The Ruler should act like a bee which collects honey without causing pain to the plant. —Mahabharata

Tax administration law covers an enormous number of issues. An essay that attempted to cover each issue in detail would run on for volumes, rather than pages. To keep the discussion at a reasonable length, this chapter offers only an introduction to some of the issues involved. Where there is little theoretical controversy, the chapter only outlines the issues and provides discussion of a basic nature.

The result is a chapter that, like Gaul, is divided into three parts. The first part consists of a relatively general discussion of the nature of the law of tax administration and procedure and how it relates to other laws.

The second part consists primarily of a checklist of those elements that should be included in a tax administration law, except for matters concerning tax compliance, which are covered in the third part. The discussion in the second part is more limited, primarily because, at least concerning most of these points, there is little debate of a theoretical nature that would profit by additional exposition here. This checklist offers an introduction to the fundamental mechanics of the law for those who might not be completely familiar with tax procedure.

The third part discusses compliance. It begins by discussing how substantive tax laws should be drafted to further compliance goals and then addresses the question of taxpayer sanctions, how they work, and what is required to make them effective. Much of the current research into sanctions has suggested some counterintuitive conclusions, which are often largely at odds with sanctions extant in many tax laws. For this reason, this part includes a relatively detailed discussion of these issues.

I. Structure of Tax Administration Law

A. Organizing Principles of Tax Administration Law

It is a frequently heard complaint that tax administration laws are complex, confusing, and arbitrary. To some degree this is probably unavoidable. Administration of a tax system must cover an enormous and diverse number of rules. Unlike the substantive laws of taxation, there is no basic “principle” of administration. In contrast, an income tax law, a property tax law, or a value-added tax (VAT) law each has unifying themes. While each may include some complex definitions, simplifications, or exceptions to these themes, there are at least a limited number of principles, typically related, around which the law can be structured and to which both policy analysts and drafters can return when creating the law.

In contrast, it is not easy to encapsule a few guiding themes for a tax administration law. There are a number of very broad principles that should apply in each administrative rule, such as fairness and efficiency, but these do not really help much in guiding substantive design. In some ways, tax administration law is constituted by a hotchpotch of rules, some related, some not very closely related, some expressing clear policy, and some based rather largely on arbitrary considerations. This is due to the fact that tax administration law is first and foremost the elucidation of a bureaucracy. Public administration, in and of itself, is not easily determined by reference to a small number of principles. Instead, it is much more a question of designing acceptable answers for myriad practical bureaucratic problems.1

Nevertheless, some order can be brought to the organization of tax administration law. Three organizing principles can be identified: organization according to function, temporal organization, and organization by legal category. In combination, the three can make for a coherent legal structure that corresponds to the bureaucracy and procedure with which the tax administration law deals.

1. Functional Categories

It is not surprising that the most important way in which the better tax administration laws are organized mirrors the way in which tax authorities themselves are, or at least should be, organized. In other words, the laws are primarily organized around the different bureaucratic functions necessary for the administration of a tax system. For this reason, this essay will refer to this form of organization as “functional.”2 Sections in tax administration laws corresponding to functional categories cover regulations and rulings, record keeping and returns, audits and investigations, dispute settlement, recovery of monies owed to the government, internal investigations, and taxpayer ombudsperson.3

Functional organization of tax administration law makes it easier for the taxation authority, as well as for other government officials involved in the taxation process, to follow and interpret the law. Each department in the taxation authority can concentrate primarily on a single part or parts of the law.4 Of equal importance, while organizing by function may initially appear to reflect the world of tax administration from the viewpoint of the bureaucracy, so doing automatically reflects administration from the taxpayer’s perspective as well. The taxpayer’s involvement in each aspect of tax administration can largely be described by his or her interaction with different departments of the taxation authority. Therefore, organizing tax administration law in functional groups also helps the taxpayer better understand the rules and the process.

2. Temporal Organization

The quality of making the law easier to understand for the taxpayer as well as for the administration can be accentuated if the functional categories are themselves organized so that they follow in logical, temporal sequence. Temporal organization for the law of tax procedure makes sense because tax procedure inevitably follows a time sequence, given that each tax obligation is based on a tax period, with a subsequent possibility of the redetermination of tax, appeal of the redetermination, and ultimate determination of tax liability for the period. A temporal organization would mean that the law should begin with those rules concerning the elucidation of the law (regulations and rulings), followed by the incurring of a liability by a taxpayer (which would typically require the taxpayer to secure an identification number, keep records, and file returns), and then continue through each step through remittance of tax (or information), enduring an audit, disputing the assessment decision, and suffering a collection action. While not every taxpayer would be involved in each possible step in administration, both taxpayers and administrators would know to skip over intervening possibilities until the next relevant issue was reached. In the process, both would be reminded of those possibilities.

3. Legal Categories

Some rules, which represent a common legal category, cut across both functional and temporal lines. They cannot easily be organized on a temporal basis because they apply at a number of possible temporal steps. These are best placed together, usually at the beginning and the end of a tax administration law. Among the most important of these categories are definitions that apply to terms found generally throughout the law, the legal rights of taxpayers,5 penalties (both civil and criminal) for a taxpayer’s failure to comply with his or her obligations, penalties for a failure on the part of the administration to comply with its obligations, and interest (due both to government on underpayment and the taxpayer on overpayment).6

There is no organizational imperative to collect the relevant rules in such categories separate and apart from functional divisions. For example, every right, as well as each liability for any penalty or interest, whether relating to taxpayers or to the administration, with few exceptions arises only in the context of a rule described in a functional category. In the broadest sense, the rights of a taxpayer can be understood to include not only what is normally thought of as “rights” (e.g., the right to secrecy, the right to representation), but also essentially everything that is not specifically required of him or her (e.g., the right not to keep unnecessary records, or the right not to have a levy enforced on him or her for monies not lawfully due). Similarly, both penalties and interest, whether owed by taxpayer or government, can be understood to arise wherever among the rules in the functional categories the obligation arises (e.g., a penalty for failure to keep records, a penalty for requiring the taxpayer to keep unnecessary records, a penalty for failure to submit to a lawful levy, or a penalty for forcing submission to an unlawful one, and interest due on any underpayment or overpayment).

There are important benefits to collecting certain general definitions, taxpayer rights, interest, and penalties in separate categories. By and large, the rights, interest, and penalties included in separate categories are those that are broadly applicable throughout much or all of the administrative process. Therefore, rather than repeating each, it is easier to put them in one place and to make clear that they refer to more than one aspect of the administration law. There are also general rules applicable to all penalties (e.g., a reasonable cause exception) that can usefully be grouped with the specific penalty rules. Also, placing certain definitions, taxpayer rights, interest, and penalties into separate sections makes it more likely that the design and therefore the application of each will be more uniform.

B. Interrelation of Tax Administration Law with Other Laws

1. Nontax Law

Tax administration law is intimately connected with various laws (including the state constitution) not specific to taxation. For example, laws concerning the operation of the executive branch may affect the structure and function of the tax administration.7 Administrative law may affect how regulations and rulings are issued.8 The civil procedure code may have considerable relevance to numerous aspects of tax administration, including rights to notice of government action, rights to counsel during proceedings, procedures for dispute settlement in civil courts (including the application of civil fines and of appeals), and rules concerning the recovery of debts.9 Criminal procedure rules typically govern the application of criminal penalties in the tax area.10

As a general rule, unless there is a specific and compelling reason, it is probably best not to provide special rules only for tax matters. William of Occam, the great fourteenth-century English natural philosopher, admonished “[d]o not multiply entities unnecessarily.”11 To do so makes things more complicated than they need to be. Where rules relevant to tax procedure are contained in other laws, it may be beneficial to make cross-references to these laws in the text of the tax administration law. In certain cases it may be preferable to modify existing law to fit the unique problems inherent in tax administration. Where possible, it is probably best to make clear what nontax rule is being modified, and to limit the modification to the minimum necessary to effect the specific tax administration purpose.12

2. Substantive Tax Law

There is no clear line separating substantive tax law and tax administration law. There are a number of ways of drawing a line, however. One would be to include in the law of administration any rule that is primarily administrative in nature. Another would be to put into each substantive tax law the administrative rules that are peculiar to that tax, and to put in the general administrative law any rule that applies to more than one type of tax. Most jurisdictions apply a mix of both.13

3. Location of Tax Administration Law

The practice of countries differs greatly in terms of where the tax administration provisions are located. In some countries, each substantive tax law contains all the provisions necessary for its administration. In countries that organize all their tax laws into one code, the tax administration provisions can be one or more titles of this code. Yet other countries have what may be called a tax administration law or a general law on taxation. The tax administration provisions may also be contained in more than one law. For example, there may be a law on the tax system, which contains many of the general rules of procedure, and a law on the state tax service, which primarily governs the organization of the state tax service, but also deals with some of its powers as against the taxpayer.14 The last approach can be confusing, particularly if the same matters are dealt with in both laws.15

II. Matters to Be Included in a Tax Administration Law

A. Compilation and Publication of All Tax Laws

All legislation concerning taxation, including tax laws, regulations, administrative interpretations, and court decisions, should be compiled and generally made available. Unless this issue is already covered elsewhere, the law on tax administration and procedure should so require. The greater the availability of such information, the easier it is for taxpayers, tax administrators, and adjudicators to research the law and to make sure that they have uncovered all the relevant information on any particular subject.

B. Definitions of General Applicability

There is considerable difference among different legal traditions as to the desirability of definition sections in statutes or as to their appropriate scope. For example, legislation based on the U.K. tradition often includes monumental definition sections, while statutes based on the French civil code tradition often have no definition sections at all.16 However, where appropriate to the particular legislative tradition, definitions of general applicability can be of considerable use to avoid confusion in interpreting the law.

Among the most important definitions is that of “taxpayer.” Any reference to a taxpayer in a law on administration and procedure should, unless otherwise indicated, include any physical or legal person who is required under the tax administration laws to collect or remit tax (plus any related interest or penalties) or information. The definition of taxpayer would, therefore, include both third-party withholding agents and those physical persons who are responsible for effecting the collection or remission of tax or information owed by legal persons. An acceptable alternative, which is followed by some laws, is to make a terminological distinction between taxpayers and such persons as withholding agents, who are responsible for paying the taxes of another.17 If such a distinction is made, then care should be taken to draft the law so that both taxpayers and responsible persons are subject to the relevant procedural requirements of the law.

Which other definitions might be included in a definition section would depend on the particular legal terminology in use in the particular jurisdiction. Some examples are discussed in chapter 3, section V(D).

C. Regulations and Rulings

In many jurisdictions, administrative regulations and rulings are an important instrument for interpreting tax law. Depending on the jurisdiction and its legal traditions, rulings can be of general application or can apply only to specific taxpayers, and they can apply prospectively or retrospectively. In the case of rulings of specific application, they can be issued in advance of a transaction or following the transaction. The treatment of these matters in a number of jurisdictions is discussed in chapter 2, sections VI(D) and (E). What follows is an outline of what might be considered appropriate to include in a tax administration law in a typical jurisdiction.

1. Regulations and Rulings of General Applicability

The government should solicit outside comment and obtain a broad range of opinion by holding public hearings on proposed regulations. Persons can testify orally or submit written testimony. The opportunity for all parties to be heard will assist the government in uncovering beforehand any unintended benefits or hardships its proposed action will produce. Hearings are required in some countries by a law on administrative procedure.18 In the absence of a similar law, the matter can be addressed specifically by the tax administration law.

In general, the tax authority should be bound by its regulations and rulings of general applicability. Of course, the tax authority must be permitted to reverse a position when necessary, but this should normally be done only on a prospective basis.19

2. Rulings of Specific Application

Although the legal traditions of some jurisdictions restrict their use, there is considerable benefit to having a procedure whereby rulings of specific application may be issued at the request of a taxpayer.20 These rulings are typically based on fact patterns presented to the taxation authority.21 In jurisdictions such as the United States, such rulings are limited only to the taxpayer in question and cannot be used as precedent by any other taxpayer.22 To ensure accuracy and fairness in rulings, all rulings should probably be approved by a central rulings office or by another appropriate higher-level authority.23

The authority to issue legally binding rulings should be specified in the law, and the procedure specified either in the law or in delegated legislation.

D. Returns and Record Keeping

1. Returns

With respect to income tax and some other taxes, many jurisdictions have systems where the taxation authority assesses the amount of tax due based on information provided to it, usually by the taxpayer. However, efficiency concerns are increasingly motivating jurisdictions to adopt self-assessment systems, where the taxpayer determines tax owed. With regard to income taxes, self-assessment systems do not necessarily require that each individual prepare a return and determine tax owed, as several types of income can be taxed through final withholding taxes. If some taxpayers, say, small businesses, cannot currently be relied upon to determine their own tax, self-assessment for income tax can be introduced in stages, starting with larger enterprises and extended to others as they gain the necessary skills.

Whether a self-assessment system is in effect or not, the taxpayer must provide essential information to the taxation authority in the tax return so that it can either determine the amount of tax owed or check on the taxpayer’s calculations.

Tax returns must spell out in detail the information required of the taxpayer. In a self-assessment system, the information will be in the form of a series of steps that the taxpayer must undertake in calculating the tax. The accompanying instructions to the return should provide comprehensive guidelines for filling out the return, taking the taxpayer logically from one line of the form to the next.

The general rules relating to returns (e.g., who is required to sign the return and procedures for extending the time to file) can be contained in the general tax administration law.24 The specific tax laws should specify the deadline for filing and who is required to file.25

2. Information Returns

An information return is a declaration by a person who, though not necessarily liable to withhold tax, has economic information about one or more potential taxpayers.26 The law must give broad powers to the tax administration to establish a filing requirement for information returns and to define the format to be used.27

3. Conditioning a Tax Benefit on Identification of the Payee

The law could require that in order to obtain certain tax deductions or credits that are triggered by a payment, a taxpayer must identify the payee. By identifying the payee, the taxpayer provides the tax administration with valuable information that can be used in auditing the payee. Because it relates to the determination of the tax base for a specific tax, this type of rule is generally included in the specific tax laws.28

4. Record Keeping

The law or regulations should specify taxpayers’ obligations to keep books of account and other records necessary for determining tax liability.29 These would include the content and form of invoices, what taxpayers must use them, and under what circumstances.

E. Audits and Investigations

1. Relationship Between the Taxation Authority and Investigative Agencies

Many, if not most, jurisdictions have found it appropriate to segregate the civil functions of tax administration from the enforcement of criminal law, so that procedural protections for citizens are not undermined. The tax administration should not, for the purpose of civil tax investigations, rely on search powers given to the police. The tax administration’s search powers should be specified in the tax administration law, subject to constitutional constraints.30

It is common for countries to provide a separation between the functions of civil and criminal investigation. Once the tax authorities have determined that there appears to be sufficient evidence of criminal behavior, the case should be turned over to the public prosecutor, and the procedures for criminal investigation should be applied from that point on.

2. Access to Third-Party Records and the Power to Issue Summonses

The tax administration should have access to the records of anyone who has financial dealings with taxpayers and who can provide relevant information on taxpayers’ income and the accuracy of their tax declarations and books and records.31

3. Indirect Methods of Assessment

The law should specifically authorize the tax administration to use alternative methods to establish or verify the amount due, whether the tax involved is income, VAT, or another tax.32 The taxation authority should be permitted to use these alternative forms whenever the taxpayer fails to provide the records otherwise required in a complete and accurate form.

F. Dispute Settlement

1. Compromises

To further efficiency, throughout the dispute-settlement process the tax authority should be allowed discretion to settle issues of controversy with the taxpayer.33 The tax authority should consider the likelihood that the authority would prevail in an adjudication and the costs of pursuing the authority’s position. The authority should also have the discretion to reduce civil penalties, but not interest due.

2. Payment of Tax During Dispute

Countries differ on whether taxpayers are required to pay any tax subject to dispute in order to pursue a dispute.34 Some consider it unfair to impose such a requirement. Others impose it to discourage frivolous disputes. An intermediate position would be to allow tax authorities or the court to waive the requirement on a case-by-case basis. Another possibility is to require payment of a portion of the tax (e.g., 50 percent).

3. Disputes Within the Taxation Authority

Disputes between tax authority and taxpayer must be resolved in as fair, timely, and efficient a manner as possible. A single method of resolving disputes, from registration of taxpayer disagreement with an assessment up to resolution of a final appeal, is preferable.

The first forum for dispute settlement should be with the taxation authority officials who first issued the assessment. The taxpayer should be given the opportunity, within a limited prescribed time, to disagree with the assessment, either in writing, in person, or both.35 If the assessing official or officials agree with the taxpayer on any point, a new assessment can be issued.36 This assessment should be reviewed by a superior official.

If a dispute continues, the taxpayer should have the opportunity, within a prescribed period of time, to appeal to a special administrative appeals board. To ensure impartiality, this unit should be completely independent of other divisions of the taxation authority. It could report to someone outside of the tax administration authority, perhaps the general counsel (chief lawyer) of the finance ministry.37

4. Tax Adjudications

If agreement is not reached, the taxpayer should have the opportunity, within a prescribed period of time, to appeal to a court. Some jurisdictions allow appeal to the ordinary courts. These are typically not well suited to adjudicate tax matters, so consideration should be given to establishing a special tax court.38 Depending on the legal traditions of the jurisdiction, this tax court can be set up inside or outside the regular court system. Also depending on the particular legal traditions of the jurisdiction, judges on the court might include both tax professionals and laypersons. The tax court would then hear evidence from both the taxpayer and the tax administration and would reach a decision in a trial-like setting. However, the regular rules of evidence need not necessarily be applicable to the tax court. In addition, experts other than lawyers or advocates, such as accountants, might be permitted to represent taxpayers before the tax court.39

In some jurisdictions, appeal is to a court specializing in appeals from administrative decisions, although not specializing in tax cases.40

Both the taxpayer and the tax administration should be permitted to appeal a decision of the tax court, ordinary court, or administrative court to a court of appeal.41 Such appeals should be based only, or at least primarily, on matters of law, not of fact.

5. Procedures in Tax Adjudications

An adjudicative proceeding should minimize surprise and give taxpayers every opportunity to know of and rebut the case against them. First, each party should fully inform the other regarding the issues being contested.42 Both parties should exchange all relevant documents within an adequate time period before the adjudication.

Depending on the jurisdiction, rules of discovery may vary depending on the stage of adjudication and the nature of the particular forum. For example, in the United States, discovery rules vary depending on whether the case is before an administrative officer at the Internal Revenue Service, the Tax Court, the Court of Claims, or the District Court. At the administrative level, the general disclosure provisions of the Administrative Procedure Act allow the taxpayer to request information in his or her file, unless that information falls within a number of exceptions, including internal communications and information relating to a law enforcement action.43 Discovery is more limited in the Tax Court than in the Claims Court, and more limited in the Claims Court than in the District Court, each of which has its own rules of procedure.44 The Administrative Procedure Act also entitles the taxpayer to a copy of any testimony given during the case.45 Absent fraud or an attempt to conceal, any document not exchanged within this time period should be barred from consideration during the procedure.

6. Burden of Proof in Tax Adjudications

In an administrative or judicial proceeding that involves a civil tax issue, the taxpayer should generally have the burden of proof. In some jurisdictions, as a general matter the burden of proof lies with the party normally in possession of the relevant evidence.46 In tax matters, this party is typically, but not always, the taxpayer. For example, the tax department would have the burden of proof in matters such as comparable gross profits ratios. However, it may be preferable to state explicitly that the burden lies with the taxpayer, except in such instances where the tax department has sole access to the necessary evidence. Absent such instances, there should also be a presumption that an assessment issued by the tax department is correct. The taxpayer then has the burden of rebutting this presumption by demonstrating the inaccuracy of the assessment.47

G. Recovery

A basic choice must be made as to whether tax debts are to be collected under the same procedures as for all other debts against the government, or for civil judgments generally, or whether special rules should apply in the tax area. Whatever the decision, some reference may be made to nontax laws, such as the civil procedure code, the civil code, or other laws governing the sale of property in satisfaction of a judgment, and some of the provisions described below may not in all cases need to be repeated in the tax laws.

1. Tax Liens

The law should provide that a tax assessment is a charge or lien that constitutes a security interest in the taxpayer’s property in favor of the government. The lien should be against all property and rights to property, whether movable or immovable, belonging to the taxpayer as of the date of assessment or subsequently acquired during the existence of the lien.48

2. Seizure of Property

In the case of taxpayers who fail to pay, the law should provide for the seizure of property so that the proceeds from the sale can be applied to their tax liability.49

3. Sale of Seized Property

With the exception of negotiable instruments such as currency or marketable securities, the law should require seized property to be sold at public auction.50

4. Recovery of Debts Owed the Taxpayer by Third Parties

The law should allow the government to reach persons who might owe money to the delinquent taxpayer, such as employers or creditors.51 The law should provide for the seizure of such property through a notice of seizure upon the third party and a requirement that the third party pay over to the government the amount owed to the taxpayer.

5. Installment Payment Arrangements

The law should authorize the tax authorities to enter into an agreement with a taxpayer, which would allow the taxpayer to pay the tax over time, in cases where the taxpayer cannot pay the amount of assessed tax immediately.52 The extension of time to pay the tax should not affect the accrual of interest on unpaid amounts.

6. Receivership

If it is necessary to seize a business for nonpayment of tax, the tax administration should have the right to ask the court to appoint a receiver for the purpose of administering the business and paying the taxes.

7. Property Transferred Without Full Consideration

The government should be given a security interest in any property that was fraudulently conveyed for less than fair consideration. Such provisions are often found in civil or commercial codes and apply generally to all creditors.53

8. Compromise and Write-Offs

The law should permit tax officials to write off uncollectible accounts and to enter into a compromise with a taxpayer whereby part of the tax liability is canceled.54

H. Internal Investigations

Provisions should be made in the law for an internal investigations department within the tax administration. The goal of this department should be to ensure that the tax authority acts fairly and honestly. The internal investigations department could, like the administrative appeals board, report to someone outside the tax administration authority, perhaps the general counsel of the finance ministry.

I. Taxpayer Ombudsperson

Some countries have established a department of the taxpayer ombudsperson. The role of the ombudsperson would be to assist taxpayers in solving complaints about the tax authority.55 Although this department could be set up within the tax authority, in order to ensure that it is not overly influenced by other personnel in the authority, it should perhaps report directly to the authority’s head.

J. Interest

Interest must be assessed on every late payment of tax56 or penalty,57 as well as on every payment due from the treasury to the taxpayer. It should be stressed that interest is not the same as a penalty due for noncompliance. Interest reflects the time value of money and should therefore never be waived or subject to compromise. An interest rate that reflects the full cost of money, including inflation, should be specified, typically by reference to the central bank discount rate, a rate on treasury obligations, or the like.58 To discourage “borrowing from the government,” and to encourage the settling of disputes, the interest rate should exceed the basic rate given debtors in the economy. In part because the government is presumably a better credit risk than a defaulting taxpayer, it may be appropriate to provide a lower rate of interest on overpayments than on underpayments.59

K. Taxpayer Rights

Provisions guaranteeing procedural protections to taxpayers can be gathered into a separate section of the tax administration law, or included in the appropriate places in a law organized on temporal or functional lines.60 This section would collect these taxpayer rights common to all tax laws in a single place, either in the tax administration law or the basic law. At the commencement of any assessment or audit, the tax department should deliver a comprehensive description to taxpayers of their rights. However, these rights need not be unique to tax administration. Wherever possible, they should be accorded with other procedural rights guaranteed under law. A list might include some or all of the following rights.

1. Confidentiality

Taxpayers should have the right to have their personal financial information accorded the greatest possible confidentiality within the taxation authority.61 This confidentiality should be breached only (1) during criminal investigations, when criminal investigators outside the taxation authority must view the information, (2) when so required during adjudication of a controversy, when an adjudicator must view the information, and (3) in certain other cases provided by law (e.g., disclosure of information pursuant to a treaty to the competent authority of a foreign government).

2. Notice

Taxpayers should have the right to be notified of an assessment, a decision on an adjudication, or any collection action against the taxpayer’s assets.62 The exception is the jeopardy assessment, when there is an imminent danger of the taxpayer disposing of the asset.63

3. Reasonable Audits

Taxpayers should have the right to have audits held at a reasonable time, in a reasonable place, and within reasonable limits.64

4. Explanation

Taxpayers should have the right to an explanation of why their tax is being assessed the way it is and to an explanation of the reasons for a decision by an adjudicator.65

5. Counsel

Taxpayers should have the right during any dealings with the tax authority to be represented by a qualified professional.66

6. Record

Taxpayers should have the right to record their meetings with the tax authority and to have all adjudications recorded.67

7. Discovery

Taxpayers should have the right to advance access to the government’s evidence in the case of an adjudication.68

8. Hearing

Taxpayers should have the right to a hearing before a decision is taken on an adjudication.69

9. Appeal

Taxpayers should have the right to an independent administrative appeal and a final judicial appeal.70

10. Limitations

There should be a limitation on the period during which an assessment may be made.71 However, this limitation should be waived in the event of fraud on the part of the taxpayer.72 The relevant rules should be specified in the tax administration law.

III. Taxpayer Compliance and Sanctions

Both substantive and procedural tax laws should always be directed toward improving taxpayer compliance. It is generally agreed that improving taxpayer compliance has many aspects to it, including making the law fair and equitable, easy to comply with, and difficult to evade. Another important aspect of improving compliance is the provision of effective sanctions for failure to comply. Typically, sanctions can be of a civil or a criminal nature, and most jurisdictions provide for both, although in some jurisdictions criminal sanctions would be included in a separate criminal code.

However, there is much dispute as to what factors contribute most effectively to taxpayer compliance. Perhaps the most contentious area of controversy lies in the nature and function of both civil and criminal sanctions. Considerable variation exists with regard to the design of sanctions among different jurisdictions. However, social scientists have made considerable progress in understanding the various aspects of compliance and, most particularly, how sanctions work. Such information has made it possible to reach tentative conclusions as to preferred ways of designing them.

A. Existing Research into Compliance Issues

There is a large (and growing) amount of literature on taxpayer compliance, much of it focused on a single type of tax, for example, income, VAT, property, customs, and so on. This discussion will, where possible, consider some principles applicable to all of the above, but by and large most of the citations to research will be from studies on income tax compliance. Although the reasons for this focus are varied, the principal reason is that the majority of studies have been done in the United States, and the income tax is the most important and most studied tax there.

Much empirical research has been done on sanctions and income tax compliance in the developed world, and, as suggested, the body of literature on compliance in the United States in particular is quite substantial.73 Unfortunately, while some theoretical and anecdotal work on compliance has been done in the developing world, there is little empirical work to guide policy planners. Therefore, if one is to use empirical research as a guide to designing rules for developing countries, it is necessary to rely excessively on studies from developed countries as a guide.74

In addition, unfortunately, much research into taxpayer compliance even in developed countries like the United States has been of rather dubious empirical value. Studies have tended to two different types: those undertaken using official taxpayer data and those using self-administered questionnaires, although some studies have also used national accounts data. Unfortunately, taxpayer data are often incomplete and can rarely distinguish very well among types of compliance and noncompliance, while self-reporting is often inaccurate. In-depth interviewing and participant observation, while probably the most accurate way of learning about taxpayer behavior, is also the most difficult to undertake.75 It may be that anecdotal evidence is the best that can be hoped for regarding many taxpayer compliance issues.

However, even though much of the social science work done may not be conclusive, the analysis can be helpful in thinking about the issues. Both theoretical and empirical studies in the general area of legal compliance, and the specific subcategory of taxpayer compliance, can give indications as to what the issues are and what might work, even if they cannot prove what might work beyond a reasonable doubt. Also, if one relies on social science research only as a general guide to understanding, it becomes easier to apply insights learned from such studies in developed countries to the different circumstances of developing countries.

Successfully developing rules that improve compliance requires at the outset two steps. First, a taxonomy of compliance needs to be developed, so that the relevant issues can be identified. Second, as the issues are identified, theories have to be formulated that can be tested against experience and that can then be used to create tax compliance principles. How the principles are implemented can be guided by specific examples, but the design of laws in a particular jurisdiction is likely to be sui generis, based upon the unique characteristics of that jurisdiction. For this reason, it is helpful to be able to return to the principles and their analysis as a guide for how to proceed.

B. Design of the Substantive Tax Law

Many taxonomies of tax compliance are reported in the scholarly literature. The most important breakdown is probably between unwilling and willing. First looking at the unwilling, people can fail to comply because (1) they do not know how, (2) it takes too much effort to do so, or (3) it is too expensive to do so.76 Related to this distinction is the ease with which a law may be avoided or evaded. A law that is easy to comply with may also be relatively difficult not to comply with. The most successful law will be one where these two properties coincide in a single law. In either case, it is the underlying substantive law of taxpayer obligations that must be designed so as to make compliance easy and noncompliance difficult.

The first and most obvious technique of filling both criteria is to reduce the total number of people who must make tax calculations, file returns, or pay money to the government. For administration of a VAT, this can mean restricting the collection of tax only to those who have a turnover of a certain size. For property taxes, it can mean using exemptions for properties of certain types or sizes.

For administration of an income tax, for example, it can mean exempting people below a certain income level from paying the tax and having as much tax as possible collected through a withholding system or pay-as-you-earn (PAYE) system. Withholding and PAYE form one of the central aspects of the administration of any income tax system. They can reduce significantly the actions that must be taken by taxpayers and reduce the number of persons who must ultimately file tax declarations.77 In these instances, an intermediary, usually the person who employs the individual taxpayer or the person who makes periodic payments to the taxpayer, determines the tax and pays it.78 That intermediary can also typically advise the taxpayer as to the applicable rules.

As a by-product of having intermediaries compute, withhold, and remit tax, the interaction between individual physical taxpayers and the administration and the associated paperwork are reduced. This is in effect a reduction in the cost of compliance for those taxpayers who have their obligations fulfilled, or intermediated, by others.79 Success is reflected in less work for the individual taxpayer and a reduction in the total number of individual taxpayer declarations that need to be filed. Of course, in these instances, the intermediary also becomes a taxpayer; withholding or PAYE becomes a separate obligation enforceable by law.80

One of the frequently repeated dogmas of tax administration is that “legal simplification” also reduces the costs of taxpayer compliance.81 A simple tax law may fulfill the twin requirements of being easy to obey and hard to disobey because it both allows the taxpayer, or taxpayer intermediary, to know more easily what is expected of him or her and also reduces manipulability of the law, thereby reducing the possibilities of tax avoidance.82 Complexity, and the chance to avoid tax obligations, can come from a number of sources. Inconsistency within the coverage of the law certainly can be one. Exceptions or special rules that provide for reduced obligations in certain circumstances not only add complexity, but they also create an incentive for taxpayers to try and fit into those circumstances. The converse is also true: the more special circumstances where taxpayers have increased obligations, the more those circumstances will be avoided. However, when one considers policy justifications for particular taxes, legal simplification can turn out not to be a very simple task.

What constitutes simplicity and consistency in a tax law, at least with regard to tax avoidance opportunities, will depend on the policies behind the specific type of tax involved. The more that obligations can be designed to treat taxpayers and circumstances alike, the more consistent will be the law, and the fewer special circumstances will result. Second, in treating taxpayers and circumstances as alike as possible, the fewer and the more internally consistent the principles of such treatment, the better. With like treatment and limited and consistent principles, and language clearly reflecting those principles, interpretation of rules should be easier. The opportunity to avoid tax obligations would be reduced.

Unfortunately, these general principles can be very difficult to effect in practice. In fact, rules designed to make withholding and PAYE easier may frequently violate these principles, as do special provisions designed for administrative ease, such as exempt amounts and schedular withholding in the income tax. However, such complications and inconsistencies in tax rules should be approved only when there is a net administrative benefit.83 While there may be no set of rules of thumb to decide when this will be the case, this should be the explicit goal of the drafters of substantive tax laws. It would generally be wise for the drafters of substantive tax laws to consult with tax administration experts to ensure such a result.

C. Sanctions

1. Purpose of Sanctions

Sanctions are perhaps one of the most overrelied-upon, and poorly understood, tools for enhancing tax compliance.84 Sanctions can also have more than one purpose. First, the most important component of sanctions is their ability to deter unwanted behavior, so as to bring about greater compliance.85 Therefore, sanctions should be applied only to behavior that is reasonably capable of being deterred. Second, sanctions must be fair under the general jurisprudential criteria in effect in a particular jurisdiction. Under the jurisprudential principles of most jurisdictions, this means that sanctions should apply only when the sanctioned person is somehow at fault and should not be unduly harsh or disproportional, or imposed in violation of principles of due process. When the principles of deterrence and fault are combined, this leaves a general principle that faultless or reasonable behavior by taxpayers, even if it results in an underpayment of tax, should not be punished by sanctions. Only negligent or unreasonable behavior resulting in an underpayment should result in sanctions.86

In addition to their deterrence component, sanctions may also have an important financial component. Financial sanctions may raise revenue, while prison sentences may increase expenditures. Financial sanctions may even be designed in such a way that they cover the tax administration’s expenses in pursuing a case, from investigation through final collection. Fines may also be designed to reduce administrative costs by encouraging the early settlement of disputes between administration and taxpayer.87 This principle may appear to be self-evident. However, there are a number of jurisdictions, particularly among transition economies, where the principle is not followed under current law.88

In countries where government resources are extremely limited, these financial aspects of sanctions policy may be especially important, and the financial costs and benefits are of serious legitimate concern. However, in most instances, it will be difficult to fashion appropriate sanctions that fulfill both deterrence and other goals. Deterring behavior—which includes “encouraging” certain behaviors, such as the efficient settlement of disputes, or the remedying of violations, such as the failure to file or the failure to pay—should determine the design and severity of sanctions; the raising of revenues should be left to the taxes themselves.

Both financial and penal sanctions may also be designed to punish, not for the purpose of directly affecting the behavior of the person punished, but for the purpose of retribution or to indicate that society seriously disapproves of particular behaviors. As will be discussed below, the severity of sanctions may play a role in affecting people’s attitudes toward the particular crime. For these two reasons, certain taxpayer activities that are viewed as particularly heinous, such as intentional evasion through fraud, are usually punished more harshly than less serious avoidance or error. Some jurisdictions punish such heinous behavior through both the civil system (e.g., increased fines for fraud) and the criminal system (additional fines and even prison terms for fraud).89

Although rarely discussed, another goal of sanctions policy should be not to cause (or worsen) other problems, that is, those outside the direct realm of tax administration. In at least two important instances, this goal will suggest that sanctions should be relatively limited in degree. This will be discussed at greater length below.

2. Operation of Deterrence

Leaving aside for the moment some of the subsidiary issues, the principal goal of sanctions is based on a simple premise—the threat of punishment deters unwanted behavior. If the likely punishment is sufficient to outweigh the prospect of gain, a rational person will not undertake the activity that will result in that likely punishment.90 Even this most basic of statements of deterrence function depends on two important assumptions. The first is that it is possible to determine and create a likely punishment that appropriately outweighs the prospect of gain. The second is that people will act in a way that is measurably and understandably rational.

For a likely punishment to be accurately determined by an individual, he or she must first understand both what choices he or she has and what any possible adverse consequences of any choice will be. Only in such cases can the individual weigh the risks of detection and its consequences against the expected benefits of violation. Rational choice economists might state this as an equation: the taxpayer will commit evasion if the benefit the taxpayer receives is greater than the total punishment provided for violation multiplied by the likelihood of punishment.91

Rational choice theory does not simply mean that people are “rational” in the sense that they act in accordance with a single set of norms, for example, that they are profit maximizers.92 People may act differently in response to a like set of circumstances and still act rationally; this is because they have different preferences or private utilities. Part of the individual’s private utility function may be unrelated to specific statutory sanctions, such as informal sanctions of the community or individual preferences for obeying all laws or only certain laws.93 However, looking first at the effect of official, state-sponsored sanctions, the private utility function of one individual is still likely to vary from that of another. In other words, to achieve a like effect, deterrents would have to vary from individual to individual precisely because individual utility functions differ.94 For example, some taxpayers may have a greater preference for money now (say, through tax evasion) over later (say, when taxes and fines are finally due) than do other taxpayers.

The reasons for these different utility functions may be relatively more “objective” or “subjective.” For example, some taxpayers may be poor or spendthrift and may be relatively indifferent to monetary sanctions because they are judgment-proof (i.e., immune from a money judgment because of insolvency, lack of property within the jurisdiction, or other reasons).95 Others may simply prefer to live for the moment. That would mean that even if a sanction were certain to be applied and the cost (in present value terms) were greater than the amount initially saved, some taxpayers might still choose to evade, and yet could still be acting rationally.

Therefore, first, with regard to poor or judgment-proof taxpayers, there should be provisions in the law that allow debts to the government resulting from monetary sanctions to remain in effect even in the event of bankruptcy. Second, for both these taxpayers and those who are, for other reasons, indifferent to monetary sanctions, the most effective deterrent may be prison rather than fines. Therefore, while jurisprudential rules concerning the imposition of prison terms may restrict their application only to cases of criminal fraud, they may in certain circumstances act as a deterrent to these taxpayers. How these general policy conclusions might be implemented in a statute is discussed below.

However, other taxpayers might care very much about monetary sanctions, while still others, perhaps because of the lack of social stigma in their particular community, care little about prison sentences. It would be impracticable, and would presumably violate the principle of equality before the law, to provide completely individualized sanctions.96 It would also probably violate the principle of equality if sanctions were designed to be infinitely strong so as to deter those least able to be deterred.97 Therefore, for rational choice theory to be implemented in the design of sanctions, general rules must be created that are reasonable in the particular jurisprudential setting and that provide the greatest average deterrence. This would include the full panoply of monetary sanctions applied as effectively as possible, plus (in certain cases) nonmonetary sanctions such as prison terms.

Even allowing for some differences in individual utility functions, basic rational choice theory suggests that if sanctions are to work, they must be severe enough, and their chance of application likely enough, that the product of these two exceeds the benefits of noncompliance for a sufficiently high proportion of the members of the target group. One direct corollary is that an increase in the perception of likelihood of being caught and punished results in an increase in the deterrence effect. This has been observed in the United States as a matter of general compliance with laws.98 It has also been observed repeatedly in studies of tax compliance.99

The perception of likelihood of suffering sanctions depends on a number of factors, the first being the actual risk. That in turn will depend on (1) the ease of detecting noncompliance, (2) the ease of proving noncompliance, and (3) the administrative effort put into detection and proof.100 Again, not surprisingly, tax studies in the United States have repeatedly shown that the easier it is to cheat and to hide the cheating, the more cheating will occur.101 Any perceived risk is likely to differ from the actual risk and will depend on objective factors, such as how well the actual risk is publicized, and subjective factors, such as how well risk is processed and internalized by the taxpayer. Again not surprisingly, studies show that taxpayers who perceive higher probabilities of being subjected to legal sanctions are more likely to comply.102

Some research has suggested a number of less obvious refinements to the proposition that people are more likely to cheat if it is less likely that they will be caught. While rational choice theory may suggest that a reduction in probability of detection and punishment can be offset by an increase in severity of sanctions, there is considerable evidence that people are not very good at analyzing probabilities.103 The first consequence of this fact is what criminologists refer to as the “tipping” effect.

Criminologists working in developed countries have found that within a given group of people there is a critical level of probability of punishment before which a marked deterrent effect is seen.104 That point at which the likelihood of punishment “tips” into a compliance effect may not, in fact probably will not, fit with actual probabilities. As a general proposition, tipping seems to occur because people often tend to discount low probability events. It has been suggested, but not proved, that there may be a general tendency for people to think relatively more rationally about economic crimes than about other types of crime and, perhaps, that such rationality increases as the person’s income increases.105 Therefore, one might conclude that, at least with regard to other types of crime, tax evasion has a lower tipping point. Further, as a general matter the tipping point for a particular individual is perhaps likely to be lower (in other words, the individual is more likely to comply) if that individual’s wealth or income is relatively greater. Because the wealthy, or those with higher incomes, are more likely to owe more in tax, one might also conclude that the tipping point is often inversely proportional (again, meaning that the individual would be more likely to comply) to increases in the amount of potential evasion. Nevertheless, this thesis does not suggest, at least in any absolute sense, where that tipping point would be.

The point at which a taxpayer would perceive that the probability of being caught is high enough that he or she would begin to comply may be influenced by different factors that would increase the taxpayer’s awareness of punishment certainty.106 This is because the tipping thesis depends on taxpayer awareness of the likelihood of punishment, and not just upon actual probability of punishment.107 If taxpayers incorrectly believe that evasion is relatively common but is still not punished, the result could be tipping points higher than if the truth were better known. Another related point has to do with the percentage of people in the population who regularly comply or fail to comply. If a person violates a rule and is not caught, he or she may be more likely to commit another violation.108 And, if many have this experience, they will be able to relate their success at evasion to larger sections of the population, resulting in a greater perception of the lack of consequences for failure to follow the law. The effects of perceived common evasion, plus one’s own positive experiences of evasion, would both raise the tipping point and thereby reduce the effectiveness of sanctions.

There are a number of practical lessons to be learned from this analysis. First, in those areas of tax administration where there is sufficient successful detection of noncompliance and application of sanctions to exceed a tipping point, there could be considerable benefit in publicizing such detection and application of sanctions. Tax administrations can publish data on the number of taxpayers caught (or possibly even their names) and the sanctions applied. Active promulgation of such information should have a positive effect on future compliance.109

However, in areas where it is clear that a tipping point has not been reached, it would actually be counterproductive for a tax administration to advertise its relative ineffectiveness. In such instances, until detection of noncompliance and application of sanctions are sufficiently high, the tax administration should not publicize its efforts. For example, it may be possible that the application of sanctions to VAT noncompliers, or to those who do not comply with income tax withholding requirements, is much higher than to those who do not comply with income tax rules for self-employment income. In such a case, it might be wise not to publicize the tax administration’s experience with self-employment compliance.

In making such determinations, however, the tax administration should recall that rates of detection of noncompliance and application of sanctions, and of tipping points, can vary among different populations. Therefore, for example, it may be that tax administration efforts may be more substantial, and more successful, among the wealthiest income tax payers. If so, publicity about successes in this subgroup could be beneficial in improving compliance among members of the subgroup.

Therefore, the tax administration should have the ability to collect the necessary information required to determine both rates of detection of noncompliance and subsequent implementation of sanctions as well as likely estimates for tipping points. It should also have both the authority and the means to publicize this information selectively.

It may seem obvious that increasing detection of noncompliance, and of applying sanctions to the noncompliant, would raise the likelihood that the tipping point would be exceeded. However, increasing the perceived likelihood of the imposition of sanctions may result in effects that, although rational, are counter to the standard deterrence effect hypothesized by rational choice theory. At his or her tipping point, the taxpayer assumes that the risk of detection is high enough that he or she will be subject to sanctions. At this point, the sanctions need be just severe enough for the taxpayer to comply.

But if the taxpayer can take effective evasive action, he or she might be able to reduce the (self-perceived) probability of suffering from sanctions. Such action can include additional tax evasion or measures to hide the evasion. As the likelihood of being caught increases, either because of the nature of the substantive law (hard to avoid or evade) or of effective administration (more work in fighting avoidance and evasion), certain taxpayers may actually be driven into even greater acts of avoidance or evasion. Only if the taxpayer’s evasive action is unlikely to be successful, or if the cost of the action to the taxpayer is likely to exceed the taxes saved, will the taxpayer comply.110

Putting more resources into effective tax administration may cost the exchequer money that could be better spent elsewhere.111 However, it is also important to remember that more avoidance or evasion on the part of taxpayers carries a number of welfare costs. Both legal strategies for avoidance and illegal actions to evade are likely to have negative effects on resource allocation.112 Therefore, even if net revenues (amounts collected minus costs in collection) were to increase with additional sanctions, there could be a net welfare loss to the economy in general, because of the changed nature of noncompliance.113 Increasing the likelihood of being caught and subjected to sanctions works best if the taxpayer cannot easily take evasive action.

Although it is probably important for the perceived probability of being caught to be sufficiently high for there to be any deterrence effect from sanctions, there is not much evidence to suggest that compliance varies directly with the degree of severity of the sanctions. There has been an inconclusive debate among criminologists over why this is the case, although there is no clear evidence to suggest that the conclusion is wrong.114 There are theories that people simply do not act rationally or that the theory of rationality must be correct and the data wrong.115 Some criminologists have suggested that while the severity of formal sanctions in economic crimes like tax compliance may have little deterrence effect, informal sanctions may have greater effect.116 However, increasing the severity of sanctions may also result in evasive action, under the same mechanism as that described previously regarding increased certainty.

The albeit insufficient empirical evidence might suggest that because an increase in severity does not change the point at which taxpayers become really afraid of being caught, it is unlikely to improve their compliance. However, such an increase in severity might nevertheless inspire them to take some extra precautions in their avoidance and evasion to ensure that they do not suffer those increased penalties. In fact, increasing the severity of sanctions without increasing their certainty to a tipping point would probably only have the detrimental effect of creating more avoidance and evasion, with the ensuing general loss to welfare.117

An additional cost of very high sanctions is that their imposition may well be unfair.118 The sanctions, if disproportionate to the offense, would also be unfair. They would be unfair in their application if they reached only a small number of violators, since the violators who were caught would be much worse off than those who were not.119 In tax administrations prone to corruption (including both the taking of bribes and the use of administrative powers against political opponents of the regime), the existence of unduly severe but not universally applied sanctions can constitute a dangerous weapon in the hands of corrupt officials. The general conclusion is that it is better to deal with noncompliance by imposing moderate sanctions more frequently than by having draconian sanctions that are rarely applied.

While fines and imprisonment may be the principal statutory responses to taxpayer noncompliance, they are not the only sanctions that figure in a taxpayer’s rational choice calculus. Criminologists have recognized that the informal sanctions of social disapproval from peers have, in certain populations and for certain types of crime, also deterred criminal acts.120 In many cases, the cultural climate regarding compliance may be even more important than the legal sanctions themselves. This may tend to be the case more with wealthier or more socially prominent subgroups, where social status is important and where social ostracism for criminal activity is more likely. Officers of prominent corporations, for example, may be particularly affected.121

While the literature on the subject is not completely convincing, the argument is powerful enough to suggest that it may be beneficial to make public the names of wealthy, powerful, or influential taxpayers who are punished for noncompliance. As discussed earlier, taxpayer confidentiality is an important right. However, once a taxpayer is actually sanctioned, any reason for protecting the name of the taxpayer is considerably weaker. Of course, if government in general or the tax system in particular is seen as corrupt or unfair, failure to pay may not be perceived as bad, and publicizing the names of those who did not pay would not result in social ostracism and greater deterrence.

Social ostracism is likely to increase with the relative heinousness of the crime. A study done for the U.S. Internal Revenue Service suggested that tax noncompliance, particularly tax evasion, might have less social stigma attached to it because the public often perceives it to be a “victimless crime.”122 This attitude may vary depending on who commits the crime. Experiences in places as diverse as New York State and India suggest popular support for government enforcement against the wealthy or powerful.123 Presumably, the more the public understands tax evasion as a crime of moral turpitude, or one that injures the public at large, the greater the effect of social ostracism against tax evaders. Therefore, it may make sense for the tax administration to publicize adverse effects of tax noncompliance on other taxpayers, such as a general increase in tax liability for those taxpayers who do comply. Again, if government in general or the tax system in particular is clearly corrupt or unfair, no amount of publicity is likely to change the public’s perception of these problems.

People also have propensities to obey laws for reasons other than the likelihood of punishment.124 Not surprisingly, social scientists have suggested that people are likely to follow rules that they feel have a strong moral justification.125 A large number of studies suggest that people who support the government in general and the tax laws in particular are more likely to comply with tax laws. If such morals are widely found within a group, then individual moral influences and the informal sanctions of the group are likely to coincide, reinforcing each other.

These individual and group views of morality can be affected by the existence of official sanctions. If a particular legal structure is viewed as legitimate by the target population, and as a general matter the severity of sanctions increases with the severity of crimes, then statutory sanctions may influence both personal and group views.126 There may also be other ways for the government to influence people’s propensity to obey the laws, from improving the perception of its own legitimacy to improving the perception of the legitimacy of particular laws. Wherever possible, these ways should be explored.

Another problem with increasing sanctions without increasing their certainty is that people may begin to view the system of tax administration as arbitrary and unfair. This problem can be magnified if sanctions are perceived to be enforced primarily against political enemies of the government in power. In addition, if sanctions are enforced largely against the less wealthy and powerful, or if the powerful and wealthy are known to escape sanctions, the public’s perception of the justice inherent in the legal system can be substantially weakened. This could reduce compliance even further.127

3. Design of Civil Sanctions

A. Deterrence

As noted previously, in general, civil sanctions should be designed with two purposes in mind: (1) to deter certain unwanted behavior and encourage desirable behavior, and (2) to punish other, more heinous behavior. Looking first to the question of general deterrence, sanctions that are easily understood by taxpayers, and that are therefore easily applied and determined, are more certain in their outcome and more likely to affect a taxpayer’s behavior, given that person’s utility function.128 Also, sanctions that are easily applied and determined are likely to take fewer administrative resources and are less likely to be subject to arbitrariness. Therefore, as a general principle, financial sanctions should be imposed as automatically as possible. Perhaps the most effective way to do this is to assess a general deterrence penalty, calculated as a percentage of the amount involved, for negligent or unreasonable failure either to (1) report the correct amount of income (or other tax base) on the return or to (2) pay tax when due.129 Of course, a judgment will always be required to determine whether a failure is based on negligence. However, the degree of judgment required can be constrained. For example, there can be a presumption that a failure to pay an amount due is unreasonable and that the taxpayer has the burden of proving reasonableness.130

In addition to being easier to apply and determine, assessing penalties on the basis of the amount of underpayment makes sense within the logic of deterrence theory. As discussed previously, a penalty designed to deter should, within the constraints of tipping points and the irrationality of humans, raise the average cost of noncompliance so that it exceeds any average savings from noncompliance. Because the literature suggests little additional compliance as sanctions increase, there is likely to be little benefit to increasing formal sanctions beyond this point. In fact, greater sanctions may only increase avoidance and evasion activities, resulting in no greater overall tax compliance and perhaps in a net loss of social welfare to the economy at large. A monetary sanction should then be equal to an appropriate percentage of the benefit of noncompliance, although what percentage is “appropriate” is not always clear. In some instances it may be appropriate to apply flat-rate penalties. These instances are discussed below.

As discussed earlier, the tax administration should publicize rates of detection of noncompliance and implementation of sanctions whenever such rates are above a tipping point for that particular noncompliance. The tipping point should be measured within a particular identifiable subgroup of taxpayers. When appropriate, the deterrence effect of social ostracism may be exercised by publicizing the names of taxpayers who have been punished for noncompliance. Finally, the tax administration should undertake, where appropriate, public information campaigns to emphasize the injuries suffered by complying taxpayers when others fail to comply.

B Encouraging Resolution of Disputes

There are two ways to reduce or eliminate the incentive to drag out settlement of a tax dispute. The first is to require the taxpayer to pay all disputed amounts and penalties at the outset. The second is not to require payment at the outset, but instead to charge interest on both until they are paid. In the former case, the possibility that the taxpayer will later run out of resources to pay is reduced. However, if the administration is mistaken, such a rule can unfairly force the taxpayer to borrow substantial amounts, perhaps at high interest rates, and put the taxpayer in financial jeopardy, even if the administration must eventually pay interest to the taxpayer on overpayment. In extreme cases, the taxpayer may not only be unable to borrow the needed amounts, but may be unable to contest the assessment at all.

These problems are exacerbated by the fact that government and taxpayer are likely to have different credit ratings, meaning that risk premia are likely to be higher for the taxpayer as borrower than for the government as borrower. Even if the taxpayer receives interest on an overpaid amount, the interest may be much lower than what the taxpayer must pay on funds borrowed to make an initial payment of the amount in dispute. Another problem with regard to risk premia exists if the taxpayer is allowed to defer payment until the end of the dispute process. Unless interest is charged based upon the least creditworthy taxpayers, it will benefit those taxpayers to “borrow” from the government.

Different jurisdictions have reacted to this conundrum in different ways. Some require payment before any dispute settlement can begin, while others do not. Interest rates on overpayment and underpayment also differ among jurisdictions.131 One possible compromise is to allow an impartial adjudicator to determine whether the taxpayer is required to pay. Taxpayers can then be encouraged to pay earlier by setting relatively high rates of interest on overpayments as well as on underpayments.

Another way to encourage the settlement of disputes is to reduce the penalty if early settlement is reached. For example, the penalty could be reduced by 50 percent if agreement is reached during the administrative stage and by 25 percent during the first litigation stage.132

C. Punishment

To punish particularly heinous behavior, an additional civil penalty can be charged for underpayment attributed to willful evasion, fraud, or reckless indifference.133 This penalty can be determined as a percentage of the portion of the underpayment that is due to such willful evasion, fraud, or reckless indifference.

D. Flat-Rate Penalties

Although sanctions should in general be fixed as a percentage of the deficiency, in some instances it may be desirable to fashion penalties in an even easier, more predictable, and more automatic way than assessing a percentage against the amount of underpayment of tax. This will be the case when there is at best an indirect connection between an action and an underpayment of tax. For example, flat amounts can be charged for each instance of failure or error and can include a flat penalty for failure to file a required document (tax return, information return) or for filing certain documents incorrectly (information returns).134

4. Severity of Civil Sanctions

Perhaps the most important lesson to be found in the research literature on sanctions is that they are ineffective unless taxpayers believe that there is sufficient likelihood that they will be caught and that the sanctions will actually be applied. As noted earlier, the most important activity that a tax administration can undertake to ensure the voluntary payment of taxes is to ensure that noncompliance is readily discovered, that the discovery results in the application of sanctions, and that the public is made aware of the difficulty of escaping noncompliance detection. These measures will lower the tipping point, resulting in greater compliance behavior by more taxpayers. In short, the most effective sanction is not the sanction per se, but the rate of enforcement. As a general matter, therefore, those tax laws that are relatively easy for the tax administration to check or audit (meaning also that they are hard to avoid or evade) will have the greatest potential for compliance. Such laws are also likely to be the easiest for the taxpayer to comply with.135

For tax laws that are harder for the tax administration to audit, it will be more difficult to ensure a low tipping point. However, unless the tipping point is low enough, increasing sanctions is unlikely to result in a decrease in evasion. The risk that greater sanctions may result in a net loss of social welfare is more substantial when the law is more difficult to enforce and when additional avoidance and evasion behavior is more likely to be successful. In addition, if the tax administration appears to be punishing relatively few noncompliers with very harsh sanctions, the law will be perceived as arbitrary, as well as a vehicle for politically motivated prosecutions. Therefore, regardless of whether the particular tax law is easy or difficult to administer, sanctions need not be substantial. Sanctions that are already established at significant levels should not be increased in severity as compliance decreases, although this approach may seem counterintuitive.

It is difficult to determine both the actual tipping points for particular populations and the chances of being caught and forced to pay. It is even less likely that these determinations can be made for different types of avoidance or for different population subgroups. It can be said, however, that the exact percentage of the amount of underpayment that should be charged would vary from jurisdiction to jurisdiction, although the variation should not be too substantial. Also, the difficulty of determining these factors when combined with the benefits of uniformity suggests that penalties should be uniform for all taxes. The most important point, however, is that efforts should focus on making compliance with the laws easier, making avoidance easier to detect, improving tax administration enforcement efforts, and, when those efforts are successful, publicizing those efforts.

The size of the penalties would vary from jurisdiction to jurisdiction depending upon the factors discussed above. However, in most cases, penalties should probably not exceed 25 percent of the amount of underpayment for negligence or 50 percent for intentional underpayment.136

In the case of additional civil penalties for the more heinous activities of evasion, the additional goal of punishment comes into play. The level of such sanctions should first be based on the legal traditions of punishment viewed as desirable in the particular jurisdiction. However, this starting point should be adjusted to take into account the adverse affects previously discussed of having particularly high sanctions. However, it is unlikely that a civil penalty for fraud should exceed 100 percent of the amount of underpayment.137

5. Rules to Increase the Effectiveness of Civil Sanctions

Because of the economic nature of tax noncompliance, monetary sanctions are not always effective deterrents. One example is where the noncomplier is judgment-proof, meaning the taxpayer has no resources to pay any amount due, including underpaid tax, interest, and any monetary sanction. First, legal rules should provide that the government has appropriate priority over other creditors. In particular, bankruptcy laws should be drafted so that tax liabilities are not extinguished in the bankruptcy of a physical person and that tax liabilities are given priority in the reorganization or winding up of a legal person.138 Second, there must be legal rules that allow the taxation authority automatically to secure liens against the taxpayer’s unsecured assets, to garnish wages, and to levy property. Also needed are rules against the conveyancing of assets to others in order to avoid government claims.

Although the rules of priority, bankruptcy, lien, attachment, execution, and fraudulent conveyancing are designed to protect the government’s claim, there may still be instances where the taxpayer is relatively judgment-proof. Some have suggested that in these instances nonfinancial penalties, such as prison terms, should be added. However, most legal systems would not tolerate the imposition of prison terms for civil offenses.139 Civil sanctions could, however, include the temporary suspension of certain privileges, such as to practice as a chartered accountant. Some jurisdictions revoke business or other licenses from delinquent taxpayers.140 Revoking such privileges, while acting as a deterrent, may actually reduce the ability of the taxpayer to pay off his or her government debt and may have the undesired effect of damaging the economy, and increasing unemployment, by essentially prohibiting a business from operating.

It may make sense to allow such nonmonetary sanctions to be applied only after a taxpayer fails in good faith to make payments under a payment plan. The taxation authority can increase the effectiveness of informal sanctions by publicizing tax violations. Such informal sanctions would apply to all delinquent taxpayers, but may have particular importance for those debtors who are judgment-proof. Sanctions are also ineffective when someone other than the noncomplier will step in and pay the fine, for example, when legal persons, such as companies, are taxpayers. The physical persons who undertake the execution of the tax law liabilities of legal persons may not be subject to an adequate financial sanction if only the legal person, and therefore the legal person’s owners, suffers. Therefore, financial sanctions must be addressed not only to the legal person, but also to the physical person. If the physical person is indemnified against any sanctions by the legal person or by others, the sanction will also not work.

These problems can be at least partially addressed by including a “responsible physical person” penalty. Such a penalty would make those physical persons who are responsible for collecting and paying taxes for legal persons liable for failure. Jurisdictions that have such responsible person penalties usually restrict them to certain types of noncompliance. In the United States, for example, such a penalty exists only for the failure to withhold the appropriate amount of taxes on payments to third parties and on failure to pay the withholding over to the government. The “penalty” for such failure is equal to the amount of underwithholding or underpayment; the responsible persons are jointly and severally liable, along with the legal person.141 However, responsible persons are not penalized for failure to pay income tax amounts due. This is probably due to the fact that, as discussed previously, withholding rules are very simple to implement and failure to withhold can be attributed clearly to a limited number of people.

The same cannot be said for the determination of income tax liability. Other taxes, such as VAT and excises, may also be easy enough to implement, so that a failure to collect and remit these taxes could also subject the responsible person to penalty. A 100 percent responsible person penalty suggests that the purpose of the penalty is not simply to extend deterrence to the physical person in charge, but perhaps also to help collect the amount of underpaid tax when the legal person is judgment-proof. In the United States, the responsible person penalty makes any person required to pay tax over to the IRS liable for the tax.142 Because the taxes chosen for such a penalty are easy to determine and collect, and easy for the tax administration to check or audit, the tipping point for the deterrence effect can be made fairly low, which means that sanctions would be more likely to be effective and high sanctions would be less likely to have additional negative welfare effects through increased avoidance and evasion behavior. The nature of these taxes also suggests that high penalties are not necessary. Therefore, the total amount of tax collected from all responsible persons through this penalty should not exceed 100 percent of the tax due, plus interest and other applicable penalties.143

6. Criminal Offenses by Taxpayers

Fraud or evasion is usually considered a crime,144 but it is often a difficult crime to prove. Some countries have therefore set forth other acts that may be part of a scheme of fraud, but that, in themselves, constitute crimes and that may be easier to prove than a fraudulent scheme. The punishment for these crimes is usually less than the punishment for fraud. If the taxpayer is convicted of fraud, these other crimes should not apply. These crimes include submitting false documents and interference with tax administration through libel, slander, or other means designed to influence official action either positively or negatively.145

Criminal offenses would be in addition to civil penalties. They can be subject to flat fines and even to terms in jail. However, as with all penalties, high criminal penalties may only result in taxpayers’ taking greater care to disguise their fraud.146 Depending on the country’s legal tradition, the provisions imposing a criminal penalty could be included in the tax administration law or in the criminal code. Wherever located, the general rules of criminal procedure should apply.

7. Tax Administrator Penalties

Unless the matter is governed adequately in the other laws, special civil and criminal penalties can be applied for tax administrators. They can include civil penalties for negligent failure to follow accepted procedures or to respect taxpayer rights. Depending on the jurisdiction, these penalties could be imposed as damages based on taxpayer suits as part of (or an addition to) the general law of civil liabilities. Permitting taxpayer suits in these cases could act as a substantial deterrent to corrupt behavior on the part of officials. Criminal behavior of tax administrators may already be penalized under the criminal code.147 Some laws also provide for criminal penalties for private income tax preparers who disclose confidential taxpayer information.148

Note: Victor Thuronyi, Melinda Milenkovich, Milka Casanegra de Jantscher, U.I. Dharmawansa, V.S. Rekhi, Reinier Kraakman, Ward Hussey, Leif Mutén, and Pamela Sak each provided extremely helpful comments and corrections.


Interestingly enough, Freud described the borderline between psychosis and neurosis as when the mind acts like a bureaucracy. Sigmund Freud, An Outline of Psychoanalysis 48 (1911).


For a tax administration, the opposite of functional organization is organization according to substantive taxes. While many tax administrations are organized in this way, the general trend is toward a functional organization. See Aldo Schlemenson, Organizational Structure and Human Resources in Tax Administration, in Improving Tax Administration in Developing Countries 343 (Richard M. Bird & Milka Casanegra de Jantscher eds., 1992). See also Richard M. Bird & Milka Casanegra de Jantscher, The Reform of Tax Administration, in id. 1, 9.


For example, The United Kingdom’s Taxes Management Act includes statutory subdivisions on returns, assessment and claims, dispute settlement, recovery of monies owed, and penalties for tax offenses. See GBR TMA §§ 7–12, 29–3, 48–56, 60–70, 93–106. The U.S. Code, in Subtitle F on procedure and administration, includes sections on returns, assessment, collection, dispute resolution, tax offenses and penalties, and taxpayer ombudsperson. See USA IRC §§ 6001–7873. Belgium’s code includes sections on returns, audits and investigation, assessment, dispute settlement, and penalties. See BEL CIR arts. 297–463.


An additional functional category sometimes found in tax administration laws, one that forms the necessary precursor to the creation of the administrative bureaucracy, provides for the creation of the tax administration itself. However, in some jurisdictions the rules concerning the creation of departments or administrative authorities may be found in a separate law. See, e.g., DEU FVG.


See the discussion of who or what constitutes a taxpayer, infra sec. II(B).


See, e.g., KAZ TC arts. 142, 161–63 (articles regarding taxpayer rights, penalties for overdue tax payments, fines for late filing of returns, penalty for understatement of taxes, current payments, and objects of taxation); FRA CGI arts. 1725–56 septies (articles regarding penalties); GBR TMA §§ 86–106 (sections on interest on overdue tax and tax penalties).


See, e.g., Grundgesetz arts. 104a–115 (DEU) [hereinafter GG]; 5 U.S.C. §§101–306 (USA).


See, e.g., the Administrative Procedure Act, 5 U.S.C. §§ 551–83 (USA) [hereinafter APA]; and the Verwaltungsverfahrensgesetz (DEU).


See, e.g., Code of Civil Procedure §§ 166–213a, 511–44, 803–71 (DEU) (procedure for service of process, appeals, and execution levied on movable property and real property); Code of Civil Procedure arts. 411–20, 542–60, 651–94 (FRA) (procedure regarding representation, notice, and appeals).


See infra sec. III(C)(6).


William of Occam, Quodlibeta Septem (1320), quoted in John Bartlett, Familiar Quotations 143 (15th ed. 1980). This is the original statement of Occam’s Razor.


Examples of such provisions are included in the tax laws of the United States, France, and Germany; these codes include sections regarding tax liens and seizure of the taxpayer’s property. See USA IRC §§ 6321–27, 6331–44; FRA CGI arts. 1920–29 septies; DEU AO §§ 281–308.


For instance, in Australia provisions regarding returns and assessments are contained in the Income Tax Assessment Act, while in the United Kingdom such provisions are included in the Taxes Management Act. See AUS ITAA §§161–77; GBR TMA §§ 7–18, 29–40. Provisions regarding assessment are also contained in Germany’s income tax law. See DEU EStG §§ 25–28.




In Australia, for example, the Taxation Administration Act 1953 and the Crimes (Taxation Offenses) Act 1980 both concern themselves with tax offenses. See AUS TAA §§ 8B–8Z; AUS CTO.


See GBR ICTA § 831 et seq.; AUS ITAA § 6 et seq.; FRA CGI; CIV CGI.


See, e.g., VEN COT §§ 19–29. French tax law draws a distinction between a contribuable (taxpayer) and a redevable. The former is the person in whose name the tax obligation is legally established; the latter is a person from whom the law may authorize the tax authorities to require payment of the tax obligation (e.g., a withholding agent or a person jointly liable for payment of the tax). See Gilbert Tixier & Guy Gest, Droit fiscal 222–23 (3rd ed. 1981). Confusingly, however, the term redevable is also used in a meaning synonymous with that of “taxpayer,” in the cases of taxes such as the VAT, the wealth tax, and the taxe professionelle. See Précis de fiscalité ¶¶ 2000, 4890, 6176 (1994); FRA CGI Titre II, ch. 1, sec. VI (Redevables de la taxe).


For example, in the United States, the Administrative Procedure Act [APA] applies to any “authority of the Government of the United States” that is not Congress, the courts, or a military authority. 5 U.S.C. § 551(1) (USA). Under the APA, the Treasury Department may issue regulations and rulings relating to internal revenue laws only after publication of a notice of proposed rule making, followed by public hearings. Id. § 553. A failure to comply would result in the regulation being invalid. See American Standard, Inc. v. United States, 602 F.2d 256 (Ct. Cl. 1979).


For further discussion of this issue, see supra ch. 2, sec. (IV)(D).


See, e.g., Treas. Reg. § 601.210(a), (e) (as amended in 1983) (USA). See also ch. 2, sec. IV(E).


See Treas. Reg. § 601.201(a)(2) (USA).


See id. § 601.201(1).


See, e.g., Id. § 601.201 (a)(2).


See, e.g., DEU AO §§ 149–53.


See, e.g., DEU EStG §§ 25, 1; DEU KStG § 49.


See, e.g., USA IRC §§ 6041, 6041A, 6042, 6044, 6045, 6049.


See id.


For example, a VAT input credit is typically allowed only if the taxpayer has an invoice from the supplier.


See, e.g., USA IRC § 6001; CAN ITA § 230; DEU AO §§ 140–48.


See, e.g., USA IRC § 7608; GBR TMA § 20C; CAN ITA § 231.3.


See, e.g., USA IRC § 7609; CAN ITA § 231.2.


See infra ch. 12.


For example, U.S. law permits the Internal Revenue Service (IRS) to “compromise any civil or criminal case arising under the internal revenue laws prior to reference to the Department of Justice for prosecution or defense.” USA IRC § 7122; see also Treas. Reg. § 301.7122-1 (as amended in 1960) (USA).


In the United Kingdom, France, and Germany, payment may be suspended when the assessment is under appeal, while in Italy suspension of payment is not permitted. Organization for Economic Cooperation and Development, Taxpayers’ Rights and Obligations: A Survey of the Legal Situation in OECD Countries 99 (1990).


In the United Kingdom, the taxpayer may, in writing, appeal a tax assessment within 30 days after the date of notice of the assessment. GBR TMA § 31.


See id. § 32.


See infra sec. II(K)(9) concerning taxpayer rights to appeal.


See, e.g., USA IRC §§ 7441–75.


See id. § 7452; see also infra ch. 5.


For example, in France, general principles of administrative law require that the legality of an administrative decision can always be reviewed by an administrative judge. See Judgment of Feb. 7, 1947, Conseil d’Etat, 1947 Recueil des arrêts du Conseil d’Etat [Lebon], No. 79128, at 50. Such decisions are then themselves appealable to the Conseil d’Etat. See Judgment of Oct. 19, 1962 Conseil d’Etat, 1962 Lebon, No. 58502, at 552.


See, e.g., id. See also USA IRC § 7482.


For example, in the United States, a written report [called a Revenue Agent’s Report or RAR] concerning the proposed changes to the taxpayer’s return, including explanations, is prepared after each examination. See Internal Revenue Manual 4237, Report Writing Guide for Income Tax Examining Officers § 231, MT 4237-17 (Apr. 23, 1987) (Basic Report), cited in Michael I. Saltzman, IRS Practice and Procedure ¶ 8.06[8] n.121 (2nd. ed. 1991). Such a report is beneficial to both the taxpayer and the tax administration. If the case is not settled, the Appeals Office prepares a memorandum discussing its decision. Id. ¶ 9.05[3]. The taxpayer may obtain a copy of this memo under the Freedom of Information Act. Id. n.5.


See 5 U.S.C. § 552(b)(5), (b)(7) (USA).


See Tax Ct. R. Prac. & Proc. 70(a)(1) (USA); Cl. Ct. R. 26 (USA); Fed. R. Civ. P. 26 (USA).


5 U.S.C. § 555(c) (USA).


For example, under French administrative law, each party must prove its case based upon the materials available on file with the court. However, a failure to reply, whether to the rapporteur or to the tax authority, allows the court to draw the inference that the party in default has no case to make in answer to the question. The effect is that the burden of proof shifts to the party who has the materials. See Judgment of May 28, 1954, Conseil d’Etat, 1954 Lebon, Nos. 28238, 28493, 28524, 30237, 30256, at 308.


For example, in the United States, under the decision of the Supreme Court in Welch v. Helvering, 290 U.S. 111, 115 (1933), the assessment of the Internal Revenue Service is presumed to be correct. Rule 301 of the Federal Rules of Evidence states that “in all civil actions… a presumption imposes on the party against whom it is directed the burden of going forward with evidence to rebut or meet the presumption....;” The taxpayer has the burden of proving the assessment wrong. Helvering v. Taylor, 293 U.S. 507, 514 (1935).


See, e.g., USA IRC § 6321.


See, e.g., GBR TMA § 61; DEU AO § 281.


For example, in the United Kingdom, if the taxpayer does not pay the sum due within five days of the seizure of property, the seized property will be sold at public auction to pay sums owed. GBR TMA § 61(4), (5), Provisions for public auction of seized property are also contained in the tax laws of Germany and the United States. See DEU AO § 298; USA IRC §§ 6335, 6336.


See, e.g., DEU AO § 309.


For example, the U.S. Code authorizes the IRS to enter into a written agreement with the taxpayer allowing the taxpayer to make payment in installments if such agreement would facilitate collection of the tax owed. USA IRC § 6159. See also GBR ICTA § 5(2), which in certain cases permits the taxpayer to satisfy his or her tax liability in two equal installments.


See, e.g., Code of Civil Procedure art. 1167 (FRA); Mass. Ann. Laws ch. 109A (Law. Co-op. 1995); Bankruptcy Code, 11 U.S.C. § 548 (USA).


See, e.g., USA IRC § 7122.


Countries with an office of ombudsperson include Australia, Austria, Denmark, France, and the United States; Canada, Germany, Italy, and Japan are among the countries without an ombudsperson. See OECD, supra note 34, at 20, 76–77. See also USA IRC § 7811.


See generally DEU AO §§ 233–39 (provisions regarding assessment of interest, including when interest payments are assessed and how they are calculated).


See USA IRC § 6601(e).


For example, in the United States, the rate is determined using the rate on treasury obligations. See id. §§ 6621, 1274(d).


For example, in the United States the overpayment rate bears an interest rate 1 percentage point lower than the underpayment rate. Id. § 6621(a)(1), (a)(2).


Some countries even have official documents outlining the taxpayer’s rights; examples of such documents include Canada’s Declaration of Taxpayer Rights (1985), France’s Charte du contribuable (1987), New Zealand’s Statement of Principles (1986), and the United Kingdom’s Taxpayer Charter (1986). OECD, supra note 34, at 70. See also ch. 2, sec. II(F).


See, e.g., DEU AO § 30; USA IRC § 6103.


See, e.g., DEU AO § 122; USA IRC §§ 6323, 6331(d), 6335.


See USA IRC § 6331(d)(3).


See id. § 7605.


See supra note 42.


In the United States, this right is guaranteed under the APA (USA).


See, e.g., USA IRC § 7521.


See supra sec. II(F)(5).


See USA IRC § 7458.


See supra sec. II(F)(4).


See GBR TMA § 34 (assessment of tax generally may not be made later than six years after the chargeable period); USA IRC § 6501(a) (assessment of tax made within three years after tax return was filed).


In the United Kingdom, if tax was not paid on account of the taxpayer’s fraudulent or negligent conduct, assessment may be made within 20 years of the chargeable period. GBR TMA § 36(1), If a U.S. taxpayer files a fraudulent return, files no return, or in any other way willfully attempts to evade tax, assessment may be made at any time. USA IRC § 6501(c)(1), (c)(2).


Much of the pre-1990 taxpayer compliance research has been summarized and reviewed in two indispensable volumes: 1 Taxpayer Compliance (Jeffrey A. Roth et al. eds., 1989) and 2 Taxpayer Compliance (Jeffrey A. Roth & John T. Scholz eds., 1989).


”Despite the huge economic literature on tax evasion, which fundamentally focuses on the appropriate role and implementation of penalties … there seems to be no empirical study of penalty design and effectiveness in developing countries. What studies do exist, mostly for the United States, appear to be both model- and country-specific … and cannot easily be generalized to the quite different circumstances of developing countries.” Bird & Casanegra de Jantscher, supra note 2, at 5 n.8 (citations omitted).


See the discussion of this issue in Robert Kidder & Craig McEwen, Taxpaying Behavior in Social Context: A Tentative Typology of Tax Compliance and Noncompliance in 2 Taxpayer Compliance, supra note 73, at 47, 64–66.


Terms often used by writers on the topic include “unknowing noncompliance,” “lazy evasion,” or “lazy noncompliance.” Taxpayer Compliance, supra note 73, at 20. Other different, but essentially redundant, terms are also used. See the reviews of the literature in Kidder & McEwen, supra note 75, at 50–62. Some rules may be particularly difficult to comply with. For example, in the United States, a special task force of the Internal Revenue Service singled out the complexity of the estimated tax rules that “frustrate[] taxpayers, particularly individuals, to the extent that penalties are an acceptable alternative to compliance (emphasis added).” Commissioner’s Executive Task Force, Report on Civil Tax Penalties 15 (Feb. 27, 1989).


Besides making it easier for a taxpayer to comply, withholding also makes it more difficult for a taxpayer not to comply. See vol. 2, chs. 14, 15. This was also the conclusion of the United States Internal Revenue Service Commissioner’s Executive Task Force, Report on Civil Tax Penalties. Commissioner’s Executive Task Force, supra note 76, at 37–42.


For example, banks can be required not only to withhold and remit taxes on periodic payments, such as interest, but also to prepare and file declarations for income other than payments made by the bank itself. See Carlos A. Silvani & Alberto H.J. Radano, Tax Administration Reform in Bolivia and Uruguay, in Improving Tax Administration in Developing Countries, supra note 2, at 19, 29, 54–56; Charles E. McClure, Jr. & Santiago Pardo R., Improving the Administration of the Colombian income Tax, 1986–88, in id. 124, 132. Of course, without a self-assessment system, some of the effort of preparing tax declarations and computing liability can be shifted to the administration. However, given the limited amount of administrative resources in developing countries, this hardly seems to be a wise idea.


This also has the effect of shifting costs to those intermediaries. See Jaime Váiquez-Caro, Comments, in Improving Tax Administration in Developing Countries, supra note 2, at 145, 150.


See the discussion of the definition of the term “taxpayer,” supra sec. II(B).


See, e.g., Henry J. Aaron & Harvey Galper, Assessing Tax Reform 42–44 (1985). Their call for legal simplification has been quoted often in tax compliance literature.


Joel Slemrod discusses these issues in Complexity, Compliance Costs, and Tax Evasion, in 2 Taxpayer Compliance, supra note 73, at 156, 157–74. Of course, not all avoidance (or the organizing of one’s affairs so as to reduce obligations without also evading) constitutes a compliance problem. Administration is concerned only with avoidance that results in a reduction in normative (or “correct”) taxpayer obligations.


See generally Richard K. Gordon, Tax Administration Concerns in the Reform of Substantive Personal Income Tax Law in Emerging Economies, 46 Bulletin for International Fiscal Documentation 163 (1992).


Sanctions can be divided into types using a number of different criteria: civil and criminal, fines and imprisonment, per violation or per amount of tax forgone, or culpability of violation.


There are two basic types of compliance failures: failures to provide accurate information when due and failures to remit the correct amount of money when due. The former (information) is important to the administration only in that it allows the latter (money) to be properly computed and remitted. Nevertheless, it can sometimes be easier to identify and assess a failure to provide information necessary for the proper remittance of tax simply by reference to the actual failure itself, rather than by attempting to put a money value on it.


In the United States, for example, failure to file a return or to pay tax will result in sanctions unless such failure is due to reasonable cause and not willful neglect. See USA IRC § 6651 (a)(1), (a)(2). In Belgium, if the taxpayer fails to file a return or files an incomplete or inexact return, he will be sanctioned. However, in the absence of bad faith, the taxpayer’s sanction may be waived. BEL CIR art. 444.


Tax codes encourage the taxpayer to pay his or her taxes by having the amount of the fine increase the longer the taxpayer withholds payment. See USA IRC § 6651 (a)(1), (a)(2); GBR TMA § 93; BEL CIR art. 414; FRA CGI art. 1727.


For example, the laws of both the Russian Federation and Kazakstan impose 100 percent penalties for understatement of tax, regardless of whether the taxpayer is at fault. RUS TS art. 13; KAZ TC art. 163.


Penalties for such crimes as willful tax evasion or fraud may include monetary fines or imprisonment or both. See USA IRC §§ 7201–16. Willful tax evasion, for example, is considered a felony in the United States and is punishable by a fine of up to $100,000 or imprisonment of up to five years, or both, along with the costs of prosecution. Id. § 7201.


Jeremy Bentham, the father of English utilitarianism, saw the choices between possible modes as based on a calculation of risks of pain and pleasure. This view is still a basic premise of most discussion of deterrence. See Johannes Andenaes, Does Punishment Deter Crime? 11 Crim. L. Q. 76, 79(1968).


The legal scholar and economist Steven Shavell describes the equation as one of the magnitude and probability of harm versus the magnitude of benefit. Steven Shavell, Criminal Law and the Optimal Use of Nonmonetary Sanctions as a Deterrent, 85 Colum. L. Rev. 1232, 1236–38 (1985). A general discussion of rational choice theory in the context of taxpayer compliance can be found in Alfred Blumstein, Models for Structuring Taxpayer Compliance, in Income Tax Compliance: A Report of the ABA Section of Taxation, International Conference on Income Tax Compliance 159, 160–61 (Phillip Sawicki ed., 1983). This was also the conclusion of the United States Internal Revenue Service Commissioner’s Executive Task Force, Report on Civil Tax Penalties, supra note 76. at 13–14.


Most analyses of economic crimes do assume that criminals are principally profit maximizers. See, e.g., the discussion in Michael Gerken & William R. Gove, Deterrence: Some Theoretical Considerations, 9 Law & Society Rev. 497, 497 (1975). With the exception of tax protesters, it is probably a good guess that tax evaders almost always do so to maximize monetary profit. Therefore, the differences in individual utility functions might be relatively less than among other criminals, and sanctions might be relatively easier to design. See generally Robert V. Stover & Don W. Brown, Understanding Compliance and Noncompliance with Law: The Contributions of Utility Theory, 56 Social Science Quarterly 363, 374–75 (1975), Unfortunately, as discussed below, even this simplification turns out to be unlikely.


These two specific issues are addressed in greater detail at the text accompanying notes 120–24.


To maximize social utility, the severity of sanctions against a particular behavior would have to increase as a particular individual’s utility (in that behavior) increases. Samuel Kramer, An Economic Analysis of Criminal Attempt: Marginal Deterrence and the Optimal Structure of Sanctions, 81 J. Crim. L. & Criminology 398, 399 (1990).


Or, as Professor Shavell so succinctly puts it, “it is impossible to deter a person with no assets by the threat of monetary sanctions.” Shavell, supra note 91, at 1237. This means that monetary sanctions can rarely be enough to act as a sufficient general deterrent to undesired behavior.


See supra ch. 2, sec. II(A). This issue is also emphasized in the United States Internal Revenue Service Commissioner’s Executive Task Force, Report on Civil Tax Penalties. See supra note 76, at 13.


It would also probably be counterproductive. See the discussion concerning possibly counterproductive aspects of relatively high penalties, infra at text accompanying notes 114–17.


A brief survey of some of the literature in the American context can be found in Scott H. Decker & Carol W. Kohfeld, Certainty, Severity, and the Probability of Crime: A Logistic Analysis, 19 Policy Studies Journal 2, 3–6 (1990).


Some examples of research confirming this point include Robert Mason & Lyle D. Calvin, A Study of Admitted Income Tax Evasion, 13 Law & Society Rev. 73, 85, 87 (1978); Harold G. Grasmick & Donald E. Green, Legal Punishment, Social Disapproval, and Internalization as Inhibitors of Illegal Behavior, 71 J. Crim. L. & Criminology 325, 327 (1980).


Many of these studies, each of which is concerned with income tax compliance in the United States, are summarized in 1 Taxpayer Compliance, supra note 73, at 97–110.


See id. 27, 30, 88, 97, 107–10; Steven Klepper & Daniel Nagin, The Anatomy of Tax Evasion, 5 Journal of Law, Economics & Organization 1 (1989).


See 1 Taxpayer Compliance, supra note 73, at 100.


This is not terribly surprising when considering the popularity of lotteries. It is not just a question of whether some people are risk-averse while others are not, but a more general, systemic inability to judge risk accurately.


This is discussed in Charles R. Tittle & Alan R. Rowe, Certainty of Arrest and Crime Rates: A Further Test of the Deterrence Hypothesis, 52 Social Forces 455, 456, 458 (1974). The article looked at general crime rates in the U.S. state of Florida.


See generally Philip J. Cook, The Economics of Criminal Sanctions, in Sanctions and Rewards in the Legal System: A Multidisciplinary Approach 50, 54 (Martin L. Friedland ed., 1989).


For example, smaller groups with a high level of group communication may have an earlier tipping point and perhaps an increased tipping effect. Don W. Brown, Arrest Rates and Crime Rates: When Does a Tipping Effect Occur? 57 Social Forces 671, 680 (1978).


There is a dearth of research into tipping and tipping points in tax compliance, even in the literature concerning income tax compliance in the United States. However, there is some. See the review in 1 Taxpayer Compliance, supra note 73, at 111.


See the discussion in Raymond Paternoster et al., Perceived Risk and Social Control: Do Sanctions Really Deter? 17 Law & Society Rev. 457, 458 (1983).


Depending on the jurisdiction, publication of the names of those taxpayers who are subjected to sanctions may raise issues of taxpayer confidentiality. For example, the U.S. Internal Revenue Service may only disclose to the general public a taxpayer’s name and assessed penalties when a compromise is reached with the taxpayer before the case is referred to the Justice Department for prosecution. USA IRC § 6103(k)(1).


If the taxpayer is uncertain as to the effectiveness of future evasive action, taxpayer costs will increase, in that uncertainty of success increases costs, while return remains the same. See James Alm et al., institutional Uncertainty and Taxpayer Compliance, 82 American Econ. Rev. 1018, 1018–20 (1992). The author’s understanding of these issues was greatly enhanced by a number of discussions with Professor Reinier Kraakman of the Harvard Law School.


Despite the statements of some revenue authorities to the contrary, most have accepted that the goal of tax administrators is not to maximize net take, but to maximize total welfare. See Richard Goode, Some Economic Aspects of Tax Administration, 28 IMF Staff Papers 249 (1981).


See Jonathan Skinner & Joel Slemrod, An Economic Perspective on Tax Evasion, 38 Nat’l Tax J. 345, 350(1985).


There are other possible effects. As taxpayers take greater evasive action, they may be inspired to commit even greater crimes. Also, if one method of avoidance or evasion becomes too difficult, they may abandon it but turn to others. See the discussion with regard to general deterrence theory in Jeffrey Grogger, Certainty vs. Severity of Punishment, 29 Economic Inquiry 279 (1991).


See the survey of data in Decker & Kohfeld, supra note 98; see also the discussion in the context of tax compliance in Steven Klepper & Daniel Nagin, The Criminal Deterrence Literature: Implications for Research on Taxpayer Compliance, in 2 Taxpayer Compliance, supra note 73, at 126, 135–36, 143–44.


See, e.g., Harold G. Grasmick & George J. Bryjak, The Deterrent Effect of Perceived Severity of Punishment, 59 Social Forces 471, 472, 475 (1980). They state that “the conclusion directly challenges the basic premise of deterrence theory that man is a rational actor.” Id. at 472. They then go on to posit that this conclusion cannot not be correct, and that the data or analysis leading to the conclusion must be wrong. Id. at 473, Some might argue that this places theory above empiricism, an assertion that most scientists, at least since Newton, would fault.


See infra text accompanying notes 120–21.


There may also be a form of substitution effect where taxpayer compliance improves in those areas where a violation might draw the administrator’s attention, but whose savings are less than those where certainty has not reached a tipping point. See the discussions of substitution effects in Michael J. Graetz et al., The Tax Compliance Game: Toward an interactive Theory of Law Enforcement, 2 Journal of Law, Economics, and Organization 1 (1986); Klepper & Nagin, supra note 101, at 18–20.


Also, it is often the case that the greater the severity of sanctions, the less likely that they will be fully applied. See Vito Tanzi & Parthasarathi Shome, A Primer on Tax Evasion, 40 IMF Staff Papers 807, 812 (1993).


This point was also emphasized in U.S. Internal Revenue Service Commissioner’s Executive Task Force, Report on Civil Tax Penalties, supra note 76, at 13.


See Grasmick & Green, supra note 99, at 327–29.


See Sally S. Simpson & Christopher S. Koper, Deterring Corporate Crime, 30 Criminology 347, 367 (1992).


See Yankelovich, Skelley & White, Inc., Taxpayer Attitudes Study: Final Report (Public Opinion Survey Prepared for the Public Affairs Division, Internal Revenue Service, 1984). The study seems to suggest that perhaps large-scale evasion would not be seen as a victimless crime.


The Leona Helmsley case in New York, where a wealthy hotel heiress was prosecuted for tax evasion, generated considerable support for the state tax administration. In India, arrests for tax evasion of major industrialists have also been popular. See Richard K. Gordon, Jr., Income Tax Compliance and Sanctions in Developing Countries, in Taxation in Developing Countries 455, 461 (Richard M. Bird & Oliver Oldman eds., 4th ed. 1990).


Scholars may disagree as to whether the existence of personal morality is part and parcel of rational choice theory or describes another theory of human interaction. Some, for example, contrast rational choice models with “psychiatric models” of moral inhibitions or internalized norms. Andenaes, supra note 90, at 78–79. However, one can also include morality and internalized norms as aspects of individual utility functions. The important issue, however, is that people do not act solely to maximize dollar profit.


That is, to avoid doing things that are wrong in themselves, as opposed to wrong because prohibited. See James J. Teevan Jr., Subjective Perception of Deterrence (Continued), 13 Journal of Research in Crime & Delinquency 155, 157 (1976).


See Gerken & Gove, supra note 92, at 502.


See Gordon, supra note 123, at 462. See also Virendra Singh Rekhi, Notes on Legal Methods of Combating Corruption: Lessons from the Indian Experience (Nov. 15, 1995) (on file with the Legal Department, International Monetary Fund).


See Commissioner’s Executive Task Force, supra note 76, at 13–15.


See USA IRC § 6651 (a); BEL CIR art. 444; DEU AO § 152. See also Commissioner’s Executive Task Force, supra note 76, at 67–68.


Upon failure to pay tax, the U.S. Code presumes that the failure to pay is unreasonable; a fine will be imposed unless the taxpayer can show that such failure was due to reasonable cause. USA IRC § 6651(a)(2).


In almost all of the OECD countries, the one exception being New Zealand, interest is imposed if the taxpayer does not pay his or her taxes on time. OECD, supra note 34, at 18–19, 62–66 (1990). Most of the OECD countries also compensate the taxpayer for overpayment with interest. Id. at 20, 83–84.


The Colombian tax law is an example of such a system. In Colombia, “[i]f the taxpayer agrees to settle at the time of the initial field audit by the tax authorities, the penalty is 20 percent … of the underpayment. If a formal demand for supplementary payment made by the tax administration is accepted by the taxpayer before the case goes to court, the penalty is 40 percent. If the taxpayer agrees to the increased assessment after the case goes to court, but before the final judicial determination of liability, the sanction is 80 percent,” McLure & Pardo, supra note 78, at 136. Thus, Colombian tax laws encourage settlement of disputes.


See USA IRC §§ 7201–202 (willful tax evasion and failure to collect or pay tax are felonies); DEU AO § 370(3)1 (penalty for tax evasion is increased if committed out of gross self-interest and is of a large scale).


See, e.g., USA IRC § 7203 (penalty imposed for willful failure to file a tax return); FRA CGI art. 1725 (penalty imposed for failing to file required documents). Similar points were raised in the United States Internal Revenue Service Commissioner’s Executive Task Force, Report on Civil Tax Penalties. See supra note 76, at 15, 42–43.


See also Rekhi, supra note 127.


For example, the U.S. Internal Revenue Service Commissioner’s Executive Task Force, Report on Civil Tax Penalties concluded that the civil penalty for underpayment owing to negligence (i.e., no reasonable care) should be 20 percent, with a de minimis rule, while the penalty for intentionally underpaying (although without fraud) should be 50 percent. Supra note 76, at 67–68.


For example, the U.S. Internal Revenue Service Commissioner’s Executive Task Force, Report on Civil Tax Penalties concluded that the civil penalty for underpayment due to fraud should be 100 percent. Id. at 68.


See, e.g., Bankruptcy Code, 11 U.S.C. §§ 507(a)(7), 523(a)(1) (USA).


No OECD country provides for prison terms for civil tax offenses. See OECD, supra note 34, at 19. While in most countries imprisonment is one possible sanction for tax crimes, it is very rarely imposed.


For example, in the United States and the United Kingdom, tax deficiencies may result in a loss of privileges, including the ability of attorneys and accountants to practice their trade. Id. at 19, 67–69.


USA IRC §§ 3402, 3403, 3505, 3509, 6672; see also AUS ITAA §§ 222AOA–222AOD.


USA IRC § 6672.


This is the case in the United States; while the IRS may assess 100 percent penalties against all responsible persons, it may enforce such assessments only until it has collected an amount equal to the tax liability. Gens v. United States, 615 F.2d 1335 (Ct. Cl. 1980).


See, e.g., USA IRC §§ 7201, 7202.


For example, in the United States, the Internal Revenue Code includes the following acts as crimes punishable by fines and prison: making fraudulent statements in a tax return or information return, making fraudulent statements under penalty of perjury, and removing or concealing information with intent to defraud. USA IRC §§ 7204–207.


See discussion supra sec. III(C)(2).


For example, in the United States, the Internal Revenue Code includes the following acts as crimes punishable by fines and prison: numerous acts by revenue officers or agents, including extortion, bribery, conspiracy to defraud, and failure to report the illegal acts of others. USA IRC § 7214. The Internal Revenue Code supplies a cross-reference to a penalty provided in the U.S. criminal law relating to officers of the United States who trade in public funds, debts, or property. Id. § 7214(c).


See, e.g., id. § 7216.