Mitsubishi Motors Australia Pty Ltd Northpark Berwick Investments Pty Ltd v Begovic

SUPREME COURT OF VICTORIA

COURT OF APPEAL

S EAPCI 2021 0063

MITSUBISHI MOTORS AUSTRALIA PTY LTD (ACN 007 870 395)	First Applicant
NORTHPARK BERWICK INVESTMENTS PTY LTD (ACN 075 238 121)	Second Applicant
v
ZELKO BEGOVIC	Respondent

---

JUDGES:	EMERTON P, McLEISH and MACAULAY JJA
WHERE HELD:	Melbourne
DATE OF HEARING:	24 May 2022
DATE OF JUDGMENT:	5 August 2022
MEDIUM NEUTRAL CITATION:	[2022] VSCA 155
JUDGMENT APPEALED FROM:	[2021] VSC 252 (Ginnane J)

---

CONSUMER LAW – Misleading or deceptive conduct – Fuel consumption label affixed to new vehicle offered for sale – Affixing of label required by law – Label displayed fuel consumption figures derived from standard testing of vehicle type – Purchased vehicle unable to substantially achieve label figures under standard test – Whether vehicle manufacturer and dealer engaged in actionable ‘conduct’ – Presenting vehicle for sale with label affixed actionable conduct – Presenting vehicle for sale not required by law – Whether conduct in ‘trade or commerce’ – Presenting labelled goods for sale in ‘trade or commerce’ – Whether conduct conveyed representation as to fuel consumption of vehicle – Whether label conveyed particular representation that fuel consumption figures substantially replicable in purchased vehicle – Representation conveyed – Whether representation attributable to manufacturer and dealer – Representation attributable – Whether representation misleading or deceptive – Representation falsified by testing of purchased vehicle – Leave to appeal granted – Appeal dismissed.

CONSTITUTIONAL LAW – Whether s 109 inconsistency between federal fuel consumption labelling rules and prohibition against misleading or deceptive conduct in state law – Fuel consumption rules contain no obligation to engage in misleading or deceptive conduct at issue – No inconsistency.

Australian Consumer Law and Fair Trading Act 2012, Motor Vehicles Standards Act 1989 (Cth), Vehicle Standard (Australian Design Rule 81/02 – Fuel Consumption Labelling for Light Vehicles) 2008 (Cth).

Google Inc v Australian Consumer and Competition Commission (2013) 249 CLR 435, Butcher v Lachlan Elder Realty Pty Ltd (2004) 218 CLR 592, distinguished.

---

Counsel
Applicants:		Mr B Walker SC and Mr B Gibson
Respondent:		Mr J Gottschall
Solicitors
Applicants:	Thomson Geer
Respondent:	SLF Lawyers

EMERTON P
MCLEISH JA
MACAULAY JA:

Introduction

In early 2017, the respondent, Zelko Begovic, purchased a new 2016 Mitsubishi MQ Triton 4x4 GLS motor vehicle from a car dealership run by the second applicant (‘Northpark’) in Berwick. The vehicle was manufactured by the first applicant (‘Mitsubishi’).
A vehicle standard made under the Motor Vehicles Standards Act 1989 (Cth) (‘the Standard’) required a ‘fuel consumption label’ to be displayed on the front windscreen of the vehicle.[1] The Standard prescribed the form and contents of the label. Relevantly, the Standard required that the label prominently display figures representing the ‘urban’, ‘extra urban’ and ‘combined test’ fuel consumption for the relevant vehicle type.[2] The Standard also prescribed standards and procedures for the testing to determine these figures.[3]
[1]Vehicle Standard (Australian Design Rule 81/02 – Fuel Consumption Labelling for Light Vehicles) 2008 (Cth) cl 4.1 (‘the Standard’).
[2]Ibid cll 4.5.1, 4.6, app A.
[3]Ibid cl 4.5.1, app B.
Mitsubishi conducted the requisite testing of a 2016 Mitsubishi Triton of the relevant type and caused the required label accurately displaying the results of that testing to be affixed to the windscreen of the vehicle before it was sold to Mr Begovic. The fuel consumption figures displayed on the label affixed to the vehicle he inspected were attractive to Mr Begovic. But soon after he purchased the vehicle, Mr Begovic concluded that it was consuming much more fuel than he expected.[4]
[4]Although it is not entirely clear, it seems to have been assumed that the vehicle purchased by Mr Begovic was the one he inspected prior to sale, on which he saw the label. It has not been suggested that anything turns on the point.
Mr Begovic commenced a proceeding in the Victorian Civil and Administrative Tribunal seeking a refund of the purchase price. The Tribunal found that, among other things, Northpark and Mitsubishi had engaged in ‘misleading or deceptive conduct’ within the meaning of s 18 of the Australian Consumer Law, applicable by virtue of s 8 of the Australian Consumer Law and Fair Trading Act 2012.[5] The Tribunal ordered Northpark to pay Mr Begovic a sum equal to the purchase price of the vehicle, and that the vehicle be returned to Northpark.
[5]Begovic v Northpark Berwick Investments Pty Ltd(Civil Claims) [2019] VCAT 772 (‘Tribunal’s Reasons’).
Northpark and Mitsubishi sought leave to appeal the Tribunal’s decision in the Trial Division, pursuant to s 148 of the Victorian Civil and Administrative Tribunal Act 1998 (‘VCAT Act’). A judge granted leave but relevantly upheld the Tribunal’s finding that Northpark and Mitsubishi had engaged in misleading or deceptive conduct, and dismissed the appeal.[6]
[6]Mitsubishi Motors Australia Ltd v Northpark Berwick Investments Pty Ltd [2021] VSC 252, [182] (‘Judge’s Reasons’).
Northpark and Mitsubishi now seek leave to appeal the judge’s decision, on three proposed grounds.[7] In brief, they contend that:
(a)affixing a label which complies with and is required by the Standard cannot constitute ‘misleading or deceptive conduct’ for the purpose of s 18;
(b)in any event, the label was not misleading or deceptive because it only represented the results of past testing of a test vehicle, which it represented accurately; and
(c)the label conveyed no representation that past testing results could be substantially replicated in the vehicle purchased.
[7]A further proposed ground of appeal was abandoned before the hearing.
Northpark and Mitsubishi have also filed notice of a constitutional matter, which arises only if their primary submissions fail. By that notice, they contend that, if conduct in compliance with the Standard contravened the Australian Consumer Law, then s 109 of the Constitution would have the effect that the Standard prevails to the extent of any inconsistency.
For the reasons that follow, leave to appeal should be granted but the appeal should be dismissed.

Background

The Standard

The Standard was made under s 7 of the Motor Vehicle Standards Act 1989 (Cth) (‘the Act’).[8] The Standard relevantly prescribed requirements for the ‘measurement of vehicle fuel consumption’ and the ‘design and application of fuel consumption labels’ to vehicles.[9] Three requirements are presently relevant.
[8]The Act has since been repealed. See now the Road Safety Standards Act 2018 (Cth).
[9]The Standard cl 1.1.
First, cl 4.5 imposed a testing and reporting requirement. It relevantly required that the vehicle manufacturer report the ‘three fuel consumption results’ for the vehicle type from the ‘urban’, ‘extra-urban’ and ‘combined’ tests conducted in accordance with test protocols prescribed by the Standard.
It is not necessary to describe those test protocols in any detail. It is sufficient to observe that they provide for testing to be conducted by the vehicle manufacturer, in a laboratory setting, on a test vehicle of the relevant type.
Secondly, cl 4.1 imposed a labelling requirement. It relevantly required that every vehicle subject to the Standard bear a fuel consumption label. The form and content of the label was prescribed in considerable detail, including as to its composition, formatting, colouration and text.[10] The Standard permitted two relevant additions to the prescribed form of label. A vehicle description was to appear at the top of the label, identifying the particular vehicle type by make, model, variant description, transmission type and fuel type.[11] Particular fuel consumption values corresponding to the expressions ‘Combined Test’, ‘Urban’ and ‘Extra Urban’ were to appear in a databox entitled ‘Fuel Consumption (L/100km)’.
[10]Ibid app A.
[11]Ibid app A, cll 1.10, 1.12.
Finally, cl 4.6.1 linked the testing and reporting requirement and the labelling requirement. It required that the three fuel consumption values displayed in the fuel consumption databox be the urban, extra-urban and combined values reported for the vehicle under cl 4.5.1.

The label

The fuel consumption label on the vehicle purchased by Mr Begovic (a copy of which appears in the Appendix to these reasons) was in the form prescribed by the Standard. It was entitled ‘Fuel Consumption’. It identified the relevant make (Mitsubishi), model (Triton), variant (GLX/GLS/Exceed), transmission type (Auto), and fuel type (Diesel).
It then set out the fuel consumption figures for ‘combined urban’, ‘urban’ and ‘extra urban’. Below the fuel consumption figures the label stated, in a form of words prescribed by the Standard:
Vehicle tested in accordance with [the Standard]. Actual fuel consumption … depends on factors such as traffic conditions, vehicle condition and how you drive.
Below that was a statement, also prescribed by the Standard, which said ‘More information at The website referred to was entitled the ‘Green Vehicle Guide’. One of the pages accessible on that website was entitled ‘Fuel Consumption Label’. That page said:
The figures displayed on the fuel consumption label are based on specific tests conducted by vehicle manufacturers to demonstrate a vehicle’s compliance with the [Standard]. All vehicles are tested under standardised, carefully controlled conditions in specialised vehicle emission laboratories. To ensure the quality and consistency of test results, laboratories and their facilities are subject to audit by the Australian Government.

Varieties of representation

At least until now, it has never really been in dispute that Mitsubishi and Northpark were responsible for whatever representations the label conveyed.
For the purposes of understanding the way in which this case has proceeded to this point, it is convenient to distinguish between four possible representations that the label might be thought to convey:
(a)That the fuel consumption figures on the label accurately record the results of testing of a test vehicle of the relevant type in accordance with the Standard (‘the test accuracy representation’).
(b)That the fuel consumption figures on the label precisely identify the fuel consumption performance of the vehicle on which the label is displayed (‘the actual fuel consumption representation’).
(c)That the fuel consumption figures on the label do not precisely identify but are broadly indicative of the real-life fuel consumption performance of the vehicle on which the label is displayed, allowing for variables such as driving and vehicle conditions (‘the indicative fuel consumption representation’).
(d)That the fuel consumption figures on the label could be substantially replicated if the purchased vehicle were to be tested in accordance with the test protocols in the Standard (‘the testing replicability representation’).
It has never really been in dispute that the label:
(a)conveyed the test accuracy representation; and
(b)did not convey the actual fuel consumption representation.
As explained below, the judge held that the case in the Tribunal had also raised the testing replicability representation and he upheld the Tribunal’s decision on that basis. That is the way in which Mr Begovic advances his case in this Court. It will be necessary to return below to the way in which the case has developed.

Tribunal proceeding

Before the Tribunal, it was accepted that s 18 of the Australian Consumer Law applied to Mitsubishi (as manufacturer) and Northpark (as dealer).[12]
[12]Tribunal’s Reasons [31]–[33].
It appears that it was also not in dispute that, whatever else it represented, the label represented something about the fuel consumption of the vehicle purchased by Mr Begovic. The Tribunal recorded that the parties agreed that the label represented that:
[T]he vehicle’s fuel consumption as tested in accordance with [the Standard] was [in litres per 100 kilometres] 7.6 combined test, 9.0 urban and 6.8 extra urban. The representation is qualified by the words ‘the actual fuel consumption and CO2 emissions depend on factors such as traffic conditions, vehicle conditions, and how you drive’.[13]
[13]Ibid [36].
Mr Begovic anticipated, based on that representation, that the vehicle’s fuel consumption would be superior to his previous vehicle.[14]
[14]Ibid [50].
The focus before the Tribunal was whether that representation — which did not clearly distinguish between the indicative fuel consumption representation and the testing replicability representation — was misleading or deceptive. Mitsubishi and Northpark submitted that the representation was not misleading or deceptive because Mr Begovic’s vehicle could achieve the fuel consumption performance indicated by the label.[15] Mr Begovic submitted that the representation was misleading or deceptive because it could not.
[15]Ibid [37].
After a compulsory conference in the Tribunal, the parties arranged for testing of Mr Begovic’s vehicle. Only Mr Begovic called expert evidence in respect of any of this testing.[16]

Ms Winkelmann’s evidence

[16]The applicants’ approach to the matter in the Tribunal was somewhat limited. They were not legally represented and called no expert evidence.

Mr Begovic’s expert was Ms Andrea Winklemann, an engineer whose expertise was not challenged. She tested the vehicle for fuel efficiency twice in early 2019, in a laboratory setting. By this time, the vehicle was more than two years old and had travelled nearly 50,000 kilometres.
Ms Winkelmann gave evidence that her testing demonstrated that the label information was ‘not true for the vehicle’.[17] Her opinion was based on results obtained under two testing methods, and those results were also not challenged.
[17]Tribunal’s Reasons [43].
First, she tested the vehicle in accordance with the Standard (‘the Standard test’). This test was designed to replicate the mandatory testing that yielded the results on the label. In effect, this test addressed the correctness of the testing replicability representation.
Secondly, she tested the vehicle under a different, ‘heavy, high speed’ driving cycle, with increased load (‘the heavy test’). This test was designed to simulate real world driving conditions, and effectively sought to evaluate the accuracy of the indicative fuel consumption representation.
There was a substantial discrepancy between the fuel consumption figures on the label and the results obtained under both the Standard and heavy tests, as follows:

Label

Standard test

Heavy test

Consumption (L/100km) Consumption (L/100km) Difference with label information Consumption (L/100km) Difference with label information

Combined test 7.6 9.6 26.6%

Urban 9.0 10.6 17.8%

Extra urban 6.8 9.3 36.8% 10.6 56.3%

As may be seen, the Standard test showed ‘combined test’ fuel consumption 26.6 per cent higher than the label figure, and ‘extra urban’ fuel consumption 36.8 per cent higher than the label figure. The heavy test of the vehicle showed ‘extra urban’ fuel consumption 56.3 per cent higher than the label figure.
Ms Winkelmann’s evidence was that these discrepancies were ‘unusual and excessive’. The difference between the label figures and the Standard test figures significantly exceeded any difference attributable to methodological differences permitted under Standard testing. The even larger difference between the label’s ‘extra urban’ figure and the equivalent heavy test figure was more unusual still. That difference, she said, was more than double the average difference between reported test results under the Standard and ‘real world’ fuel consumption. Indeed, the difference was so extreme that it suggested some ‘serious technical issue’ with the vehicle.
Tribunal’s reasons
The Tribunal accepted Ms Winkelmann’s evidence, and concluded that s 18 of the Australian Consumer Law had been contravened.[18]
[18]The Tribunal also concluded that Mitsubishi and Northpark had contravened the consumer guarantees as to acceptable quality and supply by description contained, respectively, in ss 54 and 56 of the Australian Consumer Law: ibid [64]; see also [78]. The judge upheld the appeal in respect of those provisions (on the basis that the Tribunal had made a factual finding that the vehicle was not defective, and because the provision had not been raised, respectively) and no issue is now taken in that regard.
The Tribunal characterised the conduct at issue as, on Mitsubishi’s part, ‘providing [the] information’ contained on the label and affixing the label to the vehicle, and, on Northpark’s part, adopting the label information in marketing and selling the vehicle.[19] It was not in doubt that Mitsubishi and Northpark were responsible for whatever representations were conveyed by the label.
[19]Tribunal’s Reasons [32]–[33]; see also [53].
The Tribunal found that the label simultaneously conveyed:
(a)the test accuracy representation (which was not misleading or deceptive);[20] and
(b)a representation regarding the purchased vehicle’s fuel consumption (which was misleading or deceptive).[21] The representation was identified with greater particularity by the judge, as explained further below.
[20]Ibid [42].
[21]Ibid [51]–[53].
The Tribunal concluded that the latter representation was not true for ‘the vehicle subject of these proceedings’.[22] It said:
The label information was false based on expert evidence. It misled Mr Begovic to believe the vehicle had certain fuel consumption characteristics it did not have … In purchasing the vehicle, he did not get what was represented to him by the fuel label.[23]
[22]Ibid [51].
[23]Ibid [52].
The Tribunal concluded that Mr Begovic was entitled to reject the vehicle and have his purchase price refunded.

Trial Division

Mitsubishi and Northpark sought leave to appeal the Tribunal’s decision to the Trial Division.
Before the judge, there was no challenge to the Tribunal’s characterisation of the relevant conduct. It remained common ground that Mitsubishi and Northpark had made whatever representations were conveyed in the label.
However, Mitsubishi and Northpark now submitted that the label conveyed nothing about the fuel consumption of the vehicle purchased by Mr Begovic, and conveyed only the test accuracy representation. For his part, Mr Begovic maintained that the label represented something about the fuel consumption of the vehicle he purchased, but he ultimately submitted that the critical representation conveyed was along the lines of the testing replicability representation rather than the indicative fuel consumption representation.[24]
[24]Judge’s Reasons [77].
The judge upheld the Tribunal’s conclusion that Mitsubishi and Northpark had engaged in misleading or deceptive conduct.
The judge explained in a series of passages to which it will be necessary to return how the case for Mr Begovic had evolved. In the judge’s view, rather than finding that the relevantly misleading or deceptive representation was the indicative fuel consumption representation, the Tribunal implicitly found (without expressly saying so) that it was the testing replicability representation.[25] He concluded:
[T]he results [reported on the label] were not repeated by the testing conducted by Ms Winkelmann … In my opinion, the Senior Member therefore found that the label fuel consumption figures contained on it, were not only test results obtained in the past under [testing under the Standard], but could be substantially replicated under further [testing under the Standard].[26]
[25]Ibid [77].
[26]Ibid.

The judge upheld that finding. He stated that:
[A] reasonable consumer reading the label would … take it to represent that, if the vehicle were retested[27] under [the Standard] it would produce similar results.[28]
[27]As senior counsel for the applicants identified, the word ‘retested’ is inapt because the Standard does not require individual vehicles to be tested before sale.
[28]Judge’s Reasons [117].
The judge considered that replicability was necessary to realise a purpose of the label: namely, to enable the consumer to sensibly compare fuel efficiency between different vehicles and vehicle types. He said it was:
[T]he representation that the fuel consumption figures could be replicated under … testing [that] enabled a comparison of … fuel economy to be made.[29]
[29]Ibid [121].
Elsewhere, he put it more strongly, saying that the ‘consumer could only make a reasonably accurate comparison with the fuel consumption of other vehicles if the testing results could be replicated’ under testing in accordance with the Standard.[30]
[30]Ibid [127] (emphasis added).
The judge also considered that replicability was something of which Mitsubishi, as manufacturer, was capable,[31] and which it bore responsibility for ensuring. He said:
Mitsubishi bore the responsibility of ensuring that the testing conducted on the 2016 Triton would produce fuel consumption figures that could be replicated.[32]
[31]Ibid.
[32]Ibid [118].
In another part of his reasons, the judge described the testing replicability representation as time-limited. It was a representation that the test results would be replicable ‘for a reasonable period, at least two years and perhaps for the life of the vehicle warranty.’[33]
[33]Ibid [168].
The judge concluded that the testing replicability representation was misleading or deceptive because Ms Winkelmann’s testing ‘established that the fuel consumption figures on the label could not be repeated under further testing [in accordance] with the Standard’.[34] In doing so, the judge rejected Mitsubishi and Northpark’s argument that producing and applying a compulsory label — even if its contents were inaccurate for the vehicle labelled — was inherently incapable of constituting ‘misleading or deceptive conduct’.[35]
[34]Ibid [118].
[35]Ibid [125].
The judge — like the Tribunal — appeared to accept that the label also conveyed the test accuracy representation,[36] and that that representation was not misleading or deceptive.[37]
[36]Ibid [117]. However, the judge also said: ‘Consumers would mostly have no idea … that the labelling did no more than report the results of testing performance for … certification’ (at [125]).
[37]There was disagreement before us as to whether the judge decided that the test accuracy representation was true, or whether that had simply been assumed. At all events, the respondent has not at any stage sought to disprove the representation.

Proposed grounds of appeal

There are three proposed grounds of appeal. The first and second proposed grounds are linked. They read (omitting references):
1.The primary judge erred in finding that by accurately recording the fuel consumption of a model as tested in the mandatory fuel consumption label, affixing that fuel consumption label to a vehicle, and selling that vehicle with the fuel consumption label affixed, as required by [the Standard], the applicants had engaged in misleading or deceptive conduct in contravention of s 18 of the [Australian Consumer Law].
2.Having accepted that the label represented that, based on testing conducted in accordance with [the Standard], the fuel consumption results were as stated in the label, and having accepted that the results of that testing were accurately recorded in the label, the primary judge ought to have allowed the appeal and dismissed the application before [the Tribunal].
By these grounds, the applicants contend that the judge erred by finding that they engaged in misleading or deceptive conduct in contravention of s 18 by engaging in conduct required and authorised by the Standard. That conduct was said to be (a) accurately recording in the fuel consumption label the fuel consumption of a test vehicle tested in accordance with the Standard, (b) affixing the fuel consumption label to a vehicle of the same type, and (c) and selling the vehicle with the label affixed.
The third proposed ground is:
3.The primary judge erred in finding that the label also represented that the fuel consumption figures recorded in the label could be substantially replicated if the vehicle was retested under [the Standard] within a reasonable period of at least two years after sale, and perhaps for the life of the vehicle warranty (‘the additional representation’) as:
(a)the additional representation was not open on the plan terms of the label, or the Green Vehicle Guide incorporated into the label by reference; and
(b)the additional representation was not raised before the Tribunal and was not the representation that the respondent claimed, or was found, to have relied on.
This ground relates to the testing replicability representation (or a version of it). The applicants contend that the judge erred in finding that the label conveyed that representation in addition to the test accuracy representation.

Applying s 18 of the Australian Consumer Law

Section 18(1) of the Australian Consumer Law provides that a ‘person must not, in trade or commerce, engage in conduct that is misleading or deceptive or is likely to mislead or deceive’.
It is uncontroversial that determining whether s 18 has been contravened involves a staged inquiry. The omission, or conflation, of these stages is apt to confuse the analysis.[38]
[38]Google Inc v Australian Competition and Consumer Commission (2013) 249 CLR 435, 465 [93] (Hayne J) (‘Google Inc’); Veda Advantage Ltd v Malouf Group Enterprises Pty Ltd (2016) 241 FCR 161, 205 [241] (Katzmann J).
Five stages of that inquiry are presently relevant.
First, it is always necessary to identify precisely the conduct at issue.[39] As Hayne J said in Google Inc v Australian Consumer and Competition Commission:
The first question for consideration is always: ‘What did the alleged contravener do (or not do)?’ It is only after identifying the conduct that is impugned that one can go on to consider separately whether that conduct is misleading or deceptive or likely to be so.[40]
[39]See, eg, Miller and Associates Insurance Broking Pty Ltd v BMW Australia Finance Limited (2010) 241 CLR 357, 364 [5] (French CJ and Kiefel J).
[40]Google Inc (2013) 249 CLR 435, 465 [89].
Secondly, it is necessary to establish that the conduct is relevantly in ‘trade or commerce’, being conduct that is an ‘aspect or element of activities or transactions which, of their nature, bear a trading or commercial character’.[41]
[41]Concrete Constructions (NSW) Pty Ltd v Nelson (1990) 169 CLR 594, 604 (Mason CJ, Deane, Dawson and Gaudron JJ) (‘Concrete Constructions’); Murphy v Victoria (2014) 45 VR 119, 143 [77], 148 [89]–[92] (Nettle AP, Santamaria and Beach JJA).
Thirdly, once the conduct at issue has been identified with precision, it is necessary to determine what that conduct means; in particular, whether it conveys particular representations said to be misleading or deceptive.[42] Whether conduct conveys a representation is to be determined objectively, in all of the circumstances.[43] This involves consideration not only of the literal meaning of any words, phrases and figures used, but also the context in which they are used, including the nature of the parties, the character of their dealings, and their respective states of knowledge.[44]
[42]Campomar Sociedad, Limitada v Nike International Ltd (2000) 202 CLR 45, 84 [100] (Gleeson CJ, Gaudron, McHugh, Gummow, Kirby, Hayne and Callinan JJ) (‘Campomar’).
[43]Braham v ACN 101 482 580 Pty Ltd [2020] VSCA 108, [159] (Tate, McLeish and Niall JJA) (‘Braham’); Global One Mobile Entertainment Pty Ltd v Australian Competition and Consumer Commission [2012] FCAFC 134, [108] (Greenwood, Logan and Yates JJ); Cargill Australia Ltd v Viterra Malt Pty Ltd [No 28] [2022] VSC 13, [3810]–[3811] (Elliott J).
[44]Braham [2020] VSCA 108, [159]; WEA International Inc v Hanimex Corporation Ltd (1987) 17 FCR 274, 280 (Gummow J).
Fourthly, it is necessary to determine whether the conduct is attributable to the persons said to have contravened s 18.
Fifthly, it is necessary to determine whether the conduct — and any representation — is misleading or deceptive or likely to mislead or deceive.[45]

Applicants’ submissions

[45]Australian Competition and Consumer Commission v TPG Internet Pty Ltd (2013) 20 CLR 640, 651 [39] (French CJ, Crennan, Bell and Keane JJ) (‘TPG Internet’).

It is convenient to treat the three proposed grounds together.
The applicants submitted that conduct mandated by and in compliance with the Standard was inherently incapable of amounting to ‘misleading or deceptive conduct’ for the purpose of s 18 of the Australian Consumer Law. The applicants submitted that an action in respect of which a person has no choice cannot sensibly be described as that person’s conduct. Alternatively, it cannot be actionably misleading or deceptive to do that which the law requires.
More specifically, the applicants submitted that they were not the authors of the label, and had no control over its message.[46] They were not making representations, or adopting the content of the label as their own, but only obeying the law.[47]
[46]Google Inc (2013) 249 CLR 435, 483 [149] (Heydon J); cf Guy v Crown Melbourne Ltd [No 2] (2018) 355 ALR 420, 528 [451]–[452] (Mortimer J); [2018] FCA 36.
[47]Gardam v George Wills & Co Ltd (1988) 82 ALR 415, 427 (French J); [1988] FCA 289; Yorke v Lucas (1985) 158 CLR 661, 666; Butcher v Lachlan Elder Realty Pty Ltd (2004) 218 CLR 592, 605 [40] (Gleeson CJ, Hayne and Heydon JJ) (‘Butcher’); Google Inc (2013) 249 CLR 435, 446 [15] (French CJ, Crennan and Kiefel JJ).
In terms of the staged inquiry described above, the applicants submitted that:
(a)The conduct in issue was not the making of any representation contained in the label, but rather affixing the label, and selling the vehicle with the label displayed.
(b)This conduct, in obedience to the law, was merely ancillary to, rather than ‘in’, trade or commerce.[48]
(c)The conduct in issue — affixing the label to the vehicle and selling the vehicle with the label affixed — did not convey any representation on the part of the applicants. But even if the conduct at issue was characterised as adopting the contents of the label, that conduct could only convey the test accuracy representation, the correctness of which was not in doubt. This was because the label identified that:
•the figures recorded results obtained in laboratory testing according to the Standard;
•the results were obtained in respect of a vehicle type: a specified make, model, variant, transmission type, and fuel type; and
•the recorded test results were something other than the purchased vehicle’s actual fuel consumption, which was governed by a range of particular factors identified in the qualifying statement.
As a result, it was said that the label could not convey any representation about the vehicle on which it was displayed. Ms Winklemann’s opinion therefore rested on a false premise — that the label represented something about the actual vehicle which could be falsified by real world testing.[49]
(d)In any event, any representation conveyed by the label could not be attributed to Mitsubishi or Northpark. This was because neither had any say in the form or content of the label. The language of the label was prescribed by the Standard. The fuel consumption figures, while produced by Mitsubishi, were not chosen by it. Instead, Mitsubishi was bound to display on the label the figures derived from the mandated testing conducted in accordance with the Standard.
(e)Finally, any representations conveyed by the mandated label could not be actionably ‘misleading or deceptive’, irrespective of whether they were inaccurate. This was because the form and content of the label prescribed by the Standard necessarily reflects a balancing of policy considerations, only one of which is ensuring total accuracy. The way in which the imperative of accuracy is balanced against other imperatives, such as simplicity and accessibility, is a matter of legislative choice, not to be impugned by the courts. The display of a label that complies with the Standard can therefore not be adjudged to be ‘misleading or deceptive’.
[48]Concrete Constructions (1990) 169 CLR 594, 604 (Mason CJ, Deane, Dawson and Gaudron JJ); Johnstone v Victorian Lawyers RPA Ltd (2003) 132 FCR 411, 416 [14] (Sundberg J).
[49]It was said that conduct not in compliance with the Standard, such as displaying a label which contained figures other than those produced in testing under the Standard, would be in a different category.
The applicants further submitted that, if s 18 of the Australian Consumer Law prohibited conduct which was required of a person under the Standard, then simultaneous obedience with both laws was impossible and s 18, being a State law, was invalid to the extent of the inconsistency by virtue of s 109 of the Constitution.
As noted, the applicants contended that the label conveyed nothing other than the test accuracy representation. By the third proposed ground, they also specifically contended that the judge erred by finding that it conveyed the testing replicability representation. They said there was nothing in the terms of the label capable of conveying the testing replicability representation.
First, it was said that the testing replicability representation was the result of the judge having ‘re‑engineered’ the Tribunal’s reasons. By substituting the testing replicability representation for the indicative fuel consumption representation, the judge was said to have misinterpreted the Tribunal’s reasons and departed from the task on appeal.
In that context, it was submitted that, since the testing replicability representation was not raised before the Tribunal, it could not properly form the basis for the decision to uphold the Tribunal’s decision in an appeal under s 148 of the VCAT Act, which was in the nature of judicial review.[50] Moreover, it was said, the evidence before the Tribunal was that the respondent had relied on a different representation, namely that the vehicle he purchased was more fuel efficient than the 2008 model he had previously owned.[51]
[50]Roy Morgan Research Centre Pty Ltd v Commissioner of State Revenue (2001) 207 CLR 72, 79 [15] (Gaudron, Gummow, Hayne and Callinan JJ); Osland v Secretary, Department of Justice (2010) 241 CLR 320, 331–2 [18] (French CJ, Gummow and Bell JJ); Risi Pty Ltd v Pin Oak Holdings Pty Ltd [2017] VSCA 317, [50]–[54], [74] (Tate, Santamaria and Hansen JJA).
[51]Coulton v Holcombe (1986) 162 CLR 1, 7–8 (Gibbs CJ, Wilson, Brennan and Dawson JJ); Vlahos Pty Ltd v Vlahos [2017] VSCA 166, [49]–[51] (Kyrou JA, Tate JA agreeing at [1], McLeish JA agreeing at [134]).
Secondly, it was said that the judge wrongly assumed that the Standard governing the testing of a standardised vehicle in laboratory conditions could be transposed to testing of a particular used vehicle. There was said to be no basis in the Standard or otherwise for this assumption.
Thirdly, it was said that the judge’s emphasis on Mitsubishi’s capacity and responsibility to ensure replicability was unsupported. The Standard was said not to fix Mitsubishi with any such responsibility. Mitsubishi’s capacity to ensure replicability is all but lost at the point of sale, after which it may no longer control many factors which may affect a vehicle’s performance in later testing, such as the way it is used and maintained, and whether it is modified.
Fourthly, it was said that it was logically wrong for the judge to view replicability as essential to enable the comparison of vehicle types. A comparison could be made across vehicle types on the basis of the label figures (whether replicable for a particular vehicle or not) as they were obtained under common test protocols.
Finally, it was said that the judge’s error in finding that the testing replicability representation was conveyed at all was compounded by finding that it persisted for two years, if not the life of the warranty.
Respondent’s submissions
The respondent submitted that the case involved a consumer transaction in which the goods purchased had not matched the label attached to them. He submitted that inaccuracy of the label in its application to a specific vehicle was not consistent with the Standard. It therefore could not be said that the conduct of the applicants in presenting a vehicle for sale with a label affixed that was inaccurate for that vehicle was compelled by the Standard. This was not a case of merely doing what the law required. Over the applicants’ objection, the respondent submitted that it could be inferred that the applicants had sold a vehicle inferior to that tested under the Standard. In any event, it was said that no inconsistency existed between the Standard and s 18.
In terms of the staged inquiry described above, the respondent submitted that:[52]
[52]The respondent also objected that the applicants’ case was now different from the case put before the Tribunal. There the common position was that the label represented something about the fuel consumption of Mr Begovic’s vehicle which could be verified or falsified by testing of that vehicle. The applicants’ present case — that the label could not have conveyed any actionable representation because it complied with the Standard — is, it was said, inconsistent with that position, was not raised before the Tribunal, and ought not be considered in an appeal under s 148 on a question of law. (It was submitted that a matter not raised in the Tribunal cannot give rise to a question of law upon which to appeal under s 148, but that is not correct as a general statement: see Commissioner of State Revenue v Mondous (2018) 55 VR 643, 661–2 [77] (McLeish JA, McDonald AJA agreeing at 700 [266]).)
(a)The relevant conduct was communicating the information contained in the label. Mitsubishi and Northpark were relevantly the ‘authors’ of the ‘operative parts’ of the label and they ‘created [its] message’. This was because, in particular, Mitsubishi conducted the testing from which the fuel consumption figures displayed were derived.
(b)This conduct was plainly in trade or commerce, as the information was presented in the context of a consumer purchase transaction.
(c)The information contained in the label conveyed more than the test accuracy representation. Accounting for context,[53] it made a representation as to the fuel consumption performance of the vehicle to which the label was affixed. This was because the label:
·was directed at potential purchasers of the vehicle, at the time of purchase;
·was affixed to a particular vehicle available for sale;
·was accompanied by the text ‘vehicle tested in accordance with [the Standard]’; and
·displayed fuel consumption figures for the precise type of vehicle for sale.
As a result, it was said, the label represented to the consumer that the fuel consumption figures were ‘the relevant attributes’ of the vehicle and would be correct for that vehicle. Put shortly, it was submitted that the vehicle sold was being said to ‘match the label’, in the sense that, if it had been tested in accordance with the testing protocols in the Standard, the results would have been ‘materially the same’ as stated in the label.
(d)The relevant conduct — communicating the label information — was attributable to Mitsubishi and Northpark as the effective ‘authors’ of that information.
(e)The representation was ‘misleading or deceptive’ even if it is assumed that the Standard was complied with. Given the disparity between the reported results and Ms Winkelmann’s results, Mr Begovic’s vehicle must have been inferior to the vehicle tested. This rendered the label figures — even if accurate for the test vehicle — inaccurate for his vehicle.
[53]TPG Internet Pty Ltd (2013) 250 CLR 640, 654 [46] (French CJ, Crennan, Bell and Keane JJ); Campomar (2000) 202 CLR 45, 84 [100] (Gleeson CJ, Gaudron, McHugh, Gummow, Kirby, Hayne and Callinan JJ); Butcher (2004) 218 CLR 592, 604–5 [37] (Gleeson CJ, Hayne and Heydon JJ).

In relation to the testing replicability representation, the respondent disputed that the judge’s formulation of that representation involved a time span. It was instead a representation made at the point of sale that the testing results could be replicated if the actual vehicle was tested under the Standard. The respondent submitted that the time between the purchase of the vehicle and Ms Winkelmann’s testing was irrelevant because no suggestion was raised in the Tribunal that the age or condition of the vehicle had any bearing on that testing.

Analysis

Identifying the conduct

Whatever else has changed over the course of this litigation, it seems to have been accepted throughout that both the applicants shared the legal responsibility for the label being affixed to the vehicle purchased by the respondent and that if the affixation of that label and the sale of the vehicle with the label affixed involved actionable misleading or deceptive conduct, the applicants also shared responsibility for any loss suffered by the respondent as a result.[54]
[54]The Tribunal did not make orders against Mitsubishi, but that seems not to have been treated as an obstacle to its appeal. The judge set aside the Tribunal’s orders and remitted the matter for further hearing and orders. It is therefore convenient in what follows to treat the applicants indistinguishably, as the parties did in their submissions before us.
On that footing, and accepting that the applicants denied responsibility for anything done by them that was mandated by the Standard, it is necessary first to identify the ‘conduct’ of the applicants that is in issue.[55] The applicants appeared to confine the relevant conduct to the affixing of the label to the vehicle and then selling it. However, that would be to ignore the important fact that the vehicle with the label so affixed was presented to prospective purchasers, including the respondent, for sale. That was part of the conduct of the applicants and it would be arbitrary to confine the inquiry to the question whether the affixing of the label to the vehicle and its subsequent sale were misleading or deceptive.
[55]Google Inc (2013) 249 CLR 435, 465 [89] (Hayne J).
The relevant conduct therefore consisted of the applicants presenting a vehicle to the respondent for potential sale, with a label affixed that displayed (relevantly) figures for fuel consumption said to have been calculated by testing done ‘in accordance with [the Standard]’. More will be said later about the content of the label and what it conveyed. The anterior issue is whether the presentation of the vehicle with the label affixed was the conduct of the applicants, notwithstanding that, if the vehicle was to be presented for sale, they were legally obliged to ensure that the label was affixed.
The applicants submitted that a person’s obedience of a legal requirement cannot sensibly be regarded as the conduct of that person for the purposes of s 18. That submission should not be accepted. The word ‘conduct’ is an ordinary English word that describes things done (or not done) by a person, whether lawful or otherwise.[56] If unlawful, a person’s actions may be characterised as ‘misconduct’, being a species of ‘conduct’. If lawful, their actions are still ‘conduct’. The applicants’ submission would exclude from the notion of ‘conduct’ actions which are compelled by law. But the fact that a person is bound by statute to act in a particular way does not deny that, whether or not they comply with that obligation, their actions remain their own. They engage in conduct, whether it is lawful or unlawful (misconduct). While questions may arise as to whether it can be ‘misleading or deceptive’ to comply with a statute, and as to whether the general language of the consumer law is to be read down to enable it to apply consistently with a more specific law such as the Standard, that is not a basis for giving ‘conduct’ a strained meaning at the outset of the inquiry.
[56]Ibid; Henjo Investments Pty Ltd v Collins Marrickville Pty Ltd [No 1] (1988) 39 FCR 546, 555 (Lockhart J).
The applicants further submitted that, in the context of s 18 and its predecessor in s 52 of the Trade Practices Act 1974 (Cth), the fact that they were not the authors of the label, and had no control over its contents, meant that they were not responsible for any representation it conveyed. This submission, which was not made below, seeks to draw too much from the cases upon which the applicants relied.
In Butcher v Lachlan Elder Realty Pty Ltd, a real estate agent provided a brochure to prospective purchasers of real estate which contained an inaccurate diagram. The brochure stated that its contents were obtained from other sources which the agent believed to be reliable, but that interested persons should make their own inquiries. The agent was held not to have engaged in misleading or deceptive conduct because he had merely conveyed what the vendor was representing, without adopting or endorsing it. Gleeson CJ, Hayne and Heydon JJ held that this conclusion flowed from the nature of the parties, the character of the transaction and the contents of the brochure.[57] The Court did not hold that some affirmative act of adoption or endorsement is required before the passing on of a third party’s statement can constitute the making of a representation to the effect of that statement. Whether that conclusion can be drawn, and whether such a statement has in fact been adopted or endorsed, depends on all the circumstances.
[57]Butcher (2004) 218 CLR 592, 605 [40] (Gleeson CJ, Hayne and Heydon JJ).
A similar issue arose in Google Inc, where the operator of a search engine (‘Google’) was alleged to have engaged in misleading or deceptive conduct by displaying sponsored advertising at the head of a page of internet search results. The High Court held that Google was not the maker, author, creator or originator of the information in a sponsored link.[58] Further, ‘ordinary and reasonable users’ of the search engine would have understood that representations made in the sponsored links were those of the advertisers and were not adopted or endorsed by Google.[59] The joint judgment confirmed that the relevant inquiry was a question of fact in every case:
It has been established in relation to intermediaries or agents that the question whether a corporation which publishes, communicates or passes on the misleading representation of another has itself engaged in misleading or deceptive conduct will depend on whether it would appear to ordinary and reasonable members of the relevant class that the corporation has adopted or endorsed that representation. It has also been established that, if that question arises, it will be a question of fact to be decided by reference to all the circumstances of a particular case.[60]
[58]Google Inc (2013) 249 CLR 435, 459 [68] (French CJ, Crennan and Kiefel JJ).
[59]Ibid 460 [70] (French CJ, Crennan and Kiefel JJ).
[60]Ibid 446 [15] (French CJ, Crennan and Kiefel JJ) (citations omitted); see also 483 [148] (Heydon J).
As Hayne J pointed out in Google Inc, the text of the statute does not require a corporation to have ‘endorsed’ or ‘adopted’ the content of an advertisement made by another in order for the corporation to be found to have engaged in misleading or deceptive conduct.[61] Similarly, it is clear from these cases that there is no principle that a person must be the ‘author’ of a representation in order to engage in misleading or deceptive conduct in respect of it.
[61]Ibid 467 [99], 471 [114].
When these authorities are applied to the present case, an ordinary and reasonable prospective purchaser would consider that the information in the label was provided by the applicants as an aid to the decision whether to purchase the vehicle. The label contained no disclaimer of the kind found in Butcher and provided no basis for supposing, as in Google Inc, that it was authored by others without the applicants’ input. To the contrary, the label clearly conveyed, through the website to which it referred, that the vehicle’s manufacturer was responsible for the testing which produced the information in the label. To an ordinary and reasonable prospective purchaser, the label contained information provided on behalf of the vendor and manufacturer at the point of sale.
Even if an ordinary and reasonable prospective purchaser might be taken to know that the applicants provided the information in the notice under compulsion, it would not follow that this would sustain an inference that they neither adopted nor endorsed the information appearing on their vehicle. The fact of compulsion, of itself, says nothing as to the provenance of the information required to be disclosed.[62] In all the circumstances, including the fact that the manufacturer conducted the testing and there was no suggestion that the applicants did not accept responsibility for the figures in the label, the reasonable consumer would still take the label to contain information put forward on behalf of the manufacturer and vendor of the vehicle.
[62]A hypothetical purchaser who knew that the provision of the information was mandatory would still be taken to know that the manufacturer was responsible for the testing that produced the information.
For these reasons, the applicants engaged in conduct which extended to making any representation conveyed by the label, but also extended to affixing the label, presenting the vehicle for sale, and selling the vehicle with the label affixed.

‘In trade or commerce’

The conduct just identified was plainly ‘in’ trade or commerce. In the circumstances, it would be futile to explore the question whether the applicants should be permitted to advance the above arguments, or the ‘in trade or commerce’ argument, for the first time in this Court.[63]

Representations conveyed by the conduct

[63]See n 52 above.

It is next necessary to decide whether the conduct conveyed the representation asserted by the respondent. As explained earlier, this is to be determined objectively, by considering what was said and done against the background of all the surrounding circumstances, including the circumstances in which, in this case, the vehicle with the label affixed was presented for sale to the public.[64]
[64]Campomar (2000) 202 CLR 45, 84 [100], 85 [102] (Gleeson CJ, Gaudron, McHugh, Gummow, Kirby, Hayne and Callinan JJ); Aldi Foods Pty Ltd v Moroccanoil Israel Ltd (2018) 261 FCR 301, 324 [74] (Perram J, Allsop CJ agreeing at 306 [1], Markovic J agreeing at 346 [169]).
As a preliminary matter, the applicants contended that the respondent’s case had changed, and that he initially asserted only a representation that the vehicle he purchased would be more fuel efficient than his previous vehicle.[65] In our view, this is too narrow a view of the case that was conducted in the Tribunal. The respondent did state that he understood the label to have the consequence that the new vehicle would be more fuel efficient than his old one, but this went to the reliance he placed on the label, rather than its meaning.[66] The label’s meaning was plainly said to be specific to the vehicle the respondent purchased, which explains his reliance on the evidence of Ms Winkelmann to falsify what he said was conveyed by the label as applied to that vehicle. That evidence took the case beyond a mere comparison between the old and new vehicles and directly addressed the applicability of the label to the vehicle the respondent purchased.
[65]See [69] above.
[66]See also Tribunal’s Reasons [50] and [52].
The judge dealt with these variations in the case and described how the respondent came to rely ultimately on the case as it is now put:
[W]hen her reasons are read as a whole, it is clear that the Senior Member found, based on the expert evidence, that the Fuel Consumption Label contained representations that were misleading and deceptive and false or untrue for the vehicle. There being no dispute that the test results shown on the label were accurate as to the testing that Mitsubishi had relied on to obtain approval for the vehicle model, the Senior Member’s finding appears clearly enough to have been that the label information was misleading and deceptive because the test results contained on it were not reflected in the fuel consumption that Mr Begovic experienced with his 2016 Triton from the time he took delivery of it after he had made an adjustment for real world driving conditions. In addition, the test results were not repeated by the testing conducted by Ms Winkelmann or by the joint testing. In my opinion, the Senior Member therefore found that the label represented that the Fuel Consumption figures contained on it, were not only test results obtained in the past under [Standard] testing, but could be substantially replicated under further [Standard] testing. This was the primary representation upon which Mr Begovic ultimately relied before this Court.[67]
[67]Judge’s Reasons [77] (emphasis added) (citations omitted).
The judge encapsulated the ‘ultimate submission’ of the respondent as follows:
That the label provided a representation of the [fuel] consumption of the vehicle when it was tested, or if it were to be tested under relevant [Standard] conditions. So a consumer seeing this label ought to appreciate from this label if the car to which it was affixed were to be tested in accordance with [the Standard], as the label states, [t]hat the result of that testing would be substantially as shown on this label.[68]
[68]Ibid [101] (emphasis added).
The judge acknowledged that this precise argument had not been put to the Tribunal, but considered that it was inherent in the respondent’s case. He explained:
As mentioned, Mr Begovic’s ultimate argument made in this Court was that [the] label represented that the attributes of the specific vehicle model were such that if tested again in accordance with [the Standard], the fuel consumption results would produce substantially the results stated on the label. This argument was not specifically put to the Tribunal, but the essence of Mr Begovic’s case was that the label’s fuel consumption figures were ‘totally wrong’ from the time that he first drove the 2016 Triton. The issue of whether, if retested under [the Standard], the Triton could replicate the label’s fuel consumption figures was in dispute before the Tribunal as the provision of the joint testing evidence, Ms Winkelmann’s evidence and the Vipac test results demonstrated. Mr Begovic also argued that the fuel consumption figures shown on the label could be used as the basis for a reliable and repeatable calculation as to actual fuel consumption. They provided a form of reference point for comparison, from which he could reliably predict actual fuel consumption.[69]
[69]Ibid [112] (emphasis added) (citations omitted); see also [135].
We agree with the judge’s analysis of the way the case evolved. In our opinion the case now put by the respondent, in accordance with his ‘ultimate submission’, was inherent in his claims in the Tribunal, and the judge was correct to treat the Tribunal as having implicitly proceeded on that basis. That is apparent from the Tribunal’s decisive reliance on the conclusion of Ms Winkelmann, and her testing of the vehicle under the Standard, which directly compared the results obtained for the respondent’s vehicle with the figures in the label and found that the figures in the label were not accurate ‘for the vehicle’. In other words, testing under the Standard could not replicate the figures in the label.
It is now necessary to identify the content of the representation said to have been conveyed by the applicants’ conduct.
The applicants in this Court placed the fact of legal compulsion at the forefront of their argument on this aspect of the case, contending that it cannot be actionably misleading or deceptive to do what the law requires. This argument was advanced as a complete answer to the case, irrespective of what meaning the label might be said to convey. Again, the respondent contended that this was a new argument which the applicants should not be permitted to raise for the first time on appeal. It is convenient to put the applicants’ argument temporarily to one side, both in light of that submission, and because it is difficult to evaluate it in a factual vacuum, without having first decided what representations the label conveyed, which it is sought to be said were made in compliance with the law embodied in the Standard. We will therefore return to the argument in the context of considering whether the representations were actionably misleading or deceptive.
The first representation that is made by the label is the one we have called the test accuracy representation. While it was not really doubted below that this representation was made, and that it was true, the respondent at times seemed to depart from that concession, raising the question whether the vehicle that was tested was truly representative of its type.[70] That contention, however, does not advance the respondent’s case. Not only does it involve factual matters not explored at all in the Tribunal, but it is not said to alter or add to the representations made by the label. The question whether the choice of test vehicle complied with the Standard must be put to one side.
[70]See [74] above.
Nor was any case raised by the respondent to the effect that the label did not accurately record the results of the testing done under the Standard. There was argument before us as to whether the judge had made a finding that the Standard was complied with, or merely assumed that it was, in accordance with the way the case was run, but nothing now turns on that question.
When we turn then to the label, it plainly conveys that testing has been conducted ‘in accordance with’ the identified Standard, with the specified results. It is also clear that those results are not asserted to be those which would be achieved by any given use of the vehicle in ordinary driving conditions. They are the results of a specific test.
In our view, the circumstances in which the label would be read and understood include the fact that motor vehicles are mass-produced and distinguished by make and model (among other things), such that classes of vehicle exist within which the performance and fuel consumption of individual vehicles can be expected to be more or less the same. The reasonable prospective purchaser would therefore not take the words ‘[v]ehicle tested’ to refer to the specific vehicle, but to a representative vehicle. To infer otherwise would assume an extraordinarily extensive and unnecessary regime of testing brand new vehicles intended to be relevantly identical.
The question then is whether the label conveyed the further meaning that the particular vehicle would, if tested, yield substantially the same fuel consumption figures. As has been seen, the judge found that this was so, identifying what we have called the testing replicability representation. The Tribunal, rather similarly, acted on the basis of Ms Winkelmann’s evidence as to the vehicle’s performance in testing under the Standard and found in accordance with that evidence that the label was ‘not true’ and so was misleading or deceptive ‘for the vehicle’.[71]
[71]Tribunal’s Reasons [43], [51].
The judge inferred that the Tribunal had found that the label represented, not only the test results obtained under the Standard, but that those results could be substantially replicated for the particular vehicle to which the label was attached.[72] He accepted that was a representation which a reasonable consumer would consider was made by the label, given that an evident purpose of the label was to enable comparisons about fuel efficiency to be made between vehicles.[73] Further, the judge held that a consumer could only make a reasonably accurate comparison with the fuel consumption of other vehicles if the results could be replicated under the testing regime set out in the Standard.[74] Ms Winkelmann’s evidence was relevant to that matter, together with the respondent’s evidence of actual fuel usage since the time of purchase.
[72]Judge’s Reasons [77], [135], [137].
[73]Ibid [121].
[74]Ibid [127].

Later in the judgment, the judge dealt with a ground of appeal concerning the fact that Ms Winkelmann conducted her testing more than two years after the respondent purchased the vehicle. This argument was raised only in the context of s 54 of the Australian Consumer Law, regarding the alleged breach of the statutory guarantee of acceptable quality of goods supplied. The judge stated that a reasonable consumer would expect that the fuel consumption figures in the label would be capable of being replicated under the Standard’s testing for a reasonable period, at least two years and ‘perhaps’ for the life of the vehicle warranty.[75]
[75]Ibid [168].
In our view, this last conclusion cannot be regarded as an aspect of the judge’s reasoning regarding s 18. Apart from not having been an issue raised in that context, the operation of s 54 is very different. That provision asks whether or not goods are ‘of acceptable quality’, not whether representations are misleading or deceptive, and no representation of acceptable quality was advanced by the respondent as part of the s 18 case. The argument which the judge was later addressing was not seeking to identify any representation; it concerned the respondent’s reliance on Ms Winkelmann’s evidence to establish the condition of the vehicle when purchased.[76]
[76]Ibid [153]–[154].
On that basis, proposed ground 3 falls away to the extent that it contests the identification of the ‘testing replicability representation’ with a time period of two years, or indeed any time period after the time of purchase. For the reasons given, the judge adopted a narrower version of that representation, by which the test results were represented as substantially accurate, not only for the representative vehicle tested, but for the particular vehicle in question. That could, of course, most obviously be determined by replicating the testing under the protocols in the Standard, which is what Ms Winkelmann did, albeit more than two years later.
The issue is then whether the evidence sustains that more confined representation concerning replicability. If so, the issue of legal compulsion raised by the applicants will need to be addressed in that context.
In our view, an ordinary and reasonable consumer reading the label, affixed to a vehicle whose purchase he or she is contemplating, would not consider the fuel consumption figures in the label to be confined to a test vehicle and irrelevant to the vehicle bearing the label. They would instead understand that the information was provided to them because it would bear on the cost of running the specific vehicle if they were to purchase it. The consumer can, for the reasons given, be taken to know that the vehicle itself has not been tested. But in circumstances where motor vehicles are known to be manufactured in makes, models and classes so as to be reasonably identical to other vehicles of the relevant type, and given the evident purpose of comparison by prospective purchasers of types of vehicles offered for sale, the information would be of limited utility if it did not apply to the actual vehicle offered for sale. Those circumstances mean that the ordinary and reasonable purchaser would take the label to be saying how the vehicle offered for sale would have performed under the standardised fuel consumption testing regime already conducted in respect of a representative vehicle of the same make, model and class. That understanding is a function of mass production of consumer goods intended to be reasonably identical. There is no reason why the hypothetical purchaser would regard the label as asserting only that some, or most, vehicles of the same type would record approximately the figures in the label if subjected to standardised testing. The label contains no such disclaimer.
It is true that the label makes it clear that actual results of real-world driving will differ. This alerts the consumer to the danger of taking the figures too literally. But it also highlights the fact that the consumer will understand the label to be describing something about the actual vehicle. It may be noted that the reference in the label to ‘traffic conditions, vehicle condition and how you drive’ does not suggest that anything inherent in a vehicle’s state at the time of manufacture might cause it not to perform according to the label if it were to be tested at the time of sale.
This does not mean that the consumer would take the label to be saying anything additional, as to what would happen if the testing were to be replicated at some later point. The representation is that the figures are substantially correct, under standardised testing, for the vehicle at the time of purchase, so that comparisons can be drawn and decisions made on that basis.
Accordingly, the label reasonably conveyed that the figures in the label were substantially the results which would have been obtained by standardised testing of the vehicle to which that label was affixed. For these reasons, the remainder of proposed ground 3, in particular, must fail.

Attribution

The question of attribution of the relevant conduct has already been dealt with. The applicants argued that the conduct could not be attributed to them because, in affixing the label, they acted under legal compulsion. For the reasons given, this does not deny the agency of the applicants in acting as they did, or their responsibility for the conduct of the testing that yielded the figures in the label. Even if compliance with the law motivated their conduct, it remained their own conduct.
That does not mean that the legal framework within which the conduct took place is irrelevant. The applicants also submitted that this framework compelled the conclusion that the conduct could not be actionably misleading or deceptive. It is convenient to turn directly to that issue.

Misleading or deceptive

The applicants did not directly seek to challenge the judge’s finding that, if the label represented that testing of the vehicle under the Standard’s protocols would produce the results in the label, that representation was misleading or deceptive, consistently with Ms Winkelmann’s conclusions. Instead, they submitted that the representation could not be ‘actionably’ misleading or deceptive, because the applicants were doing no more or less than what the law, in the form of the Standard, required of them.
If the impugned conduct were to be confined to the making of a representation about the past testing of a representative vehicle, there would be much force in the applicants’ argument that the performance of a statutory obligation, without more, cannot meaningfully be said to constitute misleading or deceptive conduct.[77] At least, the general prohibition on engaging in such conduct would have to be read in the light of the legislative judgment, in the public interest, that the conduct must be engaged in.
[77]R v Credit Tribunal; Ex parte General Motors Acceptance Corporation (1977) 137 CLR 545, 561 (Mason J).
When it comes to the wider representation about the specific vehicle, however, the position is different. A manufacturer and dealer are not under a legal obligation to offer any given vehicle for sale. While they would be required to ensure the label was affixed if offering the vehicle for sale, nothing in the Act or the Standard was said to oblige them to offer such a vehicle for sale in the first place. Still less do the Act or the Standard require that a vehicle be offered for sale in circumstances where the representation in the label is misleading or deceptive in respect of that vehicle. Instead, the Act and the Standard presuppose that testing of a representative vehicle will produce figures that are substantially those which would be obtained under testing in accordance with the Standard of any and every vehicle of the type in question.
In part, the applicants’ argument depends on characterising their conduct as being confined to the affixing of the label, but it is not possible to ignore that this is only part of the overall conduct in offering the vehicle for sale. When the conduct is considered as a whole, it cannot be said that any statutory obligation to engage in it prevented that conduct from being relevantly, and actionably, misleading or deceptive.
In the absence of any identified obligation in the Standard, or the Act, to engage in the particular misleading or deceptive conduct identified in this case, no question of constitutional inconsistency arises.
It is not strictly necessary, given these conclusions, to decide whether, as the respondent contended, this aspect of the case involved a new argument on the part of the applicants. For our part, however, we do not think that it did. The judge addressed a similar argument, albeit that he rejected it on other grounds. At the same time, no constitutional argument was raised below. In the circumstances, the applicants have not made out their grounds of appeal and nothing turns on the point.

Conclusion

We will grant leave to appeal, but dismiss the appeal.

Annexure — copy of the label

Label	Standard test		Heavy test
Consumption (L/100km)	Consumption (L/100km)	Difference with label information	Consumption (L/100km)	Difference with label information
Combined test	7.6	9.6	26.6%
Urban	9.0	10.6	17.8%
Extra urban	6.8	9.3	36.8%	10.6	56.3%

[2022] VSCA 155

5 August 2022