In my own case the words of such an act as the Income Tax, for example, merely dance before my eyes in a meaningless procession: cross-reference to cross-reference, exception upon exception—couched in abstract terms that offer no handle to seize hold of—leave in my mind only a confused sense of some vitally important, but successfully concealed, purport, which it is my duty to extract, but which is within my power, if at all, only after the most inordinate expenditure of time. —Learned Hand, The Spirit of Liberty

I. Introduction

Drafting tax laws is a subspecialty of legislative drafting in general.1 This chapter does not attempt a comprehensive treatment but serves as an introduction, focusing on questions that have been of particular concern in drafting tax laws. The focus is on drafting technique, other matters relating to drafting being considered elsewhere in the book.2 In addition, the chapter is limited to considerations that apply generally, regardless of language or jurisdiction. Because languages and local drafting styles differ, the approach to drafting a tax law will vary widely from country to country.

Those who draft tax legislation in developing or transition countries usually are not subspecialists in this area (i.e., lawyers who have specialized in legislative drafting and in particular in the drafting of tax legislation). Local officials responsible for drafting typically are tax experts in the ministry of finance, and are not often lawyers or specialists in drafting. Foreign advisors who assist them can be lawyers, accountants, or economists with a background in taxation, but do not usually have an expertise in legislative drafting. Therefore, this chapter explores some of the lessons that can be learned from specialists in legislative drafting.

The effectiveness of a tax law is enhanced if its words are meaningful, intelligible, well thought out, and well organized. Many tax laws do not come close to meeting these criteria. The tax laws of countries with established and sophisticated systems can be particularly impenetrable, as qualifications and exceptions have been heaped on top of existing rules. In this sense, those working in developing and transition countries have an opportunity to produce better laws than exist in developed countries. Poor drafting often leads to substantial problems in implementation of a new tax law that could have been avoided. A goal of this chapter is to encourage those involved in the tax legislative process to devote greater attention to drafting technique.

The discussion in this chapter is organized according to the criteria for a well-drafted law. I have identified these as understandability, organization, effectiveness, and integration. Understandability refers to making the law easier to read and follow. Organization refers to both the internal organization of the law and its coordination with other tax laws. Effectiveness relates to the law’s ability to enable the desired policy to be implemented. Finally, integration refers to the consistency of the law with the legal system and drafting style of the country. These criteria are, of course, interrelated and somewhat overlapping. Organization is important for understandability, and all the criteria contribute to the effectiveness of the law.

In the most general terms, the tax laws should be drafted so as to best fulfill their role in the tax system, which is to specify such matters as how much each taxpayer is liable to pay and what the taxpayer’s rights and obligations are.

A well-drafted tax law spells out with precision the matters that are within its scope. But precision is not enough. A law should not be precise at the expense of being complicated and impossible to understand. The easier a tax law is to understand, the lower will be the compliance costs, both for taxpayers and for tax administrators. It is particularly important that a tax law be easy to apply (compared with other public law, for example, a law governing the generation of toxic waste or one governing building codes) because the tax law applies to nearly every physical and legal person in the country with respect to countless transactions every day. The fact that tax law must be applicable to so many transactions in an efficient manner has an important influence on how the law must be drafted. In particular, there is little room for sloppiness. Finally, a tax law must be effective in achieving the policy goals of the legislator, both in terms of the amount of revenue to be raised—with an eye to equity, efficiency, and simplicity—and the items and persons to be taxed. Good drafting goes hand in hand with the specification of policy.

These criteria sometimes conflict. For example, a simple statute may be rejected as inequitable, because it does not recognize the differences in situation of different taxpayers. A statute that provides too much certainty may conflict with the goals of equity and revenue raising (because the certainty can be exploited by tax planners). In many cases, however, there is no conflict; complexity that is merely the result of bad drafting can be eliminated while at the same time providing greater certainty and a clearer articulation of the policy.

II. Understandability

A. Brevity

The shorter the statute, the less effort will be required to understand it, and the lower compliance burdens will be. Elegance, brevity, and clarity of expression are therefore to be sought. Every word in a statute should have a definite purpose and no unnecessary word should be used.3 In addition to being easier to understand, a more elegant statement often better articulates the policy of the law. For example, on reviewing an initial draft of a statute, the drafter might notice that several rules, perhaps located in different parts of the statute, could be combined into one general rule that covers what previously appeared as unconnected details. (This is one example of the close interrelationship between the development of policy and the process of drafting.)

The prescription for brevity does not necessarily mean that a shorter statute is better than a longer one. It may be determined that certain details need to go into the statute, and it takes additional words to express these details. Moreover, the expression of an idea in so few words that it becomes cryptic and understandable only after careful study also constitutes an extreme to be avoided. The point is simply that no word should be included if it does not serve a function.

How long the optimum law would be for a particular tax is an interesting philosophical question and an important one in drafting a new law. Whatever the answer in the abstract, much more important is the local situation. The local officials who will be working with the new law must make it their own. The constraint in terms of length is often how much can be absorbed by those who will be using the law. Because taxation suitable for a market economy is a recent phenomenon in transition countries, the length of the tax laws in such countries can be expected to increase as their experience with taxation grows. One implication is that the tax legislative process for these countries has only just begun.

B. Transparency

A statute is transparent if it easily allows the reader to understand the rationale for the rules.

One way of achieving transparency is to begin a law by stating its purpose.4 If the statement is very general, it is not helpful. On the other hand, a very general statement can do little harm. (E.g., an income tax law might begin: “This law levies a tax on income.”) If the statement is made more specific and operational, then it could serve a function by indicating the overall legislative purpose so as to facilitate interpretation of ambiguous provisions. In the case of some legislation, this might be helpful. For example, a piece of environmental legislation might stipulate that the purpose of the legislation is to eliminate pollution wherever technically feasible, regardless of the cost, or it might provide the opposite, that the statute should not be construed as requiring measures to be taken whose costs are disproportionate to the environmental benefits. Either philosophy, if articulated by the legislature, would give guidance to the courts as to the legislative intent.

In the tax area, however, it may be dangerous to make an overriding general statement of the legislative purpose, because the courts may interpret the provisions of the statute in light of this stated purpose and may be misled in doing so. It is difficult to avoid a misleading statement of purpose, because tax laws are highly technical and, in some cases, artificial. For example, person X might be denied a deduction because that is an administratively more feasible approach than taxing person Y on the payment received from person X.5 The result is that person Y is not taxed on the payment and person X is denied a deduction that should be available under general principles. This would hardly square with a general statement to the effect that “each taxpayer shall pay tax on the taxpayer’s net economic income.” Not only do the tax laws contain artificial provisions, but they are also prey to competing policy goals. Some deductions are allowed because they accurately reflect net income, others because of a legislative purpose to encourage a particular activity; in some cases, motives for providing a deduction are mixed. Because of the disparate and competing policies behind tax legislation, it becomes difficult to describe the general purpose of the law in operational terms. Tax laws therefore have generally not included a statement of purpose. Recently, however, in New Zealand, legislation has been introduced that states the purpose of the income tax act and provides an underlying framework for it.6 If adopted, it will be interesting to see how this works.

Instead of stating the law’s purpose, the law could begin with an exposition of its basic mechanics, that is, identification of the taxpayers, tax base, and location of provisions containing the rates. This would be more modest than attempting to state the purpose of the law, but it would at least allow the reader to see the basic structure of the law at the beginning. Whether this approach is suitable in a particular case will depend on local drafting style.

C. Avoiding Legalistic Language

Advocates of “plain language” drafting recommend avoiding legalistic language, so as to make the law easier to understand.7 Where legalisms are verbose or obscure, they can make comprehension more difficult, and they should be eliminated if possible. For example, artificial terms with meanings defined in the tax law should be used sparingly, as their use will often be confusing to a reader who is not thoroughly familiar with the statute. On the other hand, the use of legal terms or other terms of art can make the statute precise and shorter when these words have a technical meaning or a meaning determined by legal rules outside the tax law. For example, concepts such as corporation, partnership, employee, contract, mortgage, or lien may be well defined by laws outside the tax laws, and it is appropriate to use these words rather than simpler words that might be more understandable to laypersons. It is sometimes appropriate, however, to adopt a modified meaning for tax purposes.8 One should therefore not hesitate to use technical terms where appropriate, even if a layperson might as a result have greater difficulty in understanding the statute. While it is nice if portions of the tax laws are comprehensible to nonlawyers (or even to non–tax lawyers), and the law should be so drafted if precision is not sacrificed, comprehensibility to the layperson should not be an absolute requirement in an area as technical as tax law.9 A tightly drafted statute can be translated to the layperson in the form of instructions.

D. Numbering of Sections10

Most countries have adopted the practice of numbering the sections of a statute sequentially, that is, 1, 2, 3, and so on.11 While this is fine for a statute that will never be changed, most tax laws are amended frequently. Amendments create problems for sequential numbering. Either amendments have to be placed at the end of the statute, in which case they are not in the logically appropriate place; or they can be inserted in the appropriate place, but the sections of the statute have to be renumbered; or they can be inserted under a hybrid alphanumerical designation. Except when legislation is completely overhauled, renumbering is confusing and should be avoided because references to section numbers in other laws, in legal documents, in judicial decisions, in regulations, and in articles and other descriptive materials become incorrect.

Renumbering can be avoided by inserting new sections between the existing sections. However, this can lead to bizarre and confusing designations for sections.12 The solution adopted in the U.S. Code (one title of which is the Internal Revenue Code) is nonsequential numbering. The approach is to leave a gap in section numbering between each division of the statute. If new sections are added, they can be assigned the unused section numbers. For example, the first group of sections might run from 1 to 14, and the next group begin with section 20. One might object to this on philosophical grounds (the section numbered 20 is not the twentieth section, but only the fifteenth), but section designations like “238 bis” seem equally objectionable to me (how can there be a first section 238 and a second section 238?). If tradition can be overcome, nonsequential numbering offers an advantage.13

E. Section Headings

In many countries, sections do not have section headings,14 but are simply designated by numbers.15 The use of section headings makes it much easier to read and understand the law; moreover, it imposes discipline on the drafter. If the drafter cannot think of a good heading for a section, it may be because the section contains disparate subject matter, which would best be broken into more than one section. Recent legislation in a number of transition countries now contains section headings.16

If a decision is made to use section headings, the question arises whether headings for subdivisions of a section should also be used. For example, in the U.S. Internal Revenue Code, not only does each section have a heading, but each subsection and paragraph does too. This may be appropriate in a situation, such as for the U.S. Code, where even the subdivisions of a section are lengthy. In a sparser drafting style, the use of headings for subdivisions of articles would lead to clutter. A cluttered statute being more difficult for the reader to digest, there comes a point in the subdivision of a statute where the use of headings should stop. In most cases, my preference would be to do this at the level of the section, but the matter should be decided in light of the characteristics of the statute being drafted and of the country’s drafting style.

Where section headings (or marginal notes) are used, the question of their legal effect should be considered; that is, to what extent should or will courts rely on the section heading in construing the statute?17 It is generally best to keep the headings short (one to four words); this makes it clear that the heading will not capture all the nuances of the section it heads.

F. Sentence Structure

Long, complex sentences should generally be avoided because they impede understanding (in some cases, however, a lengthy sentence can express an overall thought more succinctly than shorter sentences). Some drafting traditions follow the opposite approach and actually encourage the use of longer sentences. In the U.K. tradition, a section may not contain more than one sentence unless broken down into separate subsections.18 Drafters can and do, however, get around this rule by creating run-on sentences using conjunctions or semicolons. Horribly long sentences result. The injunction against more than one sentence in a subdivision makes little sense. Often, a rule is best expressed using more than one sentence, and it is easier to understand the rule if these sentences are located in a single subdivision. If the goal is elegance and comprehensibility, the U.K. rule should be abandoned. However, tradition sometimes dies hard, and it is possible, although not ideal,19 to work within the constraints of the U.K. rule in jurisdictions that adhere to it.

Another question of sentence structure is whether a sentence should be broken down by numbering and indenting its logical components. This has been called “paragraphing.”20 Paragraphing is to be recommended on two closely related grounds: it is a means of removing ambiguity, and it makes sentences easier to read. Paragraphing reveals the logical structure of a sentence at a glance; it divides the sentence into elements that can more readily be comprehended one at a time and shows graphically the relationship between these elements.

For example, consider the following sentence:

The property income derived—

  • (a) from a foreign source; or

  • (b) from the disposal of an investment or asset generating foreign-source income

by an expatriate taxpayer is exempt from income tax.21

Paragraphing in this sentence makes clear that the condition in the flush language (“by an expatriate taxpayer”) applies to both items (a) and (b). It also allows specification on a self-contained basis of each of the elements of the sentence and allows the reader to quickly grasp the nature of the rule. If the paragraphing were removed, the sentence would possibly be ambiguous and would be more difficult to follow.

Paragraphing has its detriments, however. It makes the statute seem complex and abstract, where it might be easier to digest if the numbering and indentations were removed. This is especially the case when multiple tiers of subdivisions are used. Perhaps the most extreme example of this is the U.S. Internal Revenue Code, which regularly subdivides individual sections into several layers. Paragraphing should therefore not be overused, and the number of tiers of subdivision should be limited (more than two tiers is rarely necessary).

When sections of a statute are divided, it is desirable to adopt a uniform style for division, thereby allowing for easy identification and reference to subdivisions of a section.22 A contrast to this is the division style of the tax code of France, which is inconsistent. In some cases, articles are divided into paragraphs that are not numbered (i.e., there is just indentation).23 In other cases, the first division of an article is numbered according to Roman numerals,24 and in other cases, it is numbered with Arabic numbers.25

III. Organization

A. General Issues

Logical organization of a statute aids comprehension. If the statute is well organized, it is also easier to determine where one needs to look for the answer to a particular question, and which portions of the statute can be ignored by a particular taxpayer. Each tax law contains the same key elements (taxpayers, rates, tax base, procedure, and administration), and understanding is improved if all the tax laws of a particular country follow the same order in respect of these elements. Foreign advisors in particular should consult the local practice in this respect.

Organization requires grouping together provisions on the same topic. Moreover, each subdivision of the statute, including individual sections, should be constructed in an order that facilitates comprehension. Usually, this means stating the general rule first and following it with exceptions and special rules for particular cases.

Proper organization is as important for the drafter as it is for the reader of the statute. The organization of a statute is like the framing of a house. Organizing rules helps the drafter think them through. If the drafter is forced to think about where in the statute a particular section should go, then he or she will think more carefully about its function, which will help in understanding and formulating the rule. It might occur to the drafter, for example, that what started as a rather particular rule should be rewritten as a more general rule that goes elsewhere in the statute. Grouping rules together also helps the drafter figure out whether any pieces are missing.

There are many examples of bad organization.26 One is the failure to divide a long statute into parts, thereby forcing the reader to hunt through the entire statute in search of the relevant provisions. Another example, which is typical of the U.K. tradition, is the use of schedules. For example, there are 31 schedules to the Income and Corporation Taxes Act 1988 (U.K.). While it may seem commendable to relegate detailed provisions to a schedule in order to make the statute easier to read, the result is simply bad organization.27 Detailed provisions should either be in the appropriate place in the statute, or, if they are elaborations of general statutory rules, could be placed in regulations that are subordinate to those rules. The failure to integrate schedules with the statute not only makes the statute more difficult to follow, but also tends to undermine its logical integrity. A tightly drafted tax statute contains many explicit or implicit28 cross-references. The logical interrelations among its provisions are intricate. If some of these are removed to a schedule, there is a danger that they will not be adequately integrated into the logical structure of the rest of the statute, particularly once there are amendments.

B. Use of a Code

A number of countries have organized their tax laws into a single code.29 The use of a code facilitates the elimination of duplicative provisions. For example, without a code, definitions or administrative provisions might be repeated in separate laws or, even worse, might differ in two different tax laws because of historical accident.30 Consolidating common provisions into a code facilitates their rationalization because it forces one to think about what the general rule should be. Putting all the tax laws into one code also facilitates compliance, because taxpayers know that they have all the tax laws in front of them when dealing with a particular problem. In the absence of a code, people can waste time searching for tax laws in an effort to ensure that they have a complete set. This function of a code—gathering all the tax laws into one document—is an important benefit and argues in favor of using a code if at all possible.

Organization of all the tax laws into a single code is consistent with the civil law tradition. Some scholars in this tradition hold that only the general rules of taxation should be embodied in a code, with the more specific and ephemeral rules contained in specific tax laws, which can be expected to be changed more frequently.31 Whatever is included in a code, in the tax area the use of codes does not correlate with whether a country has a civil or common law system. While France has a tax code, many other civil law countries do not. And even though the United States is a common law country, it does have a code.

C. Organization of Tax Laws in the Absence of a Code

The above considerations suggest that it is desirable to place all tax laws in a code. However, this might not be possible in a particular country because it would require consensus on a substantial legislative project of codification. When a code is not used, it is possible, albeit with some effort and discipline, to achieve virtually the same result by carefully organizing the separate tax laws, making sure they fit together properly, using cross-references where appropriate to eliminate duplicative provisions, and collecting provisions of general application into one law. In many countries, the tax laws are well organized, despite not being formally embodied in a code. For example, in Germany, the Abgabenordnung (Fiscal Statute) contains many of the provisions that apply to the tax laws generally; a number of other countries follow a similar approach of placing general rules into one law.32 This avoids repetitive or inconsistent rules for different taxes.

Where all the tax laws are contained in one code, amendments are automatically consolidated into it, since they take the form of adding sections to it or repealing or replacing the language to be changed. This type of amendment is called a “textual amendment” because it is an amendment to the text of the previous law. It is desirable to make amendments in this manner, because a series of nontextual amendments makes it difficult to ascertain precisely what the law is, often necessitating a tedious task of ex post consolidation of amendments.33 One advantage of a code is that it encourages the legislature to make textual amendments.34

There is no hard and fast rule as to how many tax laws there should be. The same arguments that favor a code also suggest that the fewer tax laws the better, although one can in principle achieve close to the same result with more tax laws, as long as they are carefully coordinated. For example, Germany has two income tax laws, one for corporations and one for individuals, but the corporate income tax law is much shorter and incorporates much of the individual income tax law by reference.35 In practice, however, such coordination is difficult to achieve with a multiplicity of laws; coordination is more likely to occur if several tax laws are merged into one.

Particularly problematic is the inclusion of provisions relating to a particular tax in more than one piece of legislation, together with nontax provisions, as often happens, for example, when tax provisions are contained in foreign investment laws or in laws designed to regulate particular industries. The interaction of the various rules for a particular tax is apt to be neglected when they are spread over more than one law.

IV. Effectiveness

The fundamental test of whether a tax law is drafted properly is if it implements the desired policy in an effective manner. In trying to make sure during the drafting process that the law will be effective, it is necessary to reflect on the policy and on the anticipated implementation of the law, including its interpretation in regulations and by the courts, and on how taxpayers and tax administrators will act in applying the law.

A. Relation Between Policy and Drafting

To properly manage the drafting of tax legislation, it is important to understand the relation between policy formulation and drafting. It is, of course, necessary to make some tentative general decisions about the policy to be implemented before sitting down to draft specific legislative language. Yet in substantial ways policy does not precede drafting, the two being developed concurrently. In the first place, policy shifts as the political process of producing a tax bill unfolds. Initial responsibility for producing a draft bill might lie in a department of the ministry of finance. Often, governmental process calls for producing a complete draft even before the minister makes certain policy decisions. Changes in policy may then be made at several stages, as a tax bill undergoes consideration by the government and then by the legislature. Policy decisions can be changed up to the last point when a bill is finally adopted as law.

Second, it is impossible to decide whether a policy is wise without considering the text of the bill. The drafting process involves a constant refinement of the policy decisions. This is because drafting forces the policy to be specified more and more precisely. As this specification takes place through the consideration of tentative legislative language, numerous questions arise for the consideration of those who are responsible for the setting of policy. If certain tentative policy decisions are made before drafting begins, the drafter must thoroughly understand not only what those decisions are, but the reasons behind them. Only by fully understanding the policy choices and the reasons for making them can the drafter propose legislative language to accomplish the policy.

In the end, there is only the language of the law. Policy may still exist, in the sense of what various individuals may intend or hope for the bill to accomplish, but the state of mind of various individuals does not become the law. Tax policy in an objective sense subsists only in the language of the law. Therefore, the drafting process can be seen as the development of tax policy, which is inchoate at the beginning of the drafting process and fully realized only at the end of the process. It would therefore be more accurate to say that while tentative decisions as to the general direction of a draft bill must be made before specific language can be drafted, at that point the language of the draft and the policy behind it typically proceed hand in hand through the process until the final language is adopted.

B. Anticipating Application and Interpretation

During the drafting process, consideration should be constantly given as to whether the statute is complete. To be effective, the statute should set forth all the rules needed to determine tax liability, or should provide authority for regulations that will contain these rules. To achieve this, the drafters must try to imagine all possible situations in which the statute will be applied. As part of this exercise, the drafters would do well to consult the accumulated experience of other countries and identify the issues that have come up in applying particular kinds of provisions. A choice must be made as to (1) whether rules go into the statute or into regulations36 and (2) what level of detail is appropriate. Both choices may appropriately be made differently in different countries and for different kinds of issues.

Another aspect of thinking about how the law will be interpreted and applied is being on the lookout for ambiguity. Unnecessary ambiguity should be eliminated, although given that language is inherently ambiguous, it is impossible to eliminate all ambiguity. It is appropriate to take a practical view here and to eliminate ambiguity that would be of concern to a judge attempting to interpret the tax law. In some cases, a degree of ambiguity is desirable, because it provides flexibility for the tax administration to respond to unanticipated cases. For example, in drafting an income tax, one could list all the types of deductible business expenses. This would provide certainty, but not enough flexibility. It is better to provide a general rule, such as that all expenses incurred in the realization of income subject to tax are deductible, with specified limitations. Such a rule involves some ambiguity, but on balance is preferable to a rule that attempts to list all allowable deductions in that it accomplishes the goal of taxing net income.

More generally, in drafting it is important to anticipate the administrative or judicial resolution of disputes between taxpayers and the tax authorities. For example, suppose that it is decided to allow an income tax deduction for business entertainment expenses only if the expenses are reasonable in amount. If the statute is drafted in these terms, the drafters should consider who is going to decide whether an expense is reasonable. Will this be determined according to guidelines provided by regulations, will it be left to the judgment of individual auditors, or will it be left to the courts? If the drafters focus on these questions of procedure, alternatives for how to draft the statute might occur to them. For example, instead of using a concept of reasonableness, the statute could deny deductions for entertainment in excess of specified limits or could deny a fixed percentage of entertainment deductions or could deny this deduction altogether. Each of these alternatives is progressively simpler from the point of view of tax administration and progressively harsher for taxpayers. A policy choice must therefore be made. This is another example of how policy choices are generated and facilitated by the drafting process.

C. Drafting for a Judicial Audience

Just as any piece of writing is modified to cater to its audience, the manner in which laws are drafted should take into account how courts are expected to interpret them.37 For example, the Renton Committee distinguished in general terms between the civil and common law systems.38 It found that under the civil law system, legislation tends to be drafted in the form of broad statements of principle, with the application of these principles to particular cases being left to the judgment of the court. Classic civil law drafting practices “a deliberate restraint in the proliferation of detailed rules.”39 In contrast, common law drafting tends to be much more detailed, trying to cover each possible case, with the court taking a correspondingly narrower reading of particular provisions of a statute.

Of course, these characterizations of the judicial and corresponding statutory style in the civil and common law systems are only ideal types, and practice in particular countries and with respect to particular types of legislation may vary.40 Tax laws in civil law countries are often rather detailed. Moreover, the level of detail can vary depending on the legislature’s attitude about delegating its lawmaking authority. In any legal tradition, the tax law itself can be drafted in very broad terms, as long as there is a broad delegation of authority to issue detailed regulations. Whether a legislature wishes to do this is often a political question, and also depends on tradition and the framework of administrative law.41

In addition to the question of the level of detail of a statute, in common law jurisdictions, it is important to be aware of judicial decisions on taxation, as many important principles are governed by a “common law” of taxation.42 This is less likely to be the case in civil law countries, or when the legislature has made an attempt to codify the tax laws. Even in a country like the United States, which has a code, an extensive common law of taxation has grown up under the guise of interpreting the provisions of the statute. Congress can always override judicial decisions, but U.S. courts tend to stick to the doctrines they have developed absent a clear congressional statement that they must be abandoned in a particular area.

The interpretation of law by courts can itself be governed by rules set forth in legislation. Commonwealth countries have developed a tradition of interpretation acts, which provide definitions of commonly used terms and may contain other clauses relating to the interpretation of laws.43 In the United States, similar provisions are found in Title I of the U.S. Code. An analogy in civil law countries is found in the portion of the civil code containing general provisions on the application of laws, although these tend to be less detailed than the interpretation acts of the Commonwealth.44 With respect to tax legislation in particular, guidelines of interpretation for the courts are sometimes included in the tax laws.45

D. Relation Between Statute, Regulations, and Other Explanatory Material

Because tax legislation is often difficult to understand, new tax laws are often accompanied by explanatory documents of various kinds, which provide legislators, tax officials, and taxpayers with an understanding of their purpose and intended operation. To ensure effectiveness, statutes should be drafted with a view to what will go into these documents. It is not usually considered appropriate to try to provide all the necessary details of tax legislation in the statute. To do so would make the statute unduly lengthy and difficult to understand. Moreover, because one cannot foresee all the situations in which tax laws will be applied, all the details cannot be worked out at the time the statute is enacted. Finally, even if the drafters of the statute had in front of them the detailed rules needed to implement the statute, they might choose to leave these rules to be promulgated by the administrative branch, since administrative rules can be modified more easily than the statute.

Categories of explanatory materials may be called different things in different countries, but here are some examples:

article image

These different documents may all be helpful to taxpayers and tax administrators in understanding the law, but they differ in their legal effect. Some have the force of law, some have persuasive authority, some have little binding legal effect.46

There is no hard and fast rule as to which provisions should go into the law and which into the regulations. Which provisions are viewed as essential ones that must go into the law depends on the practice in the particular country and on politics—how much power over detail the legislature is willing to delegate. There is also the problem of time: legislative drafting is a laborious exercise, and there is a limit to how much detail can be drafted within the time frame for enactment. Neither can one easily prescribe in advance how much total legislative and regulatory text there should be to give guidance to taxpayers without smothering them in detail. It is usually best to expand the mass of regulations little by little as needed. Leaving matters to regulations can also be a political tactic; it may be difficult to reach consensus on particular points, and leaving these points to regulations can facilitate passage of a bill.

There are alternatives to issuing detailed regulations. One alternative is to provide no written rules governing details, allowing the broad principles of the statute to speak for themselves. Another is to provide that certain rules of the statute apply only where the tax authorities have given their approval in the particular case. This is a useful technique in the case of rules that govern what is expected to be a small number of cases. Instead of spelling out the rules for these cases in advance, it may be easier to proceed on an ad hoc basis.

Another alternative to regulations is for the tax administration to issue commentaries on the law. These can take varied forms. For example, in the United States, the Internal Revenue Service issues revenue rulings, dealing with the application of the law in certain situations (to be distinguished from rulings issued to particular taxpayers). Revenue Canada issues interpretation bulletins. The tax authorities of most countries issue instructions on how to fill out tax forms. In practice, these may be the only material consulted by the majority of taxpayers. While their legal significance may be minimal, their practical importance cannot be overstated. The U.K. tax authorities issue “extra-statutory concessions,” explanatory booklets, and statements of practice.47 The French tax administration publishes a book called Précis de fiscalité, which is a treatise nearing 2,000 pages in length covering all the rules of taxation.48 These commentaries have varied legal effect; they are often binding on the tax authorities, but not on the taxpayer. Even if such administrative interpretations are not legally binding on taxpayers, for all practical purposes, if they are not directly contrary to the statute, taxpayers will often follow them.

Advance guidance on interpretation of the statute can also be provided in documents that are issued contemporaneously with consideration of the legislation, for example, in the form of an explanatory memorandum submitted by the government or in the form of a committee report (i.e., the report of the legislative committee considering the bill). The extent to which the courts will consider legislative history in construing the bill differs in different legal systems.49 In U.K.-based systems, the matter can be dealt with in an acts interpretation act50 or in the statute itself.51 As a practical matter, legislative history can play an important role, even if judicial doctrine does not assign it much legal force. Therefore, consideration should be given to preparing a detailed explanatory memorandum to accompany tax legislation, if acceptable as a matter of local practice.

It is important to be aware of the country’s administrative law and practice with respect to delegated legislation. While we have reviewed in general terms the practice of several countries, there are many variations, and each country has its unique practices. These may pose real obstacles to what can be done in regulations. For example, there may be a practice or requirement in the law that regulations have to be issued contemporaneously with passage of the legislation (or within a specified period of time). The drafter should become aware of any such constraints in advance. When it is important for a rule to have binding legal force, then it may be necessary to include it in the statute if administrative practice does not readily allow it to be included in another legally binding norm. Inclusion of rules in the statute is also necessary if the statute is to operate at once, before regulations can be issued.

Attention should also be paid to providing in the statute for regulatory authority, which depends on the country’s administrative law. Often a tax statute will contain a general authority for regulations. Even when this authority is provided, there are often additional grants of authority to write regulations to implement particular provisions of the statute. Multiplication of authority for regulations should be avoided on grounds of simplicity, but there is some excuse for special grants of authority when regulations that are legislative in character are called for, that is, where the regulations are providing rules out of whole cloth rather than filling in the details of rules provided in the statute. The distinction between legislative and interpretative regulations is not, however, always clear.

Finally, because regulations are not issued all at once, there is a problem of organization. The text of regulations or administrative commentaries is often longer than that of the law. This makes it all the more important to arrange them logically, so that the reader can immediately turn to the relevant portions. Where the law is logically arranged, it makes sense to arrange the regulations in the same way. The method adopted by the United States, whereby the regulations are named according to the section numbers of the statute, makes it easy to find the regulations that correspond to any particular statutory text. The approach of the Précis de fiscalité—a treatise summarizing the rules for all the taxes in France—is also user-friendly (it is well organized and has an extensive table of contents and index).

V. Integration

It is important to ensure that a new tax law is fully integrated with the rest of the legal system. Not only does the drafter need to be aware of the obvious issues of possible unconstitutionality, but also more subtle questions of conformity with local drafting style and language as well as the legal system in general need to be considered to enhance the acceptability, understandability, and effectiveness of the law.

A. Local Drafting Style

Apart from general principles of good drafting, tax statutes must be drafted in the context of the style of legislative drafting in the country in question. Different countries have developed their own drafting practices. The officials of some countries may be willing to make changes if convinced that the result would be a better statute. Others may be wedded to their traditions and reluctant to change even if the result would be more efficient or more elegant. There is much to be said for tradition in drafting style. A country’s laws should be consistent in appearance and style so as to facilitate understanding and interpretation of the laws and maintain the dignity of the legislative process. Therefore, to draft tax laws differently, the officials responsible for legislative drafting in a country may have to make a more universal change, which might require some persuasion. Those responsible for drafting tax laws must do their best within these constraints. Unless a country’s officials decide to make a change in their drafting style, the drafter must follow that style.52

B. Gender-Neutral Language

In recent years, there has been increasing awareness of the desirability of using language that does not reflect discrimination based on gender. The issue depends on the grammatical peculiarities of the language in question, as well as on the evolution of the language and its culture. In English, it has become common to avoid exclusive use of the masculine gender to refer to an antecedent of indefinite gender and to avoid nouns denoting a particular gender where an indefinite gender is intended. To the extent called for by the language and culture of the country in question, the drafter should take care that usage of words is precise and nondiscriminatory.

C. Relation Between Tax Law and Other Legislation

The context of nontax regulatory and private law is also important for the drafting of tax laws. The tax law must often refer to provisions of economic law, such as the definition of different types of business organization. Accounting norms and principles found outside the tax laws can be critical to their operation.53 The tax law is fundamentally based on legal relations determined by nontax legislation, primarily private law.54 If this legislation is nonexistent or is in a confused state, it is difficult to draft a good tax law.55

Some nontax regulatory legislation is important for the effective functioning of the tax laws, for example, legislation requiring the registration of company shares and other securities (i.e., prohibiting issuance in bearer form), legislation limiting the scope of bank secrecy, legislation governing the integrity of the civil service, and legislation providing effective civil and criminal procedure.

There is a tendency, particularly in civil law countries, for the law to be seen as a consistent whole. Concepts defined by the civil law therefore need not necessarily be redefined in the tax law. One problem arises because a term may not have an unambiguous meaning in the civil law;56 problems can also arise—whether in a civil or common law system—when a term is defined with reference to its meaning in another statute.57 There may, however, be a good reason for the meaning of a term for tax purposes to differ from the ordinary meaning. For example, the concept of “employee” may be well defined in the labor law, but the tax definition of an employee may appropriately be broader.58 Similarly, the tax law may tax as a separate entity an economic unit that does not have independent legal personality under the civil law, or conversely may provide for flow-through treatment of an entity that is a separate person under the civil code. In some cases it is also necessary to disregard the civil law forms chosen by the parties in order to minimize their tax liability.59 Where it is desired in the tax law to use a term with a broader meaning than in the civil law, there are two alternatives. One is to use the same term as that used in civil law, but to provide a special definition in tax law. For example, the term “employee” can be used in the income tax law, but defined to include persons who are not employees under the civil law. This approach can, however, be confusing, particularly to someone who does not read all the definitions carefully. An alternative is to use a different term in the tax law, where the intended meaning is different from that in the civil law. The disadvantage of this, however, can be that the term used might sound artificial or be cumbersome. Neither approach may be fully satisfactory.

D. Specific Problems of Terminology

Certain terms whose meaning is defined in the civil law are used pervasively in tax laws of different kinds and must be used or defined with care. Some examples follow.

1. Legal Person

In most civil law countries, business entities (companies, partnerships of various kinds) are generally legal persons. In countries such as France, Spain, and those with similar civil codes, the various forms of sociétés or sociedades generally are legal persons. There are, however, some business arrangements that do not give rise to a legal person and that are characterized by a splitting of the income from the business venture among the parties.60

The situation is different in Germany. Under the German civil code, Personengesellschaften (partnerships of persons) are not legal persons, while Kapitalgesellschaften (capital companies) are legal persons. This distinction in fact may not make very much practical difference for purposes of the civil law for reasons that need not be gone into here.61 But it means that given the formal distinction, and the fact that Personengesellschaften include important forms of commercial partnerships, references in the tax law to “legal persons” will not include Kommanditgesellschaften (limited partnerships) or Offene Handelsgesellschaften (general partnerships).62 In Germany and other countries where partnerships are not legal persons,63 it is necessary, where appropriate, to define taxpayers as including legal persons and certain other entities that are not legal persons.

In some countries, the status of an entity for tax purposes is determined not by the civil law but by the tax law. Thus, in the United States, whether an entity is treated as a corporation is determined by rules under the tax code.64 A similar problem comes up when the status of a foreign entity is in question. Usually, the determination is made not according to whether the entity has the status of a legal person under the law of its home jurisdiction, but according to what its status would be in the jurisdiction in question. This should be specified.

2. Employee

Whether an individual has the status of employee or independent contractor can have importance for tax purposes. In common law countries, the distinction is typically made according to the criteria of common law. This looks to the degree of control that the employer has. In civil law countries, the determination is made according to the status of the relationship generally determined under the labor code.65 In both civil and common law jurisdictions, employees will not include directors of companies, public officials, and certain other persons whom one would wish to treat as employees for tax purposes. These should also be defined as employees for purposes of the tax law.66

3. Property

Legal systems differ in their concepts and classification of property.67 Usually, different kinds of property do not need to be defined separately in the tax laws because their meaning will be given in the civil law. But care should be taken. “Real” property (in common law jurisdictions) has a similar but not identical meaning to “immovable” property in civil law jurisdictions. Some civil codes have peculiarities. For example, the Russian civil code defines immovable property as including airplanes and businesses.68 In such cases, it may be necessary to separately define immovable property in the tax laws as excluding this type of property.69 Similar care should be taken in using concepts such as tangible property, intangible property, and fixed assets. In countries with codified accounting norms, categories of assets and liabilities relevant for tax purposes are often defined in these norms.

E. Use of Models

Recent years have seen the publication of the Basic World Tax Code, authored by two American lawyers who have also worked extensively abroad, and the Draft of a Tax Code for Middle and Eastern European States, authored by a German tax professor and commissioned by the German Ministry of Finance.70 A model tax code (general tax law) was published in the 1960s, intended primarily for Latin American countries.71 Other unpublished model tax laws are in various stages of development. The IMF’s Legal Department has prepared a number of draft model laws for use in its technical assistance work; these are revised on an ongoing basis as experience with them suggests improvements. Some of this model legislation is geared to a particular country or legal or linguistic tradition; some is intended to be used on a wider basis.

Model legislation is extremely useful as a starting point for drafting. Given the complexity of tax legislation and the wealth of experience with tax laws in many countries, it would be foolish for a drafter to attempt to reinvent the wheel. On the other hand, the complexity of the laws of countries with well-developed tax systems is so great that it is inappropriate to use them as a model without a considerable degree of pruning and revision. The various model tax laws tend to be derived from the legislation of particular countries, although a considerable amount of distillation may have taken place.

Proper use of a model in drafting legislation for a specific country involves the realization that considerable adaptation, if not wholesale revision, of the language of the model will likely be required in order to meet the particular needs of the country in question. A model can only be a starting point. As long as the limitations of any model are borne in mind, a model can be a useful, even essential, tool in drafting.

Note: The author is grateful for comments from Lloyd Ator, Susan Himes, Ward Hussey, and Bertil Wiman, and to Rick Krever, who contributed to the writing of sec. IV(D).


For more general treatments, see Lawrence E. Filson, The Legislative Drafter’s Desk Reference (1992); Robert J. Martineau, Drafting Legislation and Rules in Plain English (1991); David Renton, The Preparation of Legislation: Report of a Committee Appointed by the Lord President of the Council, 1975, Cmnd 6053; G.C. Thornton, Legislative Drafting (3rd ed. 1987); Elmer A. Driedger, The Composition of Legislation (1957); Reed Dickerson, Legislative Drafting (1977); S. Namasivayam, The Drafting of Legislation (1967)(U.K.-style drafting, with particular emphasis on Ghana and Ceylon); Louis Philippe Pigeon, Drafting and Interpreting Legislation (1988)(focus on Canada and Quebec in particular, but much of discussion is of general interest); Law-Making and Development: Formulating Policies and Drafting Legislation (Seyoum Haregot ed., 1987) (collection of essays on the role of legislation, legislative drafting, and related topics, with special emphasis on developing countries); V.C.R.A.C. Crabbe, Legislative Drafting (1993) (focus on Commonwealth practice; appendix contains Canadian guidelines on preparation of legislation); William Dale, Legislative Drafting: A New Approach (1977) (comparative study of methods in France, Germany, Sweden, and the United Kingdom). The best (and funniest) short piece on drafting I have read is Ward Hussey, Homily on Drafting Style (as yet unpublished).


For example, discussion of the drafting process and legislative process in general in ch. 1, the legal framework in ch. 2, and drafting problems that relate to specific taxes throughout the book.


Driedger, supra note 1, at xxii.


See, e.g., 16 U.S.C. § 1531 (b) (USA) (purposes of Endangered Species Act). See generally Dickerson, supra note 1, at 107–108; Renton, supra note 1, at 30, 62–63.


See vol. 2, ch. 14 (discussion of different ways of taxing fringe benefits).


See Adrian Sawyer, New Zealand Introduces New Core Provisions Bill, 12 Tax Notes Int’l 539 (1996).


See Martineau, supra note 1, at 90–92.


See infra sec. V(C).


But see Martineau, supra note 1, at 91 (legislation should be drafted so that it can be understood by a person of average intelligence). The taxpayer’s contact with the law itself may be minimal. Most people will rely on explanations of the law issued by the tax authorities, and considerable care should go into writing such explanations. These should be written in language understandable by laypersons. See also Renton, supra note 1, at 112–13.


“Section” is used in this chapter to refer to the basic unit of a statute. In some countries, particularly in the civil law tradition, “article” or “paragraph” is used.


See, e.g., Thornton, supra note 1, at 59–60.


For example, the following articles exist in the FRA CGI: art. 235 ter EA (which comes after art. 235 ter E), and art. 235 ter H quater (which, obviously, comes right before art. 235 ter HA, which would, presumably, if it had been enacted after art. 235 ter H quater, have been called art. 235 ter H quinquies). So far, the Parliament has not found the need to insert an article between art. 235 ter HA and art. 235 ter HB, or, what would be even worse, an article between art. 235 ter H ter and art. 235 ter H quater.


It is not a perfect solution, in the sense that assigning the unused article numbers to new articles might place the new articles in an order that is not quite logical. Often, however, new articles will be appropriately located in the places where gaps are left, since they will be in the nature of special rules, which are appropriately placed at the end of a group of related articles.


In the U.K. tradition, the term used is marginal notes, and these are, as the name implies, placed in the margin, except in New Zealand and Canada, where they follow the section number as a heading. See, e.g., CAN ITA. See also Thornton, supra note 1, at 125–28.


E.g., France, Germany, and Spain. In such countries, however, parts of a statute (they may be called title, part, chapter, and the like) typically do have headings. E.g., FRA CGI.


For example, section headings are used in many of the recently adopted tax laws of Estonia. See, e.g., EST LOT; EST IT; EST VAT; EST LND; EST GAM. Section headings are also used in some of the Russian and Latvian tax laws, as well as in the new Kazak code. See, e.g., RUS IT; RUS PT; LVA TF; LVA EIT; LVA ET; KAZ TC.


See Thornton, supra note 1, at 123–24, 127.


See Renton, supra note 1, at 64; Thornton, supra note 1, at 57. This tradition is followed in various countries of the Commonwealth, such as Canada and Australia.


Run-on sentences can be avoided by dividing them into several subsections. However, this leads to less-than-ideal clarity of organization of sections, since sections are as a consequence broken down into too many subsections. For example, it might be ideal to divide a particular section into three subsections if the section contains three main thoughts. Each of these subsections might consist of two sentences. But if this is prohibited, then the alternative is to divide the section into six subsections of one sentence each. In this case, the benefits of breaking out the section into its three main thoughts are lost. An example of a statute drafted under the one sentence rule, but where the sentences have been kept fairly short, is LSO IT.


See Renton, supra note 1, at 64–65; Thornton, supra note 1, at 57.


LSO IT § 24.


For example, in the Internal Revenue Code (and more generally, in the U.S. Code), sections are divided into subsections (designated by lowercase letters), then into paragraphs (Arabic numbers), subparagraphs (uppercase letters), clauses (lowercase Roman numerals), and subclauses (uppercase Roman numbers). Correspondingly, there is a uniform designation of groupings of sections in the U.S. Code. The U.S. Code is divided into titles, subtitles, chapters, subchapters, parts, and subparts.


E.g., FRA CGI art. 223 J.


E.g., FRA CGI art. 238 bis HA.


E.g., FRA CGI art. 223 L.




But see Renton, supra note 1, at 68–69. There may be political or parliamentary reasons for the use of schedules, where they can be changed by a process different from statutory amendment; this makes their use understandable, but does not remove the criticism that they make the statute more difficult to follow.


”Implicit cross-reference” means the use of a term whose meaning is specified elsewhere in the statute.


For example, Cameroon, Colombia, Cote d’Ivoire, France, Gabon, Kazakstan, Madagascar, Mali, Togo, and the United States. Even in countries where a code is used, there are some tax provisions, usually of very narrow application, that have not been included in the code. For example, in the United States, provisions that allow a deduction from taxable income for federal income tax purposes for amounts deposited in a capital construction fund for merchant marine vessels are contained in the Merchant Marine Act, 46 U.S.C. § 1161, not in the Internal Revenue Code.


Use of a code does not, however, guarantee that duplicative provisions will be eliminated. For example, the Internal Revenue Code contains numerous definitions of “related person,” with little policy justification for such multiplicity. It is not enough to put the tax laws into a code; one must also “think code,” in the sense of consolidating detailed provisions into a smaller number of more general rules.


See José M. Martín & Guillermo F. Rodriguez, Derecho Tributario General (1986). This approach is followed by a number of Latin American countries. See infra note 32; 1 Carlos Fonrouge, Derecho Financiero 41, 48–63 (Susana Navarrine & Rubén Asorey, rev. 1993).




See Renton, supra note 1, at 32, 76–84. The Renton Committee considered it desirable to proceed by textual amendment wherever possible, but found that “there will be many circumstances in which the amendment of fiscal legislation by the textual amendment method will not be practicable.” Id. at 117. It is, however, difficult to see what these circumstances are. Most amendments to tax law in the United States and many other countries are textual, so it is evidently possible to proceed this way. The failure to do so ultimately leads to a mess. The United Kingdom has been moving toward a more consolidated approach in recent years. For example, most of the laws relating to income taxation have been consolidated into the Income and Corporation Taxes Act 1988, and many amendments are now being made as amendments to specific sections of this act.


Again, the use of a code does not guarantee this result. For example, in the United States some tax provisions are “off code”; that is, they are considered to be of such narrow application as not to be of general interest.


See, e.g., DEU KStG §§ 7, 8. A similar approach is followed in several other countries. See also NLD Vpb.; ESP IS.


See infra sec. IV(D).


See Renton, supra note 1, at 51–55.


Id. at 51.


The Renton Committee provides a historical perspective on statutory detail in England, with the older statutes in laconic Latin ceding in the Middle Ages to verbosity, perhaps owing to the use of conveyancers to draft legislation. See id. at 5.


See infra sec. IV(D).


See Barry Pinson, Pinson on Revenue Law 3 (1981).


See Thornton, supra note 1, at 87–93.


See, e.g., Code civil arts. 1–6 (FRA).


See supra ch. 2, sec. III.


See supra ch. 2, sec. IV.


See B. Pinson, supra note 42, at 3.


Statement in text is based on the 1994 edition, which is approximately 2,000 pages long. It is not binding on the tax administration. See supra ch. 2, note 207.


See supra ch. 2, sec. III.


For example, sec. 19 of the Interpretation Act of Ghana provided “for the purpose of ascertaining the mischief and defect which an enactment was made to cure and as an aid to the construction of the enactment a court may have regard to … any memorandum published by authority in reference to the enactment or to the Bill, …” quoted in Namasivayam, supra note 1, at 2.


E.g., LSO IT § 3(2). (“In interpreting this Act, regard should be had to the Explanatory Memorandum accompanying the Act.”)


See Martineau, supra note 1, at 121. In some countries, there is no uniform drafting style; in this case, it may be possible to justify a more modern approach by finding local precedents or by showing that the local style admits to variation.


See vol. 2, ch. 16, appendix.


Mostly in civil law countries, law is generally classified as either public law or private law, the former having to do with the state and the latter governing relations among persons (such as property and family relationships). Private law would include what in common law systems is known as the law of contracts, torts, property, and family law. For a discussion of the distinction between public and private law, see 2 International Encyclopedia of Comparative Law, ch. 2, Structure and Divisions of the Law (R. David ed.). For a discussion of the relation between private law and tax law, see Sture Bergström, Private Law and Tax Law, 23 Scandinavian Studies in Law 31 (1979); 1 Klaus Tipke, Die Steuerrechtsordnung 89–104 (1993).


For example, in several countries of the former Soviet Union, as well as in countries of Eastern Europe, it has not been clear whether certain enterprises are legal persons under civil law. Tax legislation often uses the term “legal person” in defining taxpayers, and any uncertainty about the meaning of the term can therefore lead to difficulties in the implementation of the law. Before the split-up of the Soviet Union, the uncertainty arose in discussions about whether an “enterprise” was of necessity a legal person. See M.S. Braginskii, Legal Regulation of Entrepreneurship in the Russian Federation, 19 Rev. Central and East Eur. L. 365, 377 (1993). Laws on entrepreneurship were passed with different wordings in various countries. For example, in Latvia the Law on Entrepreneurial Activity (1990) reprinted in Joint Publication Research Service, Regional Economic Issues JPRS-UEA-90-043 (Dec. 11, 1990), art. 7, states that “the legal capacity of an enterprise arises at the moment of its registration....” This suggests to some that an enterprise is necessarily a legal person. But the Latvian Law on Partnerships, art. 1, states “a partnership is not a legal entity.” The Law on Enterprises in the Republic of Kazakstan (1991, as amended in 1993), reprinted in Foreign Investment and Privatisation in Kazakhstan: Collected Legislation 103 (W.E. Butler trans., 1993), art. 1, states “an enterprise is an autonomous subject with the rights of a juridical person....” And further, in art. 5(2): “An enterprise shall be considered to be created and shall acquire the rights of a juridical person from the date of the State registration thereof.” Under provisions such as these, the status of partnerships, associations, sole proprietorships, and branches of foreign companies can be unclear or subject to doubt. (The situation may have been clarified in Kazakstan with the adoption of a new civil code in late 1994.) Under such circumstances, it is irresponsible for a drafter of tax legislation to use the term “legal person” as if it were perfectly clear which type of enterprise is a juridical person and which is not. Instead, it may be necessary to fashion a definition of this term, or an alternative term, such as “enterprise,” that will apply for tax purposes, pending resolution of the uncertainties in the civil law.


See, e.g., Bergström, supra note 54.


See Thornton, supra note 1, at 110–16.


See infra sec. V(D)(2).


See supra ch. 2, sec. III(I).


See, e.g., Code civil art. 1871 (FRA) (société en participation).


See Handelsgesetzbuch arts. 124, 161 (DEU).


See DEU KStG § l(l)(4).


See vol. 2, ch. 21.


See USA IRC § 7701.


See, e.g., Code du travail art. L 121-1 (FRA).


See, e.g., FRA CGI arts. 80, 80 ter. The issue is discussed in greater detail infra ch. 11 and in vol. 2, ch. 14.


For a historical introduction, see Boudewijn Bouckaert, What Is Property? 13 Harv. J. L. & Publ. Pol’y 775 (1990); 1 Vinding Kruse, The Right of Property (1939).


Civil code arts. 130(1), 132 (RUS); civil code arts. 117, 119 (KAZ).


E.g., KAZ TC art. 5(16).


Ward Hussey & Donald Lubick, Basic World Tax Code and Commentary: 1996 Edition (1995); Joachim Lang, Entwurf eines Steuergesetzbuchs für mittel- und osteuropäische Staaten (Bundesministerium der Finanzen 1992).


Organization of American States, Modelo de Código Tributario (1967). In addition, the Inter-American Center of Tax Administrators has planned to publish a model tax code in 1996.