Building a better GAAR.

TABLE OF CONTENTS I. INTRODUCTION 835 II. CONEQUENTIAL ATTRIBUTES OF TAX-AVOIDANCE BEHAVIOR AND A 842

TAXONOMY REFLECTING THE DIFFERENT CONCEPTS IN ECONOMICS AND

LAW III. RESPONDING DIFFERENTLY TO THE DIFFERENT TYPES OF TAX-A 852

VOIDANCE BEHAVIOR THAT ARE THE FOCUS OF ECONOMICS AND LAW

A. Rate Calibration and Behavioral Prohibition as Responses 853

to Real Behavioral Adjustments

B. Externality Theory and the Choice of Rate Calibration 861

or Behavioral Prohibition IV. TAILORING A BEHAVIORAL PROHIBITION TO APPLY DIFFERENTLY 871

TO TE DIFFERENT TYPES OF TRANSACTIONS THAT ARE CONSIDERED

TAX A VOIDANCE IN LAW

A. Under-Inclusiveness of Judicial Anti-Avoidance Doctrines 874

and GAARs

B. Economic Substance as a Primary Business-Purpose Test 882

C. Economic Substance as synthetic Replication and the 892

Limited Relevance of the Statutory Interpretation

Exercise V CONCLUSION 904

This article uses a consequential perspective to explore the design features of a target-effective general anti-avoidance rule (GAAR). As emphasized by David Weisbach, this perspective frames tax-avoidance activity in terms of its allocative and distributional effects. Most importantly, it supports the proposition that there is nothing in the consequential attributes of tax avoidance to justify its acceptance by tax policymakers. Instead, tax avoidance should be viewed as a negative externality, and externality theory suggests appropriate policy instruments that can be used to eliminate the consequential attributes of avoidance behavior. Different types of tax avoidance present, however, different identification issues that, in fact, lie at the core of the different concepts of tax avoidance that are the focus of economics and law.

It is suggested that a target-effective behavioral prohibition can only be realistically realized in the form of a GAAR. In assessing the content of a GAAR, the article makes four separate, but related, points. First, the distinction between GAARs, as currently conceived, and judicially-created anti-avoidance doctrines, is an overly sharp one, with both forms of behavioral prohibition suffering from underinclusiveness, albeit for different reasons. Second, the relevance of statutory interpretation in the identification of prohibited behavior should be much more limited than is currently the case. Third, two different concepts of economic substance provide the basis for a legislative response tailored to apply differently to different types of tax-avoidance transactions. Fourth, the judiciary has the institutional competence to apply and enforce a GAAR, or articulate anti-avoidance doctrines, as a response to tax-attribute creation and tax-attribute trading transactions; but such competence is much more limited with instances of transactional substitution. This type of tax-avoidance activity presents many of the same identification difficulties presented by real behavioral adjustments to taxation, which are the focus of the concept of tax avoidance in economics. Except for a limited range of transactional substitutions that are equivalent to synthetic replication, the associated informational requirements are not amenable to the application and enforcement of a purpose-based behavioral prohibition. Transactional substitutions beyond this limited range should be addressed by the legislative and executive branches of government in discharging their policymaking function.

1. INTRODUCTION

Among other consequences, the corporate tax-shelter debate has induced a substantial addition to an already vast tax-avoidance literature. (1) The quantity of this literature is matched, and even exceeded, by the case law on the subject. When the literature and the case law in individual countries are aggregated, the result is an apparently unmanageable volume of both. In fact, although the experience in other countries is routinely acknowledged by national policymakers, commentators, and the judiciary, the subject of tax avoidance tends to be treated as unique to country experiences. (2) Perhaps most obviously, the tax-avoidance case law tends to be viewed as the unique product of the national judiciary, with little in the way of instructive guidance for judicial reasoning in other countries. Much of the uniqueness may be attributable to differences in judicial, taxpayer, and tax-practitioner attitudes to tax avoidance, (3) which may or may not be the result of cultural differences. (4)

Putting attitudinal differences of the various actors to the side, there is a seemingly critical distinction, in both the form and the content of the law, between statutory general anti-avoidance rules (GAARs) and judicial anti-avoidance doctrines as expressions of behavioral prohibitions. For those countries without GAARs, a recurring debate revolves around the need for such a rule in the face of a perceived ineffectiveness of judicially-created anti-avoidance doctrines, which are developed as the apparent outcome of the statutory interpretation exercise. (5) For those countries with GAARs, an ongoing debate revolves around the effectiveness of these legislative provisions as interpreted and applied by the judiciary.

This article uses a consequential perspective to expose the analytical emptiness of the non-GAAR/GAAR dichotomy in the choice of mechanism to identify and prohibit tax avoidance, at least as that dichotomy is conventionally framed. It is argued that this necessary policy perspective provides the basis for a much more precise identification of the range of behavior that should be subject to a GAAR or judicially-created anti-avoidance doctrines. It also reveals that the content of a target-effective behavioral prohibition differs from existing GAARs (at least as they have been interpreted by the judiciary), as well as judicial anti-avoidance doctrines. Moreover, this difference in content requires the use of a GAAR by tax policymakers. It is utterly hopeless to leave it to the judiciary to articulate a behavioral prohibition that is neither under-inclusive nor over-inclusive in its identification of prohibited transactions. Trained as lawyers, and not public policy analysts, judges (and the lawyers appearing before them) simply do not have the institutional competence to execute this important task. (6)

The policy significance of a consequential perspective in a tax-avoidance context has been emphasized most recently in the legal literature by David Weisbach in his work framing the role of judicial anti-avoidance doctrines as responses to corporate tax-shelter transactions. (7) The consequential perspective emphasized by Weisbach focuses on the resource allocation and income/wealth-distribution consequences of tax-avoidance behavior. In this respect, the following two, relatively straightforward, propositions are posited here:

* Tax policymakers must first acquire a sense of the efficiency and income/wealth-distribution effects that make taxavoidance transactions problematic.

* Tax policymakers must then identify the range of problematic transactions at an acceptable administrative cost. In short, tax policymakers must assess the target effectiveness of anti-avoidance rules or doctrines that are posited or articulated as possible responses to the range of transactions identified as problematic because of their consequential attributes.

There is nothing special about these two propositions. They reveal the somewhat obvious point that a policy analysis of tax avoidance is no different than the analysis of any other tax-policy issue. The resolution of this first proposition is surprisingly easy. The consequential attributes of all tax-avoidance behavior require a response of some sort intended to eliminate such behavior. (8) Resolution of the identification issue is the more difficult exercise. (9) Both GAARs and judicial anti-avoidance doctrines in their current forms are under-inclusive. This characterization holds most obviously in the application of either form of behavioral prohibition to two of three general types of tax-avoidance transactions that are the daily grist of tax practitioners. (10) These are transactions that: (i) attempt to create a tax attribute; or (ii) attempt to transfer a tax attribute. The suggested cure for the illness of under-inclusiveness in respect of transactions within these two categories is a narrowly qualified reliance on a primary business-purpose standard. For these transactions, the relevance of statutory interpretation as an identification tool should be severely limited.

GAARs and judicial anti-avoidance doctrines are not all that badly off, however, when it comes to another type of tax-avoidance transaction--transactions that involve the substitution of a lower-taxed transactional form for a higher-taxed transactional form (referred to here as "transactional substitutes"). Although under-inclusiveness is also a prominent feature of both types of anti-avoidance responses as applied to this category of transactions, the under-inclusiveness is more difficult to address than it is with those transactions that either attempt to create or transfer tax attributes. The suggested cure for the illness of under-inclusiveness in respect of transactional substitutes is a definition of economic substance that targets a narrow range of such transactions and renders statutory interpretation irrelevant as an identification tool.