Huang, Vicki T --- "An Empirical Investigation of 20 Years of Trade Mark Infringement Litigation in Australian Courts" [2019] SydLawRw 4; (2019) 41(1) Sydney Law Review 105

	Home \| Databases \| WorldLII \| Search \| Feedback Sydney Law Review

In Australia, there has been little empirical research into the enforcement of trade mark rights under s 120 of the Trade Marks Act 1995 (Cth) (‘1995 TM Act’). Nor has there been empirical research into the common practice of litigating concurrent claims under passing off or s 18 of the Australian Consumer Law 2010 (Cth) (‘ACL’). This article reports on the first study to systematically identify and review all 78 trade mark infringement judgments under the 1995 TM Act over the 20-year period since its inception (1 January 1996–1 January 2016). The analysis reveals that, contrary to initial concerns, there has been a significant decline in pleading under the ‘new’ and expansive ss 120(2) and (3) provisions and an increased reliance on the classic parameters of s 120(1). Significantly, this article finds that passing off in the context of court-resolved s 120 infringement has become redundant. In addition, the analysis reveals that a concurrent claim under ACL s 18 improves the chances of a plaintiff’s net win rate by 21.7%. The implications of these results are discussed in relation to litigation practice and doctrine. This article also demonstrates the utility of empirical research in resolving speculative assumptions about the law and in facilitating better informed legal scholarship and practice.

I Introduction

A The Relevance of the Present Study

The introduction of the Trade Marks Act 1995 (Cth) (‘1995 TM Act’)^[1] led to major changes in Australian trade mark law to comply with Australia’s obligations under the Agreement on Trade-Related Aspects of Intellectual Property Rights (‘TRIPS Agreement’).^[2] In the 20 years since the enactment of the 1995 TM Act, there has been no comprehensive study of the effect of those changes. Nor has there been empirical research regarding the practice of litigating concurrent claims in passing off and/or s 18 of the Australian Consumer Law 2010 (Cth) (‘ACL’).^[3] This article is the first to systematically identify and review all 78 trade mark infringement judgments under the 1995 TM Act over a 20-year period (1 January 1996–1 January 2016).

In addition to these substantive results, another contribution of this article relates to methodology. This study supports arguments that an empirical approach can provide a ‘more accurate description and analyses of how our legal system actually operates’^[4] and facilitates ‘a surer epistemological basis to support claims or to question others’.^[5] For example, the method applied in this study provides quantitative evidence that resolves critical competing assumptions about the role of passing off in trade mark litigation.

The method applied is based on Beebe’s empirical investigation of United States (‘US’) trade mark infringement with adjustments made to suit an Australian context.^[6] Although empirical legal studies are labour intensive, they are often well received^[7] as they test the truth of more anecdotal or impressionistic accounts of the law. Therefore, the details of the methodology are set out to encourage the application of empirical methods to other branches of law.

This article is also particularly timely in light of the recent Productivity Commission Inquiry into Intellectual Property Arrangements.^[8] That Inquiry focused on the need to ‘get the balance right’^[9] between the benefits of trade marks in terms of reducing consumer search costs and the costs of trade marks in terms of the anti-competitive effects of monopolies.^[10] To that end, the Inquiry set forth numerous recommendations in relation to bettering the quality of marks being registered.^[11] Unfortunately, the Inquiry provided no detailed analyses as to how trade mark owners actually enforce their registered rights. This is probably because, unlike the case of registration, there is no single database that records detailed infringement information. The Inquiry’s recommendations should therefore be considered cautiously. Policy reform to optimise the trade mark system requires a more complete understanding of how ‘the system is working in practice’.^[12] This means understanding not just how and what is being registered but also how trade mark owners actually enforce their rights in practice. This is particularly interesting in trade mark law, in which (unlike in copyright and patent law) trade mark owners often enforce their rights outside the statutory scheme through concurrent claims. This study is also relevant to various doctrinal debates that exist in trade mark infringement law, which will be discussed in the next section.

B Doctrinal Debates

To make out a claim for trade mark infringement under s 120(1) of the 1995 TM Act, the plaintiff must prove that the defendant has used or proposes to use^[13] a substantially identical or deceptively similar sign as a trade mark^[14] in relation to goods or services^[15] in respect of which the plaintiff’s trade mark is registered. Section 120(1) reflects the ‘classic’ conception of trade mark infringement, in which the defendant is using a deceptively similar (or substantially identical) mark on goods and/or services that have been registered by the trade mark owner.

Under s 120(2), as with s 120(1), a person infringes a registered trade mark if the defendant uses or proposes to use as a trade mark a substantially identical or deceptively similar mark. However, unlike s 120(1), the impugned user is now liable for use on goods or services beyond those for which the trade mark is registered. Under s 120(2), infringement can be found if the impugned trade mark is used upon: goods of the same description as the registered goods (s 120(2)(a)); ‘services that are closely related to registered goods’ (s 120(2)(b)); services of the same description as registered services (s 120(2)(c)); or ‘goods that are closely related to registered services’ (s 120(2)(d)). In other words, s 120(2) broadens the monopoly rights of the trade mark owner beyond use on the goods or services for which the owner’s trade mark is strictly registered.

Section 120(3) provides even broader protection for some trade mark owners. However, it only applies if the trade mark owner’s mark is ‘well known in Australia’.^[16] If so, the plaintiff can assert an infringement claim where the defendant has used a substantially identical or deceptively similar mark as a trade mark on goods or services unrelated to the goods or services for which the plaintiff is registered. This is subject to the additional requirement that the defendant’s use of the mark would likely be taken to indicate a connection between the unrelated goods and services and the registered owner of the trade mark and,^[17] as a result, the interests of the registered owner would likely be adversely affected.^[18]

Sections 120(2) and (3) (and related provisions) were introduced to meet the obligations imposed by the TRIPS Agreement, particularly art 16.^[19] The addition of ss 120(2) and (3) was criticised at the time by some members of the Working Party to Review the Trade Marks Legislation as an unreasonable extension of owners’ rights.^[20] There was also concern that the introduction of the more expansive ss 120(2) and (3) were somewhat redundant or could undermine the role played by passing off and ACL s 18 (discussed further below).^[21] Whether and how ss 120(2) and (3) have been litigated will be examined in this article.

A related question in trade mark law regards the intersection of concurrent claims with s 120 of the 1995 TM Act. In a trade mark infringement case, the primary rights relied upon are the plaintiff’s statutory rights under s 120 of the 1995 TM Act. However, a s 120 claim is often pleaded with passing off and/or ACL s 18. This gives rise to the question: if a concurrent claim is pleaded, why choose passing off, ACL s 18 or both?

There are key differences between these three causes of action with respect to origins, evidentiary burdens, protected interests, purposes and defences. The 1995 TM Act provides exclusive rights to distinctive trade marks as a species of statutory property conferred on an owner by way of registration. The interest protected is the property in the trade mark. There is no need to prove distinctiveness of the mark because this is prima facie evidenced by the fact of registration (subject to a cross claim for invalidity). Further, the exclusive rights under the 1995 TM Act are powerful and national in scope. They are the primary rights relied upon in trade mark litigation.

In passing off, liability is incurred where there is a misappropriation of reputation such that the goodwill of the plaintiff is damaged. The interest protected is the plaintiff’s reputation, the existence of which must be proved by the plaintiff and is often local, rather than national, in scope.^[22] In a trade mark infringement context, ACL s 18 concerns use of the plaintiff’s trade indicia such that the public may be misled or deceived as to a commercial association between the parties.^[23] The interest protected is the consumer’s welfare and prevention of consumer confusion.^[24]

The benefits of arguing concurrently in passing off were more apparent under the earlier Trade Marks Act 1955 (Cth) (‘1955 TM Act’) and in the time before federal consumer protection statutes were introduced in the 1970s.^[25] In Australia, under the 1955 TM Act and its predecessors, infringing conduct was narrowly confined to infringing use upon goods or services for which the plaintiff’s mark was registered.^[26] This is what this article refers to as the traditional or ‘classic’ parameters of trade mark infringement. The primary infringement provision under the 1955 TM Act was s 62(1), which defined infringement as use of ‘a mark which is substantially identical with, or deceptively similar to, the trade mark, in the course of trade, in relation to goods or services in respect of which the trade mark is registered’.^[27] In essence, s 62(1) is the precursor to the modern provision: s 120(1) of the 1995 TM Act. Passing off was able to fill the gap where the infringing conduct fell outside the scope of the goods and services covered by the registration or where the mark could not be registered.^[28] As discussed earlier, it has been thought that this gap-filling role is undermined by the current Act because ss 120(2) and (3) of the 1995 TM Act allow infringement claims for uses beyond the scope of the plaintiff’s registration.^[29]

A second role of passing off was to act as a catch-all for infringement of unregistrable marks. This was thought important because the 1955 TM Act had a narrower definition of a registrable ‘sign’. Section 6 of the 1955 TM Act provided registration for a distinctive ‘device, brand, heading, label, ticket, name, signature, word, letter or numeral, or any combination thereof’. It was thought that passing off filled the gap where a mark was unregistrable, such as for a product’s ‘get-up’^[30] (known as ‘trade dress’ in other jurisdictions).^[31] A significant passing off jurisprudence developed around the protection of ‘get-up’ — that is, the ‘shape, size and colouring of [a] container of packaging, the design of [a] label and to some extent, the design of the product itself’.^[32] It could be argued that this second gap-filling role has also been undermined by the modern 1995 TM Act, which added the following as registrable marks: ‘aspect of packaging, shape, colour, sound or scent’.^[33] Another point in relation to unregistrable marks is that passing off could protect marks embodying commonly used words that had developed a secondary reputation.^[34] These marks would be those having little to no inherent adaptation to distinguish. It was previously held that these kind of marks were an exception to the presumption of registrability.^[35] If the mark had difficulty being registered or could not be registered, then passing off would likely be an attractive option. However, the restructure of s 41 of the 1995 TM Act by the Intellectual Property Laws Amendment (Raising the Bar) Act 2012 (Cth) clarified that the presumption of registrability does indeed apply to s 41.^[36]

A third benefit of passing off in trade mark litigation was thought to be access to exemplary damages^[37] and the option of an account of profits or compensatory damages.^[38] However, more recent legislation has undermined this advantage by creating a head of ‘additional damages’^[39] (in addition to an injunction and compensatory damages) for trade mark infringement under 1995 TM Act s 126(2) to punish and deter.^[40]

A final benefit, and perhaps the only benefit, of passing off as a concurrent claim is as a ‘safety net’ or ‘back-up’ claim should the plaintiff’s trade mark rights be negatived.^[41] For example, a defendant may counterclaim to have the plaintiff’s mark removed from the register for want of validity. A concurrent action in passing off could assist the plaintiff should its trade mark registration be struck down.^[42] Another example relates to mitigating the risk of a court finding that a defendant has not ‘used’ the impugned sign as a trade mark.^[43] In such a situation, a plaintiff could succeed by using passing off as a ‘back-up claim’.

However, if the remaining useful role for a concurrent claim is acting as a safety net should the plaintiff’s rights fail under the 1995 TM Act, the question remains as to why argue concurrent claims in both passing off and ACL s 18? And if so, how do courts resolve those claims?

In a comparison between passing off and ACL s 18 in the context of s 120 litigation, the disadvantages of passing off are evident. From a practical perspective, the main distinction relates to the evidentiary burden.^[44] With regards to proof, passing off requires the plaintiff to show ‘sufficient reputation in a particular get-up or name’.^[45] There ‘is no such requirement’ in relation to ACL s 18.^[46] That is, the latter focuses on whether the defendant’s conduct is likely to mislead or deceive as a general proposition. Some commentators would argue that proof of the trader’s reputation assists in identifying the nature of the deceptive conduct.^[47] Nevertheless, it is generally the case that proving a misrepresentation damaging a plaintiff’s reputation is narrower (and thus more difficult) than proving misleading or deceptive conduct, such that a finding of passing off will likely lead to a breach of ACL s 18’s standard.

The only other obvious advantage of passing off is that, unlike ACL s 18, it is not bound by conduct that is in ‘trade or commerce’. However, this distinction is arguably undermined because conduct in ‘trade or commerce’ has been interpreted so broadly that cases that are actionable under passing off are almost always covered by ACL s 18.^[48] In addition, in the context of s 120 litigation, all trade mark use is, by definition, use related to trade source and is therefore always conduct in ‘trade or commerce’, given the wide interpretation of that expression.

The question remains as to whether and why litigants should argue concurrent claims in both passing off and ACL s 18 in trade mark litigation. Burrell and Handler argue that one reason for this quandary is that little thought was given to how the introduction of the federal consumer law in 1974 would affect either trade mark or business reputation infringement claims.^[49] Moreover, ‘this possibility [of overlap and redundancy] was not mentioned at any time during the legislative history’ of the Trade Practices Act 1974 (Cth) (‘TPA 1974’) (the precursor to the ACL).^[50]

Early concern about potential overlap was expressed in 1984, a decade after the inception of s 52 of the TPA 1974 (the precursor to ACL s 18).^[51] Blakeney argued (in what has been described as a ‘seminal’ article)^[52] that passing off was dominating and curbing the broad language of s 52 of the TPA 1974.^[53] His concern was that ‘[t]his may mean that justice is done between the parties but, it is submitted, damage is done to the doctrinal purity of the pro-consumer objectives of the statute.’^[54] Blakeney noted that, despite courts warning against heavy reliance on passing off, inevitably, ‘the final decision has been based on the application of some technical passing off rule’.^[55]

In stark contrast to this position, more recently it has been said that ‘passing off is on its last legs’.^[56] For example, Heerey QC and Creighton-Selvay argue that passing off in a trade mark infringement suit can be ‘avoided’ as the increased evidentiary cost (of concurrent claims) does not outweigh the benefits unless exceptional relief is sought.^[57] To resolve these competing positions, whether and how passing off and ACL s 18 claims are litigated in s 120 litigation is examined in this study.

II Method

Empirical legal studies are uncommon in Australia. This is likely because they are labour intensive and require both quantitative and legal reasoning skills. In Australia, there has been one study specifically examining empirical trends in trade mark litigation.^[58] In this 2006 study, Bosland, Weatherall and Jensen^[59] examined all trade mark enforcement decisions for the five-year period between 1998 and 2002. The authors found ‘two fundamentally different kinds of trade mark litigation’:^[60] counterfeit proceedings, where there was a high rate of success for the applicants;^[61] and non-counterfeit cases, where the success rate was ‘lower than the 50% rate predicted by standard economic models of litigation’.^[62]

This study differs from the Bosland et al study with respect to timeframe and research objectives. The Bosland et al study focused on all trade mark enforcement claims under different aspects of the 1995 TM Act over a five-year timeframe. In contrast, the present study looks only at cases where there was a substantial discussion of rights under s 120 of the 1995 TM Act over a 20-year timeframe (all cases in the population).^[63] This methodological difference will be discussed further below in relation to win/loss outcomes. This study looks only at the infringement of registered marks under s 120 of the 1995 TM Act and only discusses passing off and ACL s 18 in that context. This article acknowledges, but excludes, the large body of passing off and ACL s 18 jurisprudence that exists outside of s 120 litigation.^[64]

A Case Selection

In relation to Australian case law, the utility of referring to Beebe’s method relates to his approach to case selection. Beebe looked at ‘all district court opinions that made substantial use of a multi-factor test for the likelihood of consumer confusion’ over a five-year period from 1 January 2000 to 31 December 2004 inclusive (1252 opinions excluding appeals).^[65] From this pool, he ‘reviewed each of these opinions to determine whether it made substantial use of the multi-factor test’.^[66] Beebe ‘defined substantial use liberally as any use beyond the mere citation without analysis of the test’.^[67]

Beebe further culled cases that would skew the results in terms of representing how tests of infringement are applied. He therefore excluded cases that dealt with counterfeiting, ‘breaches of franchising, licensing, or distribution agreements’, and ‘opinions on motions to dismiss or on motions where the non-moving party failed to appear.’^[68] For example, with respect to counterfeiting opinions, Beebe noted that in counterfeit cases ‘the likelihood of confusion is very clear and the factors tend to weigh overwhelmingly in favor of the plaintiff’.^[69] Inclusion would thus skew the results of the study. Of the remaining 337 opinions, Beebe excluded six cases where the outcome of the multi-factor test was reversed.^[70] That is, he retained the district court opinion where there was an appeal, but the trial judge’s decision on infringement was affirmed or not agitated.

The cases for this study were located using broad keyword searches in the LexisNexis Australia legal database — ‘All Subscribed Australian Case Sources’ — for all Australian jurisdictions in order to capture the rare trade mark cases that are brought in state rather than federal courts. Keywords included ‘trade mark’ and ‘infringement’ dated between 1 January 1996 and 1 January 2016. This broad search revealed 2355 cases. Duplicates were eliminated. This set was cross-checked against a list derived from another database, Westlaw Australia, which unearthed a small handful of cases not reported in LexisNexis Australia legal database.^[71] The final list was cross-checked against a third online database, that of the Australasian Legal Information Institute (austlii.edu.au), but no further additions were required.

Only cases that provided a ‘substantial discussion’ of s 120 of the 1995 TM Act were included. ‘Substantial’ was defined liberally as ‘any use beyond mere citation without analysis of the test’.^[72] Excluded after inspection were cases that focused on ownership disputes (for example, prior use or consent), parallel imports, the 1955 TM Act, procedural issues, discovery issues, costs, damages, copyright, patents, design law or contract interpretation. It was found that those cases did not include a substantial discussion of the law of trade mark infringement.^[73] Two further categories of cases were excluded from the analysis: counterfeit cases and first-instance decisions that were reversed on appeal. As per Beebe’s and Bosland et al’s studies, counterfeit cases are not representative of typical infringement proceedings.^[74] The proceeding is quickly disposed of^[75] and the trade mark owner typically wins.^[76]

Appellate judgments were excluded. Appellate courts in trade mark cases are not hearings de novo and the reconsideration of evidence in relation to s 120 is not common.^[77] Of the first-instance decisions, only those that were not reversed on ultimate appeal were included.^[78] Separating out first-instance decisions (where the finding was not reversed) has particular advantages. First, this is a study that examines judicial reasoning regarding s 120. Second, from a statistical point of view, a homogenous group is preferable to make inductive arguments.^[79] Removing appeals and first-instance cases where the s 120 finding was reversed left 78 cases in the sample.^[80]

With regards to terminology, Australian courts use the term ‘case’ loosely, and it is not strictly a term of art. In contrast, ‘proceedings’ are defined as ‘all acts and events between the time of commencement and the judgment’.^[81] This can span from multiple judgments on preliminary and interlocutory matters to judgments of higher courts of appeal. A judgment is defined as ‘the final order or set of orders made by the Court after a hearing’.^[82] However, a judgment will usually contain a full set of reasons and conclude with a request that parties draft orders to be approved at a later date. Colloquially, the proceedings (and a single judgment within those proceedings) can be known as a ‘case’. In this article, a ‘case’ refers to a single written judgment or decision within a ‘proceeding’. The 78 cases in the sample were coded for ‘posture’ or procedural standing. This showed that there were 58 trial decisions, 16 applications for interlocutory relief, one application for interlocutory orders, two applications for default judgment and one application for summary judgment.

It may be argued that these different types of judgments should not be grouped together because they result from different requirements of proof, argument and reasoning. However, this argument can be refuted because of the consistent approach used to select the cases. As discussed earlier, cases were included if they provided a ‘substantial discussion’ of s 120 of the 1995 TM Act. A close reading of the cases found that there were trial decisions that provided limited discussion of s 120 and non-trial proceedings (for example, interlocutory applications) that provided extensive discussion of the issues. In other words, posture did not necessarily determine the quality or quantity of the discussion of issues under investigation. Therefore, although the 78 cases in the dataset reflect different postures, they are consistent in the sense that they all provide a substantial discussion of s 120 of the 1995 TM Act.

B Coding

The dataset was coded by the author and data was entered into an author-designed Microsoft Access database. In total, each case was reviewed at least seven times. In addition to coding the factual aspects of each case, each judgment was read to identify specific comments, obiter dicta and rationes decidendi of interest. Regarding hearing length, if the judgment and hearing were heard and delivered on the same day, this was counted as zero days. If the hearing and date of judgment were on separate days, then the hearing length was estimated as one day, unless further dates were listed in the header.^[83]

Coding a win or a loss for each case was complicated by the fact that, for any one case, there could be at least one or more trade marks in suit. Most litigants bring their best case to court and generally put forward all trade marks that may potentially be found to be infringed, even though there is only one mark of critical interest. This makes sense because, once an action has begun, pleadings can only be amended with the consent of the other party or with the leave of the court and can involve time bars and additional costs.^[84] Plaintiffs are strategically better off pleading as many trade marks as they think reasonable in any one case.

To deal with the coding issue of win/loss where more than one trade mark allegedly infringed, Beebe’s method was followed. Beebe coded per case rather than per trade mark;^[85] similarly the Australian studies of Huang et al^[86] and Bosland et al^[87] coded by proceedings and not by trade mark. Therefore, if there were multiple marks litigated and there was one ‘win’, the case was coded as a ‘win’ overall^[88] A win on a subset of marks will likely give the plaintiff an injunction against the defendant’s impugned behaviour. For most plaintiffs, it is likely that such a win would be considered a victory. Coding these types of cases as ‘wins’ risks over-representing the win rates of the plaintiff. However, a review of the cases found it was uncommon in a single case for there to be multiple marks discussed and a split decision about winning and losing across marks.

The alternative approach would be coding wins/losses per mark, but this method has the problem of over-inflating the sample size. As discussed above, it is assumed that, as rational actors, plaintiffs put forward their best case and that this means including all trade marks for which there is a chance of winning either a s 120 or concurrent claim. However, assuming courts are also rational actors, they will focus on only one or two marks that best represent the legal issues in dispute to dispose efficiently of the case. Therefore, it is argued that coding per case rather than per mark within a case produces a more accurate reflection of the dynamics of judicial reasoning.

III Results and Discussion

A Winning and Losing

As described in the method, the dataset included first-instance decisions not reversed on appeal. It is therefore important to examine the rate and nature of appeals as indicators of the stability of the dataset. For example, if a large number of cases were reversed on appeal, then the first-instance dataset will contain a large number of exclusions and thus not be representative of the type of cases that pass through first-instance courts.

Table 1 (below) shows that over the 20-year period there were 22 first-instance decisions that went to an appellate court, ie 26% of cases.^[89] This is low compared with appeals of patent cases.^[90] With regards to the 22 trade mark decisions that went to appeal, 73% (16/22) of appellate judgments affirmed the trial decision on the s 120 issue (affirming seven wins and nine losses). Six appellate judgments reversed the trial decision (reversing five plaintiff s 120 wins and one loss). In short, unsuccessful defendants did better on appeal than unsuccessful plaintiffs.^[91] Given the low proportion of appeals and the very low number of successful appeals, it can be said that the dataset is representative of first-instance outcomes.

Over a 20-year period, there were 40/78 (51%) s 120 wins to 38/78 (49%) s 120 losses. It should be emphasised that the win/loss coding relates to success on the s 120 argument, as distinct from the relief sought. This is important because in two interlocutory applications, the plaintiff won the s 120 claim on merits but was not successful in attaining an injunction against the defendant.

The 51% to 49% win:loss ratio is identical to the win:loss ratio reported in Beebe’s study of 331 US opinions.^[92] In addition, these findings approximate the influential Priest-Klein ‘divergent expectations model’, which predicts a 50:50 win:loss ratio where all parties are rational and fully informed.^[93] However, it is noted that win:loss ratios should be interpreted with caution.^[94] This is because the ‘empirical study of judicial outcomes (win rates)’ can be ‘notoriously problematic’.^[95] It is well-recognised that these observations can be affected by externalities, such as the ‘case-selection effect’.^[96] This ‘effect’ posits that the cases that come to court represent a ‘biased sample from the mass of underlying disputes’ because most disputes are settled prior to a hearing.^[97] Therefore, the cases and issues that proceed to litigation are not necessarily reflective of the nature of disputes in the marketplace.^[98] An additional issue may be that some ‘irrational’ litigants pursue close or losing cases to defend their mark in the marketplace at all costs.^[99] Therefore, at its highest, this win/loss finding can be said to support the proposition that this method of case selection yields a homogenous sample of ‘close cases’ representing (on average) rational parties with divergent expectations. From a methodological perspective, this is a useful finding in and of itself.

For example, the win:loss ratio for the sample in this study differs from Bosland et al’s findings for non-counterfeit Australian trade mark cases in the five-year period between 1998 and 2002. In that study, the authors found the plaintiff won in 9/30 cases (30%) and lost in 20/39 cases (67%), and there was a partial win in 1/30 cases (3%). The authors considered these results surprising and different to those expected under the divergent expectations model.^[100] They suggested that trade mark owners could be overly optimistic or prepared to risk a loss for strategic gains.

It is suggested here that the difference between the two studies relates to the methods of case selection. In this study, cases were included only if there was a substantive discussion of s 120 issues by the court. To that end, cases that dealt with enforcement of rights via ownership issues or contract disputes were excluded. The Bosland et al study explored enforcement of trade mark rights more broadly, which explains why their sample of 30 cases over a five-year period is proportionally larger than the sample of 78 cases over a 20-year period in this study. It is suggested here that the results in this study better reflect win/loss rates with respect to s 120 reasoning, and conclusions are limited and drawn within the confines of this study’s more homogenous dataset.

B Sections 120(1), (2) and (3)

This section explores in more detail the rate of pleading and reasoning under the three subsections of ss 120(1), (2) and (3) over two separate decades: Decade One (1 January 1996–31 December 2005), 33 cases; and Decade Two (1 January 2006–1 January 2016), 45 cases. By examining the 20-year period as two halves, a 36% increase in the volume of s 120 litigation from Decade One to Decade Two (from 33 to 45 cases) can be observed (Table 2 below). This increase approximates Australian economic growth between these two periods, which is estimated at 44%.^[101] In other words, the increase in the number of litigated cases in Decade Two may be explained by overall growth in the Australian economy, rather than a real increase in the volume of trade mark litigation under s 120.

Table 2 (below) shows that, irrespective of decade, s 120(1) has been the most dominant claim, accounting for 82% and 93% of trade mark infringement suits for Decade One and Decade Two, respectively. Across decades, a relative decline in pleading s 120(2) can be seen, from 33% of cases in Decade One to 29% in Decade Two. A decline can also be observed for s 120(3) claims, from 15% in Decade One to 2% in Decade Two. In sum, claims under s 120(1) (accounting for 93% in Decade Two, up from 82% in Decade One) have become predominant under ss 120, while s 120(2) and (3) are suited to a minority of cases.

Decade One: January 1996–December 2005, 33 cases
Claim	s 120(1)	s 120(2)	s 120(3)
Number of claims	27	11	5
Total cases (1996–2005)	33	33	33
Proportion of claims made per case	82%	33%	15%

Decade Two: January 2006–January 2016, 45 cases
Claim	s 120(1)	s 120(2)	s 120(3)
Number of claims	42	13	1
Total cases (2006–16)	45	45	45
Proportion of claims made per case	93%	29%	2%

These results are interesting given the history of the 1995 TM Act. As discussed earlier, s 120(2) and s 120(3) were introduced to meet obligations imposed by art 16(1) of the TRIPS Agreement.^[102] At the time, the Working Party to Review the Trade Marks Legislation expressed concern that these infringement provisions (which permit liability for use on goods and services beyond those for which the plaintiff was registered) could be considered an unreasonable extension of the trade mark owner’s rights.^[103] This article shows that excessive reliance on these ‘extension’ provisions has not occurred in the context of this dataset.

A potential explanation for why plaintiffs are not relying on ss 120(2) and (3) is because they just do not have to. It may be the case that the bar for registration has been set too low under the 1995 TM Act, such that trade mark owners have their marks registered across an ‘excessive’ range of goods and/or services. This, plus the presumption of registrability, may have induced over-registration of marks across goods and services where there may only be a spurious intention to use.

The issue of ‘cluttering on the register’ has been recently discussed by academics^[104] and in a report by the Productivity Commission.^[105] Cluttering can refer to the situation where the specification includes goods and/or services where the owner will not use the mark (over-broad registrations) or where the owner has registered a mark, but fails in its intention to actually use the mark as registered.^[106] Cluttering may also refer to disaggregated registrations where a brand owner registers sub-parts of their packaging to create a mosaic of rights over a single package. If cluttering according to these definitions exists, then a potential infringer’s conduct is likely to fall under s 120(1) for allegedly using a substantially identical or deceptively similar mark upon goods and services for which the plaintiff has broadly registered itself. There is no need to assert rights for infringement on similar or related goods and/or services under s 120(2) or (3) because the trade mark owner has registered their marks for all the goods and services that relate to their business (however tenuous). Assessing the veracity of this theory is beyond the scope of this article, but a project is underway to examine the registration data of the marks in this dataset to test this hypothesis.

C Concurrent Claims in Passing Off and ACL s 18

As discussed earlier, in Australia it is common for parties to plead a trade mark infringement case under s 120 of the 1995 TM Act alongside common law passing off and/or breach of ACL s 18. For this article, cases pursued under all three causes of actions/claims are termed ‘treble’ pleadings, cases pleaded under two heads are labelled ‘double’ pleadings and those claimed under any or all subsections of s 120 are ‘single’ pleadings.

Table 3 (below) breaks down the various combinations of pleadings that were brought across the 78 cases in this study. Plaintiff success on any claim in a pleading (for example, s 120 passing off or ACL s 18) was coded as win, loss or neutral, respectively. ‘Neutral’ was required because if a court found a win on a s 120 claim they often did not consider collateral claims in a substantive way. In these cases, there was no express win or loss attributed to the collateral claim. In addition to individual claims, the plaintiff’s overall success in its litigation was recorded. If the plaintiff had success on any one of its claims, the case was recorded as a ‘net win’ for the plaintiff. To determine the ‘net win’, each case was examined. If there was a win on any one of s 120, passing off or s 18, the outcome was coded as a ‘net win’.

In terms of popularity of claims, the top line of Table 3 shows that a s 120(1) treble pleading was the predominant action, pursued in 36/78 cases or 46%. The next most popular proceeding was a single pleading under s 120(1) (11/78 cases or 14%) and the third most popular was a treble pleading under ss 120(1)–(2)/passing off/

‘Average Hearing Days’ shows the average number of hearing days for each form of pleading. The average number of hearing days can be considered a reasonable proxy for the cost of the hearing, which in turn is a rough proxy for the cost of the entire proceeding.^[107] The length of a hearing was calculated in units of days and estimated from the information given in the header of the judgment.

What can be observed about the top three forms of pleading? The top panel in Table 3 (below) shows treble claims (that is, where a party pleaded three causes of action). The average hearing days for treble claims were high at 4.0 days per case, which is slightly higher than the average of 3.7 days for the total population. The top two lines of that panel show that s 120(1) treble claims and s 120(1) and (2) treble claims took 4.5 and 4.4 hearing days, respectively. In contrast, single claims took 3.0 and 3.4 days to hear, which is under the 3.7 day average. This makes sense: the more claims to be considered, the more time needed to present evidence and arguments and the more time it will take the court to consider them.

Regarding win rates, treble claims of any nature (as shown in the top panel) made up 52/78 cases (or 67%) of the sample, but the win rate on the s 120 component was on average 46%. However, factoring in wins from passing off and s 18, the ‘net win’ rate rose to 56%. In other words, there were five cases that lost on the s 120 claim, but won on passing off or ACL s 18.^[108] In contrast, for the s 120 single actions (17/78 cases, or 22% of the population), s 120 win rates were very high, accounting for 59%. This differs from the 46% win rate on s 120 issues in treble claims. The hearing days were also much shorter, averaging 3.2 days for s 120 single pleadings (compared with the 3.7 day average of the population as a whole).

		Cases (n = 78)		Hearing days			Win rate
	Pleaded actions	No.	%	Total		Av.	Net	s 120	PO	s 18
TREBLE CLAIMS	s 120(1), PO, s18	36	46.2%	162	4.5		50%	41.7%	27.8%	33.3%
	s 120(1)–(2), PO, s18	8	10.3%	35	4.4		50%	37.5%	25%	75%
	s 120(2), PO, s18	3	3.8%	3	1.0		100%	66.7%	66.7%	66.7%
	s 120(3), PO, s18	2	2.6%	2	1.0		100%	50%	100%	100%
	s 120(2)–(3), PO, s18	2	2.6%	4	2.0		50%	50%	0%	0%
	s 120(1)–(3), PO, s18	1	1.3%	0	0.0		100%	100%	100%	100%
		52	66.7%	206	4.0		55.8%	46.2%	30.8%	34.6%

SINGLE CLAIMS	s 120(1)	11	14.1%	33	3.0		54.5%	54.5%	n/a	n/a
	s 120(1)–(2)	5	6.4%	17	3.4		60%	60%	n/a	n/a
	s 120(2)	1	1.3%	5	5.0		100%	100%	n/a	n/a
		17	21.8%	55	3.2		58.8%	58.8%

DOUBLE CLAIMS	s 120(1), s18	5	6.4%	18	3.6		60%	60%	n/a	20%
	s 120(1)–(2), s18	2	2.6%	9	4.5		50%	50%	n/a	50%
	s 120(2), s18	1	1.3%	1	1.0		100%	100%	n/a	100%
	s 120(1)–(3), s18	1	1.3%	3	3.0		100%	100%	n/a	n/a
		9	11.5%	31	3.4		55.6%	55.6%		33.3%
	TOTAL	78	100%	292	3.7

The bottom third panel of Table 3 (above) refers to cases that presented a double pleading under s 120 and s 18. Note that the sub-sample size only includes nine cases, making inferences difficult. For these nine cases, the s 120 win rate was 56% and the net win rate (which factors in the effect of s 18 wins) was also 56%.

The data in Table 3 prompts the question: why are treble pleadings (at 4.0 days long) so popular when they take longer to hear than single s 120 claims (at 3.2 days) and have relatively poor success in winning on the s 120 portion of the pleading (46% compared with 59% and 56% for single and double pleadings)? The answer to this may lie in the increased chance of getting a ‘net win’ overall. For example, on average, a treble claimant won on the s 120 claim 46% of the time. However, when factoring in wins on either passing off or ACL s 18, the win rate for the plaintiff rose to 56%. That is, litigants who lost on the s 120 claim could still win under either passing off or s 18. This 10 percentage point increase represents a 21.7% rise in ‘net wins’ and may justify the additional costs of pursuing a concurrent claim, particularly in ‘close’ cases or where the plaintiff is vulnerable to a cross claim for invalidity.

These results can also be interpreted to mean that more confident plaintiffs may pursue s 120 single claims or s 120/ACL s 18 claims to receive the benefit of a shorter hearing and lower costs. Conversely, the inference is that less confident (or perhaps more determined) plaintiffs pursue treble claims for an increased chance of a ‘net win’, which may offset the increased costs of an extended hearing.

This section refers to the subset of 52 cases in Table 3 (first panel) in which there was a treble pleading. The data reveals that the plaintiff in treble concurrent claims won the s 18 portion of the pleading in 19/52 cases. Among these 19 cases, the plaintiff won the passing off claim in 17/19 cases, while the passing off claim in 2/19 cases was unclear or not considered. The plaintiff lost the s 18 claim in 28 cases. Of these, the plaintiff also lost the passing off case in 27/28 cases, while in one case the passing off claim finding was ‘neutral or unclear’.^[110] There were five cases in which the s 18 claim was not discussed in the judgment. Of these, in 4/5 cases the passing off claim was also neutral or not discussed, while in one of these cases the passing off claim was lost.^[111] Prima facie, it appears that passing off and ACL s 18 outcomes are synchronous. However, a close reading of the cases shows that the outcome of the concurrent portion of the case tends to be driven by ACL s 18 reasoning with the passing off action passively following. These cases will be discussed in Part IIIC(3) below.

The decline in the reliance of passing off as a cause of action in concurrent claims can also be seen over time (Table 4, below). The data was split into two groups: Decade One (1 January 1996–31 December 2005), 33 cases; and Decade Two (1 January 2006–1 January 2016), 45 cases.

A comparison of Decade One and Decade Two shows that reliance on passing off and ACL s 18 as concurrent actions declined in relative terms. That is, in Decade One, passing off presented in 76% (25/33) of trade mark cases, while in Decade Two, it presented in only 60% (27/45) of cases. Cases arguing under ACL s 18 presented in 85% of cases in Decade One (28/33), but in only 73% of cases (33/45) in Decade Two. Passing off claims declined more sharply than did concurrent claims under ACL s 18. This decline in the popularity of passing off is supported by the finding that there were no double claims in s 120/passing off. Instead, litigants who pursued a double claim did so as a s 120/ACL s 18 claim.

It is difficult to identify the reasons for the relative overall decline in concurrent claiming from Decade One to Decade Two. It may be that plaintiffs grew more confident in litigating under their registered rights under s 120 after the

1995 TM Act ‘settled in’ after the first 10 years. Alternatively, a speculative inference could be that the 2008 ‘collapse of the US sub-prime housing bubble’^[112] and decline in economic conditions lowered the tolerance for wider-scale concurrent litigation.

Regardless, this relative increase in reliance on s 120 (that is, statutory trade mark rights) has significant implications. It may indicate that more recent plaintiffs are not as concerned about vulnerability to a cross claim for revocation of their statutory rights; or it may be reflective of ‘cluttering’ of the register. Proving these propositions is beyond the information provided by this data. However, speculative inferences will be discussed further below.

When looking at concurrent claims within treble pleadings (Table 3, above), it was apparent that ACL s 18 reasoning was more prevalent than passing off reasoning. This supports some speculation in the literature that passing off is becoming less relevant than ACL s 18 in trade mark litigation.^[113] A close reading of the judgments shows that of the 52 cases where both passing off and s 18 were pleaded together, there were only 24 cases where comments were made about concurrent claims in passing off or ACL s 18. This level of commentary may seem low, but it must be recalled that in trade mark infringement litigation the primary rights relied upon are rights under the 1995 TM Act.^[114] In contrast, rights under passing off and ACL s 18 require some proof of right. If a court determines the case under a head of s 120, it does not have to explore the concurrent claims.

Examining these 24 cases in depth, in 11 cases^[115] the outcomes of the concurrent claims were driven by ACL s 18 reasoning, while the passing off claim was a secondary consideration. For example, in Lift Shop,with regards to passing off Buchanan J said ‘[t]he applicant accepted that this cause of action [passing off] added nothing to the claim under ACL s 18 and did not press it. In my view, that was an appropriate position to take’.^[116] Somewhat dismissive comments along these lines were not uncommon. For example, in Australian Postal Corporation v Digital Post Australia Pty Ltd (No 2) it was held that ‘[t]he parties did not contend that, in the event the ACL claims were dismissed, there was any likelihood of the claim under the tort of passing off succeeding’.^[117] Similarly, in Louis Vuitton, it was held that:

Although, as I have said, the applicant’s case included claims in passing off, counsel accepted that the relief ... under the ACL would be no less efficacious for it ... I was not addressed in any detail on the matter of passing off, and I do not propose to make any findings in relation to it.^[118]

In Virgin Enterprises Ltd v Virgin Home Loans Pty Ltd, Hely J stated that ‘it is not necessary to distinguish between [ACL s 18] ... and passing off. If the applicants could not sustain a claim based upon s 52, I cannot conceive any different result applying ... passing off’.^[119]

In contrast, in only three cases did the passing off reasoning lead the reasoning on ACL s 18.^[120] For example, in Pierson’s Pro-Health Pty Ltd v Silvex Nominees Pty Ltd (No 2), Lucev FM held ‘[h]aving regard to the court’s findings with respect to ... misrepresentation for the purposes of passing off ... similar considerations apply with respect to ... misleading representations ... under ss 52 and 53(c) of the TP Act’.^[121]

In four out of the 24 cases,^[122] courts made an effort in their judgment to distinguish passing off and ACL s 18 claims, but the courts’ comments did not have a substantive effect. That is, they were more descriptive of the differences between the actions, but did not result in distinctions in outcome. For example, in AMI v Bade Medical (No 2), Flick J said ‘there is a further distinction between passing off and s 52. For the purposes of passing off, it may be that deception must continue to the “point of sale”’.^[123]

These findings support recent arguments that courts in trade mark litigation fail to provide a ‘strict demarcation of the causes of action’.^[124] Examining the reasoning in these cases reveals that, while some judges do try to acknowledge the differences between passing off and ACL s 18,^[125] overall there is a strong tendency to let s 18 determinations lead the reasoning on passing off claims. That is, any stated determinations on passing off are cursory at best. The empirical data indicate no obvious benefit from arguing both passing off and ACL s 18 in the context of s 120 litigation.

Given the above, what is the relevance of passing off as a concurrent claim in a s 120 case? As discussed earlier, passing off’s perceived remedial advantages, such as exemplary damages, have now been made irrelevant under the expanded damages head under s 126 of the 1995 TM Act. Its other advantages — such as coverage of unregistered marks and its role as a safety net — can also be found under the law of ACL s 18 and, to some extent, the expanded scope of the 1995 TM Act. The ACL s 18 claim also has the distinct advantage of not requiring ‘reputation’ as a formal element of determining liability. Recall that to mount a passing off claim, the plaintiff must demonstrate reputation or goodwill as embodied by the plaintiff’s trade insignia. In contrast, the statutory claim requires a demonstration of misleading conduct leading to potential or actual consumer deception.

Given the absence of a clear benefit to pursuing passing off, this article argues that it would be a poor use of litigant and court resources to enter a passing off claim as part of a s 120 infringement suit. This is supported by the data that shows litigants — if choosing a double claim — increasingly do so under s 120 and ACL s 18 (and not s 120 and passing off). In addition, when analysing a treble claim, courts favour resolving the concurrent portion by focusing their reasoning under ACL s 18 and not passing off. It may be that pursuing multiple claims prior to litigation has its advantages in terms of settlement. However, pursuing passing off as a concurrent claim in court provides no apparent benefit.

IV Conclusion

This article reports on a comprehensive study of 20 years of substantive trade mark infringement litigation in Australian courts under s 120 of the 1995 TM Act. The findings are based on a unique dataset coded to examine not only rates of litigation and success, but also the relationship between concurrent claims under trade mark, passing off and consumer protection laws. The methodology adopted can be used to quantitatively test any area of law where reliance on ‘anecdata’ invite deeper inquiry. For example, this article has found empirical support for Heerey QC and Creighton-Selvay’s anecdotal assertions regarding passing off.

A critical finding is that the expansive ss 120(2) and (3) provisions have been of declining relevance to s 120 litigation. The data also reveals an overall decline in the pursuit of concurrent claims and a relative increase in the volume of s 120(1) litigation. However, it was found that concurrent claims can still provide net benefits in some cases. Where concurrent claims were pursued in a treble action, a plaintiff secured a higher rate of ‘net wins’ (46–56%). It is not possible to know the strategic intent of the litigating parties, but an inference made here is that a treble concurrent claim is useful for ‘close’ cases or where there is likely risk of a cross claim for invalidity. Nevertheless, in cases where there was a concurrent claim, the reasoning of the court was determined by the law of ACL s 18 and not passing off. This is contrary to Blakeney’s concerns in 1984^[126] and supports the position put forward by some practitioners^[127] that the ‘old tort’ has ‘passed on’, at least in the context of s 120 litigation.

One explanation for the change in dynamics could be that trade mark owners have their marks registered across an ‘excessive’ range of goods and/or services. This would allow them to assert their rights under s 120(1), rather than s 120(2), or under concurrent claims. Whether ‘cluttering’ has had an effect on the enforcement of rights is a critical question and addresses key concerns of the Productivity Commission and their recommendations for registration reform. To pursue this question, a cluttering project has begun in relation to the 78 cases in this infringement dataset.

In conclusion, the empirical study reported here has provided a robust and rigorous picture of trade mark enforcement over the last 20 years and tested the validity of numerous assumptions underlying trade mark law and practice. This study has also demonstrated the utility of empirical studies and a method for quantitative analysis of cases in an Australian context. The methodological design of this study can be applied to analyse any branch of law to enhance traditional legal scholarship.

^[∗]Senior Lecturer, Deakin Law School, Melbourne, Australia. I thank the following people who have commented on excerpts of this work in draft form: Associate Professor Elizabeth Adeney (Deakin University); Associate Professor Jason Bosland (University of Melbourne); Associate Professor Christopher Dent (Murdoch University); Associate Professor Dev Gangjee (Oxford University); Professor Michael Handler (UNSW); Janice Luck (Melbourne University); Dr John Morss (Deakin University); Professor Megan Richardson (Melbourne University); and Associate Professor Amanda Scardamaglia (Swinburne University). I also thank Alexander Kwong and the anonymous referees of this article for their helpful suggestions. Any errors are the author’s own.

^[1] Registrar of Trade Marks v Woolworths Ltd [1999] FCAFC 1020; (1999) 93 FCR 365, 371 [19]: ‘the 1955 Act was repealed and replaced by the Trade Marks Act 1995 (Cth) which applies to the present case. It also repealed the Trade Marks Act 1994 (Cth) which had been enacted to comply with Australia’s international treaty commitments but had not come into force when the 1995 Act was passed’. See also Working Party to Review the Trade Marks Legislation, Recommended Changes to the Australian Trade Marks Legislation (Australian Government Public Service, 1992) 74; Mark J Davison and

Ian Horak, Shanahan’s Australian Law of Trade Marks and Passing Off (Lawbook Co, 6^h ed, 2016) 648–9 [85.25].

^[2] Marrakesh Agreement Establishing the World Trade Organization, opened for signature 15 April 1994, 1867 UNTS 3 (entered into force 1 January 1995) annex 1C (‘Agreement on Trade-Related Aspects of Intellectual Property Rights’).

^[6] Barton Beebe, ‘An Empirical Study of the Multifactor Tests for Trademark Infringement’ (2006) 94(6) California Law Review 1581. See specifically in relation to case selection and the coding of wins and losses.

^[7] Hall and Wright, above n 5, 74, where the US authors reviewed the content analysis projects published during the 1990s and argued that ‘[c]ontent analysis projects appear somewhat more likely to generate discussion and citation than law review articles more generally’.

^[8] Productivity Commission, Intellectual Property Arrangements (Inquiry Report No 78, Productivity Commission, 23 September 2016) 372.

^[13] Section 7(4) of the 1995 TM Act describes ‘use of a trade mark in relation to goods’ as meaning ‘use of the trade mark upon, or in physical or other relation to, the goods (including second-hand goods)’ (emphasis in original). Note that this provision reflects more the physical application of the mark, rather than its meta-physical ‘use’, which is dealt with in case law.

^[14] A trade mark is defined under 1995 TM Act s 17 as ‘a sign used, or intended to be used, to distinguish goods or services dealt with or provided in the course of trade by a person from goods or services so dealt with or provided by any other person’.

^[15] Under 1995 TM Act s 6 ‘goods of a person means goods dealt with or provided in the course of trade by the person’ (emphasis in original).

^[16] 1995 TM Act (Cth) s 120(4): ‘in deciding, for the purposes of paragraph (3)(a), whether a trade mark is well known in Australia, one must take account of the extent to which the trade mark is known within the relevant sector of the public, whether as a result of the promotion of the trade mark or for any other reason’ (emphasis in original).

^[20] Working Party to Review the Trade Marks Legislation, above n 1, 74 states that

[d]oubts were expressed by the Working Party about the desirability of adopting the wider infringement provisions, based largely on perceived difficulties in determining the scope of protection as it relates to ‘similar’ goods and services, and in dealing with an unreasonable extension of rights gained by existing trade mark registrations ...

Further, ‘the Institute members of the Working Party remain concerned about the practical and commercial consequences of the recommended broadening of the infringement test’: at 74–75.

^[21] See, eg, Davison and Horak, above n 1, describing the 1995 TM Act as effecting a ‘radical change to the concept of infringement’. Under prior Acts, ‘[r]eputation and goodwill were rarely relevant nor was the likelihood of deception or confusion, except to the limited extent required to consider deceptive similarity of marks’: 648–9 [85.25]. Under the new Act, this remains so in the case of s 120(1) but ss 120(2) and (3) ‘introduce passing off like concepts into infringement actions although the precise meanings of some of the terms used in the subsection are uncertain and there are significant differences between those subsections and passing off’: 648–9 [85.25].

^[22] For a discussion of the history of passing off and the concept of reputation see Conagra Inc v McCain Foods (Aust) Pty Ltd [1992] FCA 159; (1992) 23 IPR 193.

^[23] Cadbury Schweppes Pty Ltd v Darrell Lea Chocolate Shops Pty Ltd [2007] FCAFC 70; (2007) 159 FCR 397, 418–19 [99], (Black CJ, Emmett and Middleton JJ) (‘Cadbury v Darrell Lea’):

whether or not there is a requirement for some exclusive reputation as an element in the common law tort of passing off, there is no such requirement in relation to Pt V of the Trade Practices Act. The question is not whether an applicant has shown a sufficient reputation in a particular get-up or name. The question is whether the use of the particular get-up or name by an alleged wrongdoer in relation to his product is likely to mislead or deceive persons familiar with the claimant’s product to believe that the two products are associated, having regard to the state of the knowledge of consumers in Australia of the claimant’s product ...

^[24] Hansen Beverage Co v Bickfords (Australia) Pty Ltd [2008] FCAFC 181; (2008) 171 FCR 579, 588–9 [44] (Tamberlin J) (citations omitted) states that

s 52 of the TPA imposes no requirement that any particular reputation must be established before a breach of s 52 can be made out. In speaking of the relationship between s 52 and the tort of passing off, Stephen J observed in Hornsby Building Information Centre Pty Ltd v Sydney Building [1978] HCA 11; (1978) 140 CLR 216 at 226 that the remedy under the TPA: ... ‘will not, as in passing off, be founded upon any protection of the trader’s goodwill but, being directed to preventing that very deception of the public which is injuring his goodwill, it will nevertheless be an effective remedy for that of which he complains’ ... ‘if what is in question is truly a contravention of s 52(1); that is to say, is conduct which is misleading or deceptive. It is only this with which s 52(1) is at all concerned. It is not concerned, as such, with any unfairness of competition in trade as between two traders.’ Accordingly, the sufficiency of the reputation which is required to be shown may be less in proceedings under the TPA than in proceedings alleging passing off.

^[25] See, eg, Wingate Marketing v Levi Strauss [1994] FCA 1001; (1994) 49 FCR 89, 117 (Gummow J) in relation to an infringement action under 1955 TM Act:

it is generally accepted that the tort of passing-off is concerned with the protection of the business goodwill of the plaintiff against damage by the misrepresentations made by the defendant. Neither protection of goodwill nor deceptive conduct are the primary concern of the action for trade mark infringement under the present legislation. However, arguments were propounded before us which assumed that the contrary was the case ....

^[26] For an overview of the historical progression of the related actions up to the 1995 TM Act, see Wingate Marketing v Levi Strauss [1994] FCA 1001; (1994) 49 FCR 89, 117–24 (Gummow J).

^[28] See, eg, Campomar Sociedad, Limitada v Nike International Ltd where Campomar had registered the name NIKE as part of its registrations for perfumery. Nike International (the sportswear company) tried to have the registration revoked even though they were not in the cosmetic business: [2000] HCA 12; (2000) 202 CLR 45. As the High Court pointed out, even if their case under the 1955 TM Act failed, ‘there is available to Nike International the law with respect both to the tort of passing-off and to misleading or deceptive conduct within the meaning of s 52 of the TP Act’: at 75 [67].

^[30] In Reckitt & Colman Products Ltd v Borden Inc [1990] UKHL 12; (1990) 17 IPR 1 at 7 (the ‘Jif Lemon’ case), in reference to the impugned lemon shaped bottle containing lemon juice, Lord Oliver of Aylmerton described get-up in a passing off context as follows:

first, he must establish a goodwill or reputation attached to the goods or services which he supplies in the mind of the purchasing public by association with the identifying get-up (whether it consists simply of a brand name or a trade description, or the individual features of labelling or packaging) under which his particular goods or services are offered to the public, such that the get-up is recognised by the public as distinctive specifically of the plaintiff’s goods or services.

^[31] In the US, it has been said that ‘[t]rade dress is also useful to convey more than merely the presentation of an article. It covers innovative trading styles, especially for restaurants and retail outlets and might also include a particular format of conducting business. In other words, the term “trade dress” encompasses the whole visual image presented by a trader to customers’: Trevor Stevens, ‘The Protection of Trade Dress and Colour Marks in Australia’ (2003) 93(2) The Trademark Reporter 1382, 1383 discussing J Thomas McCarthy, McCarthy on Trade Marks and Unfair Competition (West Group USA, 4^th ed, 1994) 8.1.

^[34] See the discussion of the role of passing off and distinctive marks under the 1955 TM Act in Oxford University Press v Registrar of Trade Marks [1990] FCA 175; (1990) 24 FCR 1, 15 (Gummow J).

The above examination of subsections (3) to (6) of s 41 of the Act demonstrates that the section limits the scope of operation of s 33(1) of the Act so far as the question of capacity to distinguish is concerned. ... It is thus not the case, as might otherwise be concluded from the terms of s 33(1), that if the registrar is uncertain whether a trade mark is capable of distinguishing the applicant’s goods or services from the goods or services of other persons then he or she must accept the application.

^[36] IP Australia, Trade Marks Office Manual of Practice and Procedure (1 August 2018) <http://manuals.ipaustralia.gov.au/trademarks/Part_22.pdf> pt 22, 8: ‘[s]ection 41 was repealed and re-enacted by the Intellectual Property Laws Amendment (Raising the Bar) Act 2012. The changes to section 41 clarify that the presumption of registrability does apply to section 41.’

^[37] Exemplary damages could be had for ‘conscious and contumelious disregard for the wronged party’s rights and to deter the wrongdoer from committing like conduct again’: XL Petroleum (NSW v Caltex Oil (Australia) Pty Ltd [1985] HCA 12; (1985) 155 CLR 448, 471 (Brennan J). See Taleb v GM Holden Ltd [2011] FCAFC 168; (2011) 286 ALR 309, 317 [41] (Finn and Bennett JJ) (citations omitted):

it is accepted in this country that the circumstances of a passing off may be such as to make it appropriate to punish a respondent for conduct showing a conscious and contumelious disregard for the wronged party’s rights and to deter the wrongdoer from committing like conduct again. Such awards have not commonly been made, the apparent reason for this being that the passing off has occurred in conjunction with a copyright infringement for which substantial ‘additional damages’ have been awarded under s 115(4) of the Copyright Act 1968 (Cth): for example, Deckers Outdoor at [115] ....

^[38] For a discussion of additional and exemplary damages, see Futuretronics.com.au Pty Ltd v Graphix Labels Pty Ltd (No 2) [2008] FCA 746; (2008) 76 IPR 763.

^[39] Intellectual Property Laws Amendment (Raising the Bar) Act 2012 (Cth) s 29.

^[40] For a discussion of the intersection of exemplary damages and the Copyright Act 1968 (Cth), see Futuretronics.com.au Pty Ltd v Graphix Labels Pty Ltd (No 2) [2008] FCA 746; (2008) 76 IPR 763. See generally Ricketson and Creswell, The Law of Intellectual Property: Copyright, Designs & Confidential Information, (Thomson Reuters, 2^nd ed, 1999) [13.900]–[13.940].

^[41] Note that this discussion relates to passing off as a collateral claim in the context of s 120 TM 1995 Act litigation. It is acknowledged that passing off still retains independent relevance in misleading conduct claims in areas such as advertising.

^[42] For example, in Toddler Kindy Gymbaroo Pty Ltd v Gymboree Pty Ltd both sides sought removal of their opponent’s mark for want of validity and the plaintiff ultimately found success on its passing off claim: [2000] FCA 618; (2000) 100 FCR 166 (‘Gymbaroo v Gymboree’).

^[43] See, eg, CSR Ltd v Resource Capital Australia Pty Ltd [2003] FCA 279; (2003) 128 FCR 408, where the mark was used for cyber-squatting purposes.

^[44] For a discussion of passing off and ACL s 18 more generally, see Robert Burrell and Michael Handler, Australian Trade Mark Law (Oxford University Press, 2^nd ed, 2016) chs 12–13.

^[45] Cadbury v Darrell Lea [2007] FCAFC 70; (2007) 159 FCR 397, 419 [99] (Black CJ, Emmett and Middleton JJ), states that, for the statutory claim

[t]he question is whether the use of the particular get-up or name by an alleged wrongdoer in relation to his product is likely to mislead or deceive persons familiar with the claimant’s product to believe that the two products are associated, having regard to the state of the knowledge of consumers in Australia of the claimant’s product.

^[46] Ibid 418–19 [99] (Black CJ, Emmett and Middleton JJ): ‘whether or not there is a requirement for some exclusive reputation as an element in the common law tort of passing off, there is no such requirement in relation to Pt V of the Trade Practices Act’.

^[47] Burrell and Handler, above n 44, 477 state that ‘it will be for the plaintiff to demonstrate sufficient reputation to ground its action for passing off or breach of the statutory prohibition on engaging in misleading or deceptive conduct’.

^[48] Ed Heerey QC and Peter Creighton-Selvay, ‘Trade Marks and Passing Off — Has the Old Tort Passed On?’ (2013) 94 Intellectual Property Forum 25, 30: ‘it is possible that where misleading or deceptive representations are made, and those representations are not made in trade or commerce, an action for passing off (or trade mark infringement) might succeed, whereas an action for contravention of s 18 of the ACL might fail’. The meaning of a misrepresentation outside of trade or commerce was discussed in the case of Attorney-General; Ex rel Elisha v Holy Apostolic & Catholic Church (1989) 37 NSWLR 293. The authors discuss the judgment of Young J, which found that the impugned conduct associated with the conduct of a church had not occurred in trade and commerce and was not actionable under s 52 of the TPA 1974 (now ACL s 18).

^[51] Note that the original language under s 52 of the TPA 1974 read: ‘(1) A corporation shall not, in trade or commerce, engage in conduct that is misleading or deceptive.’ In 2010, ‘corporation’ was replaced by ‘person’ under s 18 of the ACL. As a law of the Commonwealth, the ACL references to ‘persons’ applies to ‘corporations’: Competition and Consumer Act 2010 (Cth) s 131(1).

^[53] Blakeney argued that s 52 was modelled on s 5 of the US Federal Trade Commission Act 1914, inspired ‘by the enviable record in consumer protection’: Michael Blakeney, ‘Old Wine in New Bottles: Influence of the Common Law on the Interpretation of Section 52 of the Trade Practices Act’ (1984) 58(6) Australian Law Journal 316, 316. Blakeney laments that the intentionally ‘broad language of the section’ has been ‘interpreted more narrowly by Australian judges than by their American counterparts’: at 316.

^[58] See also Vicki Huang, Kimberlee Weatherall and Elizabeth Webster, ‘The Use of Survey Evidence in Australian Trade Mark and Passing Off Cases’ in Andrew T Kenyon, Ng-Loy Wee Loon and Megan Richardson (eds), The Law of Reputation and Brands in the Asia Pacific (Cambridge University Press, 2012) 181 (‘Huang et al study’).

^[61] Ibid 358: ‘two thirds of the counterfeit cases 66% were resolved within 10 court hours and within less than a year from the date of issue’.

^[64] See, eg, Google Inc v Australian Competition and Consumer Commission (2013) 249 CLR 435, where the High Court determined whether Google could be liable for misleading or deceptive conduct under s 52 of the TPA 1974 (now ACL s 18) for its sale of sponsored links that were triggered by keywords where a competitor had purchased the trade marked names of a rival as keywords. For an example related to the application of passing off to protect an advertising campaign, see Cadbury Schweppes Pty Ltd v Pub Squash Co Pty Ltd where the Privy Council held that a passing off claim is not limited to a trade mark, but ‘may extend to other descriptive material such as the component parts of an advertising campaign’: [1980] 2 NSWLR 851, 851. For an example of passing off applied to celebrity indicia, see Hogan v Pacific Dunlop Ltd (1988) 12 IPR 225, affirmed by Pacific Dunlop Ltd v Hogan [1989] FCA 185; (1989) 23 FCR 553, where the applicants relied on passing off and breach of the

TPA 1974 ss 52, 53(c) and (d) to claim the respondent’s TV commercial for shoes (a parodic spoof of a scene from Crocodile Dundee (directed by Peter Faiman, Rimfire Films, 1986)) represented a false connection between the applicant and the goods of the respondent.

^[71] See, eg, Société Des Produits Nestlé SA v Christian (No 14) [2014] FCCA 2968 (19 December 2014), where the absence was reported and rectified.

I excluded a small minority of fact patterns that led courts to apply the multifactor test in ways that could skew the results of the study. In most counterfeiting opinions, for example, the likelihood of confusion is very clear and the factors tend to weigh overwhelmingly in favor of the plaintiff. The same is true of opinions involving an alleged breach of a franchising, licensing, or distribution agreement. These opinions were thus excluded from the sample. For similar reasons, I also excluded opinions on motions to dismiss or on motions where the non-moving party failed to appear. I retained and noted opinions involving claims of reverse confusion, and fact patterns in which the defendant repackaged plaintiff’s goods.

^[75] See Huang et al, above n 58, 189. On average, counterfeiting cases took 1.1 hearing days, compared with 2.4 days for passing off and trade mark infringement, which illustrates the less complicated nature of the former.

^[77] Ibid 357, where the authors held this to be significant, in contrast to patent cases, which involved difficult questions of claim construction: ‘trade mark infringement actions turn largely on issues of fact’ and ‘are inherently impressionistic’ leaving appellate courts reluctant to set aside a trial court’s findings.

^[78] Beebe, above n 6, 1650: ‘This resulted in a sample of 337 opinions. I excluded the six opinions in which the outcome of the multifactor test was reversed, which yielded a final sample of 331 opinions’.

^[79] As described by Hall and Wright, ‘conventional legal scholarship analyzes issues presented in one case or a small group of exceptional or weighty cases, content analysis works by analysing a larger group of similarly weighted cases to find overall patterns’: above n 5, 66.

^[80] A full list of the 78 cases and information relating to coding is available from the author.

^[83] Note that the Bosland et al study, above n 59, had the benefit of more detailed estimates of time from FEDCAMS, a database that is no longer accessible.

^[88] Louis Vuitton Malletier v Sonya Valentine Pty Ltd, where there was a win only against one allegedly infringing mark out of two litigated marks: [2013] FCA 933; (2013) 222 FCR 45 (‘Louis Vuitton’); Hills Industries Ltd v Bitek Pty Ltd, where there was a win only in relation to a subset of goods within a class: [2011] FCA 94; (2011) 214 FCR 396.

^[89] Twenty-six percent is calculated as a percentage of the total number of 84 cases, ie 22/84. The total sample size for this question is 84 because the working sample size of 78 excludes six trial decisions that were reversed on appeal.

^[90] Weatherall and Jensen, above n 12, 266 finding the proportion of patent matters appealed to the Full Federal Court was 59%: ie 17 out of 29 proceedings for the period 1997–2003.

^[91] Note that this does not mean cases that did not seek an appeal were necessarily correct. Parties may choose not to appeal based on factors such as cost or time and not just on the merits of the case. Inferences as to the ‘correctness’ of the decisions in the 78 cases are not being made here. Rather, these cases are the best population to represent the nature of reasoning at first-instance courts.

^[92] Note that a similar method to Beebe’s method of case selection was adopted for this study. See Beebe, above n 6, 1596–7.

^[94] Clermont and Eisenberg argue that, while it is recognised that win/loss rates can be useful at a descriptive level, the descent ‘to the inference level’ can ‘lead the observer astray’: Kevin M Clermont and Theodore Eisenberg, ‘Do Case Outcomes Really Reveal Anything About the Legal System? Win Rates and Removal Jurisdiction’ (1998) 83(3) Cornell Law Review 581, 588. For critiques of the Priest-Klein model, see, eg, Donald Witman, ‘Is the Selection of Cases for Trial Biased?’ (1985) 14(1) Journal of Legal Studies 185; Steven Shavell, ‘Any Frequency of Plaintiff Victory at Trial Is Possible’ (1996) 25(2) Journal of Legal Studies 493; Daniel Kessler, Thomas Meites, and Geoffrey P Miller, ‘Explaining Deviations from the Fifty-Percent Rule: A Multimodal Approach to the Selection of Cases for Litigation’ (1996) 25(1) Journal of Legal Studies 233.

^[96] Clermont and Eisenberg, above n 94, 581. Note that there are other selection biases that are possible, such as publication bias. However, trade mark cases are generally pursued in the Federal Court of Australia, which is a superior court of record, Federal Court of Australia Act 1976 (Cth) s 5.

^[98] Ibid 589. The win/loss rates should not be used for inferences as to the ‘underlying mass of disputes and cases’ at large.

^[99] It is beyond the scope of this article to determine more complex motivations (eg, a litigant’s internal risk assessment) beyond the prima facie assumption of protecting their mark.

^[101] This was calculated by the author by determining the relative increase between the average real gross national income for January 1996–December 2005 ($230 156 million) and January 2006–March 2014 ($331 706 million) (data beyond this timeframe was not available). See Australian Bureau of Statistics, 5206.0—Australian National Accounts: National Income, Expenditure and Product (1994) Table 1 <http://www.abs.gov.au/AUSSTATS/abs@.nsf/DetailsPage/5206.0Mar%202014?OpenDocument> .

^[104] See, eg, Robert Burrell, ‘Trade Mark Bureaucracies’ in Graeme B Dinwoodie and Mark D Janis (eds) Trademark Law and Theory A Handbook of Contemporary Research (Edward Elgar, 2008) 95, 102–7; Burrell and Handler, above n 44; Christine Greenhalgh and Elizabeth Webster, ‘Have Trademarks Become Deceptive?’ (2015) 6(2) World Intellectual Property Organization Journal 109.

^[105] Productivity Commission, above n 8 [12.2, ‘Improving Effectiveness by Reducing the Scope of Cluttering’].

^[106] G von Graevenitz, C Greenhalgh, C Helmers and P Schautschick, Trade Mark Cluttering: An Exploratory Report (UK Intellectual Property Office, 2012). This definition was cited by the Productivity Commission: ibid. Note that cluttering in a US context was discussed by Beebe and Fromer, who described cluttering as a concept related to, but distinct from, trade mark depletion and trade mark congestion: Barton Beebe and Jeanne C Fromer, ‘Are We Running Out of Trademarks? An Empirical Study of Trademark Depletion and Congestion’ (2018) 131(4) Harvard Law Review 945, 950–51.

^[107] The Bosland et al study, calculated length of the proceeding from date of file till the date of judgment, as well as hearing length in hours. ‘These figures are of interest, in part because the amount of time taken to resolve a case is a proxy for the cost of the proceedings ... In terms of the efficient resolution of disputes, these figures compare favourably.’: Bosland et al above n 59, 357.

^[109] No. = number; av = average; PO = passing off; s 120 = 1995 TM Act s 120; s 18 = ACL s 18; Net win rate = s 120/PO/s 18; n/a = not applicable.

^[110] Lift Shop Pty Ltd v Easy Living Home Elevators Pty Ltd (2013) 103 IPR 511 (‘Lift Shop’).

^[112] Rick Battellino, Deputy Governor Reserve Bank of Australia, ‘Twenty Years of Economic Growth’ (Speech delivered at Moreton Bay Regional Council, Moreton Bay 20 August 2010) <http://www.rba.gov.au/speeches/2010/sp-dg-200810.html> .

^[114] This is because the plaintiff’s property rights are self-evident by way of their trade mark registration.

^[120] Bob Jane Corporation Pty Ltd v ACN 149 801 141 Pty Ltd (2013) 97 ACSR 127; SMA Solar Technology AG v Beyond Building Systems Pty Ltd (No 5) [2012] FCA 1483 (21 December 2012); Pierson’s Pro-Health Pty Ltd v Silvex Nominees Pty Ltd (No 2) [2010] FMCA 121 (12 March 2010).

^[123] [2009] FCA 1437; (2009) 262 ALR 458, 505 [194]. See also Mars v Sweet Rewards, in which the Court re-iterated that ‘passing off protects the goodwill of the applicant. Section 52, by contrast, protects consumers from being misled or deceived’: [2009] FCA 606; (2009) 81 IPR 354, 361 [27]. As another example, in Cody Opal (Australia) Pty Ltd v Dimasi there was some discussion of the ‘provisions of ss 52 and 53(c) of the TPA ... along with the sections providing for remedies for breach of these sections’ and it was said that these ‘overlap the area of operation of the tort of passing off’: [2004] SASC 357; (2004) 64 IPR 378, 411 [179].

^[125] See, eg, Verrocchi v Direct Chemist Outlet Pty Ltd [2015] FCA 234; (2015) 112 IPR 200, 210 [57]: ‘[w]hile there is a great deal of practical coincidence between the tort of passing off and contravention of the misleading or deceptive conduct provisions, the two claims have distinct premises’. See also Mars v Sweet Rewards [2009] FCA 606; (2009) 81 IPR 354, 361 [27]:

Passing off protects the goodwill of the applicant. Section 52, by contrast, protects consumers from being misled or deceived. In a passing off case, the existence of conduct which damages the applicant’s goodwill by wrongly suggesting a connexion between the respondent’s wares and the applicant’s will often also be misleading and deceptive. It is usual therefore in such cases for there to be an overlap between the passing off claim and the claim under s 52.

See also AMI v Bade Medical (No 2): ‘[i]n this respect, it may be that there is a further distinction between passing off and s 52. For the purposes of passing off, it may be that deception must continue to the “point of sale”. Whether or not there is such a requirement in respect to the tort, s 52 does not impose any such constraint’: [2009] FCA 1437; (2009) 262 ALR 458, 505 [194] (citations omitted).

Sydney Law Review

Huang, Vicki T --- "An Empirical Investigation of 20 Years of Trade Mark Infringement Litigation in Australian Courts" [2019] SydLawRw 4; (2019) 41(1) Sydney Law Review 105

I Introduction

A The Relevance of the Present Study

B Doctrinal Debates

II Method

A Case Selection

B Coding

III Results and Discussion

A Winning and Losing

B Sections 120(1), (2) and (3)

C Concurrent Claims in Passing Off and ACL s 18

IV Conclusion

Trial outcome	s 120 win	s 120 loss	Total
Reversed	5	1	6
Affirmed	7	9	16
TOTAL	12	10	22