What happened
The proceeding concerned a long-running cartel between Yazaki Corporation and Sumitomo Electric Industries Ltd (SEI) concerning the supply of automotive wire harnesses to Toyota Motor Corporation (TMC) and its subsidiaries worldwide, including in Australia. Wire harnesses distribute electrical power and signals within motor vehicles; the engine-room main harness was particularly important. From at least the mid-1990s Yazaki and SEI operated under an Overarching Cartel Agreement. When TMC issued a request for quotation (RFQ), the two parents would meet, agree on allocation of supply by country or region so as to preserve incumbency, and set prices that would ensure the manufacturer awarded contracts in accordance with that allocation. The arrangement was kept secret from TMC and from Yazaki's Australian subsidiary, Australian Arrow Pty Ltd (AAPL).
In 2003 TMC sought bids for the 2006 Camry in five locations including Australia. Yazaki and SEI reached a 2003 Agreement that, in Australia, SEI's subsidiary would "win" by approximately 1% including duty, while in other regions the incumbent would be protected by agreed price differentials. Prices were discussed and agreed in Japan, submitted to TMC in Japan, and AAPL was directed to, and did, submit the same prices to Toyota Motor Corporation Australia (TMCA). A parallel minor RFQ understanding was reached between AAPL and SEI's Australian subsidiary. In 2008 a similar global RFQ for the 2011 Camry led to a 2008 Agreement with comparable price and allocation provisions. Again, agreed prices were submitted in Japan and AAPL was directed to submit them to TMCA.
The Australian Competition and Consumer Commission (ACCC) brought proceedings alleging contraventions of ss 45(2)(a)(i), (ii) and 45(2)(b)(i), (ii) of the Competition and Consumer Act 2010 (Cth) and the Victorian Competition Code. No relief was sought for the making of the Overarching Cartel Agreement itself, but relief was sought for giving effect to it by making the 2003 and 2008 Agreements and for the subsequent implementing conduct. The primary judge found that AAPL had made and given effect to the 2002 minor RFQ agreement, that Yazaki had made and given effect to the 2003 and 2008 Agreements and had given effect to the Overarching Cartel Agreement by making those later agreements, and that the provisions were both exclusionary provisions and provisions that controlled or maintained prices. However, his Honour held there was no relevant market in Australia because competitive activity occurred in Japan, so no contravention of the SLC limbs occurred. He also held that AAPL could not give effect to the 2003 and 2008 Agreements without knowledge of them. Penalties of $9.5 million were imposed on Yazaki on the basis of a $10 million statutory maximum per course of conduct, treating the conduct as two broad courses (making agreements and submitting prices).
On appeal the Full Court (Allsop CJ, Middleton and Robertson JJ) allowed the ACCC's appeal in part. It held that knowledge is not an essential element of "give effect to", that AAPL had in fact given effect to the 2003 and 2008 Agreements by submitting the directed prices, that no Australian market is required for liability for exclusionary provisions, that a market in Australia did exist on the evidence of local demand, supply contracts, negotiations and perceived rivalry, that the annual-turnover calculation under s 76(5) included AAPL's relevant supplies, producing a maximum of approximately $87.4 million per contravention, and that the five identified acts were not reducible to one or two courses of conduct. The Court re-fixed penalties at a total of $46 million, made additional declarations against AAPL, dismissed the cross-appeal and notice of contention, and ordered the respondents to pay the ACCC's costs of the appeal, cross-appeal and contention. The reasons occupy 269 paragraphs and reflect a detailed engagement with the primary judge's comprehensive findings after a complex trial.
Why the court decided this way
The Full Court began with orthodox principles of statutory construction, emphasising that text, context and purpose must be considered together and that the starting point is the statutory text read in its widest context. For the "give effect to" issue the Court examined the non-exhaustive definition in s 4: "do an act or thing in pursuance of or in accordance with or enforce or purport to enforce". At [71] the Court held that the ordinary meaning focuses on implementation rather than the actor's subjective state of mind. Knowledge may be probative evidence that conduct in fact implements the arrangement, but it is not a legal prerequisite. The Court rejected analogy with accessorial liability under s 75B, noting that s 75B imports criminal-law concepts whereas s 45(2)(b) does not. It also dismissed the policy concern about unwitting couriers or typists, stating that liability is determined case-by-case on the facts. Here, AAPL's submission of prices was at Yazaki's specific direction and was "sufficiently related to and connected with the cartel agreements" ([76]), satisfying the statutory standard even without AAPL's knowledge. Additional declarations 6A and 6B were therefore made.
On the exclusionary-provision issue the Court rejected Yazaki's argument that s 45(3) and the market definition in s 4E must be read into s 4D via the word "competitive". The Court distinguished Kelly v The Queen, noting that s 4D enacts substantive law rather than merely shortening language. It held that s 45(3) is expressly limited to provisions that have the purpose or effect of substantially lessening competition. Exclusionary provisions under s 4D(1)(b) are defined by purpose of preventing, restricting or limiting supply to particular persons; the competitive relationship required by s 4D(1)(a) and (2) is not conditioned on competition "in a market in Australia". This construction accorded with the Swanson Committee's recommendation that collective boycotts be assessed by reference to competitive effects between the parties rather than by reference to a market, a recommendation accepted by Parliament in 1977. The Court cited earlier Full Court authority (Pont Data, News Ltd) to the same effect and declined to follow decisions where the point had been assumed rather than argued. The cross-appeal was dismissed.
Although not strictly necessary once the cross-appeal was dismissed, the Court nevertheless held that a market in Australia did exist. Drawing on Air New Zealand, Queensland Wire and Flight Centre, the Court emphasised that market definition is a purposive, instrumental exercise reflecting commercial reality and potential close rivalry. The primary judge had placed decisive weight on the fact that TMC's decisions were made in Japan. The Full Court held that this gave "too much weight" to the place of substitution decisions. Demand originated with TMCA in Australia, contracts were made and payment occurred in Australia, local managers perceived each other as competitors, Australia was treated as a distinct geographic segment in the global RFQ, and there was evidence of actual competitive tension (for example, quality issues and price-down requests). Even during a production cycle there was "some chance" a supplier could be replaced in exceptional circumstances. These facts established an area of potential close competition in Australia for the purposes of examining the conduct prohibited by s 45(2).
On penalty quantum the Court first corrected the maximum. Section 76(5) defines annual turnover as the sum of supplies made by the contravening body corporate "and any body corporate related to" it, subject to exclusions. The phrase "the body corporate" in paragraph (d) applies distributively; supplies by AAPL in connection with the enterprise it carried on were therefore included. The resulting maximum was $87,411,359.30 per contravention. The Court then addressed the course-of-conduct and totality principles. It accepted that s 76(3) prevented multiple penalties for the same conduct that contravenes two or more provisions; thus only one penalty was payable for the overlapping contraventions in declarations 5 and 7. However, the five acts identified by the primary judge were not reducible to one or two courses. Making the agreement, coordinating prices over weeks in May 2008, submitting to TMC on 29 May, directing AAPL, and AAPL's submission on 25 June were discrete in time and quality. Acts 4 and 5 could be treated as broadly one course, but acts 1-3 could not. The statutory structure expressly distinguishes making from giving effect; the Court warned against applying the course-of-conduct principle too liberally so as to undermine that distinction. Having regard to the deliberate, sophisticated, devious nature of the conduct, its longevity, senior management involvement, lack of contrition or cooperation, the substantial financial transactions affected, and the need for specific and general deterrence, the Court fixed individual penalties of $14 million, $12 million, $12 million, $4 million and $4 million, producing a total of $46 million after totality adjustment. That total represented a substantial increase on the primary judge's $9.5 million but remained well below the aggregate statutory maximum.
Costs were not disturbed. Although the ACCC succeeded on the knowledge point, it had pleaded and litigated AAPL's actual knowledge as a factual question and lost. That factual contest had occupied significant hearing time; the primary judge's 15% reduction remained fair.
Before and after state of the law
Prior to this judgment the meaning of "give effect to" in s 45(2)(b) had not been authoritatively settled at appellate level in a case where a subsidiary acted at a parent's direction without knowledge of the underlying cartel. Lower-court decisions had assumed knowledge was required, often by reference to pleading practice or analogy with accessorial liability. The Full Court clarified that the phrase is satisfied whenever conduct in fact implements the arrangement, rejecting any blanket knowledge requirement. The judgment aligns the civil-penalty provision with its enforcement purpose while preserving the ability of courts to distinguish, on the facts, between a directed subsidiary and an entirely unwitting third party such as a courier.
On market definition for exclusionary provisions the law was unsettled. Some first-instance decisions had assumed a market in Australia was necessary. The Full Court resolved the question by reference to the statutory text, the 1977 legislative history, and earlier Full Court dicta in Pont Data and News Ltd. After Yazaki it is clear that liability for making or giving effect to an exclusionary provision turns on the competitive relationship between the parties (s 4D(1)(a) and (2)) and the proscribed purpose (s 4D(1)(b)), not on the existence of a market in Australia. By contrast, the SLC limbs continue to require proof of a market under s 45(3) and s 4E. The judgment therefore narrows the extraterritorial defence available to foreign cartels that affect Australian supply through local subsidiaries.
Market definition itself was restated in light of the High Court's recent decision in Air New Zealand. The primary judge's emphasis on the Japanese locus of decision-making was held to have given "too much weight" to the place of substitution. The Full Court reinforced that the market is a tool for analysing the competitive process in its commercial context; local demand, local contracting, local negotiations and perceived local rivalry can locate a market in Australia even if global parents make overarching decisions. This functional approach narrows the circumstances in which a cartel affecting Australian supply can escape Part IV liability on market-definition grounds.
For penalties the judgment resolves conflicting approaches to s 76(5)(d). The primary judge had read "the body corporate" as referring only to the contravening entity, thereby excluding most of AAPL's turnover. The Full Court held the phrase applies distributively. Annual turnover therefore captures the commercial scale of the entire group enterprise connected with Australia. The maximum penalty is now more likely to reflect the size of multinational respondents. The judgment also clarifies that the course-of-conduct principle cannot be used to collapse the statutory maxima for multiple distinct acts into one or two notional maxima. Each contravention carries its own statutory cap; the principle is a discretionary sentencing tool applied after maxima are identified and subject to the statutory distinction between making and giving effect.
After Yazaki, cartel enforcement against global supply chains is strengthened. Foreign parents cannot readily insulate themselves by withholding knowledge from Australian subsidiaries that implement cartel prices. Exclusionary cartels need not affect an identified Australian market. Maximum penalties better reflect group turnover. Sentencing must respect the separate statutory wrongs of making and giving effect to a cartel.
Key passages with plain-English translation
At [71]: "In our view, its ordinary meaning is not ambiguous in that regard. Conceptually, it is a term whose principal concern is the existence of certain conduct and whether such conduct implements, enacts, or otherwise administers an agreement." Plain English: "Give effect to" is about whether the action actually carries out the cartel deal. It does not automatically require that the person knows the full story.
At [76]: "This was sufficiently related to and connected with the cartel agreements, and was a direction given referable to the operation of those cartel agreements." Plain English: AAPL's price submission was close enough to the cartel plan, and was done on the parent's specific instruction about the cartel, so it counts as implementing it.
At [118]: "Section 4D in its terms does not refer to 'market' but only to competition. It does not say that the persons must compete in the same market or in the market to which the arrangement applies." Plain English: The ban on exclusionary deals looks at whether the parties are rivals, not at whether there is a neatly defined Australian market.
At [121]: "In our opinion, it does not mean that the word 'competitive' in s 4D means competitive in a market as defined in s 45(3) for the purposes of s 45." Plain English: The special definition of competition that mentions markets applies only to the parts of s 45 that talk about substantially lessening competition, not to the boycott-style exclusionary provisions.
At [141]: "Identifying a market and defining its dimensions is 'a focusing process', requiring selection of 'what emerges as the clearest picture of the relevant competitive process in the light of commercial reality and the purposes of the law'." Plain English: Deciding what the market is means looking at the real commercial world and the law's goals, not applying a rigid formula. Here that picture included Australian buyers, contracts and rivalry.
At [231]: "the 'course of conduct' principle does not have paramountcy in the process of assessing an appropriate penalty. It cannot of itself operate as a de facto limit on the penalty to be imposed..." Plain English: Treating several breaches as one "course" is just a helpful way of thinking; it cannot be used to slash the legal maximum penalty that Parliament set for each breach.
At [257]: "only a very sizeable penalty in this case is likely to act as specific deterrent against future contraventions by Yazaki, and as a general deterrent for similarly-sized global companies contemplating similar conduct." Plain English: Cartels are secret and hard to catch. The fine must be large enough to make even a huge multinational think twice.
What fact patterns trigger this precedent
The ratio on "give effect to" will be triggered whenever a subsidiary or agent performs an act that implements a cartel arrangement at the direction or with the consent of the principal, even if the subsidiary lacks knowledge of the arrangement. The critical factual connection is that the act must be "in accordance with" the cartel provision. A direction from the parent to quote a particular price that the parent has pre-agreed with a competitor is the paradigm. Conversely, wholly incidental acts by unrelated third parties (for example, a courier delivering documents without any knowledge or direction referable to the cartel) are unlikely to satisfy the test on the facts.
The exclusionary-provision holding applies to any horizontal agreement between competitors (or persons deemed competitive under s 4D(2)) that has the purpose of preventing, restricting or limiting supply or acquisition to particular persons or classes. No proof of an Australian market or of actual competitive activity inside Australia is required. The fact pattern is satisfied where two suppliers agree to protect each other's incumbency by submitting cover bids or by allocating customers, even if the decision-making occurs offshore and the only Australian element is the ultimate supply contract.
The market-definition analysis will be engaged whenever supply contracts are made in Australia, payments occur in Australia, local personnel negotiate prices or specifications, or local managers perceive each other as competitors, even if global parents retain ultimate approval. The precedent is particularly relevant to global RFQs that segment demand by country; treating a country as a distinct bid segment is strong evidence of a geographic dimension to the market. The existence of "some chance" of switching suppliers mid-cycle, or of intense post-award price-down negotiations, reinforces the finding of potential close competition inside Australia.
The turnover construction applies to any contravening corporation that has related bodies corporate carrying on business in Australia. Where the contravener directs or benefits from the subsidiary's Australian sales, those sales will count toward the 10% turnover cap unless they fall within one of the express exclusions. The precedent is triggered in typical parent-subsidiary structures in which the parent treats the subsidiary's turnover as part of its own commercial enterprise.
The course-of-conduct discussion is engaged whenever multiple discrete acts implement a cartel: reaching the initial understanding, coordinating prices, submitting bids in different jurisdictions, directing a subsidiary, and the subsidiary's actual submission. Each such act can attract its own maximum penalty. The statutory distinction between making an agreement and giving effect to it must be respected; a court cannot collapse them into a single notional maximum.
How later courts have treated it
The judgment has been treated as authoritative on the construction of "give effect to". Subsequent decisions have cited [71]-[80] for the proposition that implementation, not knowledge, is the statutory focus, while emphasising that the factual inquiry remains whether the conduct is sufficiently connected to the cartel. The distinction drawn between a directed subsidiary and an unwitting stranger has been accepted as the appropriate factual filter.
The clarification that exclusionary provisions do not require an Australian market has been followed in cases alleging collective boycotts or customer allocation. Courts have cited [113]-[132] and the reliance on Pont Data and News Ltd as settling the textual and historical argument. The proposition that s 45(3) is confined to the SLC limbs is now treated as settled law.
The functional approach to market definition, emphasising potential rivalry and commercial reality over the precise locus of substitution decisions, has been applied in other supply-chain cases. References to Air New Zealand at [136]-[167] are routine when offshore parents control Australian subsidiaries. The finding that local demand, local contracting and perceived local competition can locate a market in Australia despite global decision-making has narrowed the availability of the "no Australian market" defence.
On penalty, the distributive reading of s 76(5)(d) has been accepted in subsequent turnover calculations. The proposition that the course-of-conduct principle cannot operate as a de facto cap on statutory maxima is routinely cited when respondents argue for a single penalty for multiple making-and-giving-effect contraventions. The emphasis on specific and general deterrence for hard-core cartel conduct by large multinationals, and the yardstick role of the statutory maximum, have guided penalty assessments in later international cartel cases. The $46 million total, representing a substantial uplift from the primary judge's figure, is cited as illustrating the need for penalties that cannot be viewed as an acceptable cost of doing business.
Overall, later courts have treated the decision as a comprehensive restatement of cartel liability and sentencing principles that aligns the Act with its enforcement purpose. No aspect of the ratio has been distinguished or doubted in subsequent appellate authority.
Still-open questions
The judgment leaves open the precise factual boundary between conduct that is "in accordance with" a cartel and conduct that is merely incidental. While a directed subsidiary will usually cross the line, the Court expressly contemplated that each case must be assessed on its evidence. Exactly how remote or attenuated the connection must be before implementation is not established remains for future decision.
Although the Court held that a market existed on the facts, it did not lay down a bright-line test for when global decision-making so dominates that no Australian market can be found. The interplay between global RFQs and local negotiation remains fact-sensitive. Future cases may test whether purely formal local submission of prices, without any local autonomy, is sufficient to locate a market in Australia.
On turnover, the Court did not decide whether every conceivable supply by a related body corporate must be included if the related body carries on any enterprise at all. The distributive construction was sufficient for the facts; the precise limits of "in connection with an enterprise that the body corporate carries on" may require further elaboration where the subsidiary's activities are wholly unrelated to the contravening parent's business.
The sentencing analysis treated the five acts as warranting five separate penalties (with acts 4 and 5 aggregated). The Court did not decide whether, in a longer cartel with many price submissions over years, each submission could constitute a separate contravention attracting its own maximum. The interplay between the statutory language "each act or omission" and the course-of-conduct principle in very large cartels remains open.
Finally, the costs outcome turned on the ACCC's forensic choice to plead and contest actual knowledge. The judgment does not decide whether a regulator that pleads knowledge only in the alternative, and succeeds on the legal point that knowledge is unnecessary, would still suffer a costs reduction. That question is left for cases with different pleading histories.
(Word count: 1,872)