Skattenytt nr 10 1997 s. 567
A critique of the Swedish rulings system
The Swedish model to provide advance rulings is unique, and it has worked well in Sweden since 1951.  Numerous jurisdictions have subsequently introduced advance rulings within their systems of tax administration. Why have they not followed the Swedish model? This article attempts to answer that question and to suggest successful aspects of other systems that Sweden might usefully adapt for its own use.
This article analyses the nature, objectives and key features of a rulings system. It uses this analysis to identify the advantages and disadvantages of the Swedish system. It is instructive to consider the approaches taken elsewhere, and a comparative perspective is used to explore the disadvantages, with particular reference to Australia. The article concludes with suggestions for improvement of the Swedish advance rulings system.
2 The nature and objectives of a rulings system
There are three main stakeholders in a rulings system: the government, the revenue authority and taxpayers. A successful rulings system should aim to satisfy each of the stakeholders to the greatest extent possible. A balanced critique of any system can only take place based on a clear understanding of the nature of a ruling and what it seeks to achieve. This section attempts to do this, with particular reference to the Australian system.
A ruling can be broadly defined as a decision or opinion as to how the tax authorities will interpret and apply the tax law to transactions or arrangements entered into by one or more taxpayers. Two kinds of ruling have developed: public or general rulings that are addressed to taxpayers generally, and private rulings (usually called advance rulings) that are addressed to particular taxpayers on application. The precise nature, definition and name of each type of ruling varies between jurisdictions.
Public or general rulings are administrative in nature and often do not have the force of law.  They come in a number of forms, but in many jurisdictions they determine the procedures and practices used by the revenue authorities to implement the tax law. For example, the United Kingdom Inland Revenue issues ”statements of revenue practice” that detail not only how the Inland Revenue will interpret and apply the law, but also include concessions to taxpayers to allow for the effective administration of the law.
The Australian Taxation Office takes a similar approach, but public rulings are defined by statute as the Commissioner of Taxation’s opinion as to the way the tax law applies to any person or class of persons in relation to an arrangement or class of arrangements.  The Commissioner must also state in the ruling itself when an opinion is a public ruling.  This has left it open to the Australian Taxation Office to publish public rulings in a number of forms: as broad guidelines; in short statements addressing one particular issue; in its publications, return form guides, information booklets and media releases; and in speeches of senior officers. 
Private or advance rulings provide taxpayers with an interpretation of how the law will apply to a particular taxpayer or group of taxpayers in relation to an arrangement. Usually such rulings are given in advance, before taxpayers enter into a transaction or arrangement. However, they may also be given to taxpayers after the event, but before a taxpayer files a tax return.  For example, in Australia, under the self assessment system, taxpayers may need to know how the tax law applies to a particular transaction that occurred much earlier in the tax year, in order to include it correctly in their tax return. If taxpayers record items incorrectly in their returns, they can be subject to interest and penalties on any tax shortfall.  Rulings of this kind are generally binding on the revenue authority. 
2.2.1 Defining the objectives
Introducing the new Australian binding rulings system to Parliament, the Minister Assisting the Treasurer said: 
”The new system of binding and reviewable rulings will promote certainty for taxpayers, and thereby reduce their risks and opportunity costs. The new system will also be fairer because taxpayers will be able to object to Private Rulings and have the matter reviewed by an independent tribunal or court.”
The speech identifies three main objectives for the rulings system: certainty; improvement in economic decision-making; and fairness. These objectives are common to other rulings systems, are advocated by most commentators and so form the basis for this analysis. 
Rulings are designed to give taxpayers increased certainty as to the interpretation and application of the tax law. A basic principle of most legal systems is that laws should not be retroactive.  Yet, it is a feature of tax law that there is more uncertainty in the interpretation of its provisions than is the case with much other legislation.
This is inherent in its nature, as, where most legislation is formative in a primary sense, tax is imposed as a gloss on other legislation: it relies on existing laws for its application. For example, contract law is formative, in that it determines when a contract is formed, what will cause a contract to fail and when a contract is breached. Tax law, on the other hand, depends upon transactions found in other laws, such as the existence of a contract, to impose a tax. In that sense it is not primary legislation. Thus capital gains may only arise on the disposal of an asset and a disposal may depend upon the contract as to occurrence, timing and the quantum of the gain. Income is usually taxable only because a contract exists to produce that income.
Immediately, there is a problem. Tax law does not always fit within existing legal forms. Accordingly, it often has to break transactions into their constituent parts in order to impose tax. For example, the tax treatment of convertible notes or zero coupon bonds usually cannot rest simply on the contracts creating them. Tax law has to break these instruments into their component parts in order to impose tax in a way that is consistent with the treatment of those component parts in the hands of other taxpayers. Tax law has rules governing the assessability and deductibility of interest, the treatment of debt and equity. Contracts creating hybrid instruments do not fit easily within the tax rules. The result is that tax law sometimes pulls other areas of law apart in order to impose its own rules. It is therefore necessarily complex. 
Complexity creates uncertainty. Uncertainty can give rise to effective retrospectivity in the application of the law. If taxpayers interpret the law consistently in a certain way over a period of time, only to find, subsequently, that the revenue authorities take a different interpretation and apply the law accordingly, the application is effectively retrospective. This is an inevitable consequence of complex legislation, but, if there are ways to reduce complexity, revenue authorities prefer to adopt them in order to reduce the incidence of retrospectivity. Retrospectivity in any form tends to undermine the credibility of the administration and collection process.  Rulings are a useful tool for revenue authorities to minimise retrospective application of the law and are widely used. 
Tax Reform is as important a topic approaching the year 2000 as it has been throughout most of the twentieth century.  Owens and Whitehouse note: ”tax systems in the latter half of the nineties look very different from the mid-eighties as fundamental tax reforms were implemented across the OECD”.  They identify complexity as one of the serious remaining problems.  They also warn of continued change, and change tends to compound complexity.  An obvious corollary in the face of increasing uncertainty is that rulings will become even more important to both taxpayers and revenue authorities. Reforms tend to succeed or fail depending upon the level of information provided to taxpayers.  In the context of reform, rulings are a useful medium for the dissemination of information, which is undoubtedly why their use is growing. 
Rulings aim to ensure fairness, in part through uniform interpretation and application of the law.  Where legislation is complex and tax administrations are large, it is often difficult to ensure consistency. One of the primary origins of rulings was as internal documents designed to inform revenue authority staff of the official view of the tax law as it applied to a range of common transactions.  The result was greater internal consistency and the rulings concept was extended to taxpayers to try and achieve external consistency. 
Where taxpayers perceive that revenue authorities are acting fairly and their experience of the revenue collection process is positive, compliance tends to improve.  By providing rulings, revenue authorities demonstrate to taxpayers a willingness to act consistently and to try to clarify the complexity of the law. Accordingly, rulings are an integral part of the attempt by revenue authorities to convince taxpayers that they are acting fairly.  A consequence should be, not only an increase in compliance, but also development of a perception that enforcement of the tax law is legitimate.  Where the law is clearer and more understandable to taxpayers, blatant breaches are less readily acceptable to the majority who are abiding by that law. 
2.2.4 Economic decision-making
A direct result of greater certainty, which rulings help to provide, should be a reduction in litigation over tax matters.  Where the views of the revenue authorities are widely known and can be ascertained in relation to specific transactions, there is much less room for litigation. If necessary, the structure of a transaction can be changed, if the consequences of not doing so are known in advance. Litigation is more likely where a transaction is already in place before the tax consequences are known, particularly if they are unexpected. The stakes are then commensurately higher for the taxpayers involved than they would have been if the consequences were known in advance and could have been taken into account in the decision-making process. A decrease in litigation should result in a decrease in associated transaction costs. 
Economic efficiency is identified by James and Nobes as one of the main economic criteria for a good tax system.  The unintended effects of taxation should not distort economic transactions.  That is why, at the macro level, governments often have so little real success when they attempt to manipulate investment decision-making using taxation as a broad policy tool.  Tax concessions often distort investment in ways that were not foreseen when the policies were introduced.  There is more scope for genuine efficiency at the micro level if the consequences of investment decisions can be predictable. Rulings can provide investment certainty where the law itself is not necessarily clear. In an environment where investment is seen as a crucial determinant of economic growth, the widespread use of rulings is understandable.
Most jurisdictions are flexible in providing rulings to investors where there is a clear economic or political advantage in an investment.  It is one of the main reasons why many countries provide rulings only to foreign investors. What it demonstrates, is that governments use rulings as a key part of the process for implementing economic policy, in the effort to provide investment certainty. The more sophisticated economies have extended that process into the domestic sphere.
3 Key features of a rulings system
Given the objectives of a rulings system, how can we determine whether they have been achieved? Those designing new rulings system or assessing existing rulings systems have identified certain key features which help in that process.  This section analyses each feature, comparing, in particular, the approaches taken in Sweden and Australia.
3.1 Statutory or administrative basis
In many countries, the development of a rulings practice has been gradual, and rulings form part of the administrative processes of the revenue authorities. Some countries, such as the United States and Sweden, have had a statutory basis for rulings for a long time. Others, such as Australia and New Zealand, have only recently moved to a statutory basis as part of the introduction of taxpayer self assessment. Historically, statutory rulings included relatively inflexible procedures.  This has changed over time and, where flexibility was a critical advantage of earlier administrative rulings, it is now integral also to rulings with a statutory basis.  A statutory basis provides legitimacy to the process, and stipulates more clearly the rights and obligations of taxpayers and the revenue authority in relation to rulings. 
3.2 Nature of the rulings
Systems that provide rulings usually do so in the form of private or advance rulings to individual taxpayers. Some systems also issue public or general rulings. These are discussed above at 2.1. Public or general rulings can be criticised in that, generally, they are issued by, and represent the opinion of, the revenue authorities. In Australia, the result is that the rulings have the effect of law in the majority of cases where they apply. This is because the transaction and opportunity costs of pursuing a dispute with the Australian Taxation Office make it easier, in most cases, to accept the ruling. 
3.3 Issuing body
In most countries, the revenue authorities provide the rulings. India has set up an independent authority chaired by a retired Supreme Court Judge and also comprising an officer of the Indian Revenue Service and an officer of the Indian Legal Service.  The Swedish Council for Advance Tax Rulings has a wider membership, suggesting an even greater degree of independence. A further point of difference between systems is whether rulings are issued centrally. In Australia, rulings can be issued by local offices. This approach is criticised on the basis that it can lead to inconsistencies and lapses in the quality of rulings.  Partly based on the Australian experience, New Zealand has recently introduced a centralised model.
3.4 Subject matter
Many jurisdictions restrict private or advance ruling applications to specified areas. For example, the Netherlands and India restrict rulings largely to activities involving non-residents. In contrast, Sweden and Australia provide rulings in most areas of the tax law to any taxpayer. Public rulings differ from private or advance rulings in that they are usually wide-ranging in their subject matter and apply to any taxpayer falling within the ambit of the ruling, or entering into transactions covered by the ruling.
3.5 Right of refusal
In many regimes the authority making rulings has a discretion whether to provide a ruling, although there are usually limits to the discretion. In Sweden, the ruling must concern a matter of exceptional importance to the applicant or of importance for the uniform interpretation of the law or judicial practice.  In practice, the ”exceptional importance” requirement is not interpreted strictly.
In Australia, there is a number of instances, specified in the legislation, where the Commissioner of Taxation need not make a ruling. The Commissioner also has a wide discretion to refuse to make a ruling if, in the Commissioner’s opinion, it would be unreasonable to do so, having regard to the resources available, or any other matter that the Commissioner considers relevant.  The discretion is not widely used, but it is so broad that critics query why there was any need to provide detailed legislative limitation on private rulings, when the limitations are still narrower than the Commissioner’s discretion.  The Commissioner also has a discretion to override the statutory limitations and provide a ruling anyway. 
In a rulings system, discretions to refuse rulings are generally acceptable on a number of grounds, such as: frivolous or vexatious applications; where the question is currently before the courts; in hypothetical cases; and in applications that fall outside the scope of the rulings regime. Of more concern, is a discretion to refuse an application on the grounds of resource limitations, unless the refusal is accompanied by full reasons.  Otherwise such a discretion could be used to conceal problems with resource management or with allocation of essential funding by Parliament. 
For a rulings system to succeed it needs to be flexible. There should be initial full disclosure of the relevant facts and circumstances relating to an application. The system should also allow the modification of an application in response to the views of the ruling authority, and the opportunity to accede to requests for further information.  Rulings aim to provide certainty and the process needs to facilitate clarification of ruling applications to ensure that the ruling addresses the right issues and that the applicant provides all relevant and necessary information.
The Swedish system provides a useful model, as it is the role of the secretariat supporting the Council for Advance Tax Rulings to perform those functions.  They not only ensure that the views of the taxpayer and the revenue authority are fully and properly presented with the application, but also provide their own recommendations with respect to the ruling. In contrast, in Australia, there is less emphasis on the quality of information provided or required in private ruling applications, which led the Australian Federal Court, in recent appeals against unfavourable private rulings, to find that the rulings should not have been made, as there was insufficient information to do so.  A rulings process should be comprehensive, flexible and allow all parties to clarify and provide submissions on any position.
If the aims of a ruling system are to provide taxpayers with greater certainty and fairness, as well as to make commercial decision-making easier, then timing is very important. This is recognised in Australia, where taxpayers have raised concerns over delays.  The ATO has implemented a project that aims to respond to routine requests for rulings within seven weeks and to complex requests for rulings within a period negotiated with the taxpayer.  In New Zealand there is a statutory four week period for the issue of a ruling, which may be extended only where the Commissioner of the Inland Revenue Department notifies the taxpayer as soon as is reasonable after the receipt of a ruling application, or as soon as is practicable when an earlier estimate needs to be revised.  The emphasis in most systems, is on providing rulings as quickly as possible and, usually, within two months.  In Sweden, rulings take longer, and this is discussed below in section 4.
One of the critical issues in any rulings system is the enforceability of rulings made. To satisfy the aims of any system, rulings should be binding on the parties, subject to a right of review. For example, in Australia, where a taxpayer relies on a private or public ruling, the Commissioner of Taxation is bound to apply the ruling, if it is in the taxpayer’s favour.  If a taxpayer ignores a private ruling a penalty applies of 25 % of any shortfall in tax paid, on the basis that there is a right of review of an unfavourable ruling.  If a taxpayer ignores a public ruling, it will be one of the factors that will determine whether the taxpayer has taken a reasonably arguable position when penalties are determined in the event of any shortfall in tax paid.  In Sweden, an advance ruling is binding in the assessment to tax of the applicant, to the extent that it is applicable and the applicant demands that it is applied. 
It would be even more beneficial if private or advance rulings could be relied on by third parties. There are several arguments against this.  It might make the ruling authority reluctant to provide a ruling in case of error. The time frame for rulings would extend, so as to ensure the quality and consider the effect of the ruling. It may be difficult to determine whether a third party had in fact relied on a ruling given to a taxpayer. If charges are made for rulings, it is unfair for third parties to benefit at the expense of the taxpayer who paid for the ruling. These arguments are fairly easily addressed, given the successful operation of public rulings in many jurisdictions, which operate under similar constraints. Also, third parties successfully rely on past court cases, which are paid for only by the parties. In Sweden, most important advance rulings are appealed to the Supreme Administrative Court and so become part of the published case law.
The problem remains for cases not taken on appeal, where, unlike in a court case, the factual basis for making a private or advance ruling is often confidential and cannot be published for other taxpayers to fully understand the context in which the private or advance ruling was given. It is difficult to resolve this issue in relation to specific rulings given to particular taxpayers.
A ruling may be made that is wrong in law, or the ruling authority may change its position on the interpretation of the law taken in a ruling, particularly following a court decision made subsequent to the ruling. A more obvious problem is where statutory amendments are issued which change the law that formed the basis for rulings already made. In such cases there is usually facility for withdrawal of rulings. Similar facility generally exists where an applicant makes a material omission or misrepresentation in a ruling application, or where the time period for entering into a transaction ruled upon expires.
Fairness and certainty suggest that where a taxpayer has already commenced a transaction on the basis of a ruling, the ruling should stand, except where legislation affecting the ruling is changed with retrospective effect. Otherwise, withdrawal once a transaction is undertaken undermines the basis of the rulings system. In Australia, the Commissioner of Taxation may withdraw a private ruling only for a year of income that has not commenced, unless it is with the consent of the person in receipt of the ruling.  In New Zealand, a ruling may be withdrawn provided the arrangement to which it relates has not commenced.  In Sweden a ruling has the effect of a decision of a court. This means that, provided the ruling is not overturned on appeal, or rendered inoperative by legislative amendment, it will stand.  However, the Council for Advance Tax Rulings may adopt a different interpretation of the law in subsequent ruling applications, if it changes its view of the meaning of the law or its interpretation. The approaches in these three jurisdictions all provide reasonable certainty to taxpayers who rely on rulings as to the tax consequences of transactions they wish to enter into.
OECD countries see appeals procedures as an important aspect of tax administration and as a means of securing legal protection for taxpayers.  However, appeals are often restricted to assessments of income and do not extend to earlier administrative decisions, such as the making of rulings.  This seems short sighted, given that rulings are integral to the broader assessment process. It also runs contrary to the principle of fairness and the desire to encourage investment.
Rulings are not law. Most rulings are given by revenue authorities, so it is unfair that, as administrative decisions as to the way the law will apply to an arrangement, they are not be open to appeal. It means that a taxpayer actually has to enter into the arrangement before an assessment is raised that is open to challenge, whereas the whole purpose of the ruling process is to ascertain the tax consequences beforehand. Where the taxpayer believes that an unfavourable ruling is an incorrect interpretation of the law, and there is no opportunity to challenge the ruling, in most cases the unfavourable ruling will add too much risk for the venture to proceed. This hardly represents encouragement of investment.
There are arguments against giving a right of appeal against unfavourable rulings. Some critics believe that it would create delays, but for most taxpayers a delay is not as critical as obtaining a certain outcome.  In Sweden, the appeal from advance rulings is usually faster than appeal against an assessment, as there is access directly to the Supreme Administrative Court, whereas an ordinary assessment has to go right through the ordinary court system.  In Australia, the appeal against an unfavourable private ruling follows the ordinary appeal process, but there are currently shortcomings in the operation of appeals against rulings, which are under review.  Nonetheless, the objective is to provide appeal rights to taxpayers against unfavourable rulings.
In New Zealand, one reason why there is no appeal is because the Inland Revenue Department must consult with an applicant before issuing a private ruling.  A second is that taxpayers are not penalised if they do not follow a private ruling. A third is that the courts do not want to rule on hypothetical situations.  These reasons were clearly not persuasive enough either in Australia or Sweden to deny applicants for rulings the right of appeal. However, in neither Australia nor Sweden does the right of appeal extend to a refusal by the ruling authority to make a ruling. A useful added protection for applicants would be a requirement to furnish reasons for such a refusal. In many jurisdictions this would then provide an avenue for challenge under administrative law, if the reasons contained major defects. 
Public or general rulings are published, but advance rulings addressed to individual taxpayers are seldom made public, for reasons of confidentiality, to protect the taxpayers concerned. In Sweden, the chair of the Council for Advance Tax Rulings publishes the decisions of the Council in general terms, in order to protect the identities of the applicants. Most important rulings are taken on appeal and, in that way, become part of the published case law. In Australia, the confidentiality requirements of the tax law prevent the Australian Taxation Office from publishing any details of private rulings, in case the taxpayers concerned could be identified.  New Zealand has also taken this approach. In both countries, however, if there is a matter of general concern that arises in a private ruling application, the revenue authorities will usually issue a public ruling on the matter to make their views widely known.
As a major reason for providing rulings is to ensure uniform application of the law, it would seem important that rulings are published. Sweden has managed to do this without breaching confidentiality and its example should be followed elsewhere. Where private or advance rulings are not published, as in Australia and Canada, a secret body of law can develop. Major legal and accounting firms can collate all rulings given to their clients and thereby gauge the attitude of the revenue authorities to various transactions. Smaller firms and individual taxpayers are discriminated against in that they do not have access to the details of a large number of rulings. 
There are several factors underlying a decision to charge a fee for rulings. In Australia, there is a self-assessment system and the rulings process is integral to that system. All taxpayers are responsible for lodging their own tax returns, on which they are generally assessed without query. If they are subsequently audited and are found to have understated their taxable income, penalties and interest apply. Accordingly, rulings provide certainty in a complex system. As such, they are tools to improve taxpayer compliance and so the cost is borne by the Australian Taxation Office. 
In Sweden there is a charge for rulings that aims to eliminate frivolous applications and to keep the workload of the Council for Advance Tax Rulings at a manageable level.  New Zealand provides an example of a system that charges fees to provide full cost-recovery for rulings, provided the charges are reasonable. The aim is to implement a user-pays philosophy in the provision of government services.  It also ensures that the Inland Revenue Department can cope with the rulings workload and has the resources to issue rulings in a timely manner, as is required by law. 
There is no definitive answer as to what is appropriate in relation to the charging of fees. From a taxpayers’ rights perspective, in a complex tax system, particularly where there is full or partial self-assessment, a right to a ruling with only a nominal charge is very attractive. This ensures that there is no discrimination against taxpayers who may be less well off, but who are uncertain as to the application of the law to their particular circumstances. As long as decisions as to what has to go into a tax return rest with taxpayers, it is surely their basic right to be able to understand how the tax laws apply to them. Particularly in complex systems, it is unfair if they are denied access to rulings because they cannot afford them. Even where minimal amounts of tax are involved, for a person on a low income these amounts will often represent as large a proportion of their total tax, and even their total income, as are at stake in ruling requests made by larger taxpayers.
4 Assessment of the Swedish system 
The Council for Advance Tax Rulings (the ”Council”) has nearly fifty years of experience in developing and operating a rulings process that is, in many respects, unique. Its most attractive features, for those operating within a democratic framework, are its independence and impartiality. They impact favourably on aspects of the rulings system common to other jurisdictions.
The Swedish system has a clear statutory framework, but within that framework it has considerable operational flexibility.  Its status, as an arm of the judiciary, eliminates the constraints that limit the effectiveness of a rulings body that functions as the arm of the executive, responsible for administering the tax law and collecting revenue. There is no perception or taint of bias. The Council can obtain expert opinion from a variety of sources and use that opinion to provide a correct interpretation of the law, without concern for the impact on compliance or revenue collection.
The Council reaps the benefits of operating as a centralised body. The rulings it makes are uniform in their interpretation of the law, they are consistent and are recognised within the community and the courts as being of a high quality. The quality of rulings is maintained by the broad make-up of the Council. It is an expert bench, drawn from a wide range of constituencies within the tax community, and operating within the constraints of judicial process. As a result, it has a broad perspective and is able to take a far more practical and realistic approach to interpretation of the tax law, than can most rulings bodies.
A disadvantage of having part-time members is that the work-load may prevent them from becoming familiar with all of the applications before them. This could impact upon the general quality of rulings over a period of time. However, similar criticisms can be made of any body where there are resource constraints. It is important to have systems in place that can identify when a lack of resources becomes a problem, so that additional resources can be provided before the rulings system is affected. This could become an issue in relation to efforts to speed up the ruling process. More resources could be provided in the form of an additional section to hear rulings. This might require the amendment of Section 2 SFS 1951:442.
The Swedish system is as broad as most in the subject matter covered by rulings. A liberal interpretation of what is meant by a matter of ”exceptional importance to the applicant”  has helped to make rulings accessible to a wider range of applicants than would a narrow construction. Nonetheless, the system does limit access to rulings. This is accomplished by requiring a very high standard of applications for rulings and limiting acceptable applications to questions that involve complex questions of law, or arrangements that have not previously been considered. The argument in favour of this approach is that the interpretation and application of the law is relatively straightforward in most cases, following the 1991 Swedish tax reforms. It is an application of the principle that it is only a small minority of taxpayers that enter into the complex transactions which the vast majority of the tax law is written to regulate.
This approach is perfectly acceptable if it is genuine. However, the fairness principle requires that a rulings system should be accessible to all taxpayers who would benefit from it in applying the tax law to their particular circumstances. There is no requirement for a rulings body to provide a tax advisory service. But if there are difficult questions of interpretation of the tax law that affect individual taxpayers, they should have the facility to request a ruling in exactly the same way as a large taxpayer. Charges for rulings also affect their accessibility, but the level of charges in Sweden is unlikely to act as a barrier to prevent taxpayers from applying for rulings.  Rather, it is the cost of compiling an application, with the extensive information required of applicants, that will often deter all but large taxpayers from requesting rulings. From a practical standpoint, access to rulings seems overly restrictive.
Nonetheless, the ruling process provides a model for any ruling system. Where other jurisdictions argue over consultation,  the Swedish system offers a wealth of different views. Its strength lies in the use of the secretariat to obtain a comprehensive application from the taxpayer, the views of the National Tax Board (the ”NTB”) and to provide its own expert opinion. The ruling is then made by a Council comprised of members from different sectors of the community with recognised expertise in the tax law.
A criticism of the process is that the NTB has not been able to make applications for advance rulings.  An argument against it is that, as an arm of the executive, the NTB should ensure that reforms to the law are presented to the legislature, if the interpretation of the law is unclear. Unfortunately, it is a character of language that the interpretation of law can neither be completely clear nor all-encompassing. As Hart states: 
”In every legal system a large and important field is left open for the exercise of discretion by the courts and other officials in rendering initially vague standards determinate, in resolving the uncertainties of statutes, or in developing and qualifying rules only broadly communicated by authoritative precedents.”
That is a role of the Council in relation to tax law, and it should be open to the NTB to apply for a ruling. The alternative is found in other jurisdictions, where the revenue authorities make the rulings. That is a course fraught with difficulties as it usurps the judicial function, and is discussed in 5 below. Giving the NTB the right to apply for a ruling would likely improve the rulings system. 
A major shortcoming of the Swedish rulings system has been the time taken for the Council to make a ruling, which, until recently, averaged about six months.  The principles of fairness and certainty are better served by a process that does not give significant timing advantages to one form of transaction over another. Usually timing problems relate to the availability and management of resources. The Council recognised this problem and has attempted to address it. However, although rulings are taking less time, there are many delays attributable, not to a lack of resources, but rather to the process itself. Applicants often do not provide sufficient information in their applications, or fail to provide additional information in a timely fashion when the Council asks for it. Many applicants also change or expand their questions during the process, often as a result of the opinion expressed by the NTB. The advantages arising from an extensive consultative process carry with them inevitable delays. Provided they are caused by the applicants themselves, and not the Council, the delays do not reflect badly on the ruling process.
The facility for review of rulings and, particularly, the direct access to the Supreme Administrative Court, enhances the rulings procedure. The judicial nature of the Council makes the decision-making process more accountable. The fact that rulings are reviewed by the highest tax court must also improve their quality. 
There is no necessity for review of the dismissal of a ruling application. If the dismissal were by a revenue authority, there would be strong arguments for review.  However, the Council is a judicial body following due process. The only avenue of appeal would be to the Supreme Administrative Court and it is reasonable for a judicial process to have only one formal consideration of a ruling application rather than burdening the highest administrative court with minor matters. If the Council refuses to make a ruling on a matter within its jurisdiction and provided the application is made in the proper form, a dismissal does in fact provide guidance to the applicant that the matter does not involve a problem of interpretation of the law.
Sweden is unusual in publishing details of private rulings. There is little more that could be done in this area, without jeopardising the privacy of the applicants.
5 Public or general rulings
The revenue authorities in most jurisdictions make public or general rulings. The Council does not. Guidelines are issued by the NTB from time to time. It is relevant to analyse whether, if a formal system of public rulings is introduced, the Council or the NTB should issue them.
In any democracy, the separation of powers is fundamental. In many countries, legislatures have authorised delegated arms of the executive to take on quasi-judicial roles. In most common law countries this is constitutionally acceptable provided the executive does not usurp the judicial function. The practical effect has been a narrowing of the meaning of the judicial function.  This is particular relevant to the making of public rulings by the revenue collection arm of the executive.
A rulings system is designed to interpret and apply the law. Rulings are given added authority in many jurisdictions, when they are made binding on the revenue authorities by statute. Although rulings may be classified as administrative guidelines to the operation of the law, that is a simplistic view. The purpose of a rulings system has been taken to be: (a) to fill out the law so that taxpayers can understand it and to interpret it so that it makes sense; (b) to provide administrative concessions where the legislature has erred until such time as statutory amendment can be passed; and (c) to explain how the revenue authority will exercise its discretions under the law. In a democratic system reliant upon the separation of powers, it is inappropriate that revenue authorities should exercise the first two of these functions. Consider some examples of the operation of the Australian rulings system.
First, the tax law provides that certain expenses incurred in earning assessable income are deductible. Over time, the case law has considered the deductibility of a range of specific expenses. The Australian Taxation Office (the ”ATO”) has now issued numerous public rulings detailing precisely what items of expenditure are and are not deductible in a wide range of occupations. Whether the rulings provide a correct interpretation and application of the law is irrelevant. The majority of Australians completing their tax returns treat the rulings as law. In time, there may be taxpayers who appeal against the ATO view in one or more areas, but, for the most part, the rulings stand as the accepted interpretation of the law. Yet, although they operate in practice as law, taxation rulings are neither a specie of delegated legislation nor judicial rulings. They are simply the official opinion of the Commissioner of Taxation.
Second, the Australian tax law raises some anomalies, which would be unfair to taxpayers if they were applied. The ATO has issued rulings which provide concessional treatment to taxpayers affected by those anomalies, so that the law will not apply as it stands. For example, in the tax treatment of partnership losses.  However advantageous this is for taxpayers, it means that an administrative body has substituted its own view for the law contained in a properly enacted statute.
Third, in Australia, the penalty provisions give the Commissioner of Taxation extensive discretions to remit all or part of a penalty.  How these discretions will be exercised is set out in public rulings.  In this context, the ATO is acting properly, within its delegated authority.
There are additional principles at stake in systems based upon the separation of powers. There is the principle that only Parliament can impose taxes. This is apparently the position in Sweden.)  Whether a public rulings program imposes taxes is arguable. However, the approach taken in many jurisdictions, including Australia, where the revenue authorities make public rulings, results in the effective imposition of taxes. The three Australian examples show how the ATO does more than simply interpret the existing law. It extends the law and determines when and how it will apply it. It has not been challenged for two reasons. First, in Australia, the delegated powers of the ATO are extremely broad, and rulings are legally no more than ATO opinions, which makes a constitutional challenge difficult to sustain. Second, the taxpaying community recognises the value of a public rulings system and wishes to improve it rather than challenge it.
Nonetheless, in Sweden, there may be serious constitutional difficulties with as broad ranging a public rulings system as exists in Australia.  That is why it would be better if public rulings were made by the Council, which is a judicial body that is keenly aware of the constitutional limits on a rulings programme. The Council would ensure that rulings follow directly from the applicable law and neither extend it, directly or by analogy, nor choose to ignore it if it presents difficulties.
Arguably, the NTB could issue public rulings as non-binding recommendations.  The danger in this is that even non-binding recommendations become binding in practice. Experience in Australia shows that the vast majority of taxpayers follow rulings issued, simply because the total transaction costs of entering into a dispute with the revenue authorities are so much higher than the tax involved. 
If the Council issues public rulings they would be perceived as independent and unbiased.  Ideally, the same principles should apply to a rulings body that rules on the interpretation and application of the law, as apply to any tribunal that makes similar kinds of interpretive rulings. The fact that public rulings are made to the public at large does not change their impact on individual taxpayers.
It is instructive to examine the decisions of the European Court of Human Rights (the ”Court”) on Article 6 of the European Convention on Human Rights: the right to a fair trial. They provide useful general guidelines on what are the attributes of an impartial tribunal. To be impartial, a tribunal should be independent of the executive if the executive is a party to a decision.  The executive is affected by every public ruling and so, if public rulings are to be impartial and independent, they should not be made by the NTB. Minority membership of the Council by representatives of the NTB would probably not impact upon the Council’s independence and impartiality.  Members of the Council would have to be both subjectively and objectively impartial.  Subjective partiality would occur if a member acted with personal bias.  Objective impartiality would demand that Council members inspire confidence in the public that their decisions are independent and impartial. 
Following the principles underlying the legal system in a democratic society based upon the separation of powers, if public rulings are to be made, they should be made by the Council. Experience with advance private rulings shows that it is a remarkably successful model. Although there may be legal difficulties in allowing the Council to make public rulings, every effort should be made to overcome them. Experience in other jurisdictions shows that public rulings made by the revenue authorities cannot practically remain within the boundaries of their legitimate operation. Their effect inevitably usurps, to some degree, the function of both the legislature and the courts.
Rulings are now an integral part of tax administration. As tax laws become more complex, rulings help to inject an element of certainty and fairness into the interpretation of those laws, particularly in the context of economic decision-making. Nonetheless, their benefits are tempered by inherent bias in most jurisdictions, as rulings are usually made by the revenue authorities. Sweden offers an example of a successful model of rulings that are made by an independent and impartial body.
The Swedish system stands up well when examined in the light of the criteria by which rulings systems are widely assessed. A thorough and detailed analysis of the ruling request by all the parties and the Council ensures that rulings are of the highest quality. However, there is a concern that the process is not as accessible to all taxpayers as some other systems. There are also resource constraints that have resulted in long delays in the issuing of rulings. This is being addressed. So, too, is the fact that the NTB cannot apply for a ruling.
The issue of public rulings is particularly relevant in the Swedish context. It is vital that the successful model used for private advance rulings is retained for any public ruling system if there is to be a model anywhere to show the comparative effectiveness of an independent and impartial rulings body. Otherwise, the Swedish public rulings system will inevitably result, as it has elsewhere, in the effective creation of tax law by the NTB.
Duncan Bentley heads the tax program at the School of Law, Bond University, Australia, where he is about to take up the position of Associate Dean for Postgraduate Studies in the School of Law. He is also editor of the Revenue Law Journal.
Research into the Swedish rulings system was made possible by a grant from the Australian Research Council. I am grateful to Associate Professor Anders Hultqvist and Professor Peter Melz of the University of Stockholm for their input and assistance. I have incorporated their useful comments on earlier drafts of this article. I am also grateful for their information on the operation of the Swedish rulings system to: Justice Gustaf Sandström of the Supreme Administrative Court; Mr Stefan Ersson, Chairman of the National Council for Advance Tax Rulings (the ”Council”), Mr Carl-Gustaf Wingren, Vice-Chairman of the Council, Mr Niclas Virin and Professor Christer Silfverberg of the University of Stockholm, Members of the Council; Emeritus Professor Leif Mutén of the Stockholm School of Economics; and Ms Anne Rutberg of Enskilda Juridik. Any errors are mine.
SFS 1951:442 om förhandsbesked: i taxeringsfrågor. I am grateful for the unauthorised translation into English by Torsten Fensby, formerly assistant director at the Council. India has introduced a model similar to that of Sweden, to provide advance rulings to non-residents, but arguably with less independence (Chapter XIX-B, Advance Rulings, Indian Income Tax Act 1961, effective from 1 June 1993).
In some jurisdictions, administrative principles of fairness may prevent the revenue authorities from denying an interpretation or concession contained in a ruling. In other jurisdictions, such as Australia, public rulings are binding on the revenue authorities until they are withdrawn. French practice is for the tax administration to publish circulars and letters that bind it, but another annual explanatory treatise, the Précis de fiscalité, is not binding. See, R Baconnier, ”La Pratique Du Rescrit En France” in International Fiscal Association, Advance Ruling: Practice and Legality (Kluwer, Deventer, 1994) at 5; F Vanistendael, ”Legal Framework for Taxation” in V Thuronyi (ed), Volume 1: Tax Law Design and Drafting (IMF, Washington DC, 1996) 15 at 60.
Part IVAAA Taxation Administration Act 1953 (Cth).
The Australian Taxation Office has provided guidelines as to its public ruling policy in Taxation Ruling TR 92/1. Revenue Canada issues interpretation bulletins and the United States Internal Revenue Service issues a range of rulings, instructions and other releases. They are all designed to inform taxpayers how the revenue authority will interpret and apply the tax law.
Section 5 SFS 1951:442 provides that an application for an advance ruling may not be lodged later than the date by which returns are to be filed for the purposes of the assessment in question.
Part VI Income Tax Assessment Act 1936 (Cth).
A review of the characteristics of the ruling systems in a number of OECD countries can be found in: International Fiscal Association, above n2; D Sandler, A Request for Rulings (The Institute of Taxation and IFS, London, 1994); IBFD, ”Ruling Practice: A European Survey and Comparison” (Feb 1995) European Taxation 58.
Second Reading Speech to the Taxation Laws Amendment (Self Assessment) Bill 1992.
Refer, for example: the extensive discussion in J Prebble, Advance Rulings on Tax Liability (Victoria University Press, Wellington, 1986) at ch 1; Sandler, above n 8; IBFD, above n 8; and WA Farlinger, Review of Revenue Canada, Taxation, Summary Report (Woods, Gordon, Toronto, 1985) 260.
Refer, OECD, Taxpayers’ Rights and Obligations (OECD, Paris, 1990) at 12; and a wider discussion of the topic in D Bentley, ”Formulating a Taxpayers’ Charter of Rights: Setting the Ground Rules” (1996) 25 Australian Tax Review 97.
For a detailed analysis, see GS Cooper, ”Themes and Issues in Tax Simplification” (1993) 10 Australian Tax Forum 417; S Surrey and GM Brannon, ”Simplification and Equity as Goals of Tax Policy” (1968) 9 William & Mary Law Review 915; RE Paul, ”Simplification of Federal Tax Laws” (1944) 29 Cornell Law Quarterly 285.
Bentley, above n11 at 103; in Australia the Senate Standing Committee on the Scrutiny of Bills, in its 1986/87 Annual Report (AGPS, Canberra, 1987) stated that, ”the practice of ’legislation by press release’ carries with it the assumption that citizens should arrange their affairs in accordance with the announcements made by the Executive rather than in accordance with the laws made by Parliament”.
OECD, above n11 at 13.
See K Messere, Tax Policy in OECD Countries: Choices & Conflicts (IBFD, Amsterdam, 1993).
J Owens and E Whitehouse, ”Tax Reform for the 21st Century” (1996) 50 Bulletin for International Fiscal Documentation 538.
Cooper, above n12 at 460.
RA LeBaube and CL Vehorn note this, using the example of Colombia, in ”Assisting Taxpayers in Meeting Their Obligations Under the Law” in RM Bird and M Casanegra de Jantscher (eds), Improving Tax Administration in Developing Countries (IMF, Washington DC, 1992) 310 at 314.
Developing countries are increasingly recognising the importance of rulings. See, for example, the Indian experience, Har Govind, ”Advance Tax Rulings for Non-residents” (1996) 2 Asia-Pacific Tax Bulletin 135. In the OECD context, see F Vanistendael, above n2 at 60.
In Australia, this was a stated object: Second Reading Speech to the Taxation Laws Amendment (Self Assessment) Bill 1992. Generally, see Prebble, above n10 at 22.
For example, in Australia, New Zealand and the United Kingdom.
The idea has progressed, and the emphasis on fairness is reflected in the move towards binding rulings systems.
J Li, ”Taxpayer Rights in Canada” (1997) 7 Revenue Law Journal 83; S James and IG Wallschutzky, ”The Shape of Future Tax Administration” (1995) 49 Bulletin for International Fiscal Documentation 217; L Stalans, ”Talking about Tax Audit Experiences: The Procedural Content of Socialisation”’ paper presented at the Internal Revenue Service Research Conference, Washington DC, 12–13 November 1992 and quoted in J Wickerson, ”Measuring Taxpayer Compliance: Issues and Challenges Facing Tax Administrations” (1994) 11 Australian Tax Forum 1 at 12; JA Roth, JT Scholz and AD Witte (eds), Taxpayer Compliance, Vol 1 (1989 University of Pennsylvania Press) at 129. Although, see the views of LeBaube and Vehorn, above n19 at 326.
See, in Australia, Second Reading Speech to the Taxation Laws Amendment (Self Assessment) Bill 1992 and in New Zealand, s 91A Tax Administration Act 1994 (NZ).
Prebble, above n 10 at 23.
Vanistendael, above n2 at 61; J Cleary, ”The Evolution of Tax Avoidance” (1995) 5 Revenue Law Journal 219.
Prebble, above n 10 at 23.
D Bentley, ”Problem Resolution: Does the ATO Approach Really Work?” (1996) 6 Revenue Law Journal 17.
S James and C Nobes, The Economics of Taxation (4th ed, Prentice Hall International, 1992).
S James and IG Wallschutzky, ”The Design of an Appropriate System of Tax Rulings” (1995) 5 Revenue Law Journal 175 at 180.
See the discussion in K Yelpaala, ”The Efficacy of Tax Incentives Within the Framework of the Neoclassical Theory of Foreign Direct Investment: A Legislative Policy Analysis” (1984) 19 Texas International Law Journal 365.
An example of this is the use of tax holidays by many developing countries to encourage foreign investment. Indonesia radically overhauled its tax incentive system in 1983, when it found that tax incentives alone were insufficient to achieve the levels or type of investment required: see D Bentley (1996) 70 Australian Law Journal 191 at 193. The Australian government has reduced the Research and Development concession, abolished the tax relief offered to investors in infrastructure projects and is concerned about other investment concessions. The Commissioner of Taxation stated: ”I have seen what started out as high-principled concessions flounder from their objectives because, almost inevitably, high-flying financiers and others seek to extend them”, The Australian Financial Review, 24 February 1997, 10.
This is a feature of the regional headquarter incentive programs in Malaysia, Singapore and Australia. See also, AJM Timmermans, ”Perspective of the Netherlands” in International Fiscal Association, above n2 at 45.
These features are found in whole, or in part, in International Fiscal Association, above n2 at 53; Sandler, above n8 at 73; IBFD, above n8 at 60; Prebble, above n10; James and Wallschutzky, above n31 at 177.
In the United States, the provision for rulings in the Revenue Acts of 1913 and 1919 was found to be too limited in scope. As a result, administrative or ”letter rulings” were issued to cope with demand (Rev Rul 10, 1953-1 Cum Bull 488).
Statutes in Australia and New Zealand, for example, provide a statutory basis for rulings, but the delivery, including ample discretion, is placed in the hands of the revenue authorities.
A Sawyer, ”Implementing a Regime of Binding Rulings: The New Zealand Experience”, paper presented at the UNSW (ATAX) Conference, Current Issues in Tax Administration, 11–12 April 1996 at 13.
D Bentley, ”The Commissioner’s Powers: Democracy Fraying at the Edges?” (1994) 4 Revenue Law Journal 85 at 118.
Govind, above n20 at 136.
M Smith, ”Binding Rulings”, paper presented to the 1993 Tax Conference of the New Zealand Society of Accountants, at 10.
Section 1 SFS 1951:442.
Section 14ZAN Taxation Administration Act 1953 (Cth).
PA Harris, ”Private Tax Rulings: An Advance System” (1994) Australian Tax Review 22 at 30.
Section 14ZAL Taxation Administration Act 1953 (Cth).
This is the case in Australia and New Zealand. It is the preferred approach of the United Kingdom Inland Revenue in Post Transaction Rulings: A Consultative Document (HMSO 1994) para19.
Sawyer, above n38 at 14.
James and Wallschutzky, above n31 at 188.
Sections 4, 5 and 6 SFS 1951:442.
CTC Resources NL v FC of T (1994) 48 FCR 397; Payne v FC of T (1994) 94 ATC 4191; (1994) 28 ATR 58.
National Tax Liaison Group Final Minutes for Meeting of 4 June 1996 at para 4.
Ibid at para 4.2.
Tax Administration (Binding Rulings) Regulations 1995, reg 5.
IBFD, above n8 at 60.
Sections 170BA and 170BB Income Tax Assessment Act 1936 (Cth).
Section 226M Income Tax Assessment Act 1936 (Cth).
Sections 222C and 226K Income Tax Assessment Act 1936 (Cth).
Section 11 SFS 1951:442.
These are raised in Prebble, above n10 at para 10.3, citing the arguments put forward in the United States and Canada.
Section 14ZAU Taxation Administration Act 1953 (Cth).
Sections 91DE, 91EI and 91FJ Tax Administration Act 1994 (NZ).
Section 11 SFS 1951:442.
OECD, Taxpayers’ Rights and Obligations: A Survey of the Legal Situation in OECD Countries (OECD, Paris, 1990) at 21.
Sandler, above n8 at 79; IBFD, above n8 at 60.
Prebble, above n10 at 103.
Section 9 SFS 1951:442.
Unforeseen problems in relation to the timing of rulings appeals, the jurisdiction of the courts and the scope of appeals arose in CTC Resources NL v FC of T (1994) 48 FCR 397 and Payne v FC of T (1994) 94 ATC 4191; (1994) 28 ATR 58.
Sawyer, above n38 at 9.
Ibid at 19.
In Australia, under the Administrative Decisions (Judicial Review) Act 1977 (Cth).
Section 16 Income Tax Assessment Act 1936 (Cth).
Prebble, above n10 at 86.
See, generally, Bentley, above n39.
Section 12 SFS 1951:442. See also, J Prebble, ”Canadian and Swedish Procedures for Advance Rulings in Income Tax Cases” (1987) 4 Australian Tax Forum 217 at 227; Prebble, above n10 at 82.
Sawyer, above n38 at 21.
Ibid. Fees are charged under Tax Administration (Binding Rulings) Regulations 1995 (NZ) regs 3, 4 and 5.
Except as otherwise indicated, the comments in this section are based upon SFS 1951:442 and information from discussions with those people mentioned in the acknowledgments for this article.
Section 1 SFS 1951:442.
The Council has the discretion to waive the fee in special circumstances, s 12 SFS 1951:442.
For example, see Sawyer, above n38, in the Australian and New Zealand context.
E Qwerin, ”Perspective of Sweden” in International Fiscal Association, above n2 at 41.
HLA Hart, The Concept of Law (Oxford University Press, Oxford, 1961) 132.
SOU 1996:44 Översyn av skatteflyktslagen – Reformerat förhandsbesked.
See also, Sandler, above n8 at 79; IBFD, above n8 at 60.
Prebble, above n10 at 103.
OECD, above n63 at 12.
In Australia, delegation of authority to administrative bodies is permitted under s71 of the Constitution of the Commonwealth of Australia. See further, L Zines, The High Court and the Constitution (3rd ed, Butterworths, Sydney, 1992) at ch10.
The allocation of losses among partners receives concessional treatment in Income Tax Ruling IT 2218.
For example, s 227 Income Tax Assessment Act 1936 (Cth).
Taxation Ruling TR 94/7.
A Hultqvist, Legalitetsprincipen vid inkomstbeskattningen (Juristförlaget, Stockholm, 1995) ch3.
Ibid at ch 4.
N Earle’s ”Experience Under the New System of Binding Rules” (1993) Taxation Institute of Australia, 32nd Victorian Convention Papers 23 at 24 reflects the reality of this in practice. Taxpayers are increasingly aware of the transaction costs of conflict. See D Bentley, ”Problem Resolution: Does the ATO Approach Really Work?” (1996) 6 Revenue Law Journal 17 and generally, O E Williamson, ”Transaction Cost Economics: The Governance of Contractual Relations” (1979) 22 Journal of Law and Economics 233.
Taxpayer compliance is exercising the minds of tax administrators across the world. Perceptions of fairness tend to increase compliance, which is a major reason for the introduction of rulings in the first place. See, above n24.
Democoli v Malta A 210 (1991) Com Rep.
DJ Harris, M O’Boyle and C Warbrick, Law of the European Convention on Human Rights (Butterworths, London, 1995) at 234. Nyström v Belgium No 11504/85, 58 DR 48 (1988); Versteele v Belgium No 12458/86, 59 DR 113 (1989). Although see Langborger v Sweden A 155 (1989), in which a four person tribunal with two interested lay members was found to be neither independent nor impartial, even though the other two members were judges and the presiding judge had the casting vote.
Hauschildt v Denmark A 154 para 46 (1989).
Ibid at para 47; Harris et al, above n98 at 235.
Harris et al, ibid; Fey v Austria A 255-A para 30 (1993).