Photo by Susan Yin on Unsplash

General anti-avoidance rules (GAARs) are embedded within the tax laws of many jurisdictions and this includes common law jurisdictions such as Australia, New Zealand, Canada, the United States and the United Kingdom. The wording of the GAARs that operate in these different common law jurisdictions are very different but does this mean that these GAARs operate entirely differently? My paper has come to the conclusion that despite these differences in wording, there is a noted trend towards convergence in the ways the different GAARs are applied.

In this blog, my analysis focuses on Canada, New Zealand and the United States, with references to the Australian and UK regimes where relevant. More details of the Australian and UK GAARs can be found in my paper.

For a GAAR to apply, there must first be shown that there was some transaction, scheme or arrangement undertaken by the taxpayer. There must also be a tax benefit of some kind obtained in connection with the scheme or arrangement. A GAAR also requires some mental element indicating, objectively, that the taxpayer (or persons connected with the taxpayer) entered into the scheme or arrangement for the main, or at least not incidental purpose, of obtaining a tax advantage. A GAAR contains a set of principles to enforce both the spirit and letter of the law by operating to close loopholes ‘where the economic substance is different from that reported to tax authorities’.

No matter what the specific wording of the GAAR actually is, the enquiry undertaken by courts in these common law jurisdictions ends up being effectively the same. That similar enquiry looks to the overall purpose and structure of the transactions at issue and as to whether they lack any real commercial substance.

This conclusion may seem contrary to prevailing attitudes about statutory interpretation, but the evidence reveals that no matter what specific wording is actually used in the GAAR, the identification of tax avoidance as involving artificial contrived complex arrangements which produce no real economic substance, is applied in almost exactly the same way across the different jurisdictions reviewed.


In Canada, an ‘abuse and misuse’ test is applied (subsection 245(4) of Income Tax Act (Canada) 1985). This test involves a two-stage test.

The first stage involves a contextual and purposive interpretation of the legislative provisions to determine what the object, spirit or purpose of those provisions is that the taxpayer seeks to rely upon to obtain the tax benefit. This is a question of law. The second stage involves a determination of whether the facts of the transaction fit in with the analysis of the relevant provisions or to identify if the transaction entered into by the taxpayer frustrates the object, spirit or purpose of those provisions. If they do frustrate those provisions, then an abuse of the provisions has occurred and the GAAR can be used to strike down the ‘abusive’ transaction. This is a question of fact.

The Canadian Supreme Court in Copthorne held that under the section 245 analysis, extra factors can be taken into account to determine if the transaction abused the statute as a whole and so factors such as the commercial and economic realities of the identified transaction need and should be considered. The consideration of these types of factors such as the level of artificiality and the existence of economic purpose all ‘echo’ the same sorts of factors considered in New Zealand, Australia and the United States.

New Zealand

In the Ben Nevis case, the Supreme Court of New Zealand has applied the parliamentary contemplation test to determine any commercial and economic effects of the arrangement and then to ascertain whether the transaction fits within the purpose of Parliament with respect to the relevant sections of the Income Tax Act (NZ) 2007.

This parliamentary contemplation test applies a two-step approach to determine, whether or not the tax provision has been used within its intended scope. In taking this approach, the Supreme Court of New Zealand is considering the badges of avoidance, such as the degree of artificiality and contrivance and also as to how bad the scheme smelt (applying a kind of smell test).

Despite the New Zealand GAAR not having any detailed criteria to determine purpose, New Zealand courts, in applying this parliamentary contemplation test, have generally applied similar factors such as those found in section 177D of Australia’s Income Tax Assessment Act 1936. This was certainly the approach taken in the Ben Nevis and Penny and Hooper decisions. Hence, factors such as the manner in which the arrangement was carried out, the role of the relevant parties, the commercial and economic effect of the documents and transactions, and the nature and extent of the financial consequences for the taxpayer and related parties, were all relevant in assessing the level of artificiality and hence in determining purpose in Ben Nevis.

The Ben Nevis and Penny and Hooper decisions suggest that the operation of the parliamentary contemplation test used in New Zealand is being applied in a very similar manner to the Canadian abuse and misuse test and also in a similar manner to the Australian ‘purpose’ test. In addition, the consideration of factors such as economic and commercial effect of documents and transactions and the nature and extent of the financial consequences indicates that New Zealand courts are also effectively applying factors very similar in effect to the US economic substance doctrine.

The United States

Since 2010, the United States has adopted, by amending the Internal Revenue Code of 1986 to include section 7701, a type of legislative GAAR. This amendment codified the economic substance doctrine. Under this economic substance doctrine, a US court will disregard a business transaction if it lacks economic substance and has no economic value, other than the value attributable to the tax loss, and where the transaction has no business purpose.

In Long Term Capital Holdings v United States, the court found that the transaction in question lacked any economic substance as the transaction was entirely tax motivated, had no realistic profit prospect and the transaction had far more complexity than was seen to be necessary to achieve the stated objectives. The US court, by considering the level of complexity and artificiality, effectively applied the same sort of indicative factors as are applied in Australian, Canadian and New Zealand courts.


A review of recent cases (such as in New Zealand: Ben Nevis, Penny and Hooper, and BNZ Investments; in Australia: Spotless and Hart; in Canada: Triad Gestco and Copthorne; and in the United States: Long Term Capital) indicates that the different courts in these different jurisdictions have been increasingly applying similar indicative factors supporting this noted trend towards convergence.

Therefore, it does not seem to matter whether the court uses an ‘abuse’ test (as is done in Canada) or a ‘parliamentary contemplation’ test (as is undertaken in New Zealand), or applies the ‘economic substance’ doctrine (as is applied in the United States), or applies the eight factors in section 177D of the Income Tax Assessment Act 1936 (as is done in Australia); the same fundamental enquiry is ultimately undertaken by the courts in each jurisdiction.

That enquiry focuses on similar suggestive factors such as whether the transaction is artificial in nature; whether the transaction lacks economic substance; whether the transaction involves undue complexity; whether the transaction involves the use of related parties; whether there is a difference between legal form and the economic reality of the transaction; and whether the transaction is undertaken largely for tax reasons.

This conclusion is not altogether new as other tax law scholars such as Chris Evans (2008), Stella Kasoulides Paulson (2013), Rick Krever and Peter Mellor (2017) and others have also previously identified this trend. Chris Evans (2007) has also recognised that there was ‘some degree of convergence in the jurisprudence in some of the common law jurisdictions in the approach taken by the courts to avoidance type cases, albeit through the interpretation of very different legislation’. This was also a conclusion reached in part by Judith Freedman (2005) in commenting on similar outcomes being achieved in Canada in cases such as Canada Trustco and Mathew as by courts in the UK in cases such as Barclays Mercantile and Scottish Provident, although these similar outcomes were achieved by very different routes.

The end result achieved by the courts in the different jurisdictions is that the same unacceptable behaviour (self-cancelling transactions, lack of exposure to real risk, inclusion of tax-favoured parties into transactions, and lack of arms-length dealing) is likely to be caught under each of the Canadian, New Zealand, Australian or the US provisions. This is also likely to be true of the United Kingdom provisions with respect to the more abusive type of arrangements, but the application of the UK provisions is much more restrictive as the UK provisions are targeted only at the most abusive schemes and provide for only tax avoidance which fails a ‘double-reasonableness’ test (more information about the double reasonableness test can be found on page 53 of my paper).


Although the same fundamental enquiry is undertaken, regardless of the wording used, it is conceded that there are subtle differences between the different GAARs in regard to where the threshold is set for acceptable tax planning versus unacceptable tax avoidance. An example of this is the different outcomes reached in the business restructure cases of Mochkin in Australia (which was held not to be tax avoidance) and Penny & Hooper in New Zealand (which held that there was tax avoidance).

Since substantially the same enquiry is undertaken and there is no effective difference between a GAAR with more detailed criteria than one that does not have detailed criteria for its application, my paper concluded that that there is an increasing convergence in the way each of the different GAARs operate.

Leave a comment

Your email address will not be published. Required fields are marked *