Home | Databases | WorldLII | Search | Feedback University of Melbourne Law School Research Series

Bosland, J; Weatherall, K; Jensen, P --- "Trade Mark and Counterfeit Litigation in Australia" [2006] UMelbLRS 3

Last Updated: 12 June 2008

Melbourne Law School
Research Series

[2006] UMelbLRS 3

Trade Mark and Counterfeit Litigation in Australia

Jason Bosland, Kimberlee Weatherall and Paul Jensen

(2006) 4 Intellectual Property Quarterly 347

Trade Mark and Counterfeit Litigation in Australia[°]

Jason Bosland,[*] Kimberlee Weatherall[†] and Paul Jensen[‡]

Intellectual Property Research Institute of Australia

University of Melbourne

Summary

The effectiveness of trade mark protection depends on the enforceability of rights. However, little is known about how trade mark owners actually go about enforcing their trade marks in the civil courts. The few studies which have emerged recently show a high success rate for trade mark owners. In this study, we created a database of all trade mark enforcement decisions of Australian courts for the period 1997-2003. Analysing the nature and outcomes of the trade mark litigation, we found a more complex story than previous studies: counterfeit proceedings where the trade mark owner always wins and the alleged infringer often fails to show up in court on the one hand; more contentious proceedings on the other where the trade mark owner only succeeded around a third of the time. The study raises, although it cannot answer, some interesting questions about trade mark owner knowledge and motivations in entering litigation.

Introduction

Like all forms of intellectual property (‘IP’), trade mark rights must be enforceable in order to fulfil their intended objectives. It is a frequent complaint regarding trade marks (and other forms of IP) that enforcement is burdensome, costly, and at times impracticable. And while trade mark rights are private rights, enforceable by private parties, in the face of significant IP infringement and a very significant global trade in counterfeit products – one estimate suggests that as much as 5 per cent of world trade involves counterfeit goods[1] - governments in Australia and overseas have found themselves increasingly pressed to involve themselves in IP enforcement, and to improve mechanisms for enforcement. [2]

Despite the importance of court-based enforcement to IP, little is known about how trade mark owners actually go about enforcing their registered trade marks in the courts. Such enforcement has rarely been the subject of empirical study.[3] Academic study of the law ordinarily involves tracing the development of legal doctrine, and discussing particular trade mark cases. It often focuses on those judgments which involve controversial points of law, or where uncertainty exists as to how settled legal standards apply to novel facts. It is less common for any attempt to be made to obtain an overall picture of the number, type, and progress of cases brought before the courts, in part because such information is time-consuming to collect in a rigorous and reliable way in the absence of publicly available databases.[4] This paper attempts to address this gap in the literature by providing the first empirical study of trade mark enforcement in Australian courts. We analyse trade mark litigation using a newly-created database of all trade mark enforcement judgments issued by Australian courts of superior jurisdiction in the period 1997-2003. This quantitative empirical study is intended to provide a link between trade mark theory, law and practice by giving a clear and objective representation of the trade mark litigation that is occurring in the Australian courts. Part of the benefit of such a study is, of course, simply to provide a picture of what is going on in the courts. Doing so can confirm, or dispel commonly held ‘anecdotal beliefs’ about what is happening in the courts,[5] although there is relatively little public comment, at least in Australia, regarding trade mark owner success rates before the courts.[6]

Perhaps more importantly, however, looking at the types and outcomes of the cases which proceed all the way through to judgment in the courts is evidence of the decisions that trade mark owners and alleged infringers make in the course of a dispute – decisions about whether to pursue an issue (particularly by a trade mark owner), and whether to litigate or to settle an existing dispute. This is particularly so when cases, and results, depart from the picture that economic theories of litigation would predict. As the present paper shows, trade mark enforcement litigation does not conform to the standard picture. In overall terms trade mark owners have little to complain about: the success rate for trade mark owners across all cases in Australia is not dissimilar to the overall success rate in the United States.[7] However, we have also found that digging further into the figures reveals a more complex story, and raises some interesting questions.

In effect, our study clearly identifies for the first time the two ‘worlds’ of trade mark enforcement activity in Australian courts. These two types of cases have different profiles, and very different outcomes. On the one hand, we have the counterfeit proceedings. Since such cases are legally straightforward, the real question here is why these cases have to go all the way to judgment – and whether better methods of management are available. On the other hand, we have the non-counterfeit cases. Here, as we outline further below, we have found a success rate lower even than the fifty per cent rate predicted by standard economic models of litigation.[8] This finding raises interesting questions regarding why trade mark owners are choosing to litigate proceedings that they are losing more often than they win.

In Part 2 of this paper, we describe the law of trade mark infringement and introduce the predictions that economic theory would make regarding the progress, and outcomes of trade mark litigation; we also briefly introduce the methodology used to ‘test’ those predictions. Part 3 presents a statistical picture of trade mark enforcement in Australia, focusing on the number of cases, their duration, and outcomes. From this data one theme emerges most clearly: there are two ‘worlds’ of trade mark enforcement litigation in Australia: counterfeit proceedings, and non-counterfeit proceedings. These two ‘worlds’ bear little relation to each other in duration or outcomes. Part 4 explores each of these areas in some more detail, and make some suggestions regarding the policy implications of each, or areas where further research is required. Part 5 concludes.

Methodology and Basic Results

The law of trade mark infringement

Trade mark infringement in Australian law is defined by section 120 of the Trade Marks Act (the ‘Act’). Under that section, there are three mutually exclusive categories of conduct that constitute infringement;[9] generally speaking, these categories of infringement correspond to categories in trade mark law overseas.[10] Subsection 120(1) represents the traditional scope of trade mark protection, by providing that infringement occurs where a substantially identical or deceptively similar mark is used in relation to the same goods or services as those for which the trade mark is registered. In 1995, the concept of infringement was extended in two ways.[11] Subsection 120(2) extends the trade mark owner’s rights to related markets, by making it an infringement to use a substantially identical or deceptively similar trade mark in markets closely related to the registered goods or services.[12] Owners of well-known marks are given still broader protection: subs 120(3) gives them the right to prevent use of their mark even on unrelated goods where consumers would think such use indicated a connection between the mark owner and the infringer’s goods. These extensions were controversial at the time they were introduced, with some arguing that they would lead to uncertainty regarding the scope of a trade mark owner’s rights.

The distinctions established by the Act however do not fully capture the variety inherent in trade mark litigation. A further, important distinction can be drawn between counterfeit and non-counterfeit disputes.[13] Counterfeit cases involve an infringer attempting to reproduce – and substitute for – the goods (not just the trade mark) of the trade mark owner.[14] In legal terms, ‘counterfeit trade mark goods’ are defined in TRIPs as having two key features:[15]

(a) the infringer uses an identical mark to the registered trade mark; and

(b) the infringer uses the mark in relation to the same goods as those for which the trade mark is registered.

For the purposes of Australian law, this implies that, by definition, counterfeit cases are a subset of disputes falling within the boundaries of s 120(1).[16] While a party may be liable under s 120(1) where the mark is deceptively similar, in counterfeit cases the mark used is identical or substantially so. The striking feature about counterfeit cases is that they are legally very simple: they do not involve serious disputes over the boundaries of the trade mark owner’s rights. In mimicking the goods and the trade marks, the conduct of counterfeiters clearly falls within the ambit of conduct that a trade mark owner is entitled to prevent.[17]

Civil proceedings for trade mark infringement may be brought by a trade mark owner in the Federal Court of Australia, or in State Supreme Courts. Appeals from both are heard by the Full Federal Court of Australia.[18] In addition to the right to bring proceedings for infringement, it is also possible to cross-claim for infringement in circumstances where a third party seeks to have a registered trade mark removed,[19] or sues to prevent the trade mark owner issuing unfounded threats of legal proceedings.[20] In turn, a party sued for infringement of trade mark may counter with a claim of that the trade mark should be removed from the register.[21] Remedies for trade mark infringement include damages or an account of profits, injunctions, and in some cases, delivery up of infringing goods.[22]

How would we expect trade mark cases to proceed in the court system?

The present study is not a doctrinal consideration of trade mark law, but rather, a quantitative study, aimed at ascertaining the level of trade mark litigation in Australian courts, how it progresses and the outcomes. We provide ‘the numbers’ – how often trade mark owners succeed in infringement proceedings. The question, however, is – what is the significance if trade mark owners succeed in, for example, fifty per cent of cases pursued through to judgment? It is well known that the cases that do proceed all the way to trial are only a small proportion of the disputes where proceedings are issued. Of the 391 trade mark infringement proceedings filed in the Federal Court of Australia in the five year period 1998 – 2002,[23] only 48 (approximately 12 per cent) ended with some form of judgment on the merits of the case.[24]

Nor is this set of cases likely to be a representative sample of all the disputes that actually occur. As any lawyer knows, many disputes do not lead to the commencement of proceedings: letters of demand are part and parcel of day-to-day practice in this area, and such letters often lead to cessation of allegedly infringing conduct or settlement on other terms. Even after commencement a large proportion of cases settle. There is no doubt that the cases that make it all the way to judgment are not a representative sample of all trade mark disputes. Whether a case does proceed to trial depends on a series of decisions by trade mark owner and alleged infringer. In evaluating the statistical results, therefore, we must take into account how those decisions are made.

The most prominent economic theory regarding the selection of disputes for litigation is the ‘divergent expectations’ model (or the ‘Priest-Klein’ selection model).[25] Put simply, it suggests that litigants form a rational estimate of the likelihood of success in litigation, based on the facts and the current state of the law.[26] Which disputes settle will depend on the difference between the parties’ estimates of success. Where the outcome is clear cut, parties will agree on the expected outcome of litigation and will hence be more likely to settle. Only cases where the parties disagree on the likely outcome should go to trial – with the result that, in general, cases which are ‘close to the line’ will go to trial. Furthermore, since in general both parties are equally likely to err in their estimate of whether they will win, the win rate should approach 50 per cent.[27] The divergent expectation model has certain assumptions built in: that parties are rational, not behaving strategically, have equal stakes and equal (or symmetrical) information.[28] The success rate is likely to diverge from the 50 per cent rule where parties have different stakes in the litigation, or are pursuing, for some external, strategic reason, litigation they are objectively likely to lose: for example, to deter other non-parties, or enhance their reputation for an aggressive litigation stance. Or perhaps quirks in the legal process may render settlement not feasible. By revealing general trends in the kinds of cases brought before the courts, and the outcomes of those cases, a study of this kind may assist in identifying areas where the law may not be operating satisfactorily.

A recent US study of outcomes in trade mark litigation conducted by William Landes found that trade mark owners enjoyed an average success rate over time of 67 per cent – well above the fifty per cent rate predicted by standard economic theory. Landes’ explanation was that trial stakes in IP litigation generally are asymmetric: the plaintiff IP owner has more at stake than the defendant infringer. This is more obviously true of patent litigation than trade mark litigation: in most patent proceedings, the validity of the patent is challenged, meaning that the patent owner risks not only losing their argument over infringement, but also losing their patent altogether: a risk which is not as high in trade mark litigation.[29] Nevertheless, Landes argued, a trade mark owner who loses an infringement suit may risk losses in future licensing, and other revenues because the adverse court decision has, in effect, narrowed the scope of his intellectual property, as well as provided potential licensees with greater bargaining power. The result of the differential stakes, Landes argued, was a selection bias meaning that the IP cases that go to trial are likely to have higher win rates for plaintiffs on average than all civil trials.[30]

This is not obviously correct to us, for two reasons. First, as the discussion of the law above makes clear, there are several distinct kinds of trade mark disputes, where the stakes, and the issues, may well be different. We would want to know what kinds of cases were being brought before the courts before we could draw conclusions about the matters at stake, or any potential asymmetry of stakes. Second, different kinds of cases might have different outcomes. Before drawing conclusions about the appropriateness of the results obtained, we would want more detail regarding the outcomes of the different types of proceedings. Unfortunately, the Landes study was based on very ‘high level’ court data: Landes could ascertain, that a case was a trade mark case, its overall outcome (success or defeat for the plaintiff) and the time taken for the litigation to be finalised. He could not ascertain the nature of the case. For that, a more detailed dataset is required.

Methodology of the study

In order to test these issues, and establish the pattern of trade mark enforcement actions brought before Australian courts, and their outcomes, we constructed a dataset comprising the population of all[31] civil trade mark enforcement decisions, interlocutory and final,[32] both reported and unreported, rendered by Australian courts of superior jurisdiction in relation to registered trade marks for the period 1 January 1997 – 31 December 2003. In order to provide a comprehensive picture of trade mark litigation in Australia, we used a number of publicly available case-law databases in order to capture all of the relevant decisions.[33] Researchers at the Intellectual Property Research Institute of Australia (IPRIA) read every decision, and recorded information about those decisions in a custom-built database. That information includes ‘proceedings data’ – basic information about the proceedings including important dates, the type of case, the number of hours spent in court, the people involved: parties, counsel and judges.[34] We collected data on the outcome of each decision (‘outcomes data’), separately recording the outcome on infringement and validity of each trade mark in dispute. For each trade mark enforcement decision we have also recorded, where applicable, the grounds on which a trade mark owner failed – such as the non-infringement grounds (i.e. the trade mark was not used ‘as a trade mark’ or was not ‘substantially identical or deceptively similar’ to the registered mark). By reading the judgments, and thus obtaining detailed information about the nature of the case before the courts, we are able to present a more complex picture of this litigation than the existing statistical reports on trade mark enforcement in the United States.[35]

Three exclusions should be noted. First, we have not included the enforcement of “common law” trade marks under the tort of passing off. That could be a matter for future study. Second, we have not included opposition proceedings. A registered trade mark owner who becomes aware of an application for registration of a mark that will be, as they perceive it, too close to their own, may oppose registration of that mark,[36] thus heading off at an early stage a dispute that might otherwise become a fully-fledged infringement suit. Such opposition proceedings are an important means by which trade mark owners ‘enforce’ their rights. However, they are not included as part of the current study.[37] Third, we have not studied any success which parties are having in enforcing judgments obtained: this ‘real world’ issue is beyond the scope of this paper.

A statistical picture of trade mark enforcement in Australia

Number and Type of Trade Mark Proceedings

The starting point for understanding the general pattern of trade mark enforcement in the courts is to identify the type of proceedings brought before Australian courts. The breakdown of the different trade mark proceedings that have been pursued through to judgment in Australia in the period 1997-2003 are presented in Ta[38].38

Table 1: Number and Type of Trade Mark Proceedings, 1997-2003

Three striking facts emerge from Table 1:

1. There are very few appeals in trade mark law, particularly as compared to the relatively high appeal rate in patent litigation;[44]
2. A large number of counterfeit cases proceed to judgment; and
3. The ‘new categories’ of infringement have been litigated in a mere handful of cases.

The first of these observations is easy to explain. Patent disputes involve more complex legal issues than trade mark disputes: for example, the thresholds for patent validity (particularly novelty and non-obviousness) are much harder to establish than the threshold for trade marks (distinctiveness). Likewise, establishing patent infringement involves difficult questions of claim construction requiring specialised scientific knowledge and/or evidence. By contrast, the basic questions in a trade mark infringement action tend to turn largely on issues of fact:[45] rendering appeals much less likely to succeed. The Full Federal Court will only overturn a finding of fact by a Trial Judge which they consider erroneous, making appeals less worthwhile.[46] As the figures reported below also demonstrate, trade mark owners most often lose on the basis that the mark used by the alleged infringer is not close enough to the registered mark (i.e. is not ‘deceptively similar’).[47] Such matters are inherently impressionistic, and appellate courts are cautious in setting aside the Trial Judge’s findings.[48]

The second and third observations demand more attention. According to the theories noted above, straightforward cases should not continue all the way to judgment by a court: in theory, therefore, counterfeit proceedings should ‘settle out’. The high number of such judgments is worthy of further exploration, which we do in Part 4.1 below. The third observation also raises questions about the nature of the non-counterfeit proceedings being brought: those (controversial) areas where trade mark law would be expected to be uncertain are not being litigated through to judgment. This issue is discussed in Part 4.2. The data clearly identify two, distinct worlds of trade mark enforcement. The next question is whether these proceedings differ in any significant way.

How long are trade mark proceedings?

A second fact about trade mark litigation that may be of interest is the length of trade mark proceedings in Australian courts. The characteristics of the trade mark proceedings are presented in Table 2, which sets out the average length of the proceedings, in terms of case length (number of days elapsed between the issue date of the proceedings and the last decision date) and court hours (the number of hours spent in court in the pro[49]edings).49 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 pr[50]eedings.50

Table 2: Characteristics of Court Proceedings Involving Trade Marks, 1997-2003

In terms of the efficient resolution of disputes, these figures compare favourably with the equivalent results in US courts: the average case length is over 200 days shorter in Australia.[53] Appeals of course take much less time than original proceedings:[54] indeed, in general they are resolved within a year of being filed.

More striking about Table 2 is the contrast between counterfeit and non-counterfeit cases. Counterfeit cases are much faster to resolve than other types of trade mark disputes – particularly in terms of the hours spent in court. The time is still not negligible however – the average time from issue of proceedings to judgment even in the straightforward counterfeit cases was still over 10 months. The contrast is even more striking when the data in Table 2 is presented graphically in Figure 1 and Figure 2. The data has been broken down by counterfeit (C) and non-counterfeit (O) cases (appeals are not included). From these Figures, it can be seen that outliers affect the di[55]ribution of the non-counterfeit cases:55 some non-counterfeit proceedings are particularly long and involved. Counterfeit cases, on the other hand, were much more homogeneous: 10 of the 23 cases took less than 3 hours to resolve, and only 2 c[56]es took more than 20 ho[57]s to resolve.56 Similarly only 2 cases57 were not resolved within 600 days from the date of issue of proceedings. Two thirds of the counterfeit cases (66 per cent) were resolved within 10 court hours and within less than a year from the date of issue.

Figure 1: Frequency Distribution of Trade Mark Case Length, Original Cases

Figure 2: Frequency Distribution of Trade Mark Court Hours, Original Cases

Lawyer involvement in trade mark enforcement proceedings

Another feature which differentiates the counterfeit and non-counterfeit cases is the involvement of lawyers and the ‘cooperation’ of respondents with the legal process (as measured by their appearance at hearings). Data on the representation and appearance of alleged infringers in counterfeit cases is presented in Figure 3. An alleged infringer is ‘represented’ if they have legal counsel at the hearing, and ‘appeared’ if there was someone present (in person or through counsel). The shaded box in Figure 3 represents the set of cases where there was no one present at all for the alleged infringer.

Figure 3: Representation and Appearance of the Alleged Infringer[58] in Counterfeit Proceedings[59]

APPEARANCE IN COURT

REPRESENTATION

Yes No Total

Yes

No

Total

Notably, there is a high proportion of absentee defendants (13 out of 23 or 57 per cent of proceedings) and unrepresented litigants (20 out of 23 or 87 per cent). By way of comparison, the Federal Court Annual Report for 2003-4 indicates that approximately 34 per cent of all cases filed in the Court involve at least one litigant that is not represented by legal counsel.[61] The figures reflect the nature of counterfeit operations as ‘fly-by-night’ operators who are not interested in protecting the reputation of their business; rather, they simply liquidate one operation and set up another somewhere else. In such cases, the alleged infringer has little or nothing at stake in the outcome of the case: they will either declare bankruptcy or they will disappear. Some, at least, may never pay any damages awarded, making it hard to settle the case out of court.

By way of comparison, a very different pattern of representation and appearance is observed in non-counterfeit trade mark proceedings, as shown in Figure 4. In fact, the pattern here is almost the exact opposite of that observed in counterfeit cases: the vast majority of non-counterfeit cases involve instances where the alleged infringer is both legally represented at the hearing and makes an appearance in court. This possibly reflects the fact that these cases are more complex or involve legally contentious issues, and more importantly that the defendant has higher stakes in the l[62]igation.62

Figure 4: Representation and Appearance of the Alleged Infringer in Non-Counterfeit Proceedings

APPEARANCE IN COURT

REPRESENTATION

Yes No Total

Yes

No

Total

Infringement Outcomes

Outcomes and how to measure them

While the number (and length) of cases is of interest to practitioners and policy-makers, of even more practical interest is the outcome. The value of a registered trade mark to its owner lies in its ability to stop rivals from using a substantially identical or deceptively similar trade mark. To succeed in an action for trade mark infringement, a trade mark owner must establish that they have a registered trade mark (and if challenged, that the trade mark registration is valid), and that the alleged infringer must have engaged in conduct amounting to infringement under s 120.

The data on outcomes is presented at a proceedings level. In other words, we have recorded whether the trade mark owner achieved success in the proceedings, rather than in relation to any individual trade mark.[63] Using this approach, there are three possible results, which we have expressed from the perspective of the trade mark owner:

1. Win: trade mark owner successful in proving validity/infringement;
2. Loss: trade mark owner unsuccessful in establishing validity/infringement of any of the pleaded trade marks;
3. Partial win:[64] trade mark owner is successful in proving that some, but not all, of the alleged infringing acts actually infringed the registered trade mark;[65] or, where a trade mark owner has claimed infringement of more than one registered trade mark, the trade mark owner is successful in proving infringement in some, but not all, of the registered trade marks.[66]

There is another possible outcome: ‘not determined’, where the court did not render judgment. Those cases are not included in this analysis.[67]

Infringement Outcomes

The data on outcomes on the question of infringement is presented in Table 3. We distinguish here between three kinds of determination: original, appeal or ‘ultimate’. An ultimate decision is the final decision of a court rendered in a particular dispute: whether that decision is the final appeal, or the first instance decision where no appeal was made. In other words, the ‘ultimate decision’ is the ‘last word’ of the courts on whether a trade mark has been infringed, which makes it the most important result from the perspective of the trade m[68]k owner.68

Table 3: Trade Mark Infringement Outcomes, 1997-2003[69]

Looking at the subtotals for ultimate decisions, the position for trade mark owners looks good: the trade mark owner enjoyed success in 32 of the 52 cases recorded (62 per cent)[70] and lost in 20 (38 per cent). As noted above, studies in the US have showed success rates of 67 per cent over time, making these results comparable.[71]

However, separating the counterfeit and non-counterfeit cases paints a very different picture. Looking at the ultimate decisions in counterfeit cases, we see that trade mark owners succeeding in showing infringement in every case.[72] This might seem an extraordinarily high success rate – but it is consistent with the fact that the activity of counterfeiters falls directly within the ‘core rights’ of the trade mark owner. Even more striking is the success rate in non-counterfeit proceedings, where courts found infringement in only 9 cases (30 per cent) – trade mark owners failed in 20 proceedings (67 per cent). A similarly low success rate may be seen for trade mark owners on appeal, where they won in only 2 of the 10 appeals. These results are surprising, and beg the obvious question – why? Importantly for policy-makers – does this indicate a problem with the law? We return to discussion of these points in Part 4.2.

Grounds for non-infringement

Table 3 illustrated that trade mark owners were unsuccessful in a high proportion of non-counterfeit proceedings: the next issue of interest, therefore, is the grounds on which infringement was held not to be established. Looking at this issue may help us further understand the low win rate in contentious cases. The grounds for non-infringement are presented in Table 4.

Table 4: Grounds for Non-Infringement of Trade Marks

What we see in this data is that two grounds frequently arose in original proceedings: first, whether the infringing sign was ‘deceptively similar’, and second, whether the alleged infringer’s sign was being ‘used as a trade mark’. Notably, these grounds mirror the most common grounds raised on appeal: the most frequent infringement issues raised (either successfully, or unsuccessfully) on appeal was whether the infringing sign was ‘deceptively similar’ to the registered trade mark (6 instances),[76] followed closely by the question of whether the infringing sign was used ‘as a trade mark’ (4 instances).[77] One reason why these two grounds dominate is that most other grounds are tailored to very specific circumstances[78] – the issue of deceptive similarity will usually be one which parties in a non-counterfeiting case can contest. Another reason, however, could be that there is some uncertainty about these grounds.

Outcomes on Validity

The other matter which may arise in trade mark proceedings is the validity of the trade mark registration. Data on such findings is presented in Table 5.

Table 5: Trade Mark Validity Determinations, 1997-2003[79]

The most important observation arising from Table 5 is that in Australia, the validity of the registered trade mark is often not litigated in the context of an infringement action. Of the 67 original and appeal trade mark proceedings, only 20 involved a determination on trade mark validity. This contrasts with patent litigation, where validity is challenged more often[80]han not.80 Even where a party chose to challenge validity, they were successful in only 6 of the ultimate decisions (40 per cent), with the trade mark owner successful in 9 (60 per cent). This is not surprising: the threshold for protection of a trade mark is lower than that f[81] patent:81 it is more likely that the alleged infringer will be able to distinguish themselves from the existing mark, than it is they will successfully challenge validity.

Discussion

At this point, a summary of the general statistical picture is apposite. While we have confirmed that Australian trade mark owners succeed in a similar proportion of cases overall as in the United States, the most striking theme that emerges is that in the period studied, there are two distinct ‘worlds’ of trade mark enforcement. On the one hand, we have the counterfeit proceedings: legally simple proceedings where the respondent frequently does not show up or is unrepresented, which are relatively quickly disposed of, and in which the trade mark owner is always successful in obtaining a judgment in their favour. Fully 40 per cent of the original trade mark proceedings which are pursued through to judgment are of this kind. On the other hand, we have the non-counterfeit proceedings: proceedings which in the overwhelming majority of cases relate to the most straightforward form of infringement (use of a mark on the registered goods under s 120(1)), but where trade mark owners succeeded in only just over a third of the cases brought. From an observer’s perspective, both of these ‘worlds’ raise questions of interest. In relation to each area, the question for law-makers is whether the results suggest a need for any particular legal reforms.

Counterfeit proceedings

In relation to the counterfeit proceedings, the obvious question is why so many are proceeding all the way to judgment, when the end result should be completely predictable. By definition, they involve identical (or substantially identical) marks, being used on the registered goods: conduct which lies at the core of that which a trade mark owner is entitled to prevent. According to existing theories of dispute resolution, described above, where all else is equal, clear-cut cases should settle long before they reach trial or judgment, in order to avoid the extra, unnecessary cost of court proceedings. Yet four in 10 of the original proceedings which resulted in a judgment on the merits in our study involved trade mark counterfeiting.[82] Obviously we cannot measure, simply on the basis of these figures (a) what proportion of all potential counterfeiting disputes make their way to the court, or (a) what proportion of the proceedings concerning trade mark counterfeiting which are issued proceed all the way to judgment, as opposed to settling at some earlier stage. There is a high settlement rate in trade mark proceedings: as noted above, in infringement cases, of the 391 trade mark infringement proceedings filed in the Federal Court Australia in the five year period 1998 – 2002, only 48 (approximately 12 per cent) ended with some form of judgment on the merits. It is not possible to measure what proportion of these are counterfeit proceedings.[83] But in a sense, that is not the point. In theory, all these cases should be able to be resolved.

A first reason why counterfeit cases proceed to judgment is suggested by Figure 3 and Figure 4, which showed the data on legal representation in counterfeiting cases. It will be recalled that in over eighty per cent of the counterfeit cases, the respondent was (or respondents were) not represented by lawyers, and in over half the counterfeit proceedings there was no appearance by the respondent(s) in court at all. Theories of litigation which predict that straightforward cases settle assume a rational plaintiff and rational respondent. Such models do not take account of the uncooperative respondent who is, from an economic perspective, acting irrationally in failing to seek a sensible resolution of the dispute. In circumstances where an alleged infringer fails to cooperate, the trade mark owner has little choice but to either abandon the proceedings, or pursue the proceedings through to judgment as rapidly as possible.

A second, related reason why counterfeit cases go to court is that the stakes in these cases are not equal. From a trade mark owner’s perspective, enforcement proceedings are not just about stopping the individual counterfeiter: they are also about obtaining a significant judgment, which can be used to deter other prospective counterfeiters, and to send a message to the ‘industry’. Our anti-counterfeiting plaintiffs are repeat litigators who have concerns going beyond the individual case against the individual respondent: six trade mark owners account for all the proceedings. In such circumstances, where the goals of the trade mark owner go beyond mere recovery of damages or mere injunction from the individual respondent, settlement is far less likely. The need to deter infringement raises the stakes and the potential benefits of pursuing litigation for the trade mark owner, decreasing their willingness to settle. In addition, being too willing to settle (or willing to settle for a lower amount) may make the trade mark owner appear ‘soft’, and decreases the perceived risks of infringement for all other infringers in the market.[84] Thus indirectly, readiness to settle may encourage counterfeiting by others.[85]

The deterrent aim is further evidenced by the kinds of orders sought in these counterfeiting proceedings. While additional damages for ‘flagrant’ infringement are not available under the Trade Marks Act,[86] courts do have quite wide discretion in the framing of orders such as delivery up,[87] and orders for costs – including the order for indemnity costs.[88] We examined the orders made in the counterfeit cases within the study period, where they were available (20 of the 24 proceedings[89]). In those proceedings, all 20 involved the usual remedy for trade mark infringement – an injunction – but notably, most cases involved some other, additional order:

• All 20 cases involved an order for the payment of costs by the respondent (2 of those for indemnity or full costs);
• 16 cases involved an order for delivery up of infringing copies and/or copying equipment;
• 14 involving payment of damages or account of profits. In most cases the figure was not available (an inquiry being ordered), but the figures, where available, were not insignificant: ranging between $200,000 and $653,819.[90]

What is notable is that in 12 of the 20 cases, the deterrent or punitive effect was enhanced by the ‘piling up’ of all four possible orders. Although it is debateable whether such orders are sufficient to offer deterrence – even apparently large sums such as these may not offer deterrence if the profits are high enough – such orders are strong evidence that the proceedings are serving a purpose beyond the recovery of damages from the individual counterfeiter.

A third reason why some counterfeit cases proceed to judgment is a more technical legal reason. In the discussion above, it was noted that sometimes, ‘quirks’ in the legal system will prevent settlement. Under the Trade Marks Act, Customs may seize infringing goods imported into Australia.[91] Interception at the border is a very important tool in anti-counterfeiting efforts.[92] Under the Trade Marks Act, trade mark owners may lodge a written ‘objection’ with Customs.[93] Such an objection is, in effect, a request to Customs to intercept the importation of goods which infringe the notified trade mark. Customs must then seize imported goods bearing the mark if satisfied that there are reasonable grounds for believing that the notified trade mark is infringed by the importation of the goods, notifying both the trade mark owner and the owner of the imported goods (the ‘designated owner’, in the parlance of the Act).[94] At this point, three different things may happen:

• The owner of the imported goods may consent to their forfeiture;[95]
• The trade mark owner may commence proceedings for infringement;[96] or
• No infringement proceedings may be commenced, in which case the goods must be returned to the designated owner.[97]

Where infringement proceedings are commenced, the objector must get an order from the court directing Customs not to release the goods to the designated owner.[98] In other words, to prevent the goods from being released to the designated owner, trade mark owners must obtain a judgment, unless the designated owner actually consents to the forfeiture of the goods to the Commonwealth.[99] In circumstances where respondents fail to cooperate, or fail to show up, obtaining such consent is likely to be impossible – with the result that proceeding to judgment becomes inevitable despite the straightforward nature of the case. In fact, in 8 of the 24 counterfeit proceedings which proceeded to judgment, there was an order for forfeiture of infringing copies held by Customs.[100]

These results beg the question: does the evidence in this study suggest that any legal reform is required? The management of counterfeiting matters is an important policy matter. There would seem to be good reasons why we should be concerned about providing an effective and ‘user-friendly’ enforcement regime against counterfeiting. Counterfeiting activity falls within the core of that conduct which trade mark owners are entitled to prevent. The negative effects of counterfeiting are substantial. They include (a) direct losses to the trade mark owner (ie lost sales); (b) the costs entailed by the deception of consumers, (c) indirect losses to the trade mark owner (for example, loss of reputation for quality, loss of ‘prestige’ value), (d) the costs of enforcement incurred by trade mark owners and governments alike, and (e) the ‘social losses’ – lost jobs, lost tax revenue, and lost investment in research and development that may arise as a result of the lost revenue. In some cases, there may be health and safety issues, although these issues do not arise for the kinds of goods subject to enforcement in the period of this study.[101]

Indeed, the sheer magnitude of the problem has prompted initiatives at the international level. At the 2005 G8 Summit, for example, one of the issues discussed was the need to eliminate the international trade in counterfeit goods.[102] Few of these discussions have highlighted the role of the courts: rather, the focus has mostly been on the practical mechanisms such as the stopping of importation by customs, to slow trade in such goods. However, enforcement proceedings through the courts are an integral part of the process for controlling such activities, and how such cases are being managed is hence a matter of interest for policy-makers.

It is not clear that we can significantly reduce the numbers of these cases requiring judicial resolution. Uncooperative respondents and the need to deter infringements stand in the way of removing these cases through settlement. Even in relation to the area of customs seizures, where the need for a court order is a legal quirk that ensures some cases must be pursued through to judgment, there would appear to be few alternatives. There is already a procedure to avoid court proceedings where the owner of goods consents to forfeiture. The ‘problem’ arises where the owner will not consent, and particularly, where they will not even respond to the trade mark owner. In theory, it might be possible to provide that, in cases where an alleged infringer fails to consent to, or even respond to a notice of objection and notification of seizure of goods, those goods should be forfeit without going through court procedures (rather than requiring active consent to forfeiture). This would presumably eliminate from the courts those cases where the respondent fails to appear. However, in the six year period considered here, such cases constitute only 5 judgments, so the benefits of the shift would appear to be minimal, while the costs (in terms of encroachments into property rights, and process issues) would be not insignificant.

The only alternative, then, would be resolution through another judicial forum. Since 2000, the Federal Magistrates’ Court (‘FMC’) has provided a court of lower jurisdiction in matters arising under Federal law. At the time that the court was established, it was envisaged that its advantages would include a less formal judicial culture, more streamlined procedures, and reduced costs for litigants.[103] It was intended that resolution of more simple matters in the FMC would provide more time and resources in superior courts to deal with more complex matters.[104] In recent years there has been increasing talk of having intellectual property matters heard by the FMC, in the “belief that this may provide a solution to realising a more efficient IP enforcement regime in Australia.”[105] In mid-2003 the FMC began to hear copyright infringement actions. In November 2003, the Advisory Council on Intellectual Property proposed, in a final report that the jurisdiction of the FMC should be extended to include patent, trade mark and designs matters.[106] In fact, the FMC already deals with cases of misleading and deceptive conduct, and can, under its accrued jurisdiction, deal with trade marks issues arising from a case in which such conduct is pleaded.[107] In December 2005, the Federal Government proposed a bill which would extend the jurisdiction of the FMC in matters under the Trade Practices Act (increasing the level of monetary awards that the FMC can provide, and otherwise extending the jurisdiction of the Court) and to matters transferred to it by the Federal or Family Court.[108] The amendment, while not effecting the full set of changes recommended by ACIP, would in practice, make it easier to bring trade mark matters before the FMC, as matters of accrued jurisdiction in cases brought either for copyright infringement,[109] or under the Trade Practices Act.[110]

The present study clearly demonstrates that there is a distinct set of cases which are not legally complex, which are quickly resolvable, and in which a full hearing of a superior court hardly seems warranted given the frequent failure of the respondent to even appear to be heard. This finding would seem to counter one of the main arguments which has been used by those opposing the expansion of the jurisdiction of the FMC to other IP disputes: namely, that IP matters are complex and require a certain amount of judicial expertise on behalf of judges. These factors indicate that counterfeit cases should be included in the jurisdiction of the FMS, which provides “a quicker, cheaper and more accessible forum for dealing with less complex ... general federal law disputes.”[111]

There are various ways that counterfeit cases could be directed towards the FMC:

1. Judges of the Federal Court could use the discretion vested in them under the Jurisdiction of the Federal Magistrates’ Court Legislation Amendment Bill 2005 to refer such matters down to the FMC;
2. Trade mark owners could be given the same choice given to copyright owners through the conferral of concurrent jurisdiction on both the Federal Court and Federal Magistrates’ Court;[112] or
3. An attempt could be made positively to require that counterfeiting cases, in the first instance, be brought in the FMC rather than the Federal Court of Australia.

If the Federal Magistrates’ Service were to offer a better (more efficient, quicker) forum, one would expect that the 2nd of these options would be sufficient. Nevertheless, from a party’s point of view, there may be considerable inertia. The Federal Court may be a more comfortable forum in which to litigate for repeat litigators. Further, despite the lower filing fees in the FMC, the evidence from the present study is that counterfeiting cases are dealt with relatively efficiently by the Federal Court of Australia: they take an average of just over 6 hours of court time in total to resolve (including any directions hearings), and are ordinarily resolved less than a year after filing of the initial proceedings in the court. The figures in relation to issued proceedings are even more striking: half the infringement proceedings filed in the period 1998-2002 were disposed of within 150 days (or approximately 5 months), with 10 per cent of proceedings terminated within just over a month (36 days). There is little reason to believe this can become more efficient in another court.

There may, nevertheless, be some justification for directing counterfeit cases to the FMC. While the Federal Court may be efficient from a trade mark owner’s perspective, there are not insignificant costs to society, and the court system more generally. In taking up the time of the Federal Court, such cases, even when short, arguably waste scarce judicial resources. The current burden of proceedings before the Federal Court is large. Removal of some ‘chunk’ of them would free up the time of the judges to deal with other, more complex matters. It would not be expected that the burden on the court would factor into the decisions of trade mark owners in choosing where to litigate, but from a societal perspective, it is relevant.

It would be possible, as a legal matter, to distinguish between counterfeit and non-counterfeit disputes for the purposes of determining the appropriate forum. Amendments could be made to the Trade Marks Act 1995 (Cth): for example, by introducing a separate provision for infringement involving identical goods, and a substantially identical mark. Jurisdiction under this provision could be granted to the FMC. While admittedly the effect of such a provision would be easy to avoid by a determined trade mark owner (by pleading infringement under s 120(1)), such a distinction would at least set up the expectation that counterfeit cases should not be tried in the jurisdiction.

For the most part, these are matters of broader policy; on the basis of this study it is not possible to draw a concluded view on whether trade mark counterfeit cases should be directed, in the first instance to the FMC. What is clear is that we have a subset of uncontroversial cases, where arguments about the need for expertise on the part of judges do not apply.

Non-counterfeit proceedings: why litigate a losing case?

The other surprising key finding of interest is the low success rate in non-counterfeit trade mark proceedings. As noted above, trade mark owners succeeded, in these cases, only just over a third of the time. These results are surprising, and beg the obvious question – why? Importantly for policy-makers – does this indicate a problem with the law? The answer is: perhaps, but not necessarily. We must first take into account which cases proceed all the way to judgment. Can the selectivity of the cases litigated explain the low success rate? That will depend on which cases persist so long.

Earlier in this paper we noted that the dominant ‘divergent expectations’ model of litigation assumes that (a) the law is relatively clear, and (b) that the parties are equally likely to err in predicting the result. If the law is unclear, and one side is systematically too optimistic about their chances, they will, systematically, take cases to trial (and refuse settlement) where they are objectively likely to lose. Or, they may not be too optimistic, but may consciously choose to take cases to trial despite a likely loss, for some strategic reason external to the case itself. Another assumption of the model is that the parties have equal stakes in the outcome of the trial.[113] Where stakes are not equal, the results will be biased. A party with more to lose than gain in litigation will tend to settle even in those cases where they have a chance of success: thus leading overall to a higher ‘win rate’ for those cases which do go to trial. Conversely, a party with more to gain than they might lose will be less inclined to settle even in borderline cases, leading to a lower overall win rate for that party.

How do these factors play against non-counterfeit trade mark litigation? Focusing first on uncertainty regarding the legal standard, there are two possible sources of uncertainty in current Australian trade mark law. One might be changes wrought by the 1995 Act: in particular, the new actions in the new 1995 Act – subs 120(2) and (3), which extended the scope of trade mark infringement.[114] However, the data does not support this explanation. While the model predicts increased litigation with uncertainty, in fact there have been few cases brought under these provisions.[115] Moreover, within the ultimate non-counterfeit cases the discrepancy between the wins and the losses is confined to cases where infringement is claimed under s 120(1) of the 1995 Act,[116] which substantially mirrors the previous 1955 Act. This suggests that the introduction of the new legislation is not the cause of the low infringement rates in these cases.[117]

The other possible source of uncertainty is the requirement that an infringing mark be ‘deceptively similar’. There would appear to be at least some uncertainty about this standard – it is the most frequent ground on which a trade mark owner loses on infringement, as Table 4 establishes. To some extent, uncertainty regarding this requirement is inevitable, as the question of whether one mark is deceptively similar to another is a matter of impression – and something on which reasonable minds m[118]differ.118 There is also, however, some evidence that there is some uncertainty in the legal standards – particularly in what matters are to be taken into account in assessing whether a prospective customers would be[119]nfused.119 Consider, for example, the issue of whether the reputation of a mark is relevant to the issue of deceptive s[120]larity.120 On this issue, we have had three Full Federal Court decisions, all taking slightly different [121]itions.121 Early commentary following the 1995 Act also indicated that there was some difficulty regarding the interpretation, and that the court and Trade Marks Office were not ‘at one’ on the appli[122]le law.122 The fundamental limitation with proving this theory is that it is impossible to say whether the litigation rate is high in our data and to determine exactly what proportion of all trade mark disputes actually proceed to l[123]gation.123

Another explanation for the low success rate of trade mark owners is that those owners are pursuing cases which, objectively, they are likely to lose.[124] The interesting question is why this would happen. One option is that trade mark owners are simply optimistic about their chances of success – they genuinely believe they are going to win. This is of course possible – the sense of entitlement that comes with owning a registered trade mark may lead trade mark owners to see infringement where the law says there is none. To put it another way, trade mark owners may expect the protection of the law beyond the boundaries established by the Act. On the other hand, we would expect that even if trade mark owners are overly optimistic, perhaps through a sense of entitlement, they would be disabused by their more clear-eyed lawyers.

Alternatively, trade mark owners are knowingly pursuing cases they are likely to lose. This might happen if trade mark owners have more to win from litigation than they risk losing. Unusually for IP litigation, a trade mark owner does not face a high risk of losing their trade mark: challenges to trade mark validity are relatively rare (as compared, for example, to patent litigation), and the standards for validity of a registration are low. Loss in a trade mark case (and unlike other forms of IP litigation) is not likely to lead to loss of the trade mark. On the other hand, the benefits of litigating even a questionable case are clear. By litigating, the trade mark owner may acquire a reputation as an aggressive litigator and may mark out the boundaries of the trade mark rights, creating a ‘zone of distinctiveness’ where competitors may not tread. In other words, the benefits of litigating and hence gaining broader protection may be worth the risk. In these circumstances a trade mark owner may be disinclined to settle even cases where they are correctly advised they may lose. [125]

By contrast, the best that a defendant can hope to gain in litigation may be the right to use a particular trade mark. Given the large range of possible marks, this may not be such a huge ‘stake’ unless there has already been significant investment in a mark (unlike patent cases, for example, the loser is unlikely to be shut out of a market altogether). Weighed against this benefit, the risks of litigating may be significant – in terms of their own and the trade mark owner’s costs. This, according to the models, will incline the defendant to settle even in borderline cases. The result of these differential stakes and incentives to settle could be that a greater number of disputes proceed to trial where the defendant is likely to win.

These two explanations have different policy implications. If it is the case that trade mark owners are overly optimistic, there is a mismatch between the protection that trade mark owners expect from the law, and that which the law actually provides. This might indicate the rights provided are unduly narrow – but not necessarily. The expectations of protection held by trade mark owners could be inconsistent with good policy outcomes.[126] Even if the rights are not narrow, however, if trade mark owners are overly optimistic then the scope of trade mark protection is not being properly explained – unreasonably high expectations are being generated about what protection the system will in fact give. This would suggest that, at the least, steps should be taken to correct the impression.

If, on the other hand, the main reason for the low win rate is that trade mark owners are knowingly pursuing litigation where there is a risk of loss, on the basis that the risk is worthwhile, then no policy action is suggested. A further, focused study (for example, using interview techniques) of the thinking of litigating parties in this space would be of interest.

As a final matter, it is worth noting that despite the concerns expressed by the Working Party formed originally to review the Trade Marks Act, the infrequency with which s 120(2) and (3) are litigated would seem to indicate that they have not introduced undue uncertainty into the Act. As to why the numbers are low, it may be that the low numbers of litigated disputes reflects a low number of ‘real world’ disputes. Perhaps trade mark owners are either only concerned with their ‘core’ (registered) markets, or they have trouble detecting potential infringements outside their core markets. Alternatively, it may be that the kinds of issues that might lead to disputes under subs 120(2) and (3) are being detected at the opposition stage. If trade mark owners are heading potential infringers off at this stage, these issues will never, or rarely, make it to the stage of infringement proceedings.[127] It would be of interest to know how often opposition proceedings are being launched by trade mark owners in relation to similar goods (under s 44), or even dissimilar, goods (under s 60). Finally, it could be argued that low numbers of cases under subss 120(2) and 120(3) are in fact desirable – that we want significant barriers before trade mark owners can extend their trade mark rights beyond ‘traditional’ trade mark infringement, and beyond the rights that trade mark owners have actually registered and ‘paid for’. In summary, then, while the numbers are low, this is not necessarily a cause for concern.

Conclusions

This study has highlighted some important features of trade mark enforcement litigation in Australia. First and foremost, it has highlighted that there are two fundamentally different kinds of trade mark litigation, and suggested that these might be treated differently – even to the extent of assigning jurisdiction over such cases separately. There is little reason to believe that the legally simple cases, counterfeit proceedings, with their unresponsive respondents, should be taking up the time of the Federal Court of Australia. Second, it has highlighted a even more interesting result: that trade mark owners appear to be pursuing cases that objectively they are likely to lose. While we cannot offer a full explanation for this, we have suggested that, before we become too concerned, further research is required to determine whether this result is a matter of strategy, or ignorance on the part of trade mark owners. Finally, it is worth noting that only this kind of ‘bigger picture’ quantitative study could even raise these matters for discussion. We hope that, through future studies of this kind, a better overall picture of the enforcement system can be obtained in trade mark, as it increasingly is in patent.

[°] This paper was originally published as follows: Jason Bosland, Kimberlee Weatherall and Paul Jensen, ‘Trade Mark and Counterfeit Litigation in Australia’ (2006) 4 Intellectual Property Quarterly 347. Please refer to the original for citation purposes. The authors wish to thank Glenys Fraser, Paul Bernath, Nitsa Karahalios, Sarah Moritz, Sally Pryor and Edward Sexton for all their contributions to this project.

[*] B.A. LL.B. (Hons) (Melb). Researcher, Intellectual Property Research Institute of Australia, University of Melbourne.

[†] B.A. LL.B. (Hons) (Syd) B.C.L. (Oxon.) LL.M. (Yale). Associate Director (Law), Intellectual Property Research Institute of Australia, and Senior Lecturer, Faculty of Law, University of Melbourne.

[‡] B.Ec. (Syd) Ph.D. (AGSM). Senior Research Fellow, Melbourne Institute of Applied Economic and Social Research, and Intellectual Property Research Institute of Australia, University of Melbourne.

[1] OECD, The Economic Impact of Counterfeiting (1998, Paris, OECD) at 4. While this study is often cited for the figure, it should be noted that the OECD specifically stated that ‘[t]here is no substantial aggregated data to support the high percentages, but the figures are now accepted and used to illustrate the extent of the counterfeiting problem’: at 23.

[2] In Australia, see the Hon. Philip Ruddock (Attorney-General), ‘Opening Address – Copyright Law and Practice Symposium’, 17 November 2005 (discussing copyright piracy in particular; as noted below, trade mark counterfeiting and copyright piracy are often closely linked: see below n17). Internationally, see work within the OECD, which recently launched a new 3-phase project to examine counterfeiting and piracy: see generally <http://www.oecd.org/sti/counterfeiting> .
[3] There are, however, some exceptions: Bhagat, S. and Umesh, U.N. “Do trademark infringement lawsuits affect brand value: A stock market perspective”, Journal of Market Focused Management (1997), 2 127-148; and Pfister, E. “The legal enforcement of intellectual property rights: An empirical study on French trademarks”, mimeo, University of Nancy II; William Landes, “An Empirical Analysis of Intellectual Property Litigation: Some Preliminary Results” (2004) 41 Houston Law Review 749. By contrast, there is a burgeoning empirical literature concerning patent litigation and enforcement: summarised in Weatherall, K. and Jensen, P.H. “An Empirical Investigation into Patent Enforcement in Australian Courts” (2005) 32 Federal Law Review 239.

[4] In the United States, summary statistical databases of court filings and outcomes are available for analysis by researchers. Similar levels of access are not presently available in Australia.

[5] For example, an earlier paper by two of the authors sought to examine a view, sometimes expressed in Australian circles, that Australian courts were ‘anti-patent’: Weatherall and Jensen, above n3.

[6] Compare this to patent law, where there has been a much greater degree of public debate: Weatherall and Jensen, above n4. ACIP had little to say about the actual processes or results in the courts: Advisory Council on Intellectual Property, Review of Trade Mark Enforcement (April 2004), available at <http://www.acip.gov.au/library/reviewtmenforce.pdf> [7] See below, Table 3, at 30. A US Bureau of Justice Statistics Special Report produced in October 2004, Intellectual Property Theft, 2002 (NCJ 205800) found that 2002 in matters which went to trial, the plaintiff in civil trade mark complaints was successful 61.5 per cent of the time. Similarly, a recent study produced by William Landes found that between 1978 and 2000, the plaintiff win rate at trial in trademark cases was 67 per cent (and extremely steady at that figure): Landes, above n3 at 771.

[8] George L Priest and Benjamin Klein, ‘The Selection of Disputes For Litigation’ (1984) 13 Journal of Legal Studies 1.

[9] Coca-Cola Co v All-Fect Distributors Ltd [1999] FCA 1721; (1999) 96 FCR 107; 47 IPR 481 at [44]. A fourth form of infringement exists: importation of infringing items under Part 13 of the Act. However, Part 13 does not create an independent action for infringement; rather, it provides a method for enforcement.

[10] Section 120(1) corresponds to the Trade Marks Act 1994, subss 10(1) and 10(2)(a). Section 120(2) corresponds to subs 10(2)(b). s 120(3) provides additional protection to s 10(3) of the UK Act, although it is arguable that the provision in the UK is broader (being closer to a trade mark dilution action than the Australian provision).
[11] Mark Davison, Kate Johnston and Patricia Kennedy, Shanahan’s Australian Law of Trade Marks and Passing Off (3rd ed 2003) at [18.25]. Some members of the Working Party opposed this extension; however, it was required to meet the obligations in Art 16(1) of the TRIPS Agreement. Australia was not alone in expanding liability at around that time: the United Kingdom, in its Trade Marks Act 1994 (UK) also expanded liability beyond its traditional scope: a shift required in that case by European law: see William Cornish, Intellectual Property: Patents, Copyright, Trade Marks and Allied Rights (4th ed 1999) at [17.90].

[12] More precisely, subs 120(2) states that infringement occurs where the alleged infringer is using the mark where (i) if the registered mark is for goods: in relation to goods of the same description or closely related services (s 120(2)(a) and (b)), or (ii) if the registered mark is for services: in relation to services of the same description or closely related goods (s 120(2)(c) and (d)).

[13] While there is no distinction in s 120, the Act does draw a distinction elsewhere, in that counterfeiting (falsely applying the registered mark (not a merely similar mark) to goods in the course of trade) is a criminal offence under s 146.
[14] Although they are not necessarily direct market substitutes, because a consumer interested in buying a genuine Rolex watch would generally not be persuaded into buying a counterfeit Rolex for a fraction of the price. Rather, counterfeit goods in many instances service an entirely different market to that of the trade mark owner: see Kozinski, ‘Trademarks Unplugged’ (1993) 68 N.Y.U. L. Rev. 960, 964; Gene M. Grossman and Carl Shapiro, ‘Foreign Counterfeiting of Status Goods’, NBER Working Paper No 1915 (1988) (distinguishing between ‘deceptive’ and ‘non-deceptive’ counterfeiting).

[15] TRIPS, footnote 14, Article 51. Other definitions are available: eg the OECD, which defines counterfeiting as ‘manufacturing a product which so closely imitates the appearance of the product of another to mislead a consumer that it is the product of another’: OECD, The Economic Impact of Counterfeiting (1998 Paris, OECD) at 3. See also 15 U.S.C. §1127 (US definition of counterfeit). We prefer the TRIPs definition which does not require a concept of confusion (see above n14).

[16] In the United Kingdom, counterfeit cases would fall under s 10(1) of the Trade Marks Act 1994.

[17] Many counterfeit cases involve both trade mark and copyright infringement issues: OECD, The Economic Impact of Counterfeiting (1998) at 5. Trade mark infringement proceedings may be preferred as being more factually straightforward to prove, owing to the fact that subsistence of the rights is in effect already proved through the existence of the mark on the register: Andreas Rahmatian, ‘Trade Mark Infringement as a Criminal Offence’ (2004) 67 Modern Law Review 670, 672. However, in Australia additional damages for flagrant infringement (ie, damages in the nature of a punishment rather than compensation for loss) are available under the Copyright Act, but not the Trade Marks Act. Compare this to the United States, where treble damages are available where a defendant intentionally uses a mark, knowing it is counterfeit (as defined in 15 USC §1116(d)), in connection with the sale, offering for sale, or distribution of goods or services: 15 U.S.C. §1117(b).
[18] State and Territory Courts of Appeals are not included because they do not have appellate jurisdiction in trade mark matters: see Trade Marks Act 1995 (Cth), s195(1). Trade mark disputes are not currently within the jurisdiction of the Federal Magistrates Service; see further below Part 4.1.

[19] In certain circumstances, an “aggrieved person” can apply to a “prescribed court” to have a trade mark registration cancelled or amended: Pt 8, Div 2 and Pt 9, Trade Marks Act 1995 (Cth).

[20] Trade Marks Act 1995 (Cth) ss 129-130.

[21] Trade Marks Act 1995 (Cth) ss 87, 88, 92, 94.

[22] Trade Marks Act 1995 (Cth) s 126.

[23] Some of these were actions for removal of a trade mark with a cross-claim for infringement. This figure represents the proceedings, issued in the Federal Court, which can be identified as infringement proceedings using information in the Federal Court database (FEDCAMS) recording the section of the Act under which proceedings were filed. In total, 531 proceedings were issued, which break down as follows: 391 proceedings for infringement or removal and infringement, including proceedings relating to customs seizures, 46 opposition proceedings, 8 actions for removal for non-use, 12 miscellaneous, including appeals from the Registrar of Trade Marks, 6 actions for unfounded threats, 22 for removal, including where that is coupled with a claim for unfounded threats, and 46 which could not be categorised as the section of the act was not recorded. This list may not be complete: it depends upon the Federal Court registry staff having recorded the Trade Marks Act as the first Act under which an action was brought. In particular, it may not include copyright cases involving a trade mark claim.

[24] Including those cases where there was an interlocutory injunction judgment, since these cases will generally involve the court offering some assessment of the apparent merits of the case.

[25] Also known as the ‘optimism model’: Priest and Klein, above n3. The ‘asymmetric information’ model is another, although we have not used it to analyse the data is this paper. This is because we think it is for the most part inapplicable to trade mark disputes. See also William M Landes, ‘An Economic Analysis of the Court’, (1971) 14 Journal of Law & Economics 61; Richard A Posner, ‘An Economic Approach to Legal Procedure and Judicial Administration’ (1973) 2 Journal of Legal Studies 399; John Gould, ‘The Economics of Legal Conflicts,’ (1973) 2 Journal of Legal Studies 279-300; Steven Shavell, ‘Suit, Settlement and Trial: A Theoretical Analysis under Alternative Methods for the Allocation of Legal Costs’ (1982) 11 Journal of Legal Studies 55-81. See discussion by Jean O Lanjouw & Mark Schankerman ‘Enforcement of Patent Rights in the United States’ (2003) chapter to be published in Cohen, Wesley & Steven Merrill (eds) Patents in the Knowledge-Based Economy (National Academy Press: Washington DC) forthcoming publication, 149-150, and 163.

[26] In forming such a rational estimate litigants will take into consideration the legal standard governing the dispute, any perceived institutional bias and the possibility of error caused by incomplete information.

[27] This is true even where the court is ‘biased’ against one side or another. Imagine that the standards applied by the courts favour of IP owners, so that in the whole population of possible disputes out in the real world, IP owners would win 90 per cent of the time. Parties, being aware of this, will only litigate where the case is a close one taking into account the courts’ bias.

[28] This assumes rational parties, and that plaintiffs and defendants, in close cases, are equally likely to ‘err’ in predicting the outcome: see Priest and Klein, above n8.

[29] See generally Weatherall and Jensen, above n3, 247-248.
[30] Landes, above n3, at 772.

[31] We have used a variety of methods, detailed below, to ensure we captured every decision rendered by the relevant courts. We acknowledge, however, that it is possible that we have missed some decisions.

[32] Including decisions where an application has been made for summary judgment, summary dismissal, an expedited judgement, and the restraint of unjustified threat of infringement.

[33] We searched three different electronic databases in order to generate a list of relevant decisions. We adopted this approach in preference to the other possible approach – looking only at cases reported in a set of law reports like the Intellectual Property Reports - because the editors of law reports select their cases on the grounds of their importance as precedent: a matter not relevant to our aims. For more detail in relation to the methodology of this study, see Weatherall and Jensen, above n 3.

[34] More specifically we have: the name, citation, court and court file number; the type of case (interlocutory, appellate, or original); important dates (issue date, hearing date, decision date) and court time (hours in court in the proceedings). Some of this information (in particular, issue dates and court hours) is not recorded in the written decision of the Court. We obtained this information directly from the Federal Court case management database (FEDCAMS), and from the Registries of the other courts. FEDCAMS through 2004 has been replaced by a new system, CaseTrack which records most, but not all, the information we have used here. The identity of the judge and the legal counsel is not used in any of the analyses reported here, but may be the subject of a future paper.

[35] Both the US Bureau of Justice Statistics Special Report, above n7 and the Landes study, above n3 present analyses based on the Federal Courts database in the United States, which records the subject matter of the case (eg, trade mark) and the outcome (eg, plaintiff win). That database does not, however, distinguish between types of trade mark cases, nor does it identify the issue or grounds on which a plaintiff may have failed in their action.

[36] Trade Marks Act 1995 (Cth), Part 5.

[37] Trade mark oppositions are generating increased interest worldwide: the WIPO Standing Committee on the Law of Trademarks, Industrial Designs and Geographical Indications announced in December 2005 that opposition procedures one of issues forming the basis for its future work.

[38] While s 120(2) and (3) are mutually exclusive, it is normal to plead both in case a court finds goods are not ‘related’. Cases where more than one kind of infringement was pleaded have been characterised according to either (a) a positive determination of the court (i.e. ‘this is not a s 120(3) case’), or (b) the bulk of the reasoning/discussion in the decision of the court. This decision was undertaken in order to adopt a single unit of analysis (the proceeding) throughout as much of the paper as possible.