that the tax charge does not crystallize until a later date.

Tax rates

Often, when realized capital gains are subject to tax, they are taxed at different rates from those that are applied to ordinary income. In the case of gains realized by companies, Belgium, France, and Ireland apply special rates. In all these cases, the rate differs according to the length of time that the asset has been held. In the case of gains realized by individuals, many more countries apply separate rates—again, often varying according to the holding period—or include only some fraction of the gain in the amount that is taxed at ordinary income tax rates.

Holding period criteria

As noted above, many countries treat realized gains differently according to the holding period over which the gain has accrued. When this distinction is made, gains realized on assets that have been held for longer periods are treated more favorably than short-term, “speculative” gains. In some cases, gains are exempted altogether if the asset has been held for longer than a specified period.

Capital losses

Most countries allow realized capital losses to be offset against gains in the same year. There are, however, considerable variations in the scope allowed for carrying excess losses forward (or backward), and in the possibility of offsetting them against other forms of income.

Thresholds

Capital gains realized by individuals are commonly taxable only above a certain annual (or lifetime) threshold, even in those countries where the gains are subject to the regular income tax rather than to a separate capital gains tax. The main reason that has been given for such a threshold is the administrative difficulty, and often disproportionate cost, of assessing tax on a large number of individual gains. In many countries, however—particularly those with “self-assessment” systems, such as the United States—no such threshold applies.

Inflation adjustment

Since the 1970s, a number of countries have introduced an inflation adjustment into their capital gains provisions, increasing the “acquisition cost” used in calculating the realized gain according to the increase in some specified price index since the date of acquisition. Among OECD countries, the list of those that index corporate capital gains for tax purposes now includes Austria (in cases where the asset has been held for at least 19 months), Ireland, Portugal, and the United Kingdom. Although conceptually straightforward, indexation of acquisition cost may result in formidable administrative complexities. For example, when a holding of a particular class of asset (such as a particular class of share in a particular company) has been built up over a period of years, the holding may be “pooled” and effectively treated as a single asset in an unindexed system; when indexation is introduced, however, detailed rules may become necessary to assign particular disposals from the pool to particular acquisition dates.

The Problems of Capital Gains Taxes

Underlying these widely varying tax structures for capital gains are a number of common problems.

The first is the difficulty of drawing the appropriate boundaries between accruing gains that are to be treated as income, gains that should be taxed but, for practical reasons, must be taxed on a realization basis under special provisions, and gains that should not be taxed at all. Since financial markets will usually try to exploit any differences in the tax treatment of similar types of transaction, the boundaries may need to be continuously shifted. For example, in recent years, countries such as the United States and the United Kingdom have found it necessary to devise schemes to bring into tax as ordinary income the “gain” that accrues over time on a security that pays little or no interest, but that is issued instead at a deep discount.

A second difficulty that arises, when gains are taxed at the time of realization, is that this tax treatment may tend to “lock in” the investor, discouraging him from disposing of the asset and hence crystallizing the tax charge. Empirical estimates of the magnitude of this lock-in effect that have been made in the United States differ quite widely, but there is no doubt that it can be significant in particular circumstances. It is worth noting that holding onto an asset will generally only postpone the tax charge, and not eliminate it entirely. The lock-in effect may, however, be greatly magnified if the taxpayer is able to reduce the tax charge by holding onto the asset until retirement, or to escape the charge completely by holding it until he dies.

Rollover provisions for particular types of transactions are designed to minimize the potential lock-in effect. But they do so by reducing the effective rate of tax on accrued capital gains, and hence increasing the extent of any distortion and unfairness that arises from the differential taxation of capital gains and other forms of income.

The third general difficulty arises from inflation. As noted above, the appropriate adjustments to take account of inflation are conceptually straightforward, and several countries have introduced such adjustments in their capital gains tax regimes in recent years. These adjustments weaken the case for certain other common capital gains tax provisions (such as reduced tax rates for all realized gains, reduced rates for assets that have been held for a long period, and partial inclusion of gains in the total of taxable income) that have often been justified as rough-and-ready methods of taking account of the impact of inflation. When an indexed treatment, however, is introduced for capital gains but not for other forms of income, such as interest income, whose measurement is similarly affected by inflation, new anomalies and distortions can arise in the tax system.

Debt and Equity Financing

John R. King

  • Do corporate and personal income tax systems generally favor debt over equity financing of companies?

  • Does a tax on dividend payments encourage companies to retain earnings?

  • How could the tax system be made “neutral” in its effect on a company’s choice between alternative methods of finance?

The Tax Treatment of Interest and Dividends

Corporate income tax systems treat interest and dividends in different ways. As a result, they affect the incentives that companies face to finance their investments using debt on the one hand, or equity on the other.

In almost all existing corporate tax systems, the income of a company is measured for tax purposes from the standpoint of its owners. Payments of interest to the company’s creditors are, therefore, allowed as a deduction. (As noted below, however, deductibility of interest is often subject to certain qualifications or restrictions.)

As described above, the treatment of dividends paid by a company varies much more widely. In Greece, for example, dividend payments are deductible in full, on a par with interest payments. In certain other countries (such as Iceland, Spain, and Sweden, among OECD countries), a proportion of dividends is deductible from taxable profits. In the majority of cases, however, the payment of dividends does not directly affect the amount of a company’s profits that is subject to tax.

This suggests that corporate tax systems normally favor the use of debt finance, where the return to the provider of funds takes the form of interest payments, over equity, where the return is provided in the form of dividends. Before coming to that conclusion, however, it is necessary to consider also the way in which interest and dividends are taxed in the hands of the recipient. A dividend deduction under the corporate income tax is one way in which double taxation of distributed profits may be relieved, but other methods of achieving this same result—such as crediting corporate tax paid on distributed profits against the shareholder’s liability to tax on the dividend, or excluding a part of the dividend from income that is liable to tax at the shareholder level—are rather more common. The individual saver’s choice to supply funds to the company as debt or equity finance will be influenced by many nontax considerations; but, insofar as tax is relevant, it is the total tax charge on the return to those two sources of funds that matters.

Analysis of the effect of company and personal tax systems on the incentive for companies to finance themselves using debt or equity is, therefore, complex. In most systems, however, the total tax charge on interest is lower than that on dividends.

The Boundary Between Interest and Dividend Payments

When interest payments by companies are treated for tax purposes more favorably than dividend payments, companies have an incentive to disguise the payment of a return to its owners as “interest.” Provisions in the tax law commonly seek to limit this form of avoidance, in a variety of ways. For example, a deduction for loan interest payments may be allowed only when some of the following conditions are satisfied:

  • the loan has been incurred for taxable business purposes;

  • the loan has not been obtained from shareholders, or other related parties;

  • the loan interest is not “excessive”; and

  • the amount of interest payable under the loan contract is not related to the profits of the company, or some other measure of its performance.

In recent years, tax authorities of a number of countries have become increasingly concerned with the “thin capitalization” of certain foreign subsidiary companies operating in their territories, when the ratio of equity to funds directly or indirectly lent by other companies in the same group falls short of the proportion that could be deemed normal on an arm’s length basis. The first line of defense against thin capitalization should be a withholding tax on interest paid to foreigners, but other measures may be taken as well. Some countries (such as Belgium and Italy) judge the extent of thin capitalization subjectively, on a case-by-case basis. Other countries (such as Australia, Canada, the United States, Spain, France, and Japan) have introduced an objective criterion into their legislation, in the form of a maximum debt/equity ratio which certain types of companies must not exceed if their payments of debt interest are to be allowed in full as a deduction. For example, under regulations introduced in the United States at the end of 1989, any excess of the debt/equity ratio above 1.5:1 at the end of the year may threaten the deductibility of interest.

Anti-avoidance rules of this kind may be difficult to administer effectively. It is also worth noting that they may, occasionally, create new opportunities for abuse. For example, in the United Kingdom in the early 1980s, certain companies with no current taxable profits found it advantageous to have their interest payments classified for tax purposes as dividends. By including in the loan contract a clear (but trivial) relationship between the amount of interest payable and the profits of the company, they were able to exploit the “anti-avoidance” provisions in the income tax law to achieve this result.

Alternative Views of Equity Finance

From the more favorable treatment of interest than of dividend payments in most countries’ corporate income tax systems, it does not necessarily follow that those systems result in a tax bias toward debt financing of new investments. Whether they do so has recently been a matter of controversy.

The basic difficulty arises from a company being able to obtain equity finance at a particular point in time in different ways—by issuing new equity, or by retaining earnings that would otherwise have been used to pay dividends (or perhaps to repurchase shares, in countries where this is permitted under company law). When earnings are retained, the market value of the company will increase as a result of the increase in its assets. That reward to the shareholder, in the form of an accrued capital gain, will generally be taxed at a much lower effective rate than the dividends that it replaces, and in many countries, it may not be taxed at all.

Traditionally, it was usually assumed that the tax treatment of dividends is relevant to the cost of equity finance to the company, whether that finance is obtained in the form of new share issues or retained earnings, because any tax charge on dividends must be incurred eventually by the equity-holder. But a “new view” of dividend taxation, which has emerged in academic work since the early 1970s, has challenged this assumption.94

A simple example can illustrate the basis of the “new view.” Suppose that a company finances a one-period investment project, costing 100 and yielding a return of 10 percent, by retaining earnings that would otherwise have been paid as dividends. If its shareholders have a marginal income tax rate of 30 percent, the opportunity cost to them is 70 in after-tax dividends at the beginning of the period. At the end of the period, the company can distribute to its shareholders the original 100, plus the return of 10, less corporate tax paid on that return (say at 40 percent)—that is, a total of 106. Shareholders will be liable to tax at 30 percent on these dividends, leaving them a net amount of 74.2 at the end of the period. This represents a return of exactly 6 percent on their initial “investment” of 70. Thus, the return to the shareholders is affected only by the corporate tax paid on the yield of the investment; the tax that they pay on dividends is irrelevant.

According to this view, the tax treatment of dividends only affects the cost of new equity. Once new shares have been issued, equity is “trapped” within the company: the tax charge on its dividend return can only be postponed, not avoided altogether. A corollary is that the tax charge should be capitalized in the market value of the company’s shares, which may therefore fall below the value of its assets. The new view implies that the cost of equity finance in the form of retained earnings will be substantially below the cost of new equity, particularly in the case of corporate tax systems of the classical form. Retained earnings should, thus, be the preferred source of equity finance for new investment; and the degree to which the tax system favors debt over equity should be much smaller than would be suggested by a simple comparison of the tax treatments of interest and dividends. Since retained earnings are, in practice, a much more important source of funds for investment than new share issues, this issue is of considerable importance in making a judgment of the extent to which different tax systems provide incentives for companies to finance themselves using debt rather than equity

The new view rests on some strong assumptions—for instance, that the company cannot repurchase its own shares. In principle, it is open to empirical testing; and on the whole, the tests that have been conducted so far do not support the new view, at least in its pure form. One possible reason is that accountants are well paid to find ways of extracting “trapped” equity from companies without incurring a tax charge.

Thus, the general consensus at present is that differences in the tax treatment of interest and dividends do affect companies’ choices between debt and equity financing, not only when available retained earnings are insufficient (for instance, in the case of new or rapidly growing companies), but also in the more general case where equity funds may be obtained by the company from retained earnings as well as from new share issues.

Alternative Approaches to Neutrality Between Debt and Equity Financing

Tax discrimination between debt and equity financing matters, for two reasons. First, such discrimination creates opportunities for tax avoidance, which may have serious consequences both for government revenues and for the equity of the tax system. As financial systems become increasingly adept at devising new instruments to exploit any such opportunities, these consequences may become increasingly serious in the future. Second, an encouragement to debt financing may have important real effects on company behavior. As debt/equity ratios rise, companies’ investment decisions are likely to be affected increasingly by considerations of risk; and as the incidence of bankruptcy increases, costs are imposed on the economy in the form of resources that need to be devoted to reorganizing the activities of those companies that fall into bankruptcy. Neutrality in the treatment of debt and equity financing is, thus, a desirable objective for an income tax system.

Two approaches to a more neutral system could be taken. The treatment of equity finance could be brought into line with the existing treatment of debt finance; or the treatment of debt could be brought into line with that of equity. Several alternative schemes of these two general types have been adopted or proposed.

Aligning the treatment of equity with that of debt

In practice, the approach that many countries have taken to removing or reducing tax discrimination between debt and equity finance has been through integration schemes of the kind considered in detail above. Since integration applies in practice only to corporate equity income that is distributed as dividends, this approach does not remove the distinction between debt and equity finance entirely; the return on debt is taxed (or relieved) when it accrues in the form of interest; the return on equity, on the other hand, is taxed only when it is paid in the form of dividends.

An alternative approach, which has recently been suggested by the Institute of Fiscal Studies, would put equity on a par with debt in the corporate income tax by giving an “allowance for corporate equity” (ACE), which would consist of an imputed interest charge on the company’s equity.95 If the interest charge could be set at the appropriate level—which would correspond, in principle, to the opportunity cost of funds provided by the company’s shareholders—this approach would confine the corporate income tax to pure profits, and hence would not distort the company’s decisions. Although this approach has attracted some interest since it was first proposed in 1991, no countries have adopted it in practice.

Aligning the treatment of debt with that of equity

A proposal with a rather longer history—but one which has also not been adopted in practice, except for certain special mining taxes in a few countries—is a “cash flow corporate tax” which (in one particular form) would allow no deduction for either dividends or interest.96 Combined with an immediate deduction for expenditures by the company on capital assets, this system would also have the effect of confining the corporate tax charge to pure profits and would not distort the choice between debt and equity, provided that the recipients of interest and dividends were subject to tax in exactly the same way.

Removal of the present corporate deduction for interest payments is also the central feature of the “comprehensive business income tax” that has recently been proposed by the U.S. Treasury as a method of removing discrimination between debt and equity financing.97 In this proposal, tax paid by the company would be the final tax charge on company income, irrespective of whether the income is retained, paid to shareholders as dividends, or paid to bondholders as interest.

The Concept of Cost of Capital: Marginal Effective Tax Rate on Investment

Dale Chua

  • What is a marginal effective tax on an investment?

  • What affects the marginal effective tax on investment?

  • A country study: What observations can be drawn from a matrix of marginal effective tax calculation?

In the course of doing business, a profit-maximizing firm acquires physical and intangible capital goods as new investment to replace or augment its existing, depreciated capital stock. The rule that guides the firm is to invest to the point where the incremental unit of capital provides a stream of real returns that is just enough to cover all costs, including taxes, associated with that investment. In adopting this rule, a profit-maximizing firm should invest to the point where the marginal benefit of a dollar’s worth of capital per period is equal to the cost of holding a dollar of capital for that period. Mathematically, in equilibrium, a profit-maximizing firm will set the present value of its expected future earnings from the last unit of its investment, less the accompanying future tax liability, equal to the cost it expects to pay for that unit of investment, less the present value of any capital allowance it expeas to receive from using the investment. The cost associated with the holding of a dollar of capital per period is known as the cost of capital.

In a world without taxes, the cost of capital consists of two basic components. The first is the cost of finance, which arises because capital expenditure must be paid for with funds, either borrowed or through equity participation. The second is the capital consumption cost, which arises because a newly purchased capital stock will, over time, result in a loss of value due to depreciation and technical obsolescence. Accounting for such costs will ensure that the return to shareholders is also maximized. There is no need to differentiate between the before-tax and after-tax rate of return on investment since they will be identical. With taxes, however, a firm will need to consider taxrelated factors as they impinge on the cost of finance and the capital consumption cost. The tax factor, often an important consideration, thus constitutes an additional component in the cost of capital.

In practice, capital income is taxed in many different ways. Profits before distribution are taxed at the corporate level under the corporate income tax. Distributed profits are taxed as dividends at the shareholder level under the personal income tax, which may or may not provide relief for taxes paid at the corporate level. In addition, other taxes such as property tax, special levies, and royalties will also alter the after-tax income stream received by shareholders. These taxes, which are frequently not designed with investment neutrality as an objective, will distort investment behavior and, in so doing, create inefficiencies in resource allocation. In what follows, we introduce a concept to measure such distortions.

Definition of the Marginal Effective Tax Rate

In a nutshell, the marginal effective tax rate (METR) measures distortions that are created by the taxation of capital income imposed at the corporate and personal levels on marginal investments. The latter are, by definition, those whose return is just sufficient to cover costs, including tax-induced costs. Because the distortions are created by the corporate and personal income taxes, the METR comprises the sum of the distortions created by each tax. Conceptually, therefore, the total distortion can be separated into a marginal effective corporate tax (induced by the corporate tax) and a marginal effective personal income tax (induced by the personal income tax). The METR is also defined as the difference between the before-tax (gross) rate of return on a marginal investment and the after-tax (net) rate of return on the savings that is used to finance that same marginal unit of investment.98

A measure that quantifies the impact of tax-induced distortion on investment, if judiciously applied, would yield a matrix of METR statistics that would allow a systematic assessment of investment distortions for various types of investment goods (such as machinery, buildings, and inventories) across various sectors in the economy (such as manufacturing, services, agriculture, and mining). In the 1980s, this concept was widely applied in virtually all OECD countries as well as in many developing countries, and provided policymakers with a useful tool for tax reform strategies.

The idea of METR can best be illustrated with a diagram.99 Figure IV.2 shows a standard investment market for a particular class of capital goods. The investment schedule is a downward-sloping function dependent upon the before-tax rate of return rg. This rate can be interpreted as the return to the economy since it comprises the after-tax return to the firm and the tax revenue to the government. The saving schedule depicts the level of saving that will be forthcoming for a given after-tax rate of return rn. As an after-tax return to savers (as a group), this rate can be viewed as the opportunity cost to savers for abstaining from present consumption in an effort to maximize their intertemporal utility. Both rg and rn are measured in real terms.

Figure IV.2.
Figure IV.2.

Capital Market Equilibrium with Taxes

In a world without taxes, a profit-maximizing firm will invest to the point where the return on investment is just equal to the cost of borrowing. As shown in Figure IV.2, an equilibrium is obtained at I* where rg equals rn. The corporate taxes on capital income, by distorting economic behavior, however, will cause a firm to alter its investment behavior to maximize its after-tax profits. In a similar way, the personal income tax will alter a household’s saving behavior to maximize its intertemporal utility. Assume that the tax-affected equilibrium is observed at Ie. The before-tax rate of return to investment rge and the after-tax rate of return to savers rne are now separated by a tax wedge. The wedge measured by the vertical distance t can be interpreted as the amount in taxes collected by way of corporate and personal income taxes on the marginal investment unit.100 By expressing this wedge t as a proportion of the before-tax rate of return rge, one obtains the METR. It is also evident (in Figure IV.2) that the distortion creates an efficiency cost that is equal to the tax burden triangle ABE.

In Figure IV.2, the tax system penalizes investment resulting in a positive tax wedge t. A possible reason may be that if the capital is depreciable, but the present value of tax depreciation is less than expensing, then the economic cost of capital incurred in doing business is not adequately deducted as an expense. In this case, a firm maximizing its after-tax profits will choose to use less capital and, accordingly, invest less. This is suboptimal from society’s viewpoint. In other cases, however, it is equally possible for a tax system to subsidize capital investment resulting in overinvestment. This may happen if the immediate tax saving associated with a generous investment tax allowance exceeds, in a present value sense, the future tax burden on the income derived from such an investment.

What Affects the Cost of Capital and the METR?

In what specific ways do taxes affect the METR measure? First, consider the financing requirement of a firm. To pay for new capital investment, a firm can raise funds by issuing debt (via bonds), through equity (via shares or retained earnings), or through some combination of debt and equity. Most corporate tax systems provide full deductibility for the interest cost of debt financing. As a result, the actual cost of debt financing to a firm is the risk-adjusted real market rate less deductions allowed under the corporate income tax. On the other hand, funds raised through equity are usually not deductible in the same way. Since the cost of finance is one factor in determining the cost of capital, tax codes that affect the financing operation of a firm will directly impact on the METR via the cost of capital.

Second, consider the cost of depreciation for capital goods. Unlike labor and raw materials, immediate expensing for a capital good is generally disallowed because its economic life is generally longer than one period. Longer-lived physical assets, however, do depreciate in value through repeated use. To provide for depreciation costs, all corporate systems allow some depreciation charges to be claimed against revenue. But the present value of deductible depreciation allowances rarely equals expensing, thereby causing a firm to bear part of the tax-induced cost of investment. To minimize this tax-induced cost, a firm will undertake less investment. On the other hand, if the depreciation allowances are overly generous, then the tax system will effectively be subsidizing investment at the margin. Hence, tax depreciation provisions can give a firm either insufficient or excessive incentive to invest in new capital. In either case, a tax wedge will be created. In general, other things being equal, the greater the deviation between the economic cost of depreciation and the tax depreciation allowance, the greater will be the distortion and, accordingly, a higher METR will be generated.

Third, the METR is also affected by the return to savers via the personal income tax. In equilibrium, the after-tax return to holding securities must be the same regardless of the kind of financial claims held. By affecting the after-tax return to a saver, the marginal personal tax rate will also have an impact on the METR statistic. If there were no personal income tax on returns to savers, then more lending would be forthcoming. On the other hand, a higher marginal personal tax on capital income will be more distortive as it lowers the price of present consumption and, other things being equal, will result in a larger METR.

Apart from tax factors, inflation can also have an impact on the cost of capital measure. This is due to the historical cost accounting method of valuing assets and depreciation charges for tax purposes.101 For instance, without indexation for inflation, future depreciation deductions will be less valuable to a firm. When inflation is high, a firm is penalized for having invested in long-lived capital because inflation erodes the real value of future tax depreciation from those investments. Therefore, inflation undermines the ability of a firm to recover the real economic cost of using its capital. On the other hand, a highly levered firm that finances its operation with nominal debt will benefit from high inflation because the firm will be able to write off as costs the fall in the real value of the principal due to inflation. So, inflation increases the value of nominal interest deductions. Thus, without indexation, inflation can either increase or decrease the true economic cost of investment. A firm wanting to minimize cost (maximize profits) will react by undertaking less (more) investment if it perceives that the loss from tax savings on future depreciation is greater (smaller) than the gains from nominal interest deduction. Acting through the cost of capital, inflation, like a tax parameter, will also influence the METR. The direction in which inflation will affect the METR is not possible to deduce theoretically, given the two opposite forces at work.

Computing the METR: Simple Analytics and an Empirical Application

The relationship among the various factors affecting the cost of capital (and therefore the METR), discussed above in general terms, is illustrated here using a simple analytical framework. A representative firm’s investment decisions in each period are guided by the principle of maximizing the net present value of returns to its investments. The firm will invest to the point where the real rate of return, R, of the marginal dollar of investment for one period is equal to its costs in that period. In a world without taxes, the costs to the firm will comprise (a) the real cost of financing the marginal dollar of investment, r-π, where r is the nominal cost of financing and π the expected rate of inflation; and (b) the real economic depreciation, δ If this investment is financed by debt, r would be the market rate on bonds i; if it is financed by equity, r would be its cost of equity ρ. In general, then, the nominal financing cost can be expressed as r = βi + (1-β)ρ, where β is the proportion of investment financed by debt. Hence, to be undertaken, the marginal dollar of investment must satisfy the following condition:

R=[βi+(1-β)ρ]-π+δ.(1)

With taxes, the optimization principle still applies, but the equality of real returns and costs must now be stated on an after-tax basis. In this case, the real return to the firm of the marginal dollar of investment becomes R(l-u), where u is the corporate tax rate. The costs of that same investment must now also be modified as follows: (a) the real cost of financing, rt-π, where rt denotes the after-tax cost of financing; assuming that debt is deductible but equity is not, rt = βi(1-u) + (1-β)ρ; and (b) while the real economic depreciation on the marginal dollar of investment δ remains unchanged, the effective price (net of financing costs) of that investment is now reduced by the stream of future depreciation allowances provided for in the tax system. Denote the present value of such depreciation allowances by Z.102 Then, with taxes, the firm will invest to the point where the marginal dollar of investment satisfies

R(1-u)={[βi(1-u)+(1-β)ρ]-π+δ}(1-Z).(2)

Notice that when u=0 (no taxes), equation (2) reduces to equation (1).

As is well known, R can be viewed as comprising three parts: (a) an amount which is used to maintain the real capital stock of the firm; (b) an amount that goes to the tax authority; and (c) the remaining amount as the after-tax real return to savers on the marginal dollar of investment undertaken. Since by assumption capital depreciates at the rate δ, to maintain the real value of the firm’s capital stock, the before-tax return to investment, net of depreciation, is not R but R-δ. This is the rge depicted in Figure IV.2. If the aftertax real return to savers is denoted by rne as before, then the tax wedge, t, given by rge-rne, represents the tax revenue collected by the government on the marginal dollar of investment. The METR of this investment is, therefore,

METR=1-rne/(Rδ).(3)

The after-tax real return to savers, rne, is affected by, in addition to the corporate income tax, the savers’ effective marginal personal income tax rates, which may or may not be different for different forms of corporate distributions, such as interests, dividends, and capital gains. A full treatment of these aspects of the income tax system is, however, beyond the scope of this Handbook.

With adequate macroeconomic data (such as the expected rate of inflation, the market interest rates, and the costs of equity), investment-specific data (such as the economic depreciation of different assets), firm-specific data (such as the financing mix between debt and equity), and tax variables (such as the depreciation system, and the structures and rates of the corporate and personal income tax systems), a tax wedge can thus be calculated for each type of asset (e.g., machinery) across different sectors of an economy. An example of such calculations is given below.

Table IV.10 presents the METRs by industry for several investment types for Canada in the mid-1980s.103 The first observation is that the METRs vary widely across capital assets and sectors. This implies that incentives for the allocation of resources across sectors would differ because of the different distortionary impact (intended or otherwise) of the corporate tax system. The METRs range from a tax subsidy of -14.6 for inventories in the agriculture-fishing-forestry sector to a tax burden of 42.3 for wholesale trade holding in inventories. The same matrix, however, also illustrates that, except for inventories investment in the agriculture-fishing-forestry sector (due to favorable inventory tax treatment only for this sector), inventories are the most highly taxed asset. Furthermore, other than the retail trade and service industries, investment in machinery is less highly taxed than investment in building (mainly owing to higher depreciation allowances for machinery). In general, land is also less highly taxed compared to buildings, but buildings are more highly taxed than machinery investment in some sectors.

Table IV.10.

Effective Corporate Tax Rates by Industry

article image
Source: Boadway, Bruce, and Mintz (1987), p. 87.

In the aggregate, the matrix shows a more favorable tax treatment for investment in agriculture, fishing, and forestry (via a lower small business tax rate) as well as manufacturing and services industries (via an accelerated capital cost allowance write-off) than investment in other sectors. What is less obvious from the table, however, is whether the tax system is more distortion-ary for investment across sectors or more distortionary for the investment of different assets within each sector.

Tax Incentives

Dale Chua

  • What are the argumentsfor and against tax incentives?

  • What are some of the more commonly encountered tax incentives?

  • A country study: Do tax incentives do the task they are designedfor?

Proponents of tax incentives, especially those in the business community, often argue that tax incentives stimulate investment. They assert that by offering the correct incentives, there will be increased investment in the economy, generated from within and augmented by foreign investment flows. On the other hand, they argue that with free capital mobility, if a country does not match the tax incentives offered by neighboring countries, there will be capital flight from the country as capital searches for areas offering the highest aftertax return. Although there is some validity in these arguments, they tend to comprise only a small part of the total picture on which investment decisions are made.

There are more compelling reasons besides considerations that investment flows into a particular activity in a selected area.104 For example, factors such as economic and political stability supported by adequate infrastructure, an untapped but trainable labor force, and natural resources are equally, if not more, important reasons for investors. Apart from these factors, some investors would also contend that a well administered tax system—with a low tax rate—that is certain and simple would rank even higher than tax incentives. Furthermore, whether tax incentives are effective in attracting foreign investment depends on the home country treatment of repatriated profits. If there was no prior agreement between the host and home countries with respect to special tax treatments of foreign investment, the tax revenue forgone by the host government may simply be transferred to the home country government.

Additionally, the benefits from increased investment induced by tax incentives must be judged against its costs to society. From an efficiency viewpoint, by favoring one form of economic activity over another, tax incentives distort relative prices, and therefore misal-locate resources. Tax incentives may be viewed as inequitable because they single out a particular sector for preferential treatment. Further, they undermine the sense of fairness, because a heavier tax burden must be placed on other sectors to raise a given tax revenue. Moreover, the allocation of resources is distorted further since nonpreferred sectors must pay even higher taxes so that some activities may enjoy a lower tax liability. Finally, tax incentives undermine the simplicity of tax administration by increasing monitoring costs. From an empirical standpoint, experiences tend to confirm that tax incentives are likely to lead to tax evasion as a result of the accompanying complexity of the tax system, the high degree of selectivity, discretion, and control in granting of incentives. Experience also reveals that tax incentives are often not well targeted or confined to their original objectives.105

In spite of these arguments, tax incentives are still observed in developed and developing countries as a policy option to induce investment. Countries offering such incentives believe that tax incentives, in whatever guise, must be the best and least costly way to encourage investments. We assess, in turn, three common incentives, namely, tax holidays, rate reductions, and investment tax allowances.

Tax Holidays

Tax holidays as an incentive are employed mainly in developing countries. An enterprise receiving a tax holiday is partially or fully exempt from the payment of the corporate tax over the period for which the tax holiday applies, usually in the early years of its operation. In certain cases, the renewal of a tax holiday upon expiration may be permitted or a period of lower than the normal corporate tax rate may be accorded to firms, but most countries offering such incentives require that once the holiday ends, the enterprise is treated no differently from a nonexempt business.

What are the issues associated with tax holidays? First, the usefulness of a tax holiday as an incentive depends largely on the profit status of a recipient firm. For a profit-making firm, a tax holiday raises its return on investments since profits are tax-exempt. The benefit is immediate. For an unprofitable firm, it offers little benefit since no taxes are payable anyway. This fact undermines the effectiveness of tax holidays as a policy option to encourage new investment because if the aim of the policy is to assist firms in the start-up stage, then it has clearly not achieved its objective since many firms in the starting periods are likely to be nonprofitable. On the other hand, there is no reason why firms that are profitable from the outset should be given added incentive to undertake an activity that they would undertake in any case.

From the viewpoint of a firm, however, whether or not a tax holiday is desirable will depend on the tax treatment of depreciation allowances accrued during the tax holiday period. If accrued initial and/or annual depreciation allowances cannot be carried forward, then the attractiveness of a tax holiday may be reduced because the “losses” arising from the inability to deduct depreciation charges (tax saving forgone) may outweigh the gains from a lower tax rate during the holiday period.106

A second, but equally important, weakness is that a tax holiday erodes the tax base. The issue of base erosion is potentially more serious than simply the direct revenue forgone. As is often the case in developing countries, a tax-exempt enterprise may be part of a larger holding of nontax exempt companies. It is easy to shift income from profitable but taxable companies in the group to the tax-exempt enterprise through the transfer pricing of intercompany transactions. Hence, awarding tax-exempt status to a single firm may erode the tax base more than on first approximation, since such abuses are generally difficult to police.

Third, a tax holiday removes the appeal of debt financing of capital investment. This is because the interest cost of borrowing is a fully deductible item. A tax holiday, with a preferential or zero tax rate, increases the relative cost of debt financing by removing the benefits of interest deductibility. Furthermore, in countries where dividends distributed by tax-exempt firms are also exempt from the personal income tax, the relative attractiveness of equity financing is further increased. Whether or not this is desirable is harder to judge since, on the one hand, it implies that the optimal financial structure of a firm will be influenced by tax holidays, while, on the other hand, some would argue that the upshot of this effect is a sounder financial structure for the firm and will reduce bankruptcy risks.

Fourth, tax holidays encourage short-term investment because it is the type of investment that will benefit most from tax holidays. Longer-term investment can only benefit fully from tax holidays if the tax holiday is renewable.107

In short, because tax holidays erode the tax base, create an opportunity for tax planning, increase monitoring cost, and tend to benefit only investments that are already profitable and/or short-lived, their value as a policy option to encourage investment should not be exaggerated. Furthermore, recalling that foreign capital inflows are affected by a web of factors, it should be obvious that tax holidays on their own are neither a necessary nor a sufficient condition for the encouragement of new investment.

Rate Reduction

Reducing the regular or general corporate tax rate is one of the best approaches to achieving the competing goals of tax policy design. Lending support to this view, corporate tax rates all over the world have been declining in recent times. In a world where some revenues have to be raised, a low corporate tax rate is currently seen as the best incentive. The strongest argument in its favor is that a lower tax rate increases the after-tax return to investors. Compared to tax holidays where some firms are treated more favorably than others, a single rate for all corporations is fairer, not selective, and administratively simpler. Monitoring cost will be reduced as the incentive for intercompany transfer pricing is eliminated. A stable and low rate, with minimal tampering with the tax system, has the additional benefit of promoting a sense of stability and increasing business confidence in the country’s tax system. Against these advantages is the fact that a lower rate requires a larger base to raise a given amount of revenue. This implies that in a static setting, a nonexpanding tax base may be incapable of raising sufficient revenue if the tax rate is lowered.108

Investment Tax Allowances

A common tax incentive found in many OECD and developing countries is the investment tax allowance. This incentive takes the form of a faster write-off for investment expenditures. It can appear under many guises which can be grouped as relating to (1) accelerated depreciation allowance, (2) investment expenditure allowance, or (3) investment tax credit.109

Although the main effect of these investment tax allowances—an increase in the after-tax return for the firm—is broadly the same, the time profile of tax savings for the recipient firm is slightly different under each type. As a rule, an accelerated depreciation allowance that permits a quicker write-off for a qualifying investment will affect the cost accounting of a firm in a very specific way. Therefore, how much in taxes is saved up front, as well as the implication for future tax savings, will depend on the specific design of the accelerated depreciation allowance.110 In a different way, but having the same qualitative result of increasing a firm’s after-tax return, an investment expenditure allowance allows a firm to write off a specific percentage of investment expenditure from its taxable income in the year the investment is undertaken. In this case, the relief constitutes a direct up-front tax saving for the firm. Lastly, an investment tax credit that permits a firm to reduce its tax liability by a fraction of the qualifying investment is also an up-front tax incentive. The difference between the last two is that the former is a deduction against taxable income while the latter is a credit against tax liability.

The main advantage of an investment tax or credit allowance is that its focus is specific. For example, a firm receives the benefit of the allowances against its tax liability only if it invests. Effectively, the investment tax allowance translates into a lower tax rate for the firm. Because it is narrowly based, proponents of this incentive argue that it can be used to encourage firms to take a long-term view by targeting long-lived capital goods with investment tax allowance. Furthermore, given that its focus is on current investment, there is a smaller erosion of the tax base relative to an across-the-board reduction in the corporate tax rate. As a rule, the actual value of any investment tax allowance depends largely on whether the firm is in a position to use the incentive. As in the case of tax holiday, if a firm is in a tax-loss position, it may not be able to fully benefit from the investment tax allowance. For such a firm, the value of the incentive will be small if the authorities do not refund the allowance (that is, give the firm a refund equal to the tax value of the investment allowance benefit) or permit the carryback of unused allowance against previous year taxes.

The drawbacks, however, of investment tax allowances, apart from the obvious loss of tax revenue, are (1) that they tend to favor more established firms over new firms because the latter are more likely to be in a tax-loss position and will not be able to take full advantage of this incentive; (2) that because this feature is frequently associated with non-inventory investment, the selectivity of this incentive provides no benefit to those industries where inventory investment is important; and (3) that they favor capital goods that depreciate quickly because frequent investment implies frequent claim on the investment allowance for the firm.

Country Study: Canada

To reinforce the generally accepted view that tax incentives by themselves do not always solve the problems they are designed for, we review, in this section, a recent study in Canada that assessed the benefits and costs of using a specific tax incentive to promote private sector investment. The incentive adopted was an investment tax credit (the Cape Breton Investment Tax Credit), which was set at 60 percent of investment expenditure. The investment tax credit was introduced for a subregion of the Province of Nova Scotia during the period 1985–92, a region severely affected by an increase in unemployment because of major manufacturing enterprise shutdowns. An evaluation of the economic effects undertaken by the Government of Canada reported, among other conclusions, the following:111

(1) By lowering the cost of capital, the tax credit led to a substitution of capital for labor (other things being equal, this substitution effect reduces employment).

(2) By reducing production costs in Cape Breton, the tax credit enabled firms to supply goods and services at a lower price and/or earn higher profits, which in turn stimulated production in the region.

(3) A direct subsidy to employment equal to the tax expenditure cost of the credit would have created more employment.

(4) Of the private investment that qualified for the tax credit, no more than 19 percent was estimated to be incremental.

(5) Based on interviews and on econometric evidence, it was estimated that for every dollar of federal tax forgone, 75 cents of new investment was stimulated by the credit.

(6) The tax credit improved labor productivity in the region, but decreased capital productivity, so that the credit is unlikely to contribute to the long-term improvement in the total factor productivity performance of the region.

(7) The cost of jobs created with this measure was extremely high.

(8) Economic policies that contributed to the attainment of high employment levels at the national level appear to provide a more promising method of mitigating regional income disparities than a regional investment tax credit.

The findings of the study are very specific and show that even in the case of a seemingly well-targeted tax incentive such as an investment tax credit, the cost to society is high, and that, to achieve a similar outcome, less costly alternatives might be available. Therefore, from a policy standpoint, the usefulness of tax incentives should not be overrated.

Cash-Flow Tax

Parthasarathi Shome and Christian Schutte

Introduction112

Conceptually, cash-flow taxation is based on consumption; thus, it is neutral with respect to capital formation. The practical advantages of cash-flow taxation—its definitional clarity and simplicity of measurement—were discussed in a U.S. Treasury report (1977). Perhaps they have received new interest as, more recently, tax theorists have also come to emphasize implementation and administration aspects of tax policy,113 especially for developing countries.114 The classic Haig-Simons ideal of a comprehensive income tax, based on consumption plus net accrual of wealth, seems to be rather problematic when judged by this criterion. It is argued that the hypothetical nature of the accrual concept tends to create complexities in the tax code, to increase the burden of administration and compliance, and to foster avoidance and distortion. Under these circumstances, a cash-flow tax base seems to be a promising alternative especially at the corporate level, where equity concerns about exemption of capital income are irrelevant and some technical problems are less salient.115

Opponents of the cash-flow tax question the superiority of the cash-flow tax base on equity grounds. Also, they are usually not optimistic about the administrative advantages of the tax. In the context of the corporate sector, they decry it primarily on the basis of implementation problems as well as the lack of international experience and, consequently, coordination. The doubts emanate from perceived difficulties in containing tax evasion because of transfer pricing practices or tax avoidance through intracompany leasing arrangements and because the tax may not be creditable in those countries that export capital until they themselves introduce the tax. Therefore, the tax may not be compatible with the existing international tax regime. In addition, political forces that lead to base erosion of the corporate income tax (CIT) are, of course, also likely to affect the corporate cash-flow tax (CCFT). While the cash-flow base may address some inherent shortcomings of the income tax, in and of itself, it is unrelated to the political willingness and ability to keep a tax system clean of special provisions and targeted incentives.

Conceptual elements

Conceptually, the CCFT has been discussed as a supplement to a personal expenditure tax, a personal income tax, a value-added tax (VAT), and as a tax on economic rent. The CCFT base also varies in conception, reflecting whether real transactions or real and financial transactions are taxed.

CCFT tax base. There are three types of CCFTs:

(1) The R-based CCFT (real) is one in which the tax base is net real transactions—that is, the difference between sales and purchases of real goods and services. As opposed to an income tax, the distinctive features of such a tax base are immediate expensing of capital outlays and the nondeductibility of interest payments. At the same time, interest received is no longer taxable.

(2) The RF-based CCFT (real plus financial) is one that also includes nonequity financial transactions—that is, the difference between borrowing and lending. Interest and retirement of debt would be deductible, but borrowing and interest received would be taxable: RF base = (sales + borrowing + interest received) =(purchases + interest paid + debt repaid).

(3) The S-based CCFT taxes the net flow from the corporation to shareholders—that is, S = (dividends paid + purchases of shares - issues of new shares).116 The S base is conceptually equivalent to the RF base minus the CCFT, as can be seen from a basic accounting identity: any difference between total business inflows and outflows has to be paid out either to shareholders or as tax, RF = S + CCFT. Since taxes enter into the sources and uses-of-funds statement, the rate of the RF-based tax would be tax inclusive. The S-based rate would be tax exclusive—it could well be higher than 100 percent.

Selected characteristics. The CCFT represents a silent partnership for the government in an investment. This is most clearly seen for the S base where the government, in fact, sustains tax losses from equity raised and receives revenue from distributed earnings. As outflows and inflows for the corporation are reduced proportionately by the “silent partnership,” the rate of return on an investment remains unaffected by taxation.

The “silent partnership” enables the government to appropriate a share of the above-normal returns that are generated in the economy (and a share in the cost of below-normal returns). Those above-normal returns may be economic rents from entrepreneurial activity, from nonrenewable resources, or from monopoly, but can also be investors’ compensation for risks which on average will be positive. Hence, the CCFT can also be interpreted as a tax on pure profits and on returns to risk taking.

From a theoretical perspective, there are a number of other attractive features of a CCFT:117

• The exemption of marginal returns implicit in immediate expensing does not discriminate between debt and equity. The income tax favors debt over equity by allowing deduction of interest only. Partial integration of corporate and personal income taxation does not solve this problem as long as there is a substantial spread of marginal investor-tax rates.

• Immediate expensing also ensures neutrality with respect to the rank-ordering of projects.

• Except for situations of hyperinflation (where even annual expensing falls short of a full deduction of real investment), the cost of capital is not affected by inflation under the CCFT.

• If the CCFT is introduced alongside a personal income tax, there is no need to integrate the two taxes. Because capital income is effectively exempt at the corporate level under the CCFT, the appropriate treatment would be the classical system under which the corporation is treated as a separate entity and no effort is made to attribute its earnings to equity holders.

• In addition, the CCFT is based on current transactions and hence avoids the timing-related problems of a typical income tax: expensing replaces calculation of “true economic depreciation” as well as the need for inflation adjustment of inventory and asset replacement values. The problem of capital gains is irrelevant.

For practical purposes, however, two critical assumptions underlying the theoretical neutrality results need to be stressed: it is assumed that tax rates are constant and that taxable inflows are always sufficient to offset expenses, so that an investment will actually produce an initial tax reduction.

Practical Considerations

While the CCFT cannot be strongly criticized for lacking theoretical foundations, it does come up against some practical hurdles. The important practical considerations for the CCFT fall into two categories: those that involve the transition phase and those that involve its general implementation.

Transition issues

There are numerous concerns that arise during the transition from a CIT into a CCFT. First, a “cold turkey” transition would produce windfall-tax revenue by denying companies their existing depreciation allowances. On the other hand, allowing immediate expensing of remaining depreciation could adversely affect revenue. A hybrid of allowing continued depreciation might be the only practical solution. As Sunley (1989) points out, however, such “transitional” arrangements may have to last for a number of years.

It is suggested that a short-term revenue loss is likely during the transition. Various arrangements could accommodate the amortization of old investment while new investment could generate substantial tax losses. In order to mitigate this effect, one may resort to “present value expensing”—that is, during the transitional period, deductions for new investments could be spread out over several years, grossed so that their present value would still equal the initial outlay. This would raise the issue of using the right discount rate.

The particular choice of the CCFT base and transitional provisions will obviously affect the financial position of firms. Under an R-based CCFT, leveraged firms could face financial distress because interest would no longer be deductible. Yet, continued interest deductibility for old debt might be prone to manipulation—particularly since old debt would be difficult to define. The solution, therefore, would seem to be an RF-based CCFT.

In the very short term, the tax may have undesirable announcement effects. Investment might collapse in anticipation of future expensing unless the tax can be introduced retroactively. If the prospective CCFT is RF-based, firms may increase borrowing and later repay debt by raising equity. Whether retroactive enactment is possible could depend on political factors.

General issues

Under a CCFT, issues such as the stability of revenue, a high possible incidence of tax avoidance and evasion, and international compatibility assume particular importance. These are addressed below.

Revenue implications. The CIT is an important source of revenue in many developing countries. Following a bell-shaped curve, its share of GDP and total revenue generally increase in the initial stages of development, so that for different income groups of developing countries, the CIT makes up between 11 and 23 percent of total revenue. In a few cases, the CIT accounts for more than one-fourth, even more than one-half, of total revenue.118 In industrial countries, the CIT has become relatively unimportant over time.

The revenue implications of any change in corporate taxation have to be weighed very carefully in this context—even if one may argue that, in the long run, the CCFT is likely to foster growth through increased investment and improved capital allocation and that the government would participate in such growth. If we leave aside both the purely transitional issues and the longer-term structural and dynamic effects, what can be said about the revenue implications of the CCFT?

(1) Smaller tax base? There are conflicting views about the likely differences in the size of the corporate tax base under the CIT and the CCFT. The straightforward argument against the CCFT is that full and immediate expensing seems to reduce the tax base. The government forgoes tax on the marginal returns to capital, and one would therefore expect the CCFT rate to be higher than the initial income tax rate if present-value revenue is to be sustained.

On the other hand, proponents of the CCFT have based their case partly on the massive erosion of the tax base under the CIT.119 They argue that most marginal returns escape taxation anyway. Firms find it advantageous to finance their investments through debt, as nominal interest payments are deductible. Foreign investors may choose “thin capitalization” to shield their income from host country taxation and to facilitate repatriation. Legislative rules against excessive interest deductions often are not fully effective. Firms may also avoid taxation of capital income by making use of special incentives, such as accelerated depreciation, tax arbitraging, tax-preferred activities, and investing abroad.

Empirical work on developed countries suggests that the CCFT and the current income tax base would not be very different in many cases.

In developing countries where the corporate sector is dominated by large mineral exporters, local cartels, monopolies, and debt-financed foreign corporations, the existing CIT may also be fairly close to a pure prof-its tax, as the tax base mainly consists of above-normal returns.

(2) Investment and current revenue. The tax yield under the CCFT would also tend to be very sensitive to investment. Under an income tax with “true economic depreciation,” gross returns and offsetting capital allowances on an investment follow the same time pattern. But, under the CCFT, taxable inflows from past investments are partly offset by the expensing of new outlays. Hence, current revenue will depend on the difference between the average rate of return and the rate of growth of the capital stock. During periods of rapid expansion—which, for instance, may follow structural adjustments in reforming socialist economies—revenue could dry up or even become negative, at least theoretically. In other words, revenue could drop during upswings in economic activity, making the tax procyclical.

(3) Revenue risk. As the government assumes the “silent partner” role with full loss-offset, revenue from individual projects becomes more risky—though its expected value is still positive if investors are risk-averse. Still, the “silent partnership” should not cause substantial variations in total revenue as long as independent risks of many projects can be pooled. But in the case of a small country with only a few major projects, or in the case where risks are correlated, variability of revenue as such may be an additional concern.

(4) Other effects. Two other revenue-related points deserve mention. First, introduction of the CCFT may require substantial improvements of loss-offset provisions of the existing CIT. Such improvements may be costly in terms of revenue if previous loss trading between corporations was imperfect. Second, there may also be a revenue-increasing effect if a CCFT actually improves the administration of taxes on small businesses and other hard-to-tax groups.

To conclude, the revenue effect of replacing a CIT with a CCFT remains an empirical question. Whether or not a revenue-neutral CCFT would require higher rates depends on the particular income tax laws that are to be replaced, the value of economic rents earned in the corporate sector, and the current debt-equity compositions of corporate portfolios. Transition rules might be needed to soften any adverse revenue effect of the conversion, and such rules may have to be applied for a considerable period of time.

Tax avoidance and evasion

The above discussion of revenue effects excluded possible behavioral responses on the part of corporations. As firms, however, will try to exploit possible new “loopholes” in the CCFT, revenue could be lost and tax administration might face new challenges.

(1) Gaming the system. Some tax avoidance possibilities are specific to the R-based CCFT, as there is an entire class of schemes exploiting the crucial difference between taxable real flows and tax-free financial transactions.120

• Installment sales to a tax-exempt party may understate the taxable purchase price but overstate the tax-free interest component of seller financing.

• Labor, goods, and services may be sold at low prices and assets leased at low rates to a tax-exempt party, who in return would provide a low-interest loan to the employee, seller, or lessor; prearranged defaults and loan forgiveness would be extreme cases of such low-interest loans, unless they are included as imputed flows in the tax base.

• Companies with different accounting years may reduce their tax bases by increasing purchases from each other. At the end of its accounting year, company A could make large purchases from company B and vice versa. These intercompany transactions could be debt-financed without tax consequences.

• To circumvent nondeductibility of interest payments, financing may be provided by a tax-exempt seller or lessor. Interest payments would be transformed into deductible leasing payments or purchases.

To contain these arrangements, McLure suggests that for tax deduction purposes, ceilings and floors may need to be imposed on interest rates.121

Under both the R and RF bases, taxpayers may try to shift the tax base to an affiliated low-tax party, for instance, a tax-exempt pension fund or a foreign corporation with a lower rate. Expensable capital outlays would be allocated to the high-tax party, and subsequent cash inflows would be directed toward the lowtax party. Such base shifting may take the following forms:

• transfer pricing through the purchase of inputs from the low-tax party at inflated prices, and the sale of goods at understated prices;

• low-rate leasing of capital acquired and expensed by a high-tax party to a low-tax party; and

• selling expensed assets at understated prices to the low-tax party.122

Such schemes could be operated under the income tax as well, but the incentive for them is much more powerful under the CCFT. This is because expensing makes the present value of deductions on any asset equal to the purchase price. Under any other depreciation scheme, the present value of deductions decreases with the longevity of an asset and with the discount rate used by the corporation.123 Also, because the entire deduction is available up front under the CCFT, immediate sale of the asset at an understated price becomes much more attractive. Under an income tax, the high-tax party would have to hold on to the asset to benefit from available depreciation.

A CCFT will hence increase the incentive for taxsaving leases and for mergers between corporations with different tax rates. Foreign corporations may set up subsidiaries in the CCFT country only to take advantage of expensing, then channel inflows to a lower-rate jurisdiction.

Manipulation of reported transactions may also be an important channel of tax evasion. Companies could try to overstate asset prices upon purchase to tax authorities. They could also buy equipment, take the deduction, and immediately resell, concealing or understating the price. These possibilities also arise under the income tax. Again, however, expensing, which grants tax savings up front, increases their attractiveness.

To counter base-shifting schemes, arm’s-length prices and rates for transactions between affiliates have to be enforced. But such monitoring is notoriously difficult. Perhaps certain types of transactions, such as leasing to foreigners or tax-exempt institutions, would need to be prohibited. If there is a system of wealth taxation, it may put some checks on the valuation of transferred assets. All such requirements, however, would result in considerable complications, essentially eroding the tax’s main characteristic—simplicity—on which its proponents base its attractiveness.

Some general lessons from the income tax apply even more so to the CCFT in this context. Tax treatment of activities and institutions should be uniform to reduce arbitraging opportunities. For instance, some business activities of tax-exempt institutions may be taxed. The rate structure should be flat and low to the extent possible (given revenue needs), since it determines the taxpayer’s per-dollar savings from reducing or shifting the tax base.

(2) Tax exhaustion and tax avoidance. A special case of uneven rates arises from tax exhaustion, where available deductions exceed taxable inflow. A tax-exhausted corporation has a marginal tax rate of zero, though its statutory rate may be quite high. It is unable to benefit from capital allowances.

Excess allowances (through tax losses) are likely to be much larger and more frequent under the CCFT. Especially with an R base, however, whether or not the resulting lumpy tax profile will foster additional arbitraging and mergers largely depends on loss-offset provisions. To the extent that such provisions ensure symmetrical treatment of profitable and loss-making corporations, profitable arbitraging would be curtailed.

Loss offsets are crucial to the CCFT in conceptual terms as well, and a refund would be the straightforward solution. Refunds may be problematic though, in that they aggravate the problem of “hobby farms”: businesses solely set up to generate tax losses on consumptive, nonprofit-oriented activities. Rules against such abuses under the income tax would have to be carried over to the CCFT.

To conclude this section on tax avoidance and evasion, a summary assessment of the R and RF bases is called for. The administrative advantage of the R base is that financial transactions can be entirely ignored. This actually reflects the basic concept of CCFT—equal treatment of debt and equity. On the other hand, the R base is vulnerable to the above-mentioned tax avoidance schemes. With the RF base, the tax profile is less lumpy, and incentives for base shifting and evasion are reduced. The RF base, however, has an incentive to raise capital as equity and disguise payouts as interest payments, a problem that is shared under the income tax. Therefore, the existing CIT provisions to ameliorate these problems would have to be carried over to the RF-based CCFT.

International issues

Since the CCFT is an untried tax, many legal and economic questions remain with respect to its international compatibility, especially where tax treatment of foreign investment income is concerned.

(1) Basic concerns. The introduction of a CCFT raises serious questions about international compatibility. These questions concern legal as well as economic aspects. As the CCFT might not legally qualify as an income tax, its adoption could require the renegotiation of tax treaties. Such negotiations tend to take many years and hence imply considerable transactions costs and transitory arrangements. Moreover, tax treaties for many host countries offer a stability that they may not want to risk. Reforming socialist countries that ultimately want to join the European Community (EC) may find the CCFT unacceptable simply because it would not conform to the EC requirement for a CIT.124

Host countries are worried about losing existing options for “soaking up” foreign tax credits. Also, there has been an overriding concern that home countries—in particular, the United States—might not grant foreign tax credit for the CCFT and that this might dis-courage foreign investment. The creditability problem was the major obstacle for the adoption of CCFT proposals in Canada, Mexico, Sweden, and Colombia.125

(2) Principles for taxation offoreign earned income. Briefly recalling the basic principles that are applied to the taxation of foreign income,126 the following three regimes can be distinguished:

Exemption. Home countries impose no tax at all on income earned abroad. This is sometimes also referred to as the territorial system. Income is only taxed by the host country (source principle).

Taxation upon accrual. Home countries reserve the right to tax worldwide income of their resident corporations (residence principle), and foreign income is taxed as it is earned. Taxation upon accrual is typically applied to foreign branches of resident corporations.

Taxation upon repatriation. Home countries apply the residence principle. Corporations can, however, defer their domestic tax liability by retaining earnings abroad. Because of the time value of money, deferral reduces the effective domestic tax rate. This regime is typically applied to subsidiaries.

If home countries apply the residence principle, foreign income is potentially subject to double taxation in both host and home countries. Double taxation can be mitigated in different ways. First, the home country may allow deduction of taxes paid abroad. This is actually the efficient policy, since foreign taxes represent a social cost to the home economy. If deduction is granted, resident corporations will equalize the aftertax foreign return to the pretax domestic return. This is not optimal from a global point of view, however, since capital export is discriminated against. Second, to en-sure capital-export neutrality, many home countries grant a tax credit against foreign taxes paid. Corporations will then equalize pretax rates of return. In terms of revenue, the tax credit implies that, in fact, home countries bear the foreign tax burden of their resident corporations. Unless additional tax treaties impose restrictions, the host country can “soak up” those tax credits—that is, it can tax foreign investors without deterring them. The tax credit, however, is usually limited to domestic tax liability (on the sum of domestic and foreign incomes), so that corporations ultimately face the higher of the (average) foreign or domestic tax rate. If the domestic rate is higher, the corporation will face the same effective tax rate at home and abroad. If the foreign rate is higher, the corporation may accumulate excess tax credits. Under a system of tax credit by source, such offsets are limited to income from a particular host country. Under the more generous worldwide tax credit system, excess tax credits may be used against tax on income from any host country. In this case, the corporation actually faces the higher of the domestic and the average foreign tax rate.127

(3) CCFT and foreign direct investment. Here, three regimes can be identified.

Exemption at home. The concerns about revenue and creditability are irrelevant for foreign investment from home countries which grant exemption for foreign dividend income; in this case, the host country will attract additional foreign investment until the pretax rate of return is equal to the after-tax rate of return in the home country.

A creditable CCFT What if the home country taxes foreign earnings? Let us first assume that CCFT would be creditable. Another crucial distinction has to be made between corporations whose available tax credits are less than their domestic liability on foreign earnings (their so-called excess limit position) and those who have excess credits.

If corporations are in an excess limit position, the tax credit mechanism would wash out the effects of host country tax policy. The revenue argument against CCFT is based on this point. While the investment incentive of the CCFT would be neutralized, revenue is forgone, simply to be picked up by the home country.

Note, however, that this argument would not hold in the presence of “tax sparing,” if tax sparing were to apply to the CCFT as well as to specified investment incentives under the income tax. Under tax sparing, the home country assumes that full tax has been paid in a foreign (host) country, in effect calculating foreign tax credit on the basis of regular foreign tax rates regardless of actual taxes paid (on the basis of preferential treatment). This approach protects host country tax incentives. With the notable exception of the United States, many capital-exporting countries (such as Japan and the United Kingdom) have signed tax-sparing treaties with developing countries.

Note also that under a “deferral system,” retained earnings are tax exempt in the home country. The investment incentive of the CCFT may therefore remain effective.

A noncreditable CCFT. So far, it was assumed that the home country would grant foreign tax credit for the CCFT. Possible noncreditability of the CCFT in home countries, however, especially in the United States, has been a major concern in countries considering the tax. There is no clear answer in advance to the legal question of creditability, although there seem to be policy reasons favoring its creditability. What if the CCFT were not creditable? Would double taxation discourage foreign investment? McLure (1991) points at three qualifications to this common argument. First, “mature investment” may still be attracted by the CCFT because a repatriation tax does not affect rates of return at the margin. Second, corporation may have excess tax credits and therefore be back to source-based taxation at the margin. Third, it may be argued that a tax with a marginal effective rate of zero “even when combined with a (home country) tax on repatriated earnings is unlikely to have much disincentive effect on investment in the host country.”128 A noncreditable CCFT will not distort investment at the margin, because for projects just earning the opportunity cost of capital, net CCFT payments will simply be zero.

For a project earning above-normal returns, the CCFT burden that is deductible but not creditable at home will become an additional cost. To the extent that such investment is mobile, it will be discouraged by the CCFT. One may argue, though, that above-normai returns on investments in developing countries are often earned on immobile production opportunities (for example, extraction of mineral resources or exploitation of a local monopoly by a multinational trademark). In these cases, the noncreditability of a tax on pure profits will have no effect. Note, however, that the noncreditability of CCFT is likely to entail the noncreditability of any supplementary withholding taxes. A noncreditable withholding tax would clearly introduce a distortion even for marginal investment.

Conclusions

The CCFT has the drawback of any untried tax innovation, simply that “no one does it.”129 There is no experience and administrative know-how about the possibly complex details of a transition to the CCFT, its operation, and the avoidance schemes that might emerge. No official ruling on the critical question of the creditability of the CCFT has been required so far. The uncertainty costs of experimenting with the tax may be reason enough for a developing country or an economy in transition not to implement the CCFT. The purpose of this chapter, however, is to identify potential sources of problems and to better understand the conditions for a successful experiment with a CCFT. It seems clear that, depending on the existing CIT structure, the structure of the corporate sector, the relative importance of foreign investors, and the mix of countries they come from, some countries may find the CCFT more attractive than others.

The key conclusions of the chapter may be briefly summarized with the following:

1. The theoretical pros of the CCFT seem clear. It can be interpreted as a “silent partnership” of the government in any investment, and as such, it is generally neutral with respect to financial and real decisions of corporations. The neutrality result has to be taken with a grain of salt as loss-offset provisions are likely to be imperfect and some erosion of the tax base through lobbying and similar means is probably unavoidable. Expectations of future rate changes may also modify the results. Still, in a closed economy, the CCFT would tend to increase investment and improve the allocation of capital. On the administrative level, a tax based on observable cash flows rather than on a hypothetical concept of accrual of income promises to be simpler and more robust (again, theoretically speaking). It would do away with the problems of defining “true economic depreciation,” measuring capital gains, costing inventories, and accounting for inflation.

2. Possible revenue impacts are an important aspect of the CCFT, especially in developing countries, since the CIT to be replaced is often a major source of revenue. Revenue losses are likely during the transition period, but they need not be prohibitive if the transition is carefully designed and tax rates are appropriately adjusted. A more fundamental issue, however, is that the CCFT as a tax on above-normal returns may have a significantly smaller base than the comprehensive CIT. With expensing, the tax base may also be more volatile. Nevertheless, the actual difference between the CCFT and CIT bases remains an empirical question, depending on the particular income tax laws to be replaced as well as the current compositions of corporate portfolios between debt and equity. In some cases, the two bases may be fairly similar.

3. Tax-base erosion through tax avoidance and evasion may be a serious problem for the CCFT. By choosing an RF-based CCFT over an R-based one and by carefully designing the tax code, some of the CCFT schemes could probably be contained at reasonable administrative costs.130 But the large up-front deduction that results from the expensing of capital assets would create a powerful incentive for base-shifting schemes. The administrative efforts required to contain these schemes could be considerable and would involve the enforcement of arm’s-length prices, which is notoriously difficult. The “simplicity” argument for the CCFT has to be qualified accordingly.

4. Any answer to whether international considerations favor the CCFT would be complex. To the extent that host country tax policies matter, a country with the CCFT may attract additional investment. This is most clearly the case if home countries exempt foreign earnings of their multinationals—as do many Western European countries. Other important capital-exporting countries, such as Japan, the United States, and the United Kingdom, apply the residence principle with a foreign tax credit. This system tends to neutralize host country tax policy, but not entirely. The incentive offered by the CCFT is likely to remain effective for investors who benefit from tax sparing, which is granted by many countries but not by the United States. The CCFT is also likely to attract additional “mature” investment and investors who have excess tax credits. Recently, many U.S. corporations have accumulated such credits, although this position may not be stable in the long run.

To the extent that the effect of the CCFT is washed out by the tax credit mechanism, the host country will lose. Tax forgone on marginal returns is merely picked up by the home country. In the case that the home country denies tax credit for the CCFT, some foreign investment would be discouraged. The creditability of the CCFT is an issue related mainly to the United States, since most other developed countries either exempt foreign earnings or grant some form of tax sparing.

To conclude, at this point, the CCFT remains a theoretically attractive option with some practical disadvantages. Moreover, many unanswered questions remain for its implementation by a single country—especially a developing one—in an environment that will not necessarily accommodate its smooth and effective operation.

Payroll Taxes and the Funding of Social Security Systems

Janet Stotsky

  • What is the appropriate rollfor pay roll taxes in a tax system?

  • What do we mean by funded versuspay-as-yougo social security systems?

  • What are the major design issues in a payroll tax?

  • Who bears the burden of a payroll tax?

Payroll Taxes Defined

Payroll taxes refer to taxes paid by employers and employees on the wages of employees. Payroll taxes are generally applied as a flat percentage of an employee’s gross wages up to some specified limit. These taxes have comprised a rapidly growing major tax source in recent years. In most OECD countries, payroll taxes are the principal means of supporting social insurance programs. Although there is no clear consensus on what social insurance programs government should provide, in most cases, these taxes fund social security, health insurance, unemployment insurance, and disability insurance programs. They may also fund general income support programs, but this is rare. Unlike other major taxes, payroll taxes are generally earmarked and their payment confers a right to benefits. Thus, it is sometimes argued that they are not so much taxes as prepayments for expected benefits to be received in the future.

Earmarking

Payroll taxes are unusual in that their revenues are generally earmarked for specific purposes. The argument against earmarking revenues is that it introduces rigidities into the budget process by requiring revenues to be spent on particular programs regardless of need. In some countries, payroll taxes are not only earmarked to social insurance in general, but to specific social insurance benefits, further limiting government flexibility. The argument in favor of earmarking is that taxpayers may be more willing to pay their taxes when they perceive that they are receiving something tangible in return. In addition, earmarking shields revenues from the vagaries of the legislative process, where funds for important programs could be cut in favor of less important programs for political reasons.

Pay-as-you-go Versus Funded System

Originally, many social security programs were set up with the explicit notion that they would be public pension programs and not income support programs, thereby warranting a close correspondence between tax payments and benefits. Although some social security programs were set up on a funded system, most have been converted to a pay-as-you-go system over time, although selected Asian countries have continued with “provident funds,” which operate pension schemes on a funded system. Under a pay-as-you-go system, the taxes that current workers pay cover expenditures for current beneficiaries. In contrast, under a funded system, the taxes that current workers pay are accumulated in a reserve fund to cover their future benefits. A pay-as-you-go system does not rely on the accumulation of a reserve fund. Private pension plans are required to operate their plans on a funded system, generally with government monitoring to ensure that sufficient reserves are being accumulated in the fund. A pay-as-you-go system can provide an economically superior outcome than a funded system if the return on investments is lower than the growth of the taxable base (in this case, payroll). In recent years, the opposite has occurred, however. Thus, social insurance programs have faced the problem of rapid growth in current and future liabilities as a result of an aging population and expansion of benefits. At the same time, they have confronted a significant reduction in the growth of the tax base as a result of declining productivity and population growth. It is difficult to operate a pay-as-you-go system on an actuarially sound basis under such adverse circumstances. There has thus been some tendency to move toward a funded system for social security, making it more like a private pension plan.131

In the United States, the social security system was substantially overhauled in 1983. The major changes were a significant increase in tax rates matched by some reduction in future benefits, thereby leading to the accumulation of a surplus in the social security trust fund. The accumulation of this surplus is transforming, in part, the system from a pay-as-you-go to a funded one. The generation that is paying for the accumulation of the surplus is effectively paying not only for the retirement of the current generation but also for its own retirement (at least in part). This is the out-come of converting from a pay-as-you-go to a partly funded system. The movement toward a funded system strengthens the case for earmarking social insurance contributions so that this money can be kept separate from general government accounts, where it could be spent on expenditures other than those for which the contributions were made.132

Design of Payroll Taxes

Compared to the personal income tax, payroll taxes are a model of simplicity. Nevertheless, there are still several issues regarding the choice of tax base and rates that arise in their design.

Tax base

With respect to the choice of base, most countries apply the tax to gross wages rather than to the often much narrower labor income base of the personal income tax. Some countries, however, have moved toward a similar base for payroll taxes and the income tax with a more unified system of funding social insurance programs and general government. The appeal of applying the tax to gross wages is that it enhances the vertical and horizontal equity of the tax. The disproportionate benefits that higher income taxpayers get from the use of tax preferences reduces their taxable income and hence, the vertical equity of the personal income tax structure. The uneven benefits that taxpayers get from the use of tax preferences reduces horizontal equity.

Most countries do not exempt any wages before applying the payroll tax, unlike in the case of the personal income tax. Some countries do allow a small exemption, but generally set the threshold lower than for the personal income tax. Usually, countries do, however, set an upper limit on wages subject to the tax. The argument for placing a ceiling on taxable wages is that since payroll taxes fund social insurance programs that are not designed to entirely replace private insurance and pension programs, higher-income individuals could supplement their social insurance programs with private insurance and pensions. Hence, there is no need to provide insurance beyond some reasonable level. The limit on benefits should thus correspond to a limit on taxes. Most countries still maintain ceilings on taxable wages, but have in many cases increased them (and in some cases, abolished them) in recent years, as revenue needs for social insurance programs, especially health, have grown, and governments have introduced and acknowledged a more explicit redistributive component to social insurance programs while extending their scope to the population.

Rates

There are several important issues with respect to the structure of payroll tax rates. A first issue is whether to have separate rates for each social insurance program or a single rate that covers all benefits. In most OECD countries, schedular rates applicable to specific programs are the norm. Although, in some countries, the schedular rates have been replaced by a general supplement to the income tax or a combination of both methods. The benefits of using a schedular approach depend on the degree to which earmarking of the taxes is viewed as desirable, since it makes earmarking easier.

A second issue is what form of tax rate to use. Originally, several countries structured the tax as a flat amount on wages, but virtually all have moved to an ad valorem tax over time. A related issue is whether to levy the tax with a single rate or a graduated rate. Most countries have levied a single rate, in keeping with the original intention that these taxes should fund contributory social insurance programs rather than redistributive income support programs. Only the United Kingdom levies a graduated rate.

Statutory incidence

One of the most important political issues with respect to the design of the payroll tax is how to split the liability for the tax payment, termed the statutory incidence, between the employer and employee. The most typical arrangement is for the employer and employee to share the statutory incidence equally, that is, for there to be a 50–50 split. But, any arrangement is possible, as is demonstrated by cross-country experience.

Burden of the Payroll Tax

While the statutory incidence may be important for political reasons, it is a well-known principle of tax theory that the statutory incidence is irrelevant in determining the economic incidence (see Chapter II). The burden of the tax ultimately depends on a complicated set of behavioral responses to the tax. In the short run, since wages are generally contractually fixed, employers will try to pass on any increase in payroll taxes through higher prices for output. At these higher prices, however, the quantity demanded of the output is likely to fall. The short run effects depend in part on whether the government undertakes fiscal or monetary measures to offset the macroeconomic effects of the tax change. To prevent output from falling, the government could undertake stimulative fiscal or monetary measures. A complete analysis would also account for the incidence of the benefits of social security programs.

In the long run, the burden is likely to be largely shifted back to the labor market.133 If employees have a certain productivity, which their wages reflect, this productivity is not altered by the payroll tax (the tax does not fundamentally change production relationships). Thus, the wage (inclusive of taxes) that employers will ultimately want to pay will reflect this unchanged productivity. Thus, the outcome ultimately largely depends on the elasticity of labor demand and the elasticity of labor supply with respect to the wage (see Chapter II). The economic incidence of a tax generally falls on the side of the market whose demand or supply is most inelastic. Thus, to the extent that labor supply is more inelastic than labor demand, the economic incidence of the tax, regardless of the source of statutory contribution, is likely to be shifted onto employees through a reduction in their net of tax wages. More concretely, this suggests that a 15 percent payroll tax, with statutory contribution shared equally between the employer and employee, ultimately results in a 15 percent decline in net of tax wages rather than a 7.5 percent decline in net of tax wages (and a 7.5 percent decline in the return to other factors). In contrast, if employers were unwilling to alter their demand for labor in response to changes in the tax, the economic incidence of the tax would fall on them, either by lowering the return to capital or other factors, or by lowering profits. If demand for the product were inelastic, the burden of the tax could also fall on consumers through increased prices of outputs.

Empirical evidence on the economic incidence of the payroll tax is inconclusive, given the many difficulties inherent in such an exercise. Nevertheless, the evidence supports the view that in the long run, the economic incidence of the payroll tax largely falls on employees. In the short run, however, employers may not be able to shift the burden of the tax entirely onto employees; thus, they or consumers may also bear some of the burden of the tax.

Assuming that the economic incidence of the tax is on employees, the structure of the tax leads to a regressive tax schedule (meaning that the average tax rate falls with income). When viewed in the context of total income (both labor and nonlabor), the flat rate on wages in combination with no tax on nonlabor income results in a regressive tax. Since higher income taxpayers tend to have more nonlabor income relative to labor income, the payroll tax relative to total income falls as income increases. The ceiling on taxable wages in combination with no exemption under most payroll taxes also leads to a regressive tax schedule because the effective tax rate falls for taxpayers with incomes above the ceiling, since this income is not taxed. For instance, a 10 percent payroll tax on wages up to $50,000, results in an effective tax rate of 10 percent for employees with wages up to the ceiling but for an employee with $100,000 in wages, the effective tax rate falls to 5 percent.

Economic Effects

The payroll tax has several important economic effects. First, it may lead to a decline in the overall progressivity of the tax system by introducing a regressive component that may partially offset other progressive components of the tax system. Although the payroll tax has been criticized for being regressive, its defenders argue that given the generally progressive nature of the benefits of most social insurance programs, the overall package, including both taxes and expenditures, is not necessarily regressive. Further, the income tax which typically includes tax preferences, often tends also to suffer from an effective lack of progressivity except at the lowest income level. Second, the structure of the tax may alter the level and mix of employment. The payroll tax increases the overall tax on labor relative to other inputs, which may induce some substitution away from labor to other inputs. This effect is mitigated to the extent its economic incidence has on labor. If the payroll tax is regressive, its introduction leads to an increase in the cost of lower-wage labor relative to higher-wage labor, which may result in some reduction in the number and net wages of lower-wage employees relative to higher-wage employees. On the other hand, for an existing payroll tax, increasing the ceiling increases the cost of higher-wage employees relative to lower-wage employees, which may result in a decline in the relative employment and net wages of higher-wage employees. Social insurance also has an effect on savings.134

Practices in OECD Countries

OECD countries can be divided into several broad groups with respect to the financing of social insurance programs: six countries have adopted an approach where the financing of social insurance has been integrated with the tax/transfer mechanism, thus moving away from the traditional reliance on payroll taxes to fund these programs. These countries are Australia, Denmark, Iceland, the Netherlands, New Zealand, and Sweden (in this category for employees). Five countries have adopted a hybrid system in which payroll taxes remain important but with one global rate instead of several schedular rates. These countries are Norway, Portugal, Spain, the United Kingdom and, to a lesser extent, Ireland. Thirteen countries follow the traditional approach with social insurance programs financed out of separate schedular payroll taxes with no or low thresholds and ceilings on taxable wages.

Of the countries with the schedular approach, there is a good deal of variation in rates and scope of coverage. In the United States, the social security (old age, survivors, and disability) component of the tax is 6.20 percent, levied on gross wages up to a ceiling with no floor, with the statutory incidence split equally between the employer and employee. The health care component is 1.45 percent and has a different ceiling but otherwise the same structure. The structure in Germany is similar, although the rates are among the highest in the OECD. The social security component of the tax is 9.35 percent and the health component is 6.41 percent. Both taxes have ceilings and no floors, with the statutory incidence split equally between the employer and employee. In both countries, smaller taxes fund the separate disability and unemployment insurance programs, although the structure of these taxes are similar to the structure of the social security and health taxes. Canada, in contrast, at the federal level, has a much less encompassing payroll tax system. It levies a tax on social security of 2.1 percent with a ceiling and statutory incidence split equally between the employer and employee. It levies no tax for health on the employer and a 1.95 percent tax on the employee. It has public unemployment insurance but no public disability insurance.135

Practices in Selected Developing Countries

Virtually all developing countries have some social insurance system, primarily old age, survivors, disability, and work injury insurance. As in industrialized countries, developing countries finance their social insurance systems using some combination of payroll taxes levied on employees, payroll taxes levied on employers, and supplemental government contributions. In developing countries, however, the organized sector may be relatively small, limiting the size of the payroll tax base. In addition, the payroll tax rate tends to be lower than in industrialized countries. As a result, coverage tends to be less comprehensive than in industrialized countries.136

In Latin America, all of the countries have social insurance programs, financed mainly by payroll taxes. In some countries, however, other taxes may supplement payroll tax revenues. In Argentina, for example, a substantial share of revenues from the VAT is earmarked to support the social security system. In Chile and Uruguay, general revenues are used to support the social security system.137

In Asia, most countries likewise have social insurance programs, financed mainly by payroll taxes. Most Asian countries, including China, India, Indonesia, Korea, Malaysia, Pakistan, and Singapore, have some form of old age, disability, and death coverage, financed by payroll taxes. India, Malaysia, Singapore, and Sri Lanka have provident funds that operate on a funded basis to provide pension benefits.138 These countries also cover sickness and maternity, work injury, and unemployment through pay-as-you-go payroll taxes. In some cases, other revenue sources may also be used to support the social security system.

Asset and Wealth Taxes

Business Assets and Receipts Taxes

Russell Krelove and Janet Stotsky

  • What role is there for a business assets tax or receipts tax in a tax system?

  • How can they serve as presumptive or minimum taxes?

  • What structural issues are critical in designing an assets or receipts tax?

Some countries levy a tax on the value of a business’ assets or on its receipts. These taxes are usually relatively simple compared to a business income tax. Typically, the assets tax is levied as a relatively low percentage of business assets while the receipts tax is levied as a relatively low percentage of receipts. These taxes are sometimes used to supplement or replace other business taxes in industrialized and developing countries, and they may also be used as a business minimum tax in some developing countries. This chapter examines the roles and major characteristics of these two taxes.

Rationale and Application

Countries appear to have adopted an assets tax for different reasons. One rationale for an assets tax is to substitute for a business income tax. It is commonly believed that the economic incidence of a business income tax falls on owners of assets. Thus, an assets tax and an income tax may amount to a tax on the same tax base. It is typically assumed that in a well-functioning capital market with capital mobility, capital migrates to where it receives the highest return. Thus, the average rate of return on capital should vary little from one use to another. A tax based on this average rate of return could, therefore, be a reasonable proxy for an income tax. Nevertheless, despite well-functioning capital markets, it is typically found that rates of return vary across businesses engaged in the same activities and across industries. Consequently, a tax based on an average rate of return on business assets is likely to be an imperfect proxy for an income tax at best.

Another rationale for a business assets tax is to add an mixed of progressivity to the business income tax. For the assets tax to add to the progressivity of the income tax, it is important to design the tax so that the assets tax liability supplements the business income tax so that businesses with higher income pay a higher overall tax. This is likely to be difficult in practice; thus, it is probably easier to add a progressive rate schedule to the regular business income tax.

Another rationale for a business assets tax is to serve as a business minimum income tax.139 An assets-based business minimum tax has a stronger theoretical grounding than one based on some alternative measures of the base, such as gross receipts or turnover, inasmuch as one expects economic income to bear some systematic relationship to assets.

A final reason for adopting an assets tax in the case of state businesses is to enable the state to gain some return on its capital investments. This tax might be seen to substitute for the payment of a dividend to equity owners in private businesses. This rationale for an assets tax would, however, cease to apply if these businesses were privatized.

Many Latin American countries levy assets taxes, using a variety of definitions of the tax base. Mexico, Ecuador, and Argentina levy an annual tax on gross assets.140 Costa Rica taxes only gross fixed assets. Several other countries, including Bolivia, Colombia, Panama, Peru, and Uruguay use net worth as the tax base.141 Many European countries also have some form of a business assets tax.

Countries also appear to have adopted a gross receipts tax for different reasons. In developing countries, it may be difficult to objectively verify components of revenues and costs. When components of cost are difficult to measure, a tax on cash receipts alone may provide a proxy for a business income tax, although the theoretical grounds for this equivalence are much weaker than for an assets tax. Some developing countries have adopted gross receipts or turnover taxes as minimum taxes. Gross receipts or turnover taxes are prominent in France and the francophone African countries, often with a minimum amount of tax payable.

Determination of the Tax Base

What should constitute the base?

One critical issue in the design of an assets tax is the choice of a tax base. The tax is generally imposed on a taxpayer’s gross business assets, including both current and long-term assets. Current assets may include cash and securities, receivables, and inventories, while long-term assets may include land and other fixed assets at depreciated value, and intangible assets at amortized value. It might be desirable in defining gross business assets to use net working capital (current assets minus current liabilities) instead of current assets, because the amount of current assets might simply reflect some short-term financing decisions. Alternatively, the tax could be imposed only on net assets (gross assets net of debt-financed liabilities) or only on fixed assets.

The choice of which asset measure to use for the tax base depends on a combination of theoretical and practical considerations. From a theoretical perspective, levying the tax on gross assets is appropriate, if the purpose of the tax is to provide a proxy for a broad-based income tax that does not favor debt-financed assets. Levying the tax on net assets (or equity) allows debt-financed assets to escape tax, similar to a business income tax that allows an income tax deduction for interest on debt but not for dividends. A tax on net assets thus favors debt-financed investments. Levying the tax on net assets might also induce taxpayers to align debts with assets included in the tax base to reduce the tax. For this reason, gross assets, rather than net assets, may constitute a more appropriate tax base. Levying the tax on fixed assets is simpler but has less theoretical justification. A tax on fixed assets would tend to discourage investment in fixed assets as opposed to other asset forms.

A gross receipts tax includes in the base the turnover from operations of the business, so that it is equivalent to a broad-based sales tax on business output. The base may, however, be defined in broader terms to include other items of positive cash flow, such as receipts from asset sales or equity sales, and receipts from issuance of new debt. Sales of new equity and issuance of new debt are related to the acquisition of capital, so that a tax on receipts from issuance of equity and debt is closely related to a tax on gross assets. A major difference is the timing of tax payments—a tax on new financing would be paid up front, rather than over the life of the asset.

Valuation issues

A tax on assets may be difficult to administer because of the inability to measure accurately some components of assets. Assets may pose valuation difficulties for several reasons.142 First, the value of a business’ assets may vary. Second, many assets have long life spans and are not regularly traded in capital markets. Third, inflation distorts the value of assets by creating a divergence between historic and replacement costs. Fourth, some small assets may be sufficiently costly to value that, from a practical point of view, it is not worthwhile to attempt to value them. Fifth, in some cases, it may be difficult to ascertain the owner-ship of assets. A similar set of valuation problems arise with regard to liabilities.

Current assets pose different valuation problems than long-term assets. Current assets vary more than long-term assets within a short time period. Thus, it might only be appropriate to take an average value for the purposes of the tax. Nevertheless, since current assets are relatively liquid, their value is relatively easy to establish, even in an inflationary environment. Certain components of current assets, such as inventories, are relatively illiquid and hence would pose more of a valuation problem.

Long-term assets pose valuation problems because they are less liquid and infrequently traded. Ideally, fixed assets, such as plant and equipment, should be valued at fair market value, but usually the only available measure of value is historic costs reduced by depreciation. Land, although it is not a depreciable asset, poses a similar problem. Ideally, intangible assets, such as goodwill or patents, should be valued at an arm’s-length price that would be established if the business were sold, but usually the only available measure is historic value reduced by amortization. In an inflationary environment, the book value of gross assets may be a poor proxy for the economic value of gross assets. Thus, a tax on assets without any adjustment for inflation could be quite distorting, tending to favor businesses whose assets are somewhat older on average.

Similarly, if the goal is to tax net assets (or equity), it is necessary to have a measure of the fair market value of equity. Unless the stock of the business is publicly traded, this measure will not be available. The book value of net assets may be a poor proxy for its market value.

In principle, it is possible to make some adjustments to balance sheet items to account for inflation. It is possible to adjust certain historical asset values for inflation by indexing those assets to price indexes that are appropriate for the asset. The consumer price index can be used to adjust asset values that are likely to reflect overall price levels, such as machinery, equipment, and structures, if no asset-specific indexes are available. It is not, however, appropriate to use the consumer price index for land, since its value often changes in response to demand and supply characteristics of a local market that bear little relationship to overall price changes. An index of land or real estate values is more appropriate to use in this instance. It is much more difficult to adjust the value of intangibles in any accurate fashion.

A receipts tax avoids many of the valuation problems that arise under an assets tax, especially the problems arising from the mismeasurement of cost in an inflationary environment. Receipts respond on average proportionately to prices, so that tax revenue grows with prices.

Double taxation

Double taxation of certain assets is potentially a problem, if taxable businesses own financial interests in one another. Businesses may either own equity in or make loans to other taxable businesses. If the purpose of the assets tax is to serve as a proxy for an income tax on real and financial flows, then the assets should be taxed on any flows that they generate, so that it is appropriate to include both the financial assets and the tangible assets that they finance in the tax base. On the other hand, if the purpose of the tax is to serve as a tax on real flows, then the assets should only be taxed once. There are two ways to reduce the possibilities of double taxation: either by allowing the business that owns the equity or makes the loan to other businesses to deduct those assets from its taxable base or by allowing the business that has issued equity to or borrowed from other businesses to deduct those liabilities from its taxable base. If the purpose is to tax real flows, it seems more logical to allow the business that owns equity in or makes loans to other taxable businesses to deduct those assets from its taxable base. Under the Mexican assets tax, however, the business that borrows from another taxable business is allowed a deduction. Thus, either arrangement is possible.

A variation of this problem arises if businesses own subsidiaries or other businesses that are taxable. In this case, the base should include all assets of the business, but not assets of subsidiaries or other taxable businesses. The Mexican assets tax has provisions to avoid double taxation of intercorporate ownership.

Timing of valuation

Another issue in the design of the tax is when to value the assets for the purposes of determining the tax liability. A business’s holdings of assets are likely to fluctuate over the course of a year. One possibility is to require a business to calculate average holdings of each type of asset over the year. A more precise method is to require a business to calculate a weighted average of its holdings of each type of asset, with the weight being the proportion of the year it held the asset. Another possibility is to choose one day of the year, such as the last, for determining the value of each type of asset held.

Under the Mexican assets tax, the tax base is defined as the average value of the taxpayer’s assets over the year. In general, the value of the assets is equal to the sum of monthly averages of financial assets divided by the number of months in the period plus the average of investments in land, other fixed assets, and intangible assets plus the average of inventories minus the average of liabilities after specified reductions.143

Liquidity considerations

The presumption under the income tax is that profit generates the liquidity to finance the payment of tax. Liquidity may be a greater problem under an assets tax. A business may face difficulties paying its assets tax liability if the value of assets is changing quickly and there are constraints on the taxpayer’s ability to borrow against the value of the assets. This situation is most likely to arise when the assets have low current yields. Instead of forcing businesses to sell assets to meet tax liabilities, one possibility is to allow the tax to accrue until the asset is sold or until, for other reasons, the liquidity constraints have been relaxed. It is possible to adopt an averaging provision to limit the liquidity problems, as was done with the Mexican gross assets tax.144 Finally, in certain circumstances, it may be expedient to exclude from the definition of the assets tax base those assets that are most likely to lead to liquidity problems for taxpayers. Liquidity problems are less likely to arise under a receipts tax.

Transfer pricing

Multijurisdiction companies benefit from the possibility of using transfer pricing to minimize taxes. The possibilities for such transfer pricing depend on the form of tax. Under a gross assets tax, the incentive to choose output and input prices strategically is diminished, especially when that base consists mainly of tangible assets.145146 Under a gross receipts tax, the incentives on the cost side are removed, but there is still an incentive, in internal transactions, to understate the selling price when the seller is located in a jurisdiction levying such a tax.

Determining a threshold for liability for taxation

Usually, corporate income tax is payable on taxable income, independent of the size of the firm, although small firms sometimes face a lower tax rate. Under an assets tax, businesses with a relatively low value of assets may be exempted from the tax. This feature would serve to introduce an mixed of progressivity into the tax and reduce the number of taxpayers subject to this tax. With sales taxes, for example a VAT, small traders are generally excluded for administrative reasons, usually by a turnover floor. A similar rationale can be adopted to support a turnover floor for a receipts tax, although the appropriate floor would be lower than the corresponding floor for the VAT, since turnover is larger than value added.

Determination of the Tax Rate

The tax rate that is chosen should reflect the objectives of the tax. If revenue is an important objective, the tax rate should be set high enough to generate the desired amount of revenue. If the main objective is to ensure a minimum contribution to the exchequer by businesses, then the tax rate should be set high enough to yield revenue from businesses with economic income but not so high as to impose undue burden on businesses with economic losses. This would suggest a relatively low rate.

The appropriate tax rate on an assets base depends in part on which assets are included in the base, the expected real rate of return on business assets, the regular business income tax rate, and the debt-equity ratio of the business. If the tax applies to gross assets and is intended to be a proxy for a broad-based income tax or minimum tax on income, then with an expected real rate of return of ρ and a business income tax rate of τ, the tax rate should be t = τρ. Sadka and Tanzi (1993) suggest that for setting the tax rate on a gross assets base, it might be desirable to treat the business as though it were entirely financed by debt so as to avoid biasing investment against equity financing. With an expected real rate of return of ρ and a real rate of interest of r, the presumptive profit is ρ-r per unit of gross assets (after allowing for imputed interest on equity) and the tax rate on gross assets should be t = τ(ρ-r). Similarly, if the tax is intended to be a minimum tax on equity, then the rate would depend on the debt-equity ratio of the business. For example, if the debt-equity ratio is 1, then the tax on equity might be twice the equivalent rate on gross assets, after allowing for a deduction for debt in the measurement of the tax base.

With a gross receipts tax, the tax rate can be chosen to mimic on average the impact of the income tax, by using for all firms the average ratio of sales to income. If the ratio of taxable income to receipts is b, and if the corporate tax rate is T, then the tax rate on receipts should be t =bT. But this would impose a substantial burden on loss-making companies.

Risk-Sharing Characteristics of the Various Taxes

Risk-sharing characteristics of the three taxes differ. In Chapter II, the distinction was made between income risk and capital risk, the former applying to un-certainty in output and current input prices, and the latter arising from uncertainty concerning physical depreciation of capital and also in replacement costs. As previously discussed, an income tax that allows only partial or no offset for losses tends to penalize the bearing of income risk relative to a tax that allows full loss offset. In addition, when depreciation for tax purposes is based on historical cost, the government does not share in capital risk under an income tax. Asset taxes based on market value do not share in income risk, but they can, if appropriately designed, be neutral with respect to capital risk. When asset prices rise, asset tax collections based on market value increase, and conversely, when asset prices fall, so do asset tax collections. For asset valuation based on some rule, the degree of sharing of capital risk depends on how the rule responds to changes in physical depreciation of the asset and to variations in the replacement cost of capital. For a receipts tax in the form of a tax on turnover of the firm, tax revenue fluctuates with output prices, but does not respond to variation in input prices or in the gains and losses on the capital of the firm. Thus a turnover tax does not share in all income risk, or in capital risk.

Taxation of Land and Property

Janet Stotsky and M. ZÜhtÜ YÜcelik

  • What is the rationale for property taxation? Under what forms can it be applied?

  • What techniques should be used for valuation of taxable property? Why do the effective rates differ from statutory rates?

  • What are the arguments for betterment levies?

  • What is the incidence of property tax?

Taxes on land and property are among the oldest and most common forms of taxation. Although property taxes typically constitute a minor source of revenue at the central government level, they may contribute substantially to the financing of local public services. In 1992, property taxes ranged from less than 1 percent to more than 7 percent of total revenue in industrialized countries. Their share in total revenue also varied across developing countries in a similar fashion.147

Rationale for Property Taxation

The taxation of land and property may be justified on the grounds of both the benefit and ability to pay principles of taxation. Benefit considerations point to various kinds of in rem property taxes. One may argue that the protection provided by the state for private property through the maintenance of general law and order justifies the imposition of a tax; or more narrowly, one may argue that the construction of a road adjoining the property confers a benefit for which a tax might be charged. Generally, however, the application of the benefit principle of taxation is justified when it can be shown that the value of the benefits of publicly provided goods equals the tax yield. This proposition is hard to establish, but would be most likely to hold true at the local government level.

Ability-to-pay considerations suggest that the holding of property not only implies an ad personam tax capacity to receive property income but also implies an inherent form of potential consumption. A property tax may be a useful part of the tax system to the extent that either rental income or imputed rental income from property are not captured by the personal income tax.

Forms of Property Taxation

There are three basic forms of property taxation: (1) a tax based on the annual or rental value of the property; (2) a tax based on the capital value of the land and improvements; and (3) a tax based on the site (or land) value (which is essentially a type of capital value tax). Some tax systems may use a combination of methods.

Under the annual value system, the tax is based on an estimate of the annual net rental value from the use of the property. Net rental value is usually derived from income flows, with some adjustments, or from capital values.

Under the capital value system, the tax is based on the assessed value of land and improvements. Practice varies with respect to the assessment of capital value. In some countries, land is assessed separately from improvements while in other countries, land and improvements are assessed together.

In theory, the discounted stream of net rental payments should be equal to the capital value of a property. Thus, the annual value and capital value methods should yield equivalent tax bases. In practice, the two tax bases may not yield equivalent values for several reasons. First, since the capital value is based on expected future flows of income from the property, differences in expectations may result in differences between annual value and capital value. Second, both methods use assessment practices that are frequently ad hoc and inconsistent. Thus, for the property, net rental value may be different from market rental value and capital value may be different from market value.

The site or land value system includes only the site value in the property tax base, excluding improvements, such as houses, factories, and crops. It is applied in a number of countries, including Kenya, Taiwan, Australia, New Zealand, and South Africa. “Site value” is different from “unimproved land value,” in that it includes the value of drainage, leveling, timber clearing, and similar site improvements. The main disadvantage of this form of property taxation is that it narrows the tax base compared to the other forms of property taxation and thus requires higher tax rates to produce the same revenue. It may also be difficult in the case of land upon which structures have been built; the value of the land has to be separated from the value of the structure. But the main advantage is that in rural areas, it may be simpler to administer than a property tax.

Taxation of Agricultural Land

Taxation of agricultural land represents the oldest form of property taxation, although the relative importance of this tax has declined over time. One variant of a property tax applied to agricultural land is a tax based on land area. A flat rate is applied to each unit of land area irrespective of its annual rental value or capital value. Some land taxes may apply differential rates to land distinguished by quality or the availability of irrigation, as in Nepal.148 The advantage of this approach is its simplicity. The main disadvantages are that land area may bear little relation to land value and in a highly inflationary environment, this tax may quickly lose its value unless the nominal amount of tax applied to a unit of land is regularly adjusted upward.

Valuation Problems

Property is a heterogeneous good. Its value reflects economic, social, physical, and legal factors. Economic factors include the level of income in the community, taxes and other prices, the proximity to centers of business activity, and the availability and quality of public services. Social factors relate to the availability of cultural and recreational facilities. Physical factors relate to the characteristics of the land, such as soil quality and size, characteristics of any improvements, and characteristics of the environment. Legal factors relate to public and private restrictions on the use of the property, including rent control and zoning regulations. These characteristics of the property may be partly or completely capitalized into the value of the property.

Accurate assessment is the benchmark of a good property tax administration. The purpose of assessment is essentially to determine the “fair market rental” or “fair market value” of the property. Accurate assessment under any system of property taxation requires an active property market. The annual value method requires an active rental market so as to provide information on rental values, while the capital market method requires an active market in which property is bought and sold so as to provide information on capital values. Urban areas typically provide active markets in property, while rural areas may not, thus making assessment more difficult in rural areas.

Owner-occupied property that is neither rented nor sold regularly may pose a valuation problem for both the annual and capital value methods of property taxation since no explicit rent or capital value is observed but must be imputed from information derived from property with similar characteristics. Business property may also pose valuation problems since rental rates are frequently negotiated.

In practice, the assessed value of a property is generally set below its fair rental or market value because of exclusions and exemptions from the base or because methods of assessment are used that are not based on full property valuation. In addition, infrequent assessments and poor assessment practices contribute to differences between assessed value and fair market value. Arbitrary differences between assessed values and fair market values may create inequities in the property tax and contribute to the unpopularity of the tax. In environments characterized by high property price or general price inflation, if assessments are not conducted frequently, this may lead to erosion of the tax base and make the property tax an inelastic source of revenues.

As an alternative to regular assessment, it is possible to use a price index that is appropriate for the asset as a guide to changes in value. A construction price index may serve to measure changes in the value of the structural component, while a land value index may be appropriate to use for the land component.

The existence of rent control may complicate the assessment of property value because it may reduce the level or growth of rents or assessed value. A reduction in property value from rent control may lead to a considerable loss of tax revenue.149

Under the annual rental value system, there are two main methods for measuring the tax base. The first method assesses the income-producing capacity of each class of land or property, based on standardized classifications and other information; and the second method assesses the capital value either based on sales of the property or comparable property, or based on alternative official appraisal standards, and derives the annual value by applying an assumed rate of return on the capital value.

Under the capital value system, there are three main methods for assessing property value. The most common method bases assessments on sales of the property or comparable property. This method requires an active market to provide sufficient observations on properties with different characteristics to yield a value for each characteristic. It is then possible to derive a value for properties with different combinations of characteristics by aggregating the value of each of the separate characteristics. With sufficient observations, it is possible to apply formal statistical techniques, known as hedonic analysis, to yield highly accurate assessments. A second method uses observations on annual rental streams and converts this rental stream into a capital value using an appropriate rate of discount. A third method develops an estimate of how much it would cost at current input prices to replace a piece of property in its existing condition. This method may not, however, be used for land valuation, since land is irreproducible.

For the assessment of agricultural land, the most accurate method is the capital value approach. The annual rental value method is limited by the frequent absence of rental information and the replacement cost approach is inapplicable. Taxation of agricultural lands based on capital value, in contrast to annual rental value, is less likely to be considered as a tax in lieu of income taxes in the agricultural sector. Thus, it facilitates the expansion of the income tax to the agricultural sector.150

Rate Structure

Property tax systems vary in terms of their rate structure. They may apply either a flat or progressive tax rate schedule to assessed value. A flat rate offers the advantage of simplicity and minimizes the opportunity for taxpayers to bargain with tax authorities to reduce their tax liabilities. Effective tax rates are often well below the statutory tax rates because property is typically assessed below its fair market value. The ratio of assessment to fair market value tends to vary widely from place to place and over time, thereby making it difficult to compare effective tax rates on property on the basis of the statutory rates alone. Typically, however, statutory tax rates are in the order of 1–3 percent, while effective tax rates are even lower.

Exemptions

Most property tax systems exempt properties belonging to the government, local authorities, charities, religious institutions, and foreign embassies. In some countries, owner-occupied properties are exempt from the tax under the rental value system or are subject to reduced rates. With respect to agricultural land taxation, improvements are generally exempt from the tax and preferential treatments are granted to promote the use of fertilizer, cultivation of specific crops, and land reclamation ventures.151

Betterment Levies

Betterment levies (or special assessments) attempt to apportion the cost of public investments to properties benefiting from these projects. They are generally levied for a specific purpose and their application is limited to those property owners considered to be direct beneficiaries of public investments.152 Such projects include irrigation systems, new roads, or urban renewal projects.

One variation on the betterment levy is the valorization tax, which was extensively applied in Colombia. Municipalities used this tax to recoup the cost of the municipal projects, such as the construction of new streets and sewers, widening and paving of existing roads, lighting of streets, three plantings, and so on. Rural areas used this tax to recoup the cost of highway, flood control, and land reclamation projects under-taken by departments and regional agencies.153 Municipalities and other local authorities were allowed to assess and collect the valorization tax upon approval of the projects involved. Property owners were given the right to be consulted in the conception and realization of the project and in the allocation of the proceeds of the tax. This levy contributed significantly to local government finances in Colombia and made possible several extensive infrastructure projects. For the period 1959–63, valorization tax proceeds in Colombia represented, on average, 38.6 percent of property taxes, 16.1 percent of municipal tax revenue, and 6.2 percent of municipal current revenue.154

Role of Cadastre in Property Taxation

An effective implementation of the property tax requires information on all property, including its physical size and boundaries, ownership, and the value of land and improvements. These requirements can be satisfied through a reliable set of titles, ownership records, and cadastral surveys. The “cadastre” consists of an official record of the location, size, and owner-ship of each parcel of land. A tax cadastre also includes the information needed for property taxation, such as the value of the land and improvements. The cadastre provides the tax officials with information on taxable properties and addresses of liable taxpayers. Therefore, a reliable cadastre is essential for an effective administration of a property tax.155 A reliable cadastre is a task which may take a number of years, but once prepared, and thereafter regularly updated, it will recoup its cost many times over in improved assessments and higher tax revenues. In the absence of a reliable cadastre, it is still possible to administer a property tax by using block valuation and applying a common property value to all property within a block. This method works best when property is relatively uniform within a block. It is still possible to allow variation under this method by requiring property owners to justify any deviation of their property value from the value assigned to property in the block.

Incidence of the Property Tax

The old view of the incidence of the property tax was that the component of the tax that falls on land is largely borne by owners of land since land is perfectly inelastic in supply. Since landowners tend to have higher income, this component of the tax is progressive. The component of the tax that falls on structures is borne by both the supply and demand sides of the market because structures are more elastic in supply. Since renters tend to have lower income, this component of the tax is less progressive in incidence or even regressive. Thus, under the old view, the property tax tends to be at best proportional and possibly regressive. The new view of property tax incidence is that the tax falls on all owners of capital through movements of capital from high-taxed sectors to low-taxed sectors. Since owners of capital tend to have higher income, the new view maintains that the property tax is progressive. Empirical evidence on property tax incidence is inconclusive.156

Taxation of Bequests, Inheritances, and Gifts

M. ZÜhtÜ YÜcelik

  • What are the policy objectives of transfer taxation and under what forms can they be achieved?

  • How should the rates, taxable bases, and exemptions be established to attain policy objectives?

  • What is the revenue importance of transfer taxes in selected industrial countries?

Taxes on property transferred at death have always been considered an appropriate object of taxation and are among the oldest forms of taxation. They appeared first during Roman times in 6 A.D. and were called “vicessima hereditatum.”157 These levies are imposed as a matter of social philosophy and as a policy instrument to adjust the distribution of wealth and to generate revenue for the Treasury. These taxes may be imposed as an estate tax or as a gift tax on the donor, or as an inheritance tax on the heirs. The United States, the United Kingdom, and other Commonwealth countries apply estate taxes, whereas other European countries apply inheritance taxes.158

Objectives of Transfer Taxation

A country may have various social and economic objectives in imposing death taxes: first, to limit one’s right to dispose of one’s wealth at death; second, to limit one’s right to acquire wealth by way of bequests, without “own effort,” which is considered a windfall; third, to establish an alternative to the taxation of capital income during the recipient’s lifetime; in this way, it may be possible to avoid a highly progressive taxation of capital income and to reduce its disincentive effects on saving and Investment; fourth, to redistribute wealth; and fifth, to generate revenue for the Treasury. Taxes on inter vivos transfers (gifts) may serve to prevent the transfer of wealth prior to death so as to avoid death taxes.

The relative importance of the objectives to a given country shapes the form of taxation of transfers, its rate structure, and level of exemptions.159

Forms of Transfer Taxes

Transfers at death may be taxed in three forms: (1) as an estate tax imposed on the entire estate left by the testator without any reference to inheritors; (2) as an inheritance tax imposed on individual shares of inheritors; and (3) as an accessions tax, which is a unified tax on the donee with progressive rates on the total value of bequests and gifts received during the lifetime. Ireland is the only country to impose a true accessions tax. The Irish system requires full integration with lifetime aggregation of gifts and inheritances from any donor (before 1984, only aggregation by each donor was required). France also requires full integration with lifetime aggregation of gifts and legacies only with reference to transfers between a recipient and each donor separately.160

The estate tax is fairly simple to administer as it is levied on the entire estate of the testator. It emphasizes the objective of reducing the concentration of wealth. Large exemptions and undervaluations observed in practice create a substantial tax avoidance. “Generation skipping” is another avenue of avoidance, by which the testator leaves his or her estate to grandchildren with the objective of reducing the number of times an estate changes hands, and hence the number of times the estate tax can be levied. The creation of trusts and inter vivos gifts are other ways of reducing the estate tax. A gift tax thus complements an estate tax.

Inheritance taxes are more difficult to administer as they require the valuation of individual shares. But they may be better adapted to the ability to pay of inheritors.161 Accessions taxes are the most difficult to administer as they potentially require the valuation of a series of gifts or inheritances.

Rates and Exemptions

Estate and gift tax rates are generally progressive. In the United States, estate and gift tax rates were unified in 1976 at levels from 18 percent to 70 percent. New Zealand has a single 40 percent rate. In the United Kingdom, the rates vary between 30 percent and 60 percent. The estate tax threshold is $600,000 in the United States and the equivalent of $235,000 in New Zealand and $104,000 in the United Kingdom. The estate tax treats all inheritors the same with respect to the application of the tax, although certain exemptions may apply. In New Zealand, the matrimonial home is entirely exempt for social policy considerations. In the United States, an unlimited exemption is granted to spouses. In Germany, an exemption on the first DM 40,000 is granted for spouses, children, and grandchildren; and DM 10,000 in other cases.162

Inheritance tax rates are also generally progressive and differentiated according to the beneficiary’s relationship to the deceased. Exemption thresholds may also be differentiated for the same purpose. Most commonly, three or four different rate scales exist. Japan and Ireland have a single scale with different thresholds for beneficiaries.

Spouses and children are usually the most favored class, while nonrelated beneficiaries are more heavily taxed. Their initial rates are five to eight times greater than the rate applied to spouses and children. The argument here is that for an unrelated beneficiary, an inheritance is generally a windfall, whereas for close relatives, particularly the spouse and children of the deceased, an inheritance is expected. In many countries, close relatives have a legal right to inherit some portion of the estate regardless of their wealth.163 This treatment is justified on grounds of social policy, which aims at supporting the family unit as an institution rather than by reference to the principle of horizontal equity. It is possible to derive certain equivalences between estate, inheritance, and gift taxes. For instance, if tax rates are flat, taxation of the estate or of individual shares of heirs yields equivalent tax burdens. If the rates are progressive, then these two kinds of taxes may not yield equivalent tax burdens. With progressive rates, an estate would yield a higher revenue than an inheritance tax as long as the estate was not distributed to one heir. But even with progressive taxes, the tax yield would be the same if the estate were distributed to one heir. Even equivalent tax rates may not yield equivalent tax burdens if the taxes are levied at different times, resulting in different tax burdens in present value terms.

Tax Relief for the State, Charities, and Political Parties

Property given or bequeathed to the state and to other government institutions is generally exempt from death taxes. Transfers to charities which are of public benefit are generally granted exemption or substantial relief. In some countries, relief is limited to domestic charities; in others, foreign charities also qualify unilaterally or on a basis of reciprocity. In Austria, Germany, the Netherlands, and the United Kingdom, gifts and bequests to political parties are given relief, while in Ireland, they enjoy full exemption.

Treatment of Residents and Nonresidents

Death taxes are levied on all the property of deceased persons who are resident or domiciled in the country imposing the tax. Resident beneficiaries are taxable on all property situated within or outside their country of residence that they inherit from the deceased persons resident or domiciled in the taxing country. Nonresident beneficiaries are taxable only on property situated in the taxing country.

But foreign property which forms part of the estate of a nonresident is normally not taxable even when it is received by a resident beneficiary. This rule does not apply in Germany and Japan because residents are taxable on foreign property inherited from nonresidents.

Not all countries charge all the property of a deceased resident. In Finland and Luxembourg, immovable property situated abroad is not taxable. With respect to nonresidents, the tax does not extend always to all his property situated in the taxing country. In Denmark, Belgium, and Luxembourg, it is limited to immovable property situated within their borders.164

Treatment of Selected Assets

Some exemption for household and personal effects is common. In New Zealand, all household and personal effects, jewelry, and works of art are exempt from estate tax if passed to a spouse; the same exemption applies to gifts between spouses for gift tax purposes.

The argument used for such exemptions and reliefs are the need to avoid some troublesome valuation work and hardship which may be caused by taxing nonincome-producing assets, which the survivors need for their personal use and do not intend to sell.

Works of art and collections are given exemption or favorable treatment in the public interest. Germany extends such treatment to works of art or immovable property if their preservation is in the public interest. In Ireland, these assets are granted tax relief if they remain within the country and are made accessible to the public.

Pension rights and annuities are exempt in many countries. In France, a pension commencing on death is taxable but annuity passing by survivorship is exempt.

Productive assets are granted special reliefs. Agricultural land and forestry enjoy reliefs for death tax purposes such as 50 percent reduction in value for farmland in Ireland and the United Kingdom and 75 percent reduction in France. Reliefs for transfer of agricultural assets are intended for economic and social reasons, such as preventing fragmentation of the farmland which may lead to inefficiency, as well as preserving agricultural families in their farmlands.

Specific reliefs are also granted to industrial and commercial enterprises owned as a family business. For example, in the United Kingdom, a 50 percent reduction is allowed on the value of a sole proprietor’s business or a partner’s interest in the partnership. Similar concessions are granted in Finland and the United States.165

Valuation

For transfer tax purposes, property is valued at its market value at the date of transfer. For an estate tax, the date of death is the natural rule. In case of inheritance taxes, actual date of acquisition of the inheritance is logically more appropriate, but, in practice, valuation at the date of death is the rule in many countries, except in Denmark and Ireland, where the relevant date is the actual date of transfer. Cadastral values of immovable property are used in Portugal and Spain. In other countries, immovables may be valued by the tax department.166

Life interests, annuities, and other income interests are usually valued at a multiple of the annual income according to the age of the beneficiary, based on a statutory table prescribed for the purpose.

As for the valuation of unincorporated businesses and shares, the principal factors which may be taken into account are the going-concern value of the business, the assets value, the liquidation value, the earnings yield, and the dividend yield. A liquidation basis of valuation is appropriate for shares in a bankrupt company. With respect to a profitable trading concern, the company’s estimated maintainable profits are capitalized at a rate commensurate with the risk to capital involved.

In some cases, taxpayers declare assets at their own valuation, which are then subject to review by the tax authorities. Wide divergences occur naturally between the values put upon an asset by the taxpayer and the tax authorities. The taxpayer may be tempted to declare values which are unrealistic by any standard to reduce the tax liability. The size of such evasion depends on the resources which the tax authorities are able to devote to review the taxpayers’ valuations.

In extreme cases, some assets are altogether omitted. Personal effects (including jewelry) and securities in bearer form are the most frequently and easily concealed types of assets. Practice of such tax evasions depends also on the general attitude of taxpayers toward tax evasion.167

Effects on Saving and Investment

An estate or inheritance tax may affect a donor’s saving behavior in two ways. On the one hand, it may induce donors to save more to compensate for the income effect of the tax on after-tax bequests. Alternatively, it may lead the donor to substitute lifetime consumption for bequests because the price of leaving a legacy has increased. By reducing the amount of inheritance for a potential heir, an estate or inheritance tax may stimulate the desire of inheritors to work and save both before and after an intergenerational transfer. A tax at death may also encourage entrepreneurs and investors to place much of their wealth in trusts or donate to charities. Furthermore, by increasing the demand for liquidity, death taxes might bias portfolio compositions toward more conservative investments.

Since most capital transfers occur at death, a tax at death may create fewer distortions in the type of investment decisions of a potential donor over his lifetime than a comprehensive income tax.168

Revenue Importance of Death Taxes

Death and gift taxes represent a minute percentage of GDP in industrial countries. Several countries recently reduced property transfer taxes (e.g., Canada, the United Kingdom, and the United States), on the grounds that they fell too heavily on agriculture, private business, or both. Tax revenue growth in France is explained by the substantial increases in the rates of inheritance tax by the socialist government in 1982. The increase in Japan is explained by the rapid growth in national wealth associated with the high savings ratio of its citizens.169

In the United States, the estate tax was first introduced in 1916 with rates ranging between 1 percent and 10 percent. The top marginal rate was gradually increased, reaching 77 percent in 1940. A separate gift tax was introduced in 1924 with rates between 1 percent and 25 percent. Gift tax rates also increased gradually but they always remained below those of the estate tax.170

In 1976, the following several major changes took place: (1) unifying the gift and estate taxes (with rates from 18 percent to 70 percent);171 (2) introducing a comprehensive levy on generation-skipping transfers; and (3) expanding the marital deduction.

Even after allowing for large tax reliefs and differentiated rates for social policy considerations, the revenue yield should have been higher than observed today in the United States. Of the $123 billion of wealth slated to be transferred at death in 1986, only $36 billion was included on estate tax returns and $6 billion paid in taxes, resulting in an effective tax rate of 5 percent.172

Wealthy people avoid the transfer tax through an array of estate planning techniques while maintaining full control of their assets. Placing a low value on wealth already accumulated may be a primary method of tax avoidance. To some extent, tax avoidance affects real economic behavior, such as the nature of investments or property use. In most cases, however, tax avoidance consists of hiring legal experts to arrange property rights in appropriate ways. In 1992, some 16,000 lawyers in the United States were specialized in trust, probate, and estate law. Accountants also engage in estate planning market. Moreover, taxpayers spend considerable efforts on avoidance schemes.

References

  • Aaron, Henry J., Who Pays the Property Tax? A New View (Washington: Brookings Institution, 1975).

  • Aaron, Henry J., and Alicia H. Munnell, “Reassessing the Role of Wealth Transfer Taxes,” National TaxJournal (Columbus, Ohio) Vol. 41 (June 1992).

    • Search Google Scholar
    • Export Citation
  • Alworth, Julian, The Finance, Investment and Taxation Decisions of Multinationals (Oxford: Basil Blackwell, 1988).

  • Atkinson, Anthony B., “On the Measurement of Inequality and Poverty,” Journal of Economic Theory, Vol. 2 (September 1970), pp. 24463.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Atkinson, Anthony B., Social Justice and Public Policy (Cambridge, Massachusetts: MIT Press, 1983).

  • Atkinson, Anthony B., “Income Maintenance and Social Insurance,” in Handbook of Public Economics, Vol II, ed. by Auerbach Alan J. and Martin S. Feldstein (Amsterdam: North-Holland, 1987), pp. 779908.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Auerbach,Alan J., Martin S. Feldstein eds., Handbook of Public Economics, Vol. I (Amsterdam: North-Holland, 1985).

  • Auerbach, Alan J., and James M. Poterba, “Tax-loss Carryforwards and Corporate Tax Incentives,” in The Effects of Taxation on Capital Accumulation, ed. by Martin S. Feldstein (Cambridge, Massachusetts: National Bureau of Economic Research, 1987), pp. 30542.

    • Search Google Scholar
    • Export Citation
  • Ault, Hugh J., and David F. Bradford “Taxing International Income: An Analysis of the U.S. System and Its Economic Premises,” in Taxation in the Global Economy, ed. by As-saf Razin Joel Slemrod (Chicago: University of Chicago Press, 1990), pp. 1146.

    • Search Google Scholar
    • Export Citation
  • Bagchi, Amaresh, “Tax Reform in Developing Countries: Agenda for the 1990s,” Asian Development Review (Manila), Vol. 9, No. 2 (1991), pp. 4077.

    • Search Google Scholar
    • Export Citation
  • Bahl, Roy W., and Johannes F. Linn Urban Public Finance in Developing Countries (New York: Oxford University Press, 1992).

  • Baumol, William J., Superfairness: Applications and Theory (Cambridge, Massachusetts: MIT Press, 1986).

  • Becker, Arthur P., ed., Land and Building Taxes: Their Effect on Economic Development (Madison, Wisconsin: University of Wisconsin Press, 1969).

    • Search Google Scholar
    • Export Citation
  • Benge, Matt, and Tim Robinson How to Integrate Company and Shareholder Taxation: Why Full Imputation Is the Best Answer (Wellington, New Zealand: Victoria University Press, 1986).

    • Search Google Scholar
    • Export Citation
  • Bernard, Jean-Thomas, and Robert J. Weiner “Multinational Corporations, Transfer Prices, and Taxes: Evidence from the U.S. Petroleum Industry,” in Taxation in the Global Economy, ed. by Assaf Razin and Joel Slemrod (Chicago: University of Chicago Press, 1990), pp. 12359.

    • Search Google Scholar
    • Export Citation
  • Bird, Richard M., Taxing Agricultural Land in Developing Countries (Cambridge, Massachusetts: Harvard University Press, 1974).

  • Boadway, Robin W., “The Theory and Measurement of Effective Tax Rates,” in The Impact of Taxation on Business Activity, ed. by Jack M. Mintz and Douglas D. Purvis (Kingston, Ontario: John Deutsch Institute for the Study of Economic Policy, 1987), pp. 60105.

    • Search Google Scholar
    • Export Citation
  • Boadway, Robin W., Neil Bruce, Jack M. Mintz, Taxes on Capital Income in Canada: Analysis and Policy, Canadian Tax Paper, No. 80 (Toronto: Canadian Tax Foundation, 1987), p. 24.

    • Search Google Scholar
    • Export Citation
  • Boadway, Robin W., Dale H. Chua, and Frank, Flatters “Investment Incentives and Corporate Tax System in Malaysia,” in Fiscal Incentives for Investment in Developing Countries, ed. by Anwar Shah and Robin W. Boadway (Washington: World Bank, 1992), pp. 8.18.35.

    • Search Google Scholar
    • Export Citation
  • Boadway, Robin W., Dale H. Chua, and Frank Flatters, and David E. Wildasin Public Sector Economics (Boston: Little Brown, 1984), pp. 25769.

    • Search Google Scholar
    • Export Citation
  • Boskin, Michael J., and Charles E. McLure, eds., World Tax Reform: Case Studies of Developed and Developing Countries (San Francisco: International Center for Economic Growth, 1990).

    • Search Google Scholar
    • Export Citation
  • Bradford, David F, and U.S. Treasury Tax Policy Staff, Blueprints for Basic Tax Reform (Arlington, Virginia: Tax Analysts, 2nd ed., revised, 1984).

    • Search Google Scholar