imposition of compulsory levies on individuals or entities by governments. Taxes are levied in almost every country of the world, primarily to raise revenue for government expenditures, although they serve other purposes as well.

This article is concerned with taxation in general, its principles, its objectives, and its effects; specifically, the article discusses the nature and purposes of taxation, whether taxes should be classified as direct or indirect, the history of taxation, canons and criteria of taxation, and economic effects of taxation, including shifting and incidence (identifying who bears the ultimate burden of taxes when that burden is passed from the person or entity deemed legally responsible for it to another). For further discussion of taxation's role in fiscal policy, see government economic policy. In addition, see international trade for information on tariffs (tariff).

In modern economies taxes are the most important source of governmental revenue. Taxes differ from other sources of revenue in that they are compulsory levies and are unrequited—i.e., they are generally not paid in exchange for some specific thing, such as a particular public service, the sale of public property, or the issuance of public debt. While taxes are presumably collected for the welfare of taxpayers as a whole, the individual taxpayer's liability is independent of any specific benefit received. There are, however, important exceptions: payroll taxes (payroll tax), for example, are commonly levied on labour income in order to finance retirement benefits, medical payments, and other social security programs—all of which are likely to benefit the taxpayer. Because of the likely link between taxes paid and benefits received, payroll taxes are sometimes called “contributions” (as in the United States). Nevertheless, the payments are commonly compulsory, and the link to benefits is sometimes quite weak. Another example of a tax that is linked to benefits received, if only loosely, is the use of taxes on motor fuels to finance the construction and maintenance of roads and highways, whose services can be enjoyed only by consuming taxed motor fuels.

During the 19th century the prevalent idea was that taxes should serve mainly to finance the government. In earlier times, and again today, governments have utilized taxation for other than merely fiscal purposes. One useful way to view the purpose of taxation, attributable to American economist Richard A. Musgrave, is to distinguish between objectives of resource allocation, income redistribution, and economic stability. (Economic growth or development and international competitiveness are sometimes listed as separate goals, but they can generally be subsumed under the other three.) In the absence of a strong reason for interference, such as the need to reduce pollution, the first objective, resource allocation, is furthered if tax policy does not interfere with market-determined allocations. The second objective, income redistribution, is meant to lessen inequalities in the distribution of income and wealth. The objective of stabilization—implemented through tax policy, government expenditure policy, monetary policy, and debt management—is that of maintaining high employment and price stability.

There are likely to be conflicts among these three objectives. For example, resource allocation might require changes in the level or composition (or both) of taxes, but those changes might bear heavily on low-income families—thus upsetting redistributive goals. As another example, taxes that are highly redistributive may conflict with the efficient allocation of resources required to achieve the goal of economic neutrality.

In the literature of public finance, taxes have been classified in various ways according to who pays for them, who bears the ultimate burden of them, the extent to which the burden can be shifted, and various other criteria. Taxes are most commonly classified as either direct or indirect, an example of the former type being the income tax and of the latter the sales tax. There is much disagreement among economists as to the criteria for distinguishing between direct and indirect taxes, and it is unclear into which category certain taxes, such as corporate income tax or property tax, should fall. It is usually said that a direct tax is one that cannot be shifted by the taxpayer to someone else, whereas an indirect tax can be.

Direct taxes are primarily taxes on natural persons (e.g., individuals), and they are typically based on the taxpayer's ability to pay as measured by income, consumption, or net wealth. What follows is a description of the main types of direct taxes.

Individual income taxes are commonly levied on total personal net income of the taxpayer (which may be an individual, a couple, or a family) in excess of some stipulated minimum. They are also commonly adjusted to take into account the circumstances influencing the ability to pay, such as family status, number and age of children, and financial burdens resulting from illness. The taxes are often levied at graduated rates, meaning that the rates rise as income rises. Personal exemptions for the taxpayer and family can create a range of income that is subject to a tax rate of zero.

Taxes on net worth are levied on the total net worth of a person—that is, the value of his assets minus his liabilities. As with the income tax, the personal circumstances of the taxpayer can be taken into consideration.

Personal or direct taxes on consumption (also known as expenditure taxes or spending taxes) are essentially levied on all income that is not channeled into savings (saving). In contrast to indirect taxes on spending, such as the sales tax, a direct consumption tax can be adjusted to an individual's ability to pay by allowing for marital status, age, number of dependents, and so on. Although long attractive to theorists, this form of tax has been used in only two countries, India and Sri Lanka; both instances were brief and unsuccessful. Near the end of the 20th century, the “flat tax”—which achieves economic effects similar to those of the direct consumption tax by exempting most income from capital—came to be viewed favourably by tax experts. No country has adopted a tax with the base of the flat tax, although many have income taxes with only one rate.

Taxes at death take two forms: the inheritance tax, where the taxable object is the bequest received by the person inheriting, and the estate tax, where the object is the total estate left by the deceased. Inheritance taxes sometimes take into account the personal circumstances of the taxpayer, such as the taxpayer's relationship to the donor and his net worth before receiving the bequest. Estate taxes, however, are generally graduated according to the size of the estate, and in some countries they provide tax-exempt transfers to the spouse and make an allowance for the number of heirs involved. In order to prevent the death duties from being circumvented through an exchange of property prior to death, tax systems may include a tax on gifts above a certain threshold made between living persons (see gift tax). Taxes on transfers do not ordinarily yield much revenue, if only because large tax payments can be easily avoided through estate planning.

Indirect taxes are levied on the production or consumption of goods and services or on transactions, including imports and exports. Examples include general and selective sales taxes (sales tax), value-added taxes (value-added tax) (VAT), taxes on any aspect of manufacturing or production, taxes on legal transactions, and customs or import duties (tariff).

General sales taxes are levies that are applied to a substantial portion of consumer expenditures. The same tax rate can be applied to all taxed items, or different items (such as food or clothing) can be subject to different rates. Single-stage taxes can be collected at the retail level, as the U.S. states do, or they can be collected at a pre-retail (i.e., manufacturing or wholesale) level, as occurs in some developing countries. Multistage taxes are applied at each stage in the production-distribution process. The VAT, which increased in popularity during the second half of the 20th century, is commonly collected by allowing the taxpayer to deduct a credit for tax paid on purchases from liability on sales. The VAT has largely replaced the turnover tax—a tax on each stage of the production and distribution chain, with no relief for tax paid at previous stages. The cumulative effect of the turnover tax, commonly known as tax cascading, distorts economic decisions.

Although they are generally applied to a wide range of products, sales taxes sometimes exempt necessities to reduce the tax burden of low-income households. By comparison, excises are levied only on particular commodities or services. While some countries impose excises and customs duties on almost everything—from necessities such as bread, meat, and salt, to nonessentials such as cigarettes, wine, liquor, coffee, and tea, to luxuries such as jewels and furs—taxes on a limited group of products—alcoholic beverages, tobacco products, and motor fuel—yield the bulk of excise revenues for most countries. In earlier centuries, taxes on consumer durables were applied to luxury commodities such as pianos, saddle horses, carriages, and billiard tables. Today a main luxury tax object is the automobile, largely because registration requirements facilitate administration of the tax. Some countries tax gambling, and state-run lotteries have effects similar to excises, with the government's “take” being, in effect, a tax on gambling. Some countries impose taxes on raw materials, intermediate goods (e.g., mineral oil, alcohol), and machinery.

Some excises and customs duties are specific—i.e., they are levied on the basis of number, weight, length, volume, or other specific characteristics of the good or service being taxed. Other excises, like sales taxes, are ad valorem—levied on the value of the goods as measured by the price. Taxes on legal transactions are levied on the issue of shares, on the sale (or transfer) of houses and land, and on stock exchange transactions. For administrative reasons, they frequently take the form of stamp duties; that is, the legal or commercial document is stamped to denote payment of the tax. Many tax analysts regard stamp taxes as nuisance taxes; they are most often found in less-developed countries and frequently bog down the transactions to which they are applied.

Taxes can be distinguished by the effect they have on the distribution of income and wealth. A proportional tax is one that imposes the same relative burden on all taxpayers—i.e., where tax liability and income grow in equal proportion. A progressive tax is characterized by a more than proportional rise in the tax liability relative to the increase in income, and a regressive tax is characterized by a less than proportional rise in the relative burden. Thus, progressive taxes are seen as reducing inequalities in income distribution, whereas regressive taxes can have the effect of increasing these inequalities.

The taxes that are generally considered progressive include individual income taxes (income tax) and estate taxes. Income taxes that are nominally progressive, however, may become less so in the upper-income categories—especially if a taxpayer is allowed to reduce his tax base by declaring deductions or by excluding certain income components from his taxable income. Proportional tax rates that are applied to lower-income categories will also be more progressive if personal exemptions are declared.

Income measured over the course of a given year does not necessarily provide the best measure of taxpaying ability. For example, transitory increases in income may be saved, and during temporary declines in income a taxpayer may choose to finance consumption by reducing savings. Thus, if taxation is compared with “permanent income,” it will be less regressive (or more progressive) than if it is compared with annual income.

Sales taxes and excises (except those on luxuries) tend to be regressive, because the share of personal income consumed or spent on a specific good declines as the level of personal income rises. Poll taxes (poll tax) (also known as head taxes), levied as a fixed amount per capita, obviously are regressive.

It is difficult to classify corporate income taxes and taxes on business as progressive, regressive, or proportionate, because of uncertainty about the ability of businesses to shift their tax expenses (see below Shifting and incidence (taxation)). This difficulty of determining who bears the tax burden depends crucially on whether a national or a subnational (that is, provincial or state) tax is being considered.

In considering the economic effects of taxation, it is important to distinguish between several concepts of tax rates. The statutory rates are those specified in the law; commonly these are marginal rates, but sometimes they are average rates. Marginal income tax rates indicate the fraction of incremental income that is taken by taxation when income rises by one dollar. Thus, if tax liability rises by 45 cents when income rises by one dollar, the marginal tax rate is 45 percent. Income tax statutes commonly contain graduated marginal rates—i.e., rates that rise as income rises. Careful analysis of marginal tax rates must consider provisions other than the formal statutory rate structure. If, for example, a particular tax credit (reduction in tax) falls by 20 cents for each one-dollar rise in income, the marginal rate is 20 percentage points higher than indicated by the statutory rates. Since marginal rates indicate how after-tax income changes in response to changes in before-tax income, they are the relevant ones for appraising incentive effects of taxation. It is even more difficult to know the marginal effective tax rate applied to income from business and capital, since it may depend on such considerations as the structure of depreciation allowances, the deductibility of interest, and the provisions for inflation adjustment. A basic economic theorem holds that the marginal effective tax rate in income from capital is zero under a consumption-based tax.

Average income tax rates indicate the fraction of total income that is paid in taxation. The pattern of average rates is the one that is relevant for appraising the distributional equity of taxation. Under a progressive income tax the average income tax rate rises with income. Average income tax rates commonly rise with income, both because personal allowances are provided for the taxpayer and dependents and because marginal tax rates are graduated; on the other hand, preferential treatment of income received predominantly by high-income households may swamp these effects, producing regressivity, as indicated by average tax rates that fall as income rises.

Although views on what is appropriate in tax policy influence the choice and structure of tax codes, patterns of taxation throughout history can be explained largely by administrative considerations. For example, because imported products are easier to tax than domestic output, import duties were among the earliest taxes. Similarly, the simple turnover tax (levied on gross sales) long held sway before the invention of the economically superior but administratively more demanding VAT (which allows credit for tax paid on purchases). It is easier to identify, and thus tax, real property than other assets; and a head (poll) tax is even easier to implement. It is not surprising, therefore, that the first direct levies were head and land taxes.

Although taxation has a long history, it played a relatively minor role in the ancient world. Taxes on consumption were levied in Greece and Rome. Tariffs (tariff)—taxes on imported goods—were often of considerably more importance than internal excises so far as the production of revenue went. As a means of raising additional funds in time of war, taxes on property would be temporarily imposed. For a long time these taxes were confined to real property, but later they were extended to other assets. Real estate transactions also were taxed. In Greece free citizens had different tax obligations from slaves, and the tax laws of the Roman Empire distinguished between nationals and residents of conquered territories.

Early Roman (ancient Rome) forms of taxation included consumption taxes, customs duties, and certain “direct” taxes. The principal of these was the tributum, paid by citizens and usually levied as a head tax; later, when additional revenue was required, the base of this tax was extended to real estate holdings. In the time of Julius Caesar, a 1 percent general sales tax was introduced (centesima rerum venalium). The provinces relied for their revenues on head taxes and land taxes; the latter consisted initially of fixed liabilities regardless of the return from the land, as in Persia and Egypt, but later the land tax was modified to achieve a certain correspondence with the fertility of the land, or, alternatively, a 10th of the produce was collected as a tax in kind (the tithe). It is noteworthy that at a relatively early time Rome had an inheritance tax of 5 percent, later 10 percent; however, close relatives of the deceased were exempted. For a long time tax collection was left to middlemen, or “tax farmers,” who contracted to collect the taxes for a share of the proceeds; under Caesar collection was delegated to civil servants.

In the Middle Ages many of these ancient taxes, especially the direct levies, gave way to a variety of obligatory services and a system of “aids” (most of which amounted to gifts). The main indirect taxes were transit duties (a charge on goods that pass through a particular country) and market fees. In the cities the concept developed of a tax obligation encompassing all residents: the burden of taxes on certain foods and beverages was intended to be borne partly by consumers and partly by producers and tradesmen. During the later Middle Ages some German (Germany) and Italian cities introduced several direct taxes: head taxes for the poor and net-worth taxes or, occasionally, crude income taxes for the rich. (The income tax was administered through self-assessment and an oath taken before a civic commission.) Taxes on land and on houses gradually increased.

Taxes have been a major subject of political controversy throughout history, even before they constituted a sizable share of the national income. A famous instance is the rebellion of the American colonies against Great Britain (United Kingdom), when the colonists refused to pay taxes imposed by a Parliament in which they had no voice—hence the slogan, “No taxation without representation.” Another instance is the French Revolution of 1789, in which the inequitable distribution of the tax burden was a major factor.

Wars (war finance) have influenced taxes much more than taxes have influenced revolutions. Many taxes, notably the income tax (first introduced in Great Britain in 1799) and the turnover or purchase tax (Germany, 1918; Great Britain, 1940), began as “temporary” war measures. Similarly, the withholding method of income tax collection began as a wartime innovation in France, the United States, and Britain. World War II converted the income taxes of many countries from upper-class taxes to mass taxes.

It is hardly necessary to mention the role that tax policies play in peacetime politics, where the influence of powerful, well-organized pressure groups is great. Arguments for tax reform, particularly in the area of income taxes, are perennially at issue in the domestic politics of many countries.

The development of taxation in recent times can be summarized by the following general statements, although allowance must be made for considerable national differences: The authority of the sovereign to levy taxes in a more or less arbitrary fashion has been lost, and the power to tax now generally resides in parliamentary bodies. The level of most taxes has risen substantially and so has the ratio of tax revenues to the national income. Taxes today are collected in money, not in goods. Tax farming—the collection of taxes by outside contractors—has been abolished, and taxes are instead assessed and collected by civil servants. (On the other hand, as a means of overcoming the inefficiencies of government agencies, tax collection has recently been contracted to banks in many less-developed countries. In addition, some countries are outsourcing the administration of customs duties.)

There has also been a reduction in reliance on customs duties and excises. Many countries increasingly rely on sales taxes and other general consumption taxes. An important late 20th-century development was the replacement of turnover taxes with value-added taxes. Taxes on the privilege of doing business and on real property lost ground, although they have persisted as important revenue sources for local communities. The absolute and relative weight of direct personal taxation has been growing in most of the developed countries, and increasing attention has been focused on VAT and payroll taxes. At the end of the 20th century the expansion of e-commerce created serious challenges for the administration of VAT, income taxes, and sales taxes. The problems of tax administration were compounded by the anonymity of buyers and sellers, the possibility of conducting business from offshore tax havens, the fact that tax authorities cannot monitor the flow of digitized products or intellectual property, and the spate of untraceable money flows.

Income taxation (of individuals and of corporations), payroll taxes, general sales taxes, and (in some countries) property taxes bring in the greatest amounts of revenue in modern tax systems. The income tax has ceased to be a “rich man's” tax; it is now paid by the general populace, and in several countries it is joined by a tax on net worth. The emphasis on the ability-to-pay principle and on the redistribution of wealth—which led to graduated rates and high top marginal income tax rates—appears to have peaked, having been replaced by greater concern for the economic distortions and disincentives caused by high tax rates. A good deal of fiscal centralization occurred through much of the 20th century, as reflected in the kinds of taxes levied by central governments. They now control the most important taxes (from a revenue-producing point of view): income and corporation taxes, payroll taxes, and value-added taxes. Yet, in the last decade of the 20th century, many countries experienced a greater decentralization of government and a consequent devolution of taxing powers to subnational governments. Proponents of decentralization argue that it can contribute to greater fiscal autonomy and responsibility, because it involves states and municipalities in the broader processes of tax policy; merely allowing lower-level governments to share in the tax revenues of central governments does not foster such autonomy.

Although it is difficult to make general distinctions between developed and less-developed countries, it is possible to detect some patterns in their relative reliance on various types of taxes. For example, developed countries usually rely more on individual income taxes and less on corporate income taxes than less-developed countries do. In developing countries, reliance on income taxes, especially on corporate income taxes, generally increases as the level of income rises. In addition, a relatively high percentage of the total tax revenue of industrialized countries comes from domestic consumption taxes, especially the value-added tax (rather than the simpler turnover tax). Social security taxes—commonly collected as payroll taxes—are much more important in developed countries and the more-affluent developing countries than in the poorest countries, reflecting the near lack of social security systems in the latter. Indeed, in many developed countries, payroll taxes rival or surpass the individual income tax as a source of revenue. Demographic trends and their consequences (in particular, the aging of the world's working population and the need to finance public pensions) threaten to raise payroll taxes to increasingly steep levels. Some countries have responded by privatizing the provision of pensions—e.g., by substituting mandatory contributions to individual accounts for payroll taxes.

Taxes in general represent a much higher percentage of national output in developed countries than in developing countries. Similarly, more national output is channeled to governmental use through taxation in developing countries with the highest levels of income than in those with lesser incomes. Indeed, in many respects the tax systems of the developing countries with the highest levels of income have more in common with those of developed countries than they have with the tax systems of the poorest developing countries.

Although they need to be reinterpreted from time to time, these principles retain remarkable relevance. From the first can be derived some leading views about what is fair in the distribution of tax burdens among taxpayers. These are: (1) the belief that taxes should be based on the individual's ability to pay, known as the ability-to-pay principle, and (2) the benefit principle, the idea that there should be some equivalence between what the individual pays and the benefits he subsequently receives from governmental activities. The fourth of Smith's canons can be interpreted to underlie the emphasis many economists place on a tax system that does not interfere with market decision making, as well as the more obvious need to avoid complexity and corruption.

Various principles, political pressures, and goals can direct a government's tax policy. What follows is a discussion of some of the leading principles that can shape decisions about taxation.

The principle of horizontal equity assumes that persons in the same or similar positions (so far as tax purposes are concerned) will be subject to the same tax liability. In practice this equality principle is often disregarded, both intentionally and unintentionally. Intentional violations are usually motivated more by politics than by sound economic policy (e.g., the tax advantages granted to farmers, home owners, or members of the middle class in general; the exclusion of interest on government securities). Debate over tax reform has often centred on whether deviations from “equal treatment of equals” are justified.

The ability-to-pay principle requires that the total tax burden will be distributed among individuals according to their capacity to bear it, taking into account all of the relevant personal characteristics. The most suitable taxes from this standpoint are personal levies (income, net worth, consumption, and inheritance taxes). Historically there was common agreement that income is the best indicator of ability to pay. There have, however, been important dissenters from this view, including the 17th-century English philosophers John Locke (Locke, John) and Thomas Hobbes (Hobbes, Thomas) and a number of present-day tax specialists. The early dissenters believed that equity should be measured by what is spent (i.e., consumption) rather than by what is earned (i.e., income); modern advocates of consumption-based taxation emphasize the neutrality of consumption-based taxes toward saving (income taxes discriminate against saving), the simplicity of consumption-based taxes, and the superiority of consumption as a measure of an individual's ability to pay over a lifetime. Some theorists believe that wealth provides a good measure of ability to pay because assets imply some degree of satisfaction (power) and tax capacity, even if (as in the case of an art collection) they generate no tangible income.

The ability-to-pay principle also is commonly interpreted as requiring that direct personal taxes have a progressive rate structure, although there is no way of demonstrating that any particular degree of progressivity is the right one. Because a considerable part of the population does not pay certain direct taxes—such as income or inheritance taxes—some tax theorists believe that a satisfactory redistribution can only be achieved when such taxes are supplemented by direct income transfers or negative income taxes (income tax) (or refundable credits). Others argue that income transfers and negative income tax create negative incentives; instead, they favour public expenditures (for example, on health or education) targeted toward low-income families as a better means of reaching distributional objectives.

Indirect taxes such as VAT, excise, sales, or turnover taxes can be adapted to the ability-to-pay criterion, but only to a limited extent—for example, by exempting necessities such as food or by differentiating tax rates according to “urgency of need.” Such policies are generally not very effective; moreover, they distort consumer purchasing patterns, and their complexity often makes them difficult to institute.

Throughout much of the 20th century, prevailing opinion held that the distribution of the tax burden among individuals should reduce the income disparities that naturally result from the market economy; this view was the complete contrary of the 19th-century liberal view that the distribution of income ought to be left alone. By the end of the 20th century, however, many governments recognized that attempts to use tax policy to reduce inequity can create costly distortions, prompting a partial return to the view that taxes should not be used for redistributive purposes.

Under the benefit principle, taxes are seen as serving a function similar to that of prices in private transactions; that is, they help determine what activities the government will undertake and who will pay for them. If this principle could be implemented, the allocation of resources through the public sector would respond directly to consumer wishes.

In fact, it is difficult to implement the benefit principle for most public services because citizens generally have no inclination to pay for a publicly provided service—such as a police department—unless they can be excluded from the benefits of the service. The benefit principle is utilized most successfully in the financing of roads and highways through levies on motor fuels and road-user fees (tolls (toll)). Payroll taxes used to finance social security may also reflect a link between benefits and “contributions,” but this link is commonly weak, because contributions do not go into accounts held for individual contributors.

The requirement that a tax system be efficient arises from the nature of a market economy. Although there are many examples to the contrary, economists generally believe that markets do a fairly good job in making economic decisions about such choices as consumption, production, and financing. Thus, they feel that tax policy should generally refrain from interfering with the market's allocation of economic resources. That is, taxation should entail a minimum of interference with individual decisions. It should not discriminate in favour of, or against, particular consumption expenditures, particular means of production, particular forms of organization, or particular industries. This does not mean, of course, that major social and economic goals may not take precedence over these considerations. It may be desirable, for example, to impose taxes on pollution as a means of protecting the environment.

Economists have developed techniques to measure the “excess burden” that results when taxes distort economic decision making. The basic notion is that if goods worth $2 are sacrificed because of tax influences in order to produce goods with a value of only $1.80, there is an excess burden of 20 cents. A more nearly neutral tax system would result in less distortion. Thus, an important postwar development in the theory of taxation is that of optimal taxation, the determination of tax policies that will minimize excess burdens. Because it deals with highly stylized mathematical descriptions of economic systems, this theory does not offer easily applied prescriptions for policy, beyond the important insight that distortions do less damage where supply and demand are not highly sensitive to such distortions. Attempts have also been made to incorporate distributional considerations into this theory. They face the difficulty that there is no scientifically correct distribution of income.

In discussing the general principles of taxation, one must not lose sight of the fact that taxes must be administered by an accountable authority. There are four general requirements for the efficient administration of tax laws: (tax law) clarity, stability (or continuity), cost-effectiveness, and convenience. Administrative considerations are especially important in developing countries, where illiteracy, lack of commercial markets, absence of books of account, and inadequate administrative resources may hinder both compliance and administration. Under such circumstances the achievement of rough justice may be preferable to infeasible fine-tuning in the name of equity.

Tax laws and regulations must be comprehensible to the taxpayer; they must be as simple as possible (given other goals of tax policy) as well as unambiguous and certain—both to the taxpayer and to the tax administrator. While the principle of certainty is better adhered to today than in the time of Adam Smith, and arbitrary administration of taxes has been reduced, every country has tax laws that are far from being generally understood by the public. This not only results in a considerable amount of error but also undermines honesty and respect for the law and tends to discriminate against the ignorant and the poor, who cannot take advantage of the various legal tax-saving opportunities that are available to the educated and the affluent. At times, attempts to achieve equity have created complexity, defeating reform purposes.

Tax laws should be changed seldom, and, when changes are made, they should be carried out in the context of a general and systematic tax reform, with adequate provisions for fair and orderly transition. Frequent changes to tax laws can result in reduced compliance or in behaviour that attempts to compensate for probable future changes in the tax code—such as stockpiling liquor in advance of an increased tariff on alcoholic beverages.

The costs of assessing, collecting, and controlling taxes should be kept to the lowest level consistent with other goals of taxation. This principle is of secondary importance in developed countries, but not in developing countries and countries in transition from socialism, where resources needed for compliance and administration are scarce. Clearly, equity and economic rationality should not be sacrificed for the sake of cost considerations. The costs to be minimized include not only government expenses but also those of the taxpayer and of private fiscal agents such as employers who collect taxes for the government through the withholding procedure.

Payment of taxes should cause taxpayers as little inconvenience as possible, subject to the limitations of higher-ranking tax principles. Governments often allow the payment of large tax liabilities in installments and set generous time limits for completing returns.

The primary goal of a national tax system is to generate revenues to pay for the expenditures of government at all levels. Because public expenditures tend to grow at least as fast as the national product, taxes, as the main vehicle of government finance, should produce revenues that grow correspondingly. Income, sales, and value-added taxes generally meet this criterion; property taxes and taxes on nonessential articles of mass consumption such as tobacco products and alcoholic beverages do not.

In addition to producing revenue, tax policy may be used to promote economic stability. Changes in tax liabilities not matched by changes in expenditures cushion cyclical (business cycle) fluctuations in prices, employment, and production. Built-in flexibility occurs because liabilities for some taxes, most notably income taxes, respond strongly to changes in economic conditions. A more-active approach calls for changes in the tax rates or other provisions to increase the anticyclical effects of tax receipts.

Some economists propose tax policies to promote economic growth. This approach may imply a qualitative restructuring of the tax system (for example, the substitution of taxes on consumption for taxes on income) or special tax advantages to stimulate saving, labour mobility, research and development, and so on. There is, however, a limit to what tax incentives can accomplish, especially in promoting economic development of specific industries or regions. An emphasis on economic growth implies the need to avoid high marginal tax rates and the tax-induced diversion of resources into relatively unproductive activities.

The incidence of a tax rests on the person(s) whose real net income is reduced by the tax. It is fundamental that the real burden of taxation does not necessarily rest upon the person who is legally responsible for payment of the tax. General sales taxes are paid by business firms, but most of the cost of the tax is actually passed on to those who buy the goods that are being taxed. In other words, the tax is shifted from the business to the consumer. Taxes may be shifted in several directions. Forward shifting takes place if the burden falls entirely on the user, rather than the supplier, of the commodity or service in question—e.g., an excise tax on luxuries that increases their price to the purchaser. Backward shifting occurs when the price of the article taxed remains the same but the cost of the tax is borne by those engaged in producing it—e.g., through lower wages and salaries, lower prices for raw materials, or a lower return on borrowed capital. Finally, a tax may not be shifted at all—e.g., a tax on business profits may reduce the net income of the business owner.

Tax capitalization occurs if the burden of the tax is incorporated in the value of long-term assets—e.g., a decline in the price of land that offsets an increase in property taxes. Capitalization can result where there is forward shifting, backward shifting, or no shifting. Thus, an increase in the price of gasoline resulting from higher motor fuel taxes may reduce the value of high-consumption automobiles, a tax on the production of coal that cannot be shifted forward would reduce the value of coal deposits, and a tax that reduces after-tax corporate profits may reduce the value of corporate stock. In all these cases the present owner of the asset takes a capital loss because the value of the asset will be lower by the capitalized value of the tax.

It can be difficult to determine the incidence of a tax; indeed, the tax may be partly borne by the taxpayer and partly shifted. In many cases the problem can be adequately resolved by using what economists call partial equilibrium analysis, which involves focusing on the market for the taxed product and ignoring all other markets. For example, if a small tax were to be imposed on an addictive substance, there is little doubt that it would be borne by the users of the substance, who would pay the tax rather than forgo use of the substance. More generally, the incidence of taxation depends on all of the market forces at work. In a market economy the introduction of any tax triggers a whole series of adjustments in consumption, production, the supply of productive factors, and the pattern of foreign trade. These adjustments in turn will have repercussions on the prices of various commodities, productive factors, and assets that may be far removed from the area of the initial impact. In other words, a tax levied on a certain object may affect the prices of nontaxed goods and services that are not even used in the production of the object. Thus, the initial impact of a tax does not indicate where the ultimate burden will rest unless one knows what repercussions the tax will have throughout the system of interrelated economic variables—i.e., unless recourse is made to what is called general equilibrium theory, a method of analysis that attempts to identify and incorporate the economy-wide repercussions and implications of taxation. In what follows, an attempt will be made to isolate some of the factors involved.

The direction and extent of tax shifting is determined basically by one principle: The user of a tax object can avoid the tax burden to a greater (lesser) extent the easier (the more difficult) it is to find nontaxed or less-taxed alternatives or substitutes for the tax object; the supplier of a production factor that is taxed or used in the production of a taxed good can avoid the tax burden to a greater (lesser) extent the easier (the more difficult) it is to find equivalent nontaxed or less taxed alternative employment opportunities for this factor. Because the demand for substitute goods will increase, their prices may rise, thus benefiting the producers of such goods and placing part of the tax burden on those individuals who used them before the tax was imposed. Likewise, the productive factors that seek alternative employments to avoid the tax will tend to receive lower returns in those employments, thus placing part of the burden on individuals who supplied the factors in those sectors before the tax was imposed. For example, if wine is taxed while beer is not, then—if these two beverages are regarded as perfect substitutes and the price of beer does not rise with increased demand—the tax burden will fall on the owners of land used for viticulture and on the workers engaged in it. It will fall mainly on the landowners if the soil is specific to grapevine growing and if labour has alternative employment possibilities. If, on the other hand, wine drinkers are determined to drink only wine, they will bear most of the tax burden. If some substitution of beer for wine takes place and the price of beer rises somewhat, both wine and beer drinkers will bear the burden and owners of resources specialized to the production of beer will benefit.

In addition to the substitution effect discussed above, one must take into account the income effect. When taxation reduces real income, consumption of certain goods and services (and of leisure) will be reduced, because people have less money to spend. Furthermore, if a tax causes a significant redistribution of real income and if different income classes have different propensities to save and different patterns of consumption, then the income redistribution will influence the demand for various goods, the supply of labour, and the demand for various resources.

Other considerations affect tax shifting, but they are derived from the basic principle of substitution. The extent of shifting may vary over time, depending on how long it takes to adjust consumption patterns, reallocate land and capital, retrain labour, and so on. Those users and suppliers who have the most difficulty in adjusting will bear the largest burden.

The breadth of the tax base affects tax incidence. The broader (narrower) the tax base—i.e., the more (less) inclusive the scope of the tax—the more difficult it is to escape the tax burden, since the range of nontaxed or less-taxed substitutes is narrower (wider). Thus, an excise tax on only a few alcoholic beverages allows the tax to be escaped through a change in the consumption pattern, while a tax on all such beverages does not. In a similar fashion, the returns on capital will be affected less by the taxation of corporation profits alone than by the taxation of both corporation and noncorporation profits.

The smaller the jurisdictional unit imposing the tax, the easier it tends to be for a user to obtain nontaxed or less-taxed substitutes from outside the jurisdiction and for a supplier to find nontaxed or less-taxed outside employment opportunities for his goods and services. Thus, a tax levied by a subnational government on the production of a particular good is likely to be borne by suppliers of commodities and productive factors that are immobile. This is particularly relevant to the determination of the incidence of state income taxes and local property taxes, taxes that are often thought to be “exported” to out-of-state consumers. In small communities the only really immobile factors are likely to be real estate, certain local services, and perhaps poor families.

The rigidities of imperfect markets (market) are likely to increase the uncertainty of the shifting response. Thus, a monopolist may absorb part of a tax in lower profits rather than shift all of the burden to the user of the product. In industries where there are few firms (oligopoly), the price behaviour of a firm is mainly determined by what it expects its competitors to do. It may be especially easy for regulated public utilities to shift taxes forward. Rigid product prices are likely to increase the incidence of taxes on employment, unless monetary policy allows the tax-induced changes in relative prices to take place in the setting of a generally rising price level.

All of these considerations are analytical and theoretical. Efforts have been made to measure the impact of taxation by studying the actual effects of a particular tax on income and employment. These studies reflect the obvious and inherent difficulty that the tax impact cannot be easily isolated from the economic consequences of other events. For example, studies of corporate income tax shifting vary in their results, from the conclusion that the tax is not shifted at all to the conclusion that it is shifted by more than 100 percent, depending mainly on the methods used to isolate the tax impact.