NZME Limited v Commerce Commission [2018] NZCA 389; [2018] 3 NZLR 715 (26 September 2018)

	Home \| Databases \| WorldLII \| Search \| Feedback Court of Appeal of New Zealand

BETWEEN	NZME LIMITED First Appellant FAIRFAX MEDIA LIMITED Second Appellant STUFF LIMITED Third Appellant
AND	COMMERCE COMMISSION Respondent

Hearing:	5–8 June 2018
Court:	Kós P, Miller and Asher JJ
Counsel:	D J Goddard QC, S C Keene, A S Butler and AJWO Lomas for Appellants J A Farmer QC, J D Every-Palmer QC, F J Cuncannon, PIC Comrie-Thomson and G C Spittle for Respondent
_Judgment:	_{26 September 2018 at 4.00 pm}

The parties’ businesses and the transaction [6] The decisions below, in brief [15] The structure of this judgment [21] Media plurality and quality [25] Does the Act recognise only public benefits of an economic kind? [31] Salient provisions [31] Legislative history of the Act [35] The Act’s objectives [43] The authorities on public benefit [54] New Zealand practice [55] Australian practice compared [66] Public benefit: conclusions [69] Measuring benefits and detriments [82] “Likely” benefits and detriments [83] The ‘single owner risk’ [90] Objective, transparent and rigorous reasoning required [95] Adequacy of the Commission’s methodology [99] Were the Commission and the High Court wrong on the merits? [107] Quality detriments [109] Plurality [121] The paywall issue [127] The balancing exercise [133] Result [138]

[1] NZME Ltd and Stuff Ltd (formerly Fairfax New Zealand Ltd; we will call it Fairfax for convenience) publish between them New Zealand’s major daily newspapers, Sunday papers and news websites, and a large number of community newspapers. Their business model is being disrupted by digital media and their advertising revenues and subscriber numbers are in sharp decline. They say that they wish to merge to better navigate their transition to a world dominated by digital media.

[2] The Commerce Commission found that the transaction would substantially lessen competition in markets for the supply of online national news, for Sunday newspapers and advertising in them, and for community newspapers in ten centres and for advertising in them. It denied the parties an authorisation under s 67 of

the Commerce Act 1986.¹The Commission accepted that the transaction would deliver substantial and quantifiable public benefits in the form of productive efficiency gains for the merged firm, but it found that these were outweighed by losses in quality and of media “plurality”, by which is meant, broadly speaking, the number and diversity of views offered to the public. Media plurality contributes to the quality of public discourse and the health of a democracy, and so benefits the entire community.²

The Court agreed that the transaction would substantially lessen competition in relevant markets.⁴It also denied an authorisation, agreeing with the Commission that losses in media plurality and quality outweighed the transaction’s public benefits.

[4] This second appeal is brought by leave of the High Court.⁵When granting leave Dobson J identified the following questions:

(a) Was the High Court correct, as a matter of law, to find that the Commission has jurisdiction to take into account non-economic, unquantified detriments (in the form of plurality concerns) when applying the legal test for authorisation under s 67(3) of the Act?

(b) Did the High Court err in law and fact in applying the statutory test of whether the unquantified quality detriments and plurality detriments identified by the Commission were “likely”, and were attributable to the transaction?

(c) Did the High Court err in law and fact in its approach to balancing unquantifiable detriments against the net quantified benefits of

³NZME Ltd v Commerce Commission [2017] NZHC 3186 [HC judgment]. The appellants also pursued challenges on natural justice grounds in the High Court, but those challenges are not pursued in this appeal.

⁴With one exception, the market for advertising in Sunday newspapers. Nothing turns on this conclusion.

⁵Commerce Act 1986, s 97; and NZME Ltd v Commerce Commission [2018] NZHC 216.

[5] We were told that this is the first case in which a public good, such as media plurality, has tipped the scales in an authorisation application under the Act. It poses important questions. The appellants say that the Commission and High Court could not take non-economic detriments into account at all; that the High Court wrongly took into account a detriment, the risk of a single owner exploiting the merged entity for political purposes, that it adjudged unlikely to happen; and further, that the Commission and the High Court made no adequate attempt to measure loss of plurality, contrary to what this Court said in Telecom Corporation of New Zealand Ltd v Commerce Commission (AMPS-A (CA)).⁶These are ultimately questions of law. They require that we examine the purposes of the authorisation mechanism and the Act itself. Finally, the appellants say the Commission was wrong, and so was the High Court, to decide that the transaction would not result in a net benefit to the public. This is a question of fact and degree.

[6] NZME and Fairfax accept the High Court’s concise summary of their businesses and the proposed transaction.⁷In what follows we substantially adopt that summary.

[7] NZME is a media and entertainment business. It produces print publications, namely The New Zealand Herald, the Herald on Sunday and the Weekend Herald, and digital publications, including nzherald.co.nz. It owns and operates radio broadcasting businesses, including Newstalk ZB, ZM and Radio Hauraki, and it provides other e-commerce services. NZME also has ownership interests in other newspaper and publishing companies. Its businesses are located in the North Island and include six daily newspapers, two paid weekly papers, 11 online versions of newspaper websites, two lifestyle websites, 10 radio station websites, 16 other websites, six magazines, nine radio stations, and 23 community newspapers. NZME was formerly known as

[8] Fairfax is a New Zealand subsidiary of the Australian media company, Fairfax Media Ltd (Fairfax Media), the second appellant. Fairfax produces numerous print publications, operates websites, tablet and smartphone apps for stuff.co.nz, and holds additional media interests through shareholdings in other newspaper publishers. It also operates a website providing a private neighbourhood forum for neighbours to talk and share online. Fairfax’s principal newspapers include The Dominion Post, The Press and the Sunday Star-Times. It publishes nine daily newspapers (of which four are in the North Island and five in the South Island), three paid weekly papers, seven websites, 62 community publications spread throughout the country, and

[9] The High Court recorded that the transaction would involve NZME acquiring all of the shares in Fairfax. In exchange NZME would pay NZD 55 million in cash and would issue shares equal to a 41 per cent shareholding in NZME to a wholly-owned Australian subsidiary of Fairfax Media. We note in passing that the transaction has now lapsed. The appeal is not moot, however. It will determine whether the parties may renew their agreement.

[10] The High Court observed that the internet has exposed traditional media businesses to new forms and sources of competition. It identified two. The first comprises major newspapers worldwide, which provide international news and other content that is available, free or behind a paywall, to New Zealand consumers. The Court explained that direct consumer access to international news sources means there is now no New Zealand market confined to producers of international news. Competition analysis accordingly focuses on the production of news, opinions and other information concerning New Zealand.

[11] The second comprises news aggregators, which do not generate content but collate and redistribute news and information produced by others. Facebook, a social media platform, and Google, which is principally a search engine, dominate. They collate news and information constantly and tailor the content they offer according to the viewer’s algorithmically-predicted areas of interest.

[12] Digital competition for consumer attention to news and information has disrupted traditional media, sharply reducing the circulation of printed publications. Such publications sell both information of interest to readers, such as news or opinion, and space for advertisers to market their products. As circulation falls so does revenue from sales of both the printed products and the advertising displayed in them.

[13] The High Court explained that the appellants have responded by establishing their own websites to supplement their print publications and by adopting a “digital-first” news strategy, meaning that content is posted first on their digital sites (nzherald.co.nz for NZME and stuff.co.nz for Fairfax), with a sub-set of the items posted since the last print publication being selected for publication in their daily newspapers.

[14] The appellants’ revenue losses from falling circulation of printed products have been offset to some extent by revenue from their own websites or apps. On these digital platforms the appellants’ revenue streams are unbundled. Neither appellant presently uses a paywall, so readers do not pay to view. Advertising is the only source of revenue. It has been growing at a rapid rate, but as yet it is by no means sufficient to offset falling revenue from printed publications.

[15] As mentioned above, the Commission found that there would be a substantial lessening of competition in several relevant markets.⁸For this reason it denied a clearance under s 67. That decision, which was upheld by the High Court, is not challenged in this appeal.

[16] To secure an authorisation, the appellants had to satisfy the Commission that the transaction’s likely public benefits would outweigh its likely adverse effects on competition and the public interest. When assessing effects, the Commission adopted a single counterfactual that comprised two alternative scenarios predicting how the

_{merged entity would behave compared to the separate businesses. It
recognised that}

⁸CC determination, above n 1, at [375]–[378], [514]–[518], [705]–[709], [1000] and [1011].

media markets are undergoing rapid change.⁹In the first scenario (digital and print), it assumed that the businesses (whether merged or separate) would maintain roughly the same mix of digital and print publications. In the second scenario (digital and limited print), it assumed that the businesses (again whether merged or separate) would scale back their print publications and instead focus on increasing production of digital news.¹⁰

[17] The appellants identified very substantial benefits, principally taking the form of productive efficiency gains. The Commission accepted that efficiency gains in the range of around $140 million to $210 million over a five-year period were attributable to the transaction.¹¹It subtracted quantifiable detriments attributable to a paywall that it found would likely be created. The net quantifiable benefits were around

_{$47.5 million to $200 million across the same five-year
period.}¹²_{Against that, it}

identified substantial but unquantifiable detriments in the form of losses of media quality and plurality. These detriments were “of such significance” that they outweighed the quantified net benefits of the transaction, and the conclusion was “not finely balanced”.¹³

[18] The High Court adopted the same counterfactual. It agreed with the Commission that the transaction would substantially lessen competition in relevant markets, except the market for advertising in Sunday newspapers.¹⁴It did not think it likely that a paywall would be built following the transaction.¹⁵

[19] However, the High Court agreed that the transaction would likely result in substantial losses of media quality and plurality.¹⁶It found that a particular risk associated with plurality, the risk that a single owner of the merged firm would exploit

_{its control for political purposes, was remote but nonetheless relevant.
The appellants}

⁹At [143]–[144]. The appellants abandoned a challenge to the counterfactual in the High Court:

say that it was an error of law to take a remote risk into account. The Court found that the Commission was right to decline authorisation, but its decision was more emphatic. It would have declined authorisation on quality grounds even if plurality concerns were excluded.¹⁷

[20] We record that the parties now agree, following the High Court judgment, that the range of quantifiable benefits attributable to the transaction is approximately

$130 million to $200 million over five years. This excludes detriments associated with a paywall, which would be in the range of $0 to $[55–65] million over the same period.¹⁸

[21] We approach the issues in the following way. First, we examine the important concept of media plurality, on which the first question posed by the High Court depends. It will be seen that while the parties agree on a definition they differ on what it means for the appeal, the appellants saying that plurality is non-economic and hence irrelevant, and the Commission that it is substantially an in-market consideration.

[22] Secondly, we respond to the appellants’ argument that the Act’s objectives require that “benefit to the public” be read down, confining the term to benefits of an economic nature and excluding distributional considerations. This is a policy argument founded in part on the way in which the Commission and courts have interpreted the Act since it was enacted. We approach it in this way:

(a) We identify the relevant provisions and review the Act’s legislative history. This review spans a period before and after its enactment in

(b) In light of that history, we decide whether the Act’s objectives exclude distributional considerations.

[23] Thirdly, we respond to the appellants’ arguments about the Commission’s and the High Court’s methodology. We approach them in this way:

(a) We consider whether the High Court was wrong to include in the balancing exercise a detriment the likelihood of which it found remote. (That detriment is the single owner who can and does exploit the merged firm for political purposes.) We examine what the Act requires of the Commission and courts when measuring benefits and detriments, and we consider the burden and standard of “proof” in proceedings of this kind. We then decide whether the High Court was wrong to take that detriment into account.

(b) We consider whether the Commission’s methodology was otherwise insufficiently objective, transparent and rigorous. This criticism also affects the High Court decision, since it substantially followed the Commission’s methodology.

[24] Fourthly, we decide whether the High Court and the Commission erred on the merits when balancing benefits and detriments:

(b) we examine plurality detriments, to the extent not already reflected in the discussion about quality;

[25] The Commission adopted a definition of plurality that incorporates both diversity of content and influence over content. It was expressed as an objective: plurality should ensure both that media offer diversity of information, opinions and perspectives, and that no one media owner, or voice, enjoys too much influence over

public opinion and the political agenda.¹⁹The Commission recognised that plurality has both external and internal dimensions, meaning respectively diversity across and within organisations. The definition itself is uncontroversial.

[26] Quality is closely connected to plurality, but not synonymous; it includes the range and diversity of views on offer but also refers to the breadth, depth and timeliness of investigation, analysis and presentation, especially of news. The Commission considered that the transaction would cause not only a loss of range and diversity but also a “significant” reduction in the quality of news offered to readers.²⁰

[27] We have characterised plurality as a public good, which means something that a person a) can consume without diminishing the quantity available for others and b) cannot practically be excluded from consuming. Consistent with that, the Commission recognised that plurality affects all New Zealanders whether or not they consume news content. For that reason the Commission described plurality as an out of market consideration. The Commission found that it is an in-market consideration too, because it also affects those who do consume news.²¹

[28] Mr Goddard QC, for the appellants, described plurality as a non-economic factor, as did the High Court when identifying the questions listed at [4] above. We did not understand counsel to mean by this that economics has nothing to say about the subject. Rather, he argued that plurality concerns not the economic welfare of consumers but the health and resilience of New Zealand democracy, and for that reason it must be managed at a political level. This is to characterise plurality as substantially, if not entirely, an out of market consideration (to use the Commission’s terminology) that is not appropriately managed under the Act.

[29] We accept that plurality can be characterised as non-economic to the extent that its effects are felt outside relevant markets and are not easily measured in

¹⁹The definition is that of Ofcom, the United Kingdom communications regulator: Ofcom Measurement framework for media plurality: Ofcom’s advice to the Secretary of State for Culture, Media and Sport (5 November 2015) at [2.2].

price-quality terms. That is true, for example, of the effects of a single owner acquiring influence over public opinion and the political agenda. However, we share the Commission’s view that plurality is also felt in relevant markets, although its effects may be less direct than those of price and quality. So it is not entirely non-economic, a point which should be borne in mind when assessing the argument that the Commission had no business taking it into account.

[30] Plurality cannot be quantified. The Commission followed the approach of the English communications regulator, Ofcom, under which plurality is estimated using the number of different news sources on media platforms, the number of consumers using different sources across all platforms and the frequency with which they do so, the influence of news consumption on opinion, and contextual factors such as the regulatory framework for news media.²²We return to the topic of measurement at [95] below, when examining the appellants’ arguments that the Commission ought to have analysed plurality more rigorously.

[31] We confine ourselves to the immediately relevant provisions. Section 66 provides that the Commission may grant a clearance for a business acquisition if satisfied that it will not have the effect of substantially lessening competition in a market.²³We will call this an SLC for brevity. Under s 67 it may authorise a transaction that would cause an SLC:

(1) A person who proposes to acquire assets of a business or shares may give the Commission a notice seeking an authorisation for the acquisition.

²³Section 47 of the Commerce Act prohibits a business acquisition where it would substantially lessen competition in a market, and the words “business” and “acquire” are defined in s 2 of the Act. The phrase “lessening of competition” is not defined, except to say that lessening

(3) Within 60 working days after the date of registration of the notice, or such longer period as the Commission and the person who gave the notice agree, the Commission shall—

(a) if it is satisfied that the acquisition will not have, or would not be likely to have, the effect of substantially lessening competition in a market, by notice in writing to the person by or on whose behalf the notice was given, give a clearance for the acquisition; or

(b) if it is satisfied that the acquisition will result, or will be likely to result, in such a benefit to the public that it should be permitted, by notice in writing to the person by or on whose behalf the notice was given, grant an authorisation for the acquisition; or

(c) if it is not satisfied as to the matters referred to in paragraph (a) or paragraph (b), by notice in writing to the person by or on whose behalf the notice was given, decline to give a clearance or grant an authorisation for the acquisition.

[32] The term “benefit to the public” is not defined, but s 3A provides that efficiencies are mandatory relevant considerations when deciding whether such benefit is likely to result from conduct:

Where the Commission is required under this Act to determine whether or not, or the extent to which, conduct will result, or will be likely to result, in a benefit to the public, the Commission shall have regard to any efficiencies that the Commission considers will result, or will be likely to result, from that conduct.

The purpose of this Act is to promote competition in markets for the long-term benefit of consumers within New Zealand.

[34] “[C]ompetition” means workable or effective competition,²⁵and “market”

refers to New Zealand goods and services markets the boundaries of which are gauged by applying, as matters of fact and common sense, the concept of substitutability.²⁶

²⁴Section 1A was inserted, on 26 May 2001, by s 4 of the Commerce Amendment Act 2001.

As originally enacted the Commerce Act had a long title which provided that it was “An Act to promote competition in markets within New Zealand”, but that long title was repealed by the Commerce Amendment Act. We discuss this legislative history in further detail below at [47].

The concept of “workable” or “effective” competition is traced, via Re Queensland

United States Attorney-General’s National Committee to Study the Antitrust Laws:²⁸

The basic characteristic of effective competition in the economic sense is that no one seller, and no group of sellers acting in concert, has the power to choose its level of profits by giving less and charging more. Where there is workable competition, rival sellers, whether existing competitors or new or potential entrants into the field, would keep this power in check by offering or threatening to offer effective inducements ...

[35] As is well-known, the Act is based on the Trade Practices Act 1974 (Cth), which in turn is derived from the Sherman Antitrust Act (US) and Clayton Antitrust Act (US).²⁹The Australasian statutes reoriented competition law in both jurisdictions. Predecessor legislation — in New Zealand’s case, the Commerce Act 1975 — had generally followed an English model that³⁰did not create absolute prohibitions but required that restrictive trade practices be registered and approved if they were in the public interest or restrained if they were not.³¹

[36] The Sherman Act prohibits outright every “contract, combination ... or conspiracy, in restraint of trade or commerce”.³²It was traditionally used to restrain mergers to monopoly,³³but mergers are now addressed under the Clayton Act, which

_{prohibits mergers that may substantially lessen competition or
tend to create a}

²⁸Stanley N Barnes and S Chesterfield Oppenheim Report of the United States Attorney-General’s National Committee to Study the Antitrust Laws (United States Department of Justice, March 1955) at 320. The concept is further traceable to the work of the Harvard economist Edward S Mason: see the discussion in Alan J Meese “Debunking the Purchaser Welfare

Account of Section 2 of the Sherman Act: How Harvard Brought Us a Total Welfare Standard

²⁹Sherman Antitrust Act 15 USC § 1–7 (1890); and Clayton Antitrust Act 15 USC § 12–27, 29

USC § 52–53 (1914). See Mark N Berry “New Zealand Antitrust: Some Reflections on the First

³¹Department of Trade and Industry Commerce Bill 1985: A Background to the Bill and An Outline to its Provisions (August 1985) at 7–8.

³³Northern Securities Co v United States [1904] USSC 64; 193 US 197 (1904); Standard Oil Co of New Jersey v

United States [2004] USSC 1; 221 US 1 (1911) [Standard Oil]; and United States v First National Bank & Trust

monopoly.³⁴The standard under both statutes is usually considered to be the same.³⁵

Courts have long held that while some practices are per se unlawful, others must be evaluated under a “rule of reason”.³⁶The rule of reason focuses on the impact on competition of the restraint in question, and does not permit consideration of other social or economic goals.³⁷The legislation contains no authorisation mechanism, which distinguishes it from its New Zealand counterpart.

(a) The impetus for reform was the desire to direct the economy toward more competition and greater efficiency.³⁸It was also considered important to harmonise New Zealand and Australian law.³⁹

(b) Officials recognised that small economies must sometimes tolerate levels of market concentration that would be considered anticompetitive elsewhere if they are to overcome diseconomies of scale and the costs of competing in distant markets.⁴⁰

(c) The adoption of a new model engendered resistance in the business community, which feared substantive uncertainty and unpredictability in the law. This concern was associated with transition from a prescriptive, rules-based regime, in which many restrictive trade

³⁵United States v Rockford Memorial Corp 898 F 2d 1278 (7th Cir) (1990) at 1281–1282.

³⁷National Society of Professional Engineers v United States [1978] USSC 67; 435 US 679 (1978) at 681 and 688;

National Collegiate Athletic Association v Board of Regents of the University of Oklahoma

⁴⁰Trade Practices Act Review Committee Report to the Minister for Business and Consumer Affairs (August 1976) at [11.11]. Berry notes that New Zealand markets are characterised by high concentration levels, high entry barriers and inefficient levels of production: Berry, above

to a behavioural regime in which such practices might be adjudged illegal after the fact and under broadly-expressed standards.

[38] An authorisation mechanism was incorporated to address these concerns and ensure that other public policy objectives might be balanced against competition.⁴¹

It is not a mere incident of the legislation but a central feature. It applies to both business acquisitions and restrictive trade practices, allowing firms to obtain approval in advance for transactions that would cause an SLC but offer some offsetting public benefit.⁴²It draws from English, American and Australian tradition, combining what a contemporary commentator described (among other things) as a self-regulating statute incorporating public benefit analysis before specialist tribunals.⁴³

[39] As originally implemented in Australia, an authorisation required that the arrangement concerned would result in a “substantial” public benefit that “would not otherwise be available”.⁴⁴This standard was thought too restrictive, and following the report of the Swanson Committee the legislation was amended in 1977 to require simply that the arrangement would be likely to result in a public benefit outweighing any lessening of competition.⁴⁵The Committee explained that:

11.11 ... It seems to the Committee that it is generally accepted in the Australian environment, and in regard to the size of the market and the size of economic units operating in that market in at least some industries, that there will be cases in which the community accepts that public benefit or public interest considerations should justify the existence of restrictions on competition.

the Committee emphasised that competition must remain the legislation’s primary objective but authorisation should be available where it could be shown that the

⁴³Bruce G Donald and JD Heydon Trade Practices Law (The Law Book Co, Sydney, 1978) at 11.

11.15 ... However, if in a given case it can be shown that public benefits, i.e. not merely benefits to the parties to the restrictive conduct, are available, and that those benefits outweigh the benefits to the public foregone by the absence or restriction of competition, then that conduct should be permitted to continue.

It will be seen that the Committee distinguished benefits to the public, who would experience the transaction’s loss of benefits, from private benefits to the parties.

[41] In the meantime, the Trade Practices Tribunal (the Australian Tribunal) had delivered its famous decision in QCMA, in which it explained the concept of public benefit in the broadest terms:⁴⁶

One question that arises is whether by the public is meant the consuming public. One submission to us was that, in the context of the objectives of the Act, we should direct our attention to that part of the public concerned with the use or consumption of flour in the Queensland market.

However this is not what the Australian Act says; and we cannot but think that the choice of a wider expression was deliberate, as pointing to some wider conception of the public interest, though no doubt the interests of the public as purchasers, consumers or users must fall within it and bulk large.

Another question raised is whether public benefit must be contrasted with private benefit. Can a benefit to some of the private parties to the merger — for example the shareholders of Barnes — be claimed as a public benefit? Must a benefit which accrues to the private parties be “passed on” to members of the wider community before it can be considered? The commission has expressed its view ... that the test requires “benefits to the public and not merely to the applicant or some other limited group”. While agreeing with this statement as far as it goes, we would not wish to rule out of consideration any argument coming within the widest possible conception of public benefit. This we see as anything of value to the community generally, any contribution to the aims pursued by the society including as one of its principal elements (in the context of trade practices legislation) the achievement of the economic goals of efficiency and progress.

[42] As noted, harmonisation with Australian law was among the motivations for the Act. Its authorisation test was materially identical to that in the Trade Practices

⁴⁶QCMA, above n 27, at 182–183. The names of the relevant bodies in Australia have changed names over the years. We refer to the body currently known as the Australian Competition Tribunal as “the Australian Tribunal” and the body currently known as the Australian Competition and Consumer Commission as “the Australian Commission”.

[43] Mr Goddard argued that the Commission must, and usually does, adopt a “total welfare” approach, under which a transaction’s benefits and detriments are valued equally regardless of who receives or incurs them. This argument rests on the proposition that the Act’s objectives exclude distributional considerations, by which is meant a preference for one class of beneficiary over another, or for widely dispersed benefits over those retained by the parties to a given transaction.

[44] Total welfare takes aggregate economic efficiency to be the law’s objective and seeks to maximise it. So producer surplus is treated as a welfare gain because the surplus is available for use elsewhere in the economy,⁴⁸and the distribution of gains and losses is discounted because distribution has no bearing on efficiency. Total welfare draws no distinction between gains from a transaction that are “private” in the sense that they accrue to the suppliers who are parties to it and those that are “public” in the sense that they accrue to a class of consumers.

[45] There is a longstanding normative debate about whether total welfare ought to be the objective of antitrust law.⁴⁹The alternative is the consumer welfare approach, which when narrowly defined looks to the share of gains that goes to consumers and excludes gains to producers in the same market.⁵⁰The rationale for admitting distributive considerations is sometimes expressed in fairness terms — consumers

_{should receive public benefits associated with restrictive trade
practices because they}

⁴⁷Compare as at 1986 Commerce Act, s 67(5); and Trade Practices Act, s 90(6). We observe that the Department of Trade and Industry referred to Australian case law when explaining what was meant by “a substantial lessening of competition” in practice: Department of Trade and Industry, above n 31, at 12.

⁴⁸We define producer and consumer surplus following Oliver Williamson: see Oliver E Williamson “Economies as an Antitrust Defense: the Welfare Tradeoffs” (1968) 58 Am Econ Rev 18.

⁴⁹The literature is surveyed in Herbert Hovenkamp “Implementing Antitrust’s Welfare Goals”

⁵⁰We recognise that consumer welfare has sometimes been defined in terms synonymous with total welfare, notably by Robert Bork, but the definition used here is orthodox and serves our purpose of distinguishing a policy that admits distributional considerations from one that does not. See Robert H Bork The Antitrust Paradox: A Policy At War With Itself (The Free Press,

experience the anticompetitive effects.⁵¹But the underlying explanation is simply that the law admits non-efficiency values because it reflects norms of the community that it serves.⁵²

[46] Mr Goddard is correct that the Commission normally pursues a total welfare approach. It stated in its 2013 Authorisation Guidelines (2013 Guidelines),⁵³and in its former Guidelines to the Analysis of Public Benefits and Detriments (1997 Guidelines) that distributional effects are generally irrelevant because the distribution of gains and losses does not affect efficiency.⁵⁴

[47] This emphasis on total welfare emerged in New Zealand in the 1990s. The legislation was amended in 1990 to insert s 3A; as noted above, it requires that when examining public benefits the Commission must “have regard to” efficiencies likely to result from the conduct in question.⁵⁵Hampton and Scott explain that the

1990 amendment followed a series of Commission decisions discounting benefits that were not passed to consumers.⁵⁶

[48] In 1994 the Commission first issued its Guidelines to the Analysis of Public Benefits and Detriments.⁵⁷The guidelines followed a 1992 interdepartmental review and proposed legislation — never enacted — that would have required that when assessing public benefit the Commission must treat efficiency as the primary consideration and ignore distributional effects. These norms were reflected in the guidelines.⁵⁸

[49] In 2001 the purpose statement was amended to specify in s 1A, as noted above, that the Act’s purpose is “to promote competition in markets for the long-term benefit

at 4 and 16–27. See also Richard A Posner The Economics of Justice (Harvard University Press, Cambridge, 1981).

⁵³Commerce Commission Authorisation Guidelines (July 2013) at [53] [2013 Guidelines].

⁵⁴Commerce Commission Guidelines to the Analysis of Public Benefits and Detriments

⁵⁶Lindsay Hampton and Paul G Scott Guide to Competition Law (LexisNexis, Wellington, 2013)

of consumers within New Zealand”.⁵⁹Professor Ahdar suggested that the Bill stemmed from the desire of the government of the day to emphasise consumer welfare out of concern that it had been neglected in Commission and judicial decisions.⁶⁰

The parliamentary record supports that view. Speaking to the Bill, the then Minister of Commerce, the Hon Paul Swain, said:⁶¹

[The purpose statement] makes clear that competition is not an end in itself, but a means to promote the welfare of New Zealanders. Consumers are given special mention as they are the ultimate beneficiaries of competition. However, the welfare of all New Zealanders will continue to be important. This is an important aspect of the change, and brings us closer into line with the equivalent legislation in Australia. The focus on competition in the purpose statement also does not preclude wider public benefit issues being taking into account where appropriate. It simply clarifies that there should be a presumption in favour of competition, and competition must prevail unless the efficiencies of other public benefits are shown to exceed the detriments from the lessening of competition.

[50] The Commerce Committee expressed concern that by referring to the “long term” the amendment might lead the Commission to overemphasise dynamic efficiency at the expense of more immediate benefits, but it found the term acceptable on the basis that welfare meant consumer welfare.⁶²A reference to facilitating the “efficient operation of markets” was also removed by the Committee.⁶³In the House the Chairman of the Committee, David Cunliffe, stated that:⁶⁴

Members will forgive me if I provide a little context about an academic debate that has raged for some years between those who support an efficiency test ... the so-called Chicago School, and those who seek a welfare-based test, the

New Zealand Parliament supports a welfare-based, Harvard School approach that puts the interest of consumers first. However, we have taken due account

efficiency play in the market, and for that reason it is the long-term interests of consumers that appear in the new purpose statement.

_{the High Court referred to the 2001 amendment and rejected a submission
that under}

⁶⁰Rex Ahdar “Consumers, redistribution of income and the purpose of competition law” (2002)

s 1A the only relevant public benefits are those that flow directly to consumers.⁶⁵

Rather, the Court endorsed the Commission’s by then established total welfare approach and accepted that the distributional effects of a merger are irrelevant:

[241] We are satisfied that the introduction of s 1A should not disturb the Commission’s established practice of treating as neutral any wealth transfers between New Zealand consumers and producers. Determinations of authorisation applications under the Act are properly concerned with balancing any efficiency detriments associated with breaches of the statutory competition standard, against any efficiency gains that may result from the business acquisition or contractual arrangement in question. It is the balancing of these real resource impacts on the economy that best serves the long-term interests of consumers. The inclusion of ad hoc wealth transfers, which are not losses to society, would distort the efficiency assessment by assuming additional economic harm to the public of New Zealand. In any event, consumers might well be the ultimate beneficiaries.

The Commission has since taken the view that Air New Zealand binds it to generally follow a total welfare approach when assessing public benefits.⁶⁶

[52] It is debateable whether the High Court should be taken to have held that the Commission must follow an objective that the Commission had adopted of its own volition and the Court did not itself evaluate against the legislative history. The Court inquired rather whether the new s 1A meant that only direct consumer benefits counted.

[53] In any event, this Court has never held that the Act compels a total welfare approach. It has long recognised that the Act pursues economic welfare and uses the language of economics; in Tru Tone Ltd v Festival Retail Marketing Ltd (Tru Tone), Richardson J referred to the original purpose statement and said that the Act “is based on the premise that society’s resources are best allocated in a competitive market where rivalry between firms ensures maximum efficiency in the use of resources.”⁶⁷

To make this general point, though, is not to nominate total welfare as the Act’s objective, still less to question the statutory presumption that competition is the

_{mechanism through which efficiency gains will be delivered to consumers.
There is}

⁶⁶CC determination, above n 1, at [1063], citing Air New Zealand, above n 65.

no reason to suppose that when Richardson J spoke of efficiency he had the distinction between total and consumer welfare in mind.⁶⁸

[54] We turn to examine how public benefit has been interpreted in practice. It will be seen that the term has never been limited to economic or in market considerations.

[55] We begin with Telecom Corporation of New Zealand Ltd v Commerce Commission (AMPS-A (HC)), in which Telecom sought clearance or authorisation for the acquisition of certain radio frequencies.⁶⁹It failed before the Commission and the High Court but succeeded on further appeal.⁷⁰In the High Court judgment is found what is still the leading New Zealand discussion of public benefit. The Court noted the Australian approach,⁷¹citing Re Rural Traders Co-op (WA) Ltd:⁷²

It is undesirable to attempt to fix in advance the limits of what the concept of “benefit to the public” encompasses or to exclude, in advance, from its ambit any contribution to the legitimate aims pursued by society. In the context of Trade Practices legislation, the encouragement of competition and competitive behaviour within relevant markets and the achievement of the economic goals of efficiency and progress will commonly be paramount. The fact that a particular result of a proposed acquisition may be neutral in so far as such behaviour or such economic goals are concerned does not, however, preclude it from being a relevant (and, conceivably, a determining) factor in the assessment of public benefit ...

[56] The Court recognised that in New Zealand s 3A (which had no equivalent in the Australian legislation) compelled regard for any efficiencies, but held that the

_{section left for case-by-case judgement the weight that should be
assigned to them:}⁷³

⁶⁸We recognise that there have been other changes to the Act since 1986 which we have not discussed; some of these are surveyed in Commerce Commission v Woolworths Ltd [2008] NZCA 276, (2008) 12 TCLR 194 at [65]–[72] [Woolworths]. However, it was not suggested that these differences are material in this case. Nor was it suggested that the differences in wording between s 61 of the Act and s 67 are material.

⁶⁹Telecom Corporation of New Zealand Ltd v Commerce Commission (1991) 4 TCLR 473 (HC)

⁷²Re Rural Traders Co-op (WA) Ltd (1979) ATPR 40-110 (Trade Practices Tribunal) at 18,123 [Rural Traders].

The new section compels regard to any efficiencies that will likely result from the acquisition, but what weight will be given to them, either in relation to other potential elements of public benefit or in relation to public detriment, must be a matter of judgment in the particular case. We bear in mind that efficiency has three dimensions commonly referred to as allocative efficiency, production efficiency, and dynamic efficiency. Efficiency considerations, positive and negative, are relevant in the assessment of both benefit and detriment but clearly do not exhaust society’s interest in the business conduct the subject of the Commerce Act.

Accordingly, the Australian approach to public benefit was applicable in New Zealand. The Court envisaged that distributional concerns could be of significant or even determinative weight.⁷⁴

[57] Addressing Telecom’s claim that efficiency gains need not be passed on to consumers, the Court elaborated on the concept of public benefit:⁷⁵

Thus the distinction customarily drawn is between public benefit and private benefit (or private motivation), between social value and private value, when there is market failure in an extended sense. Public benefit has been found in the very nature of the conduct for which authorisation is sought (e.g. “professional” standards of work and acceptance of fiduciary responsibility) ... in the use of a non-market process to organise an activity (e.g. cooperative enterprise) ... and in dimensions of market performance that include static and dynamic efficiency but go beyond them (e.g. the protection of children from some kinds of advertisement)...

[58] The Court went on to accept that public benefit includes efficiency gains; further, that may be true even if little or none of the benefit is passed to consumers. It cautioned, however, that in such a case it may be necessary to inquire whether the gains are durable or may be frittered away through slackness and rent-seeking activities:⁷⁶

It is important in assessing the magnitude of these benefits, in our view, to focus not so much on their immediate distribution as on their durability. Where efficiency gains are not “passed on to the consumer”, there may be a question that requires inquiry as to whether, or to what extent, the lack of competitive pressure will allow the productivity gains to be frittered away in slackness and wasteful “rent-seeking” activities.

[59] This Court allowed the appeal, but it did not take a different view of the concept of public benefit.⁷⁷In a well-known passage about methodology to which we must return, Richardson J noted that the relevant benefits and detriments were almost entirely efficiency gains and no issue arose about quantifying and weighing “disparate public interest considerations”.⁷⁸

[60] For its part, the Commission has consistently accepted that an applicant for an authorisation may invoke public benefits of any kind. It is sufficient for our purposes to cite the 2013 Guidelines applied in this case.⁷⁹They state, invoking the judgments in AMPS-A (HC), Air New Zealand, and Godfrey Hirst NZ Ltd v Commerce Commission (Godfrey Hirst (No 1)),⁸⁰that a public benefit is anything of value to the community generally, that the term includes but is not limited to efficiencies, and that the Commission recognises as a public benefit any gain to the public of New Zealand that would result from the proposed transaction regardless of the market in which that benefit occurs or who it benefits.⁸¹Benefits are assessed net of any disbenefits associated with them.⁸²

[61] We pause here to note that, as the Commission submitted, if in pursuit of economic efficiency the Act excludes non-economic considerations from s 67, then it must follow that the Commission was wrong in other cases to consider the following benefits: reduced pollution,⁸³health benefits of breastfeeding,⁸⁴safer handling of hazardous substances,⁸⁵reduced stigma for psychiatric patients,⁸⁶and social effects of

⁷⁷AMPS-A (CA), above n 6. We return to this Court’s reasoning in the case at [96] below.

⁸⁰AMPS-A (HC), above n 69, at 528; Air New Zealand, above n 65, at [319]; and Godfrey Hirst

⁸³Nelson City Council and Tasman District Council [2017] NZCC 6 at [111]–[112].

⁸⁵Refrigerant Licence Trust Board CC Decision No 735, 25 November 2011 at [77]–[81].

⁸⁶Midland Regional Health Authority and Health Waikato Ltd CC Decision No 275, 1 August

[62] The 2013 Guidelines approach detriments in an asymmetric manner, counting only “anti-competitive detriments that arise in the market(s) where we find a lessening of competition”.⁸⁸This suggests that any detriments of a non-economic or non-market nature are excluded. For this approach the guidelines cite Godfrey Hirst (No 1) and observations of Wilson J in this Court in New Zealand Bus Ltd v Commerce Commission.⁸⁹Ultimately, however, these authorities rest on the Commission’s own practice. In Godfrey Hirst (No 1),⁹⁰the Court said that this practice had been sanctioned by the High Court and this Court in AMPS-A.⁹¹We do not agree. The practice was not in issue in AMPS-A and it is drawing too long a bow to read an endorsement into the judgments.

[63] This Court returned to the topic of public benefit in its 2016 judgment in Godfrey Hirst NZ Ltd v Commerce Commission (Godfrey Hirst (No 2)), stating after reference to the legislative history and the judgment in AMPS-A (HC) that the concept includes non-economic benefits and detriments:⁹²

The legislative history shows Parliament’s intention to leave this category [authorisation of business acquisitions] open for the Commission’s expert assessment ... While the benefits are not confined to the particular market, the Commission and the courts must take account of the values or public interest at stake in that particular market when determining benefits or detriments to the wider public, especially when economic activity can have negative consequences for others and many social goods and services are now distributed through market mechanisms ...

[64] In its determination in this case the Commission relied on media plurality, which it characterised in part as an out of market detriment. It did not find this inconsistent with the 2013 Guidelines, which it described as general and necessarily

_{non-exhaustive.}⁹³_{In any event, the
guidelines had not been updated following}

⁸⁹Godfrey Hirst (No 1), above n 80, at [72]; and New Zealand Bus Ltd v Commerce Commission

Godfrey Hirst (No 2). Citing QCMA, the Commission reasoned that were it to ignore out of market detriments it might act against the public interest:

80. ... there would be a category of negative consequences of a proposed merger that we are required to ignore. For example, if a merger was to have an adverse impact on the environment, employment, privacy interests, or other constituents of social welfare which fall outside of the market(s) in which competition has been lessened or else are not efficiency related, we would be required to ignore those factors in assessing whether there was such a public benefit that the transaction should be permitted.

81. The implication of the Applicants’ approach is that we might have to authorise a merger that in our assessment was not in the public interest. That is, if we considered that there was a negative consequence that outweighed the positive aspects of a proposed merger, we might still have to authorise depending on where those negative impacts were felt.

82. It is difficult to discern a rationale for Parliament wanting the Commission to consider only some of the detriments to the public of a merger and to disregard others, and we would only adopt such an approach if compelled to do so by the statutory language, or judicial interpretation of the Act.

83. In our view, the language of the Act does not compel the interpretation that some negative consequences count for the purposes of the analysis and some do not. To the contrary, we consider that our statutory task is to determine whether the merger will be likely to result “in such a benefit to the public that it should be permitted” notwithstanding that the merger has the effect or likely effect of substantially lessening competition. Whether there is such a ‘benefit to the public’ cannot be considered divorced from outcomes that harm the public, whether or not they are economic or market-oriented in nature.

96. ... we would not be giving effect to section 67 if we disregard a material source of negative consequences. Usually the approach set out in our Authorisation Guidelines will be sufficient to capture the dynamics involved in a proposed acquisition, but the plurality issue has caused us to carefully address where the relevant negative consequences should be included in our analysis.

[66] The Australian Commission and Tribunal have not gone so far as the Commission in the pursuit of total welfare. The legislation has never contained a provision corresponding to s 3A and distributional considerations have not been discounted. That said, economic efficiency has been treated as the dominant consideration and it has been accepted in an authorisation setting that efficiencies need

not be passed on to consumers. In its 2004 decision in Qantas Airways, the Tribunal stated that:⁹⁴

In our view, the objective and statutory language of the Act, as well as precedent, support the use of a form of the total welfare standard as the most appropriate standard for identifying and assessing public benefit. We say a “form of” the total welfare standard because, as the passage cited from Re Howard Smith shows, whilst the Tribunal does not require that efficiencies generated by a merger or set of arrangements necessarily be passed on to consumers, it may be that, in some circumstances, gains that flow through only to a limited number of members in the community will carry less weight.

[67] In the decision cited in that passage, Re Howard Smith Industries Pty Ltd, the Tribunal had said, following QCMA, that:⁹⁵

The Tribunal has to determine what constitutes “the public” in order to assess whether there is likely to be a substantial benefit to the public from a proposed merger. It is not simply the public as consumers. If a merger is likely to result in the achievement of economies of scale and a considerable saving in the cost of supplying a good or service this might well constitute a substantial benefit to the public, even though the cost saving is not passed on to the consumers in the form of lower prices. Nevertheless, if such a merger benefited only a small number of shareholders of the applicant corporations through higher profits and dividends, this might be given less weight by the Tribunal, because the benefits are not being spread widely among members of the community generally.

[68] In Australian Competition and Consumer Commission v Australian Competition Tribunal (ACCC) a Full Federal Court recently held that the current legislation, the Competition and Consumer Act 2010, permitted but did not require that the Tribunal use the modified total welfare approach:⁹⁶

What the Tribunal is required to do is to assess the benefit to the public resulting, or likely to result from the merger, and to do so in “all the circumstances”. Having done that it is to decide whether that benefit warrants the grant of the authorisation. That is the statutory charter it has. Whilst certain matters may be inferred from the language of s 95AZH(1) as we have outlined above, it is nevertheless a broadly expressed provision. It is

⁹⁴Qantas Airways Ltd [2005] ACompT 9, (2005) ATPR 42-065 at [185] (the decision was released in 2004 but reasons were not delivered until 2005). See also The Hospital Benefit Fund of Western Australia Inc v ACCC (1997) ATPR 41-569 (FCA) at 43,904–43,905; and Rural Traders, above n 72, at 18,123.

a legitimate way for the Tribunal to proceed in assessing the benefit to the public resulting from a merger to adopt the modified total welfare standard identified in Qantas. But the tail must not wag the dog, and it is not that standard that s 95AZH imposes. Consequently, the reasons of the Tribunal are to be measured for their legal efficacy against only the text of s 95AZH(1) and what can be reasonably implied from it, not the test identified in Qantas.

[69] Having surveyed the legislation, its history, and the authorities, we can now answer the question whether “benefit to the public”, as used in s 67, excludes non-economic or out of market considerations. The High Court held that the Commission has jurisdiction to consider a loss of plurality resulting from the transaction.⁹⁷It accepted that out of market considerations may seldom arise and the Commission may be susceptible to challenge on the merits if it takes them into account, but as a matter of construction Parliament cannot have intended to exclude such considerations where a proposed transaction is likely to cause them.⁹⁸We agree generally with these conclusions.

[70] Drawing together several threads of the discussion above, we make several points about the legislation. We confine ourselves to observations of relevance to this appeal.

[71] The first concerns efficiency. We accept that the reference to “long-term benefit” in the purpose statement recognises that the Act values efficiency. It also admits the possibility that the pursuit of efficiency may not benefit consumers in the short term. Section 3A further presumes that efficiency gains may benefit the public and prescribes that regard must be had to them when assessing public benefit. But as a matter of construction the Act treats efficiency as a subset of public benefit; put another way, efficiency is a mandatory consideration but others are not excluded. To paraphrase AMPS-A (HC), efficiency matters but it does not exhaust society’s interest in the transaction.

[72] Second, “benefit to the public” is not defined, although it sets the standard for authorisation, which for reasons given at [37]–[40] above, is an integral feature of the

legislative scheme. So the Commission is to decide what benefits the public in the circumstances of any given case.

[73] The 2001 amendments confirm rather than detract from what was said about public benefit in the passages from AMPS-A (HC) that we have cited above.⁹⁹

In particular, the Act is not exclusively concerned with efficiency but rather allows it to be balanced alongside other public benefits that may include anything of importance to the community as a whole. Nothing in the legislation requires that public detriments be defined less comprehensively. The identification and weighting of public benefits, including efficiency gains, and detriments is left to the Commission’s judgement.¹⁰⁰

[74] Third, from the courts’ perspective this analysis is not novel. It is consistent with the authorities as we have explained them. It does not discount efficiency. On the contrary, this Court recognised in Tru Tone that workable and effective competition is prized because it delivers efficiency gains to consumers.¹⁰¹To modify, by reference to the 2001 amendments, what the Court said there, the Act rests on the premise that consumers benefit from competitive markets in which rivalry among firms maximises efficiency. Efficiency is not confined to the efficient economy-wide allocation of resources; it includes productive and dynamic efficiency in relevant markets. In this case public detriments prevailed over efficiency benefits, so it bears emphasis that the converse may be true, and often is. The legislative history confirms Parliament’s intention that authorisation should allow efficiency considerations to prevail over an SLC where the transaction concerned will sufficiently benefit consumers over time or in other ways.

[75] Fourth, it follows that the legislation permits the modified total welfare approach discussed at [66]–[67] above. We should not be taken to say, however, that the Commission must follow the modified total welfare approach in practice.

_{The Commission is equipped to develop policies and guidelines within the
statutory}

¹⁰⁰Godfrey Hirst (No 2), above n 92, at [22] and [35]. We record that Mr Every-Palmer did not ask the Court to hold that the Commission may value dispersed benefits more. He recognised that the issue had not previously arisen in this case. It is however necessary to respond to the appellants’ argument that distributional considerations, which include the dispersal of benefits, may not be taken into account.

framework, and it is responsible for doing so. The courts are responsible for assessing the Commission’s decisions, in law and on their merits, against that framework and what can reasonably be found in it.

[76] This judgment establishes that it would be an error to exclude a public benefit or detriment on the ground that the Act is concerned with efficiency alone. For the avoidance of doubt, it does not follow that the Commission would err were it to discard or discount any given public interest consideration in the circumstances of any given case. There may be good reasons for doing so. By way of example only, the case may turn on first-order effects in relevant markets, or further inquiry may be thought impractical or unreasonably costly. More than that we need not say in this case.

[77] The appellants also contend that had Parliament wanted to regulate plurality in the news media directly it would have enacted sector-specific legislation. Mr Goddard submitted that it would be surprising if the Commission were permitted to act as “the regulator of everything”, for the role would take it well outside its statutory framework and institutional competence and make the authorisation process much less predictable. Plurality is the concern not of economic welfare but of social policy. The Commission may not fill a perceived gap in media regulation, as it effectively sought to do by reasoning that existing regulation in insufficient to protect media plurality post-transaction.

[78] The High Court rejected these submissions, rightly so in our opinion. The Act manifestly is not designed to regulate the media, but the Commission did not try to do so.¹⁰²Its jurisdiction over mergers extends to media businesses, and in this case it engaged with plurality and quality because and to the extent that the transaction would affect those matters.¹⁰³It did not assume regulatory oversight of the media by considering whether loss of plurality can safely be left to other regulatory institutions

[79] We recognise that the Commission’s internal expertise is unlikely to extend to media plurality, but we agree with the High Court that its processes allow it to obtain expert assistance on the non-market consequences of a merger.¹⁰⁴That being so, this consideration does not affect our conclusion that the Commission has jurisdiction to consider plurality.

[80] We accept that non-economic detriments may complicate merger analysis and introduce an additional element of unpredictability, which is undesirable. We have noted above that fear of uncertainty was a significant consideration when the legislation was enacted. Where the law employs standards, the community looks to courts and tribunals to lend content and predictability when administering it.¹⁰⁵

However, this point does not go to jurisdiction either. Some measure of uncertainty is inherent in the legislative decision to permit authorisation on widely-defined public benefit grounds. We reiterate that in adding s 3A to the Act, Parliament made efficiencies a mandatory consideration but it did not exclude others or say anything about the weight to be assigned to them.

[81] We conclude that the High Court was correct to find that the Commission might lawfully take into account non-economic or out of market detriments when deciding under s 67(3) whether to authorise the transaction.

[82] We turn to the question whether the High Court erred in applying the statutory test for identifying and balancing relevant benefits and detriments.

[83] It is settled law that any given benefit or detriment is relevant only if it is likely in the sense that there is a real and substantial risk that it will happen.¹⁰⁶

_{The appellants say that the High Court erred by attaching some weight
to an effect}

¹⁰⁵See the discussion in Richard A Posner Economic Analysis of Law (8th ed, Aspen, New York,

that it recognised as unlikely, namely the risk that someone will exploit ownership of the merged entity for political purposes:¹⁰⁷

[275] By comparison with the extent to which major media organisations in other jurisdictions identify with political parties, the level of political influence over media owners in New Zealand is relatively slight. Given that, and given the initial listed company ownership structure, we accept the risk of a dominant ownership position of the merged entity being exploited for political purposes is somewhat remote. Of course a single ownership structure that creates even a remote risk of such serious adverse change is deserving of some weight.

Mr Every-Palmer QC also contended that “likely” is a “practical filter” rather than a statutory requirement.¹⁰⁸If this latter argument is correct, the Commission would not err if it counted a remote effect when balancing benefits and detriments under s 67(3). This leads us to survey briefly what the authorities have to say on this point.

[85] To set this discussion in context, we observe that the Commission makes authorisation decisions under conditions of uncertainty. The effects concerned are those judged likely to result in the future from a transaction that is in prospect. The Commission compares them with those of a counterfactual — what will happen in future without the transaction — that need not be the status quo. As this Court put it in Commerce Commission v Woolworths Ltd, both factual and counterfactual are “necessarily incapable of accurate assessment”.¹⁰⁹As the Court also observed in Woolworths, it matters who bears the burden of this uncertainty.¹¹⁰

(a) An effect is “likely” if there is a “real and substantial risk” or “real chance” that it will occur. It must be more than a mere possibility but need not be more likely than not.¹¹¹The likely existence of such a risk

¹⁰⁸The Commission cited AMPS-A (CA), above n 6, at 446 per Richardson J for this, but the word

¹¹²QCMA, above n 27, at 183; adopted in AMPS-A (HC), above n 69, at 512–513.

the High Court recorded counsels’ agreement that, in general terms, a real and substantial risk might be one that had at least a 30 per cent prospect.¹¹³We mention that not to suggest precision but to demonstrate that the Commission need not be satisfied that a given effect is more likely than any alternative; it must follow that more than one alternative may qualify for consideration.

(b) The Commission has inquisitorial powers and may consider information from many sources,¹¹⁴but it need not continue its inquiries until it has satisfied itself that the relevant effect is or is not likely. Rather, it may rest on the information provided by the applicant. It must also refuse an authorisation unless satisfied that the transaction should be authorised.¹¹⁵For these reasons the applicant bears a practical burden of persuasion.¹¹⁶

To say that the Commission is “satisfied” is simply to say that it has made up its mind on all the material before it.¹¹⁷

(d) What the Commission must be satisfied of is that the acquisition will, or will likely, result in such a benefit to the public that it should be permitted. This is a “balance sheet” exercise, as it was put in QCMA, in which the transaction’s likely benefits are balanced against its likely detriments, the most important detriments normally being those causing

¹¹³Woolworths Ltd v Commerce Commission [2007] NZHC 1351; (2008) 8 NZBLC 102,128 (HC) at [113]. For the avoidance of doubt, we express no view on the controversial question, which arose in Woolworths, whether this means the Commission must consider multiple counterfactuals where there said to be more than one, or may use the one it thinks more likely.

¹¹⁴Woolworths, above n 68, at [101]–[102]. The Court there recognised that it is not entirely apt to

[87] We take the opportunity to explain what was said about the balancing exercise in Woolworths. It was common ground between counsel there that the Commission must ultimately be satisfied on the balance of probabilities.¹¹⁹The Court itself used that phrase,¹²⁰while recognising that it is inapt insofar as it suggests a standard of proof. It is inapt because it may suggest that clearance and authorisation proceedings are like normal civil proceedings in which a court must decide causation. For a court causation is usually a question of historical fact. It finds the facts on the balance of probabilities, then treats those facts as certain when gauging the consequences.¹²¹

[88] By contrast, the Commission assesses benefits and detriments that may be caused in a future state of affairs. Those effects need not be proved on the balance of probabilities, and the weight assigned to a given effect may reflect not only its extent or impact but also its likelihood. To decide where the balance lies, then, is to compare one future state of affairs — or an hypothesis, to use French J’s term in Australian Gas Light Co v Australian Competition and Consumer Commission — in which benefits outweigh detriments with another in which they do not.¹²²The Commission may not authorise the transaction unless satisfied that the one state of affairs is more likely than the other.¹²³

[89] Appeals from the Commission are general in nature, meaning that the appellate court assesses the record against the same standard and forms its own view. An appellant bears the burden of satisfying the court that the decision below was wrong, but if the court reaches a different conclusion on the merits then the Commission, or the intermediate appellate court as the case may be, was wrong “in the only sense that matters”.¹²⁴There is room for deference to the Commission’s

_{or High Court’s advantages of process or expertise.}¹²⁵_{We record that we do not defer}

1609–1610 per Stuart-Smith LJ. Of course the court may assess damages on a lost chance basis if loss depends on a future event the happening of which is contingent on someone else’s actions.

¹²²Australian Gas Light Co v Australian Competition and Consumer Commission [2003] FCA

¹²³John Land and Leela Cejnar “Counterfactual Analysis in Merger Reviews” (2012) 25 NZULR

in this appeal, which turns on questions that we find ourselves able to answer on the material before us.

[90] We turn to the appellants’ argument that the High Court erred by taking into account a risk that is less than likely, namely the risk that someone will exploit ownership of the merged entity for political purposes.

[91] We have set out the Court’s conclusion above. It characterised the risk as “somewhat remote” but deserving of “some weight” because the effect would be serious if it came to pass. Mr Every-Palmer argued that by using this language the Court was merely locating the risk at the lower end of a spectrum of likely effects. (We observe that the Commission itself did not characterise the risk in that way in its determination, stating rather that there is little to prevent a change in shareholding.¹²⁶) Counsel pointed out that the Court concluded that “the importance of the likely loss of plurality” and prospect of reduced quality outweighed the transaction’s benefits, so characterising loss of plurality as likely.¹²⁷He emphasised, citing Godfrey Hirst (No 2), that it is “false scientism” to insist on an express calculation of probabilities.¹²⁸

[92] In our opinion the High Court erred to the extent that it took into account what it considered a remote risk that a single owner would exploit the merged entity for political purposes. We agree that even a remote risk of this kind is a matter of public concern. Post-transaction, NZME’s substantial presence in relevant markets would create a vulnerability that ought to worry policymakers. But unless the risk is thought “likely” it should not enter the balance under s 67(3). Before us, the Commission did not suggest that it is likely. NZME, which is to be the acquirer, is listed on the NZX and we were given to understand that it is widely held. A subsidiary of Fairfax Media Ltd, which is listed on the ASX, would own 41 per cent of NZME’s shares following the transaction. This we take to be a controlling interest in the merged entity.

_{But Fairfax Media also appears to be widely held. So it seems that
absent a substantial}

change in shareholding no one person would be in a position to impose their own political agenda on NZME.

[93] We do accept that the error has no practical significance. Because the possibility was remote it was evidently assigned little weight, as one would expect. The balancing exercise turned primarily on quality, not plurality, and the Court’s decision was not finely balanced. We also note that the Commission did not make the same error.

[94] Further, it does not follow that loss of plurality ought to have been discounted, nor adverse effects on democracy. We return to the significance of plurality below at [121].

[95] The appellants cite Richardson J’s judgment in AMPS-A (CA) for the proposition that analysis must be transparent, structured and rigorous.¹²⁹They accept that not all benefits and detriments can be quantified, but argue that the High Court erred by concluding that because loss of plurality is unquantifiable it might rely on intuitive judgement.¹³⁰The Commission and the Court ought to have employed an analytical technique such as break-even analysis, which would require that they estimate the costs of remedying the losses of plurality and quality and measure those costs against the transaction’s benefits.¹³¹The cost of remedy would be measured by estimating the cost of journalist and editorial staff needed to replace those lost to the transaction. Had that been done, it would have been apparent that the transaction would yield a clear net public benefit.

...the desirability of quantifying benefits and detriments where and to the extent that it is feasible to do so. The commission encourages applicants to quantify anticipated public benefits. In this case certain major efficiency gains were quantified for Telecom at some $75 million. While both the commission

¹³¹Cass R Sunstein Simpler: The Future of Government (Simon & Schuster, New York, 2014)

and the Court did not accept elements in that quantification, both bodies considered that there would be significant efficiency gains if Telecom had management rights over both AMPS-A and AMPS-B. In those circumstances there is in my view a responsibility on a regulatory body to attempt so far as possible to quantify detriments and benefits rather than rely on a purely intuitive judgment to justify a conclusion that detriments in fact exceed quantified benefits.

[97] In that case, as we have already noted, wider public benefit considerations were not in issue. The benefits and detriments were almost all efficiency gains and losses, and it was these that Richardson J said should be quantified so far as possible. Speaking more generally, Richardson J cautioned against speculation and intuition, urging that “the value judgment should be as informed by practical evidence as possible”.¹³³

[98] We accept Mr Goddard’s submission that Richardson J’s underlying point, which we affirm, was that the Commission’s reasoning ought to be as objective, rigorous and transparent as feasible. The Act uses economic concepts — markets, market power, substantial lessening of competition — that permit informed evaluation.¹³⁴We accept that the Commission ought to measure gains and losses where it is sufficiently feasible and instructive to do so, and otherwise ought to ensure that the value judgement required of it is informed by practical evidence and analysis. If it fails to meet this standard then it risks reversal on the merits on appeal. But quantification can convey an impression of precision that is quite misleading.¹³⁵

analysis, and some efficiencies are measurable, it is an error to reason that the authorisation process is concerned only with, or values most, those things that can be measured. If made, that error would introduce a bias in favour of measurable efficiency considerations that, as we have explained, is not found in the statutory

¹³⁶At [35]–[38]. Compare Michael S Jacobs “An Essay on the Normative Foundations of Antitrust

[99] For the most part, the appellants do not challenge the methodology adopted by the Commission and the High Court. It produced, as already noted, substantial estimated efficiency gains. They focus rather on the treatment of plurality and quality detriments.

[100] The appellants first argued that plurality ought to have been unbundled so that each of its elements was separately analysed, for two reasons: plurality has two dimensions (voice and political influence) that are not equally likely to suffer in the factual, and the Commission’s analysis of plurality was so abstract as to be meaningless. Next, it was said that the Commission and the High Court failed to identify the nature and extent of plurality detriments, as they must do in order to assign weight to them. Finally, it was said that the High Court erred by taking an intuitive approach to the balancing exercise, for even unquantifiable benefits may be subjected to cost-benefit analysis. Mr Goddard instanced break-even analysis, under which judgment is informed by closely analysing the nature and extent of the unquantifiable effect.¹³⁷He also argued that the detriment may be measured by its costs of avoidance; in this case, the cost of journalists and editorial staff positions lost in the transaction, and that cost is very much less than the efficiency gains.

[101] We preface what we have to say by observing that while some methodological practices, such as the use of counterfactuals, have become settled over time, the Act itself does not prescribe a methodology for identifying and evaluating benefits and detriments, including the SLC. It leaves the Commission to choose a methodology that seems best suited to the circumstances. In ACCC the Full Federal Court for this reason rejected an argument that the Tribunal must assign explicit and lesser weightings to public benefits that are not dispersed among consumers.¹³⁸

[102] The argument that the dimensions of plurality ought to have been unbundled derived much of its force from the proposition that the High Court erred by giving weight to the single owner risk. We have held, however, that the error was of little

practical significance. The Commission’s further response is that plurality is not susceptible to unbundling and it did what it could to gauge extent and weight; it used hard data where possible and consulted widely and employed the Ofcom measures mentioned above. We agree. The dimensions of plurality overlap. Our attention has not been drawn to a methodology that might allow them to be separated and separately measured.

[103] In AMPS-A (CA) Richardson J said that decisions should not be made in a “purely intuitive” way.¹³⁹It does not follow that the High Court erred when it stated, after noting the absence of any quantitative evidence of “the impact of a reduction in quality causing a reduction in the number of visits to an appellant’s online site”, that the competing views on quality were “therefore intuitive”.¹⁴⁰The statement was made when addressing an argument that the appellants’ need for advertising would constrain their capacity to reduce quality in news. The Court surveyed the practical evidence about that before concluding that an intuitive judgment was required. It did not say that in the absence of quantitative evidence all judgments are intuitive, or that its decision need not be informed by the evidence. The question before it turned on an effect — reduction in news consumption in response to a loss of quality — that is incapable of measurement and we think the Court was saying no more than that.

[104] Mr Goddard did not argue that the Commission must employ break-even analysis, using the technique rather to illustrate that unquantifiable detriments can be assessed in a rigorous way. We take the point so far as it goes. But it has not been shown that any alternative methodology would make any difference here. We consider that the Commission’s process was sufficiently disciplined and its conclusions generally robust.

[105] We turn lastly to the argument that the costs of avoiding harm to plurality and quality are the proper measure of these detriments. The appellants say that if it would cost less than $130 to $200 million to avoid these detriments then the harm they cause cannot sensibly be ascribed a higher value. Mr Goddard submitted that the High Court

misunderstood this argument when it responded that the benefits do not represent a sum available to New Zealand to avoid the detriments.¹⁴¹

[106] We do not accept that plurality and quality detriments can be measured by the cost of employing journalists and editors to replace those whose positions are lost in the merged firm. The harm caused may be much greater than that. If it can be remedied at a cost which is by comparison modest, then the question naturally arises whether anyone would supply the funding in order to avert the harm. There being no realistic possibility of new entry, the funding would have to come from the public purse. The Commission and the High Court did not find that a sufficiently real possibility, and as the High Court pointed out in the passage just mentioned, the appellants were not offering to share their private efficiency gains to that end. We do not suggest that they ought to do so. But neither do we agree that the High Court missed the point.

[107] We introduce this section by remarking on the factual and counterfactual. Printed newspapers are following a declining trajectory that the transaction might slow for a time but could not arrest. The appellants hope that the transaction will provide “additional runway” for adaptation over the next five years. So there is a substantial degree of uncertainty about what Mr Goddard called the delta, the difference between factual and counterfactual, and that must affect the accuracy of any assessment of benefits and detriments.

[108] The appellants rely on the delta’s uncertainty to challenge the Commission’s and the High Court’s assessments of plurality and quality detriments. Mr Goddard emphasised that they did not gauge how much of these effects would be felt within the next five years. However, the point cuts both ways, lending force to the Commission’s argument that the appellants are wrong to treat the efficiency gains as “bankable”. The gains were not detailed in a merger implementation plan. Rather, they were derived from a PriceWaterhouseCoopers report which it seems the appellants

_{themselves described during the Commission’s investigation
as an “academic}

exercise”. The Commission took advice from an accounting firm, BDO, but it did not audit the estimates. Rather, it worked with them, typically stating that it accepted the ranges offered by PwC as “reasonable estimates”.¹⁴²We take the same approach.

[109] We turn to one benefit, a productive efficiency gain through savings in editorial staff numbers, which came to the forefront in argument. It did so because counsel treated journalists’ time, and hence journalist numbers, as a rough proxy for quality in news.

[110] The appellants say, relying on the PwC report, that the transaction would result in the firm shedding some 13 per cent, of their editorial staff, in which we include journalists. This is seemingly a modest percentage. Mr Goddard argued that it is so because journalists generate the content that allows the appellants to sell attention and so drives revenue from both advertising and subscriptions.

[111] In argument before us, the Commission pointed out that the 13 per cent figure excludes some editorial staff counted by PwC, that the PwC report suggested the merged entity might shed a materially higher percentage, 15–20 per cent, and that even this figure is under-inclusive because some editorial staff were categorised differently.

[112] In its determination the Commission made a further point about the PwC estimates, noting that for confidentiality reasons PwC had not been able to share their thinking with NZME and Fairfax management. It illustrated the resulting reliability concern by noting that the report adopted the dubious assumption that general news reporters would be rationalised only where the parties overlap geographically.¹⁴³

[113] The appellants respond by identifying a number of structural and behavioural factors that offset their incentive to reduce costs by shedding editorial staff.

_{They contend that any losses would be confined to coverage of arcane
topics and a}

modest reduction in the number of perspectives offered on topics of secondary public interest. Among other points:

(a) Audience attention is their only source of revenue in both subscription and advertising markets. They maximise it by using their editorial staff to supply diverse views and quality journalism, so appealing to readers who have diverse interests and many competing claims on their attention, and in that way they maximise revenue from both advertisers (the only source of revenue on “free” digital media such as Stuff) and subscribers.

(b) There is evidence, which the Commission accepted, that consumers respond to changes in perceived quality.¹⁴⁴

(c) The appellants experience competition from other media, such as free-to-air television and radio, as well as news aggregators such as Facebook. Competition for attention is not confined to content generated by journalists; it includes reader-generated content.

(e) Editors adhere to codes of conduct and ethics, and the appellants are regulated by the Media Council.

[114] These arguments did not carry the day below. The Commission’s detailed reasoning is dispersed through sections covering effects in differing markets and scenarios, and some of it is combined with discussion of plurality, but these passages fairly summarise its conclusions:

887. In response to a reduction in competition and, therefore, less competitive pressure to attract audiences, we consider that the merged entity would be likely to have less incentive to invest in editorial resources. We consider that this would be likely to include not only the number of journalists hired, but that the merged entity could also reduce the amount it invests in journalism and/or invests in journalist training. Such decisions may

not affect the volume and variety of stories produced, but could reduce the quality of the article, for example, a journalist may be given less training or less resources for in-depth analysis, such as travel expenses.

894. We consider that the reductions of editorial staff are likely to have a significant effect on the quality of online New Zealand news that would be produced with the merger. We are of the view that NZME and Fairfax currently compete to invest in and deploy their editorial resources in a manner which they consider will best attract readers of online New Zealand news. This involves editors allocating resources to produce news that is varied in terms of the topics that are covered and the perspectives and viewpoints discussed.

897. Given these different perspectives, we consider that a rationalisation of editorial roles due to a reduction of competition is likely to result in a degradation of quality of news produced by the merged entity, as there would be a reduction in the variety of editorial decision-making. In particular, there is likely to be a concentration of editorial opinions around what topics to cover and what stance, angle or perspective to write on particular issues and what stories are given prominence.

[76] ... We consider it an inevitable part of the rationale for the merger that efficiencies pursued would include scaling back journalistic and editorial resources, as well as managerial resources. One necessary consequence of doing so would be a reduction in quality. We are therefore satisfied that there would be a material reduction in the scope of topics reported upon, and the diversity of views expressed within them. That would constitute a reduction in quality.

[77] We do not consider it a sufficient answer that the merged business would be economically incentivised to provide coverage on what readers want. Dissatisfaction with a reduction in the scope of what is published is unlikely to be conveyed either promptly or effectively enough to influence management decisions on what may appear to be duplicated resources, but which readers would prefer were retained. So too with the level of more arcane topics that are likely not to be covered post-merger. We agree with the Commission that the merged entity could reduce the quality of its output in numerous respects that would not be observed by consumers in any concerted way, so that the reduction in quality could be effected without producing any compelling demand for restoration of the qualities that existed in the competitive era.

[78] We also agree with the Commission that competition between news producers stimulates coverage by each of a wider range of topics than would be covered in a non-competitive environment. If one publication covers a

news story, then its rival will be incentivised to cover that story as well, albeit with a different perspective.

[79] We are mindful that senior editorial staff currently employed by the appellants provided assurances of their intentions to maintain variety and diversity of news content. We have no reason to doubt the integrity of the editorial personnel and respect their commitment. The commercial reality is that the extent to which they can achieve variety and diversity of news content will be dictated by management decisions.

[116] We share these views. In our opinion, in the factual the merged entity will have a powerful incentive to save costs by shedding journalists and editorial staff, with a corresponding impact on quality. The incentive is powerful because the firm will be under pressure to extend the “runway” so far as it can. We think that reductions would likely exceed those estimated by the appellants in argument before us, with a correspondingly substantial effect on quality. It is difficult to gauge, but we think that staff reductions at the upper end of the PwC range would cause quality effects of a substantial magnitude. We recognise that larger reductions would also increase efficiency gains.

[117] By way of explanation, we accept that the merged firm will continue to experience incentives to supply quality journalism. But we make four points:

(a) We agree with the Commission and the High Court, for the reasons they gave, that competition between the appellants is a major driver of quality at present. The Commission found that each of the appellants considers the other its major rival.¹⁴⁶One witness described how they monitor and mimic one another’s behaviour. We discuss below an example, the conduct of NZME when considering whether to establish a paywall. The transaction would eliminate all of that.

Such would be the merged firm’s share of news content generated and consumed in newspaper and online markets that competitive pressure

platforms, notably free to air television and radio, that are not yet full substitutes for the appellants’ newspapers, and some of them do not enjoy the trust that the appellants have established among consumers.¹⁴⁸

(d) Consumers do respond to perceived quality losses by reducing consumption, but the emphasis is on “perceived”. We agree with the Commission and the High Court that consumers do not find it easy to gauge quality and may adapt to quality changes without noticing them.¹⁴⁹

[118] We accept that increased funding for public media is likely. We were supplied with evidence of a Budget commitment to increase funding by $15 million to increase media contribution to an informed democracy, through the establishment of a Ministerial Advisory Group. We reject the Commission’s submission that it is speculative for us to have regard to those commitments, but we do nonetheless accept that publicly-owned media organisations (or policy intervention more generally) are unlikely to address our concerns about loss of quality. Existing public media organisations do not operate in most of the relevant markets. Consistent with that, the Chief Executive and Editor in Chief of Radio New Zealand, Mr Paul Thompson, explained that (in his view) Radio New Zealand was “never a direct competitor” to the appellants, and saw its relationship with the appellants as a more collaborative one.

[119] We accept, as did the High Court, that editors adhere to codes of conduct and ethics, but we share its view, and that of the Commission, that neither this nor Press Council decision-making can protect against quality reductions. The point is well explained by Dr Peter Thompson, a senior lecturer in media studies at

¹⁴⁹CC determination, above n 1, at [733]; and HC judgment, above n 3, at [288].

[T]he competitive pressure to get that story first but also to get that story right is I think very important. And if there’s a hollowing out of the news budget, I mean it’s all very well the editors saying “Oh we’ve got integrity, we protect our independence”, that may not be [the] budget choice for them to make. And they don’t know, they can’t possibly say what the future will hold under the new company structure...

[120] It follows that in our opinion the transaction would be very likely to result in quality reductions and these would likely be of a substantial nature.

[121] When assessing the transaction’s effect on plurality care must be taken to not double-count quality effects that have already been recognised when predicting how the merged firm would behave in the factual. Plurality effects will be felt generally throughout the community.

[122] Plurality matters in this case because the transaction would affect it in a substantial way. New Zealand markets are already highly concentrated by international standards and the appellants account for a very large share of the production and distribution of news. The transaction would reduce to four the number of major news providers, and the merged entity would employ many more journalists and editorial staff than Television New Zealand, Mediaworks and Radio New Zealand combined.¹⁵⁰It would account for almost 90 per cent of daily newspaper circulation (with an extensive regional presence), control the two largest New Zealand online news suppliers and the two major Sunday papers, and through NZME’s stable of nine radio stations account for a substantial share of the radio market.

[123] We accepted at [92] above that the High Court was wrong to take into account the risk of a single owner exploiting the merged entity for political purposes. However, as Mr Every-Palmer submitted, concentrated ownership is not a prerequisite to a media organisation setting an editorial agenda or campaigning for a cause. Even if its owners are politically agnostic, the merged firm is more likely to follow a uniform

competition provided under the factual is a better way of securing a diversity of views than the discretionary internal plurality that would exist in the merged entity.¹⁵¹

[124] Diversity of views is also intrinsically valuable.¹⁵²To cite John Stuart Mill:¹⁵³

Why is it, then, that there is on the whole a preponderance among mankind of rational opinions and rational conduct? ... it is owing to a quality of the human mind, the source of everything respectable in man either as an intellectual or as a moral being, namely that his errors are corrigible. He is capable of rectifying his mistakes, by discussion and experience. Not by experience alone. There must be discussion, to show how experience is to be interpreted. Wrong opinions and practices gradually yield to fact and argument; but facts and arguments, to produce any effect on the mind, must be brought before it.

[125] This “marketplace of ideas” justification recognises that the community benefits from a variety of different perspectives and from allowing people to participate in the community by expressing those views.¹⁵⁴There were a significant number of submitters to the Commission, both with backgrounds in journalism and from the wider community, who emphasised the importance of a diversity of views to the functioning of the media and society. The point is well explained by

I think probably the best way to describe it is that the mainstream media outlets in this country kind of operate ... a broader news ecosystem and we sort of, feed and prompt each other into covering stories, you know if there was only one of us there’d be a ... lot fewer stories being generated ... you know diversity would probably fall away simply because there’s not as many people out there making calls, banging doors, you know checking that sort of thing.

[126] Our findings on quality apply to plurality. We have rejected the appellants’ arguments that there will be only minor reductions in editorial staff and that any losses will be offset by other media. We find it very likely that there will be a substantial

_{loss of plurality in the factual that will not be experienced in the
counterfactual.}

¹⁵³John Stuart Mill On Liberty (first published 1859; republished Batoche Books, Ontario, 2001)

¹⁵⁴Attorney-General v Times Newspapers Ltd [1974] AC 273 (HL) at 315 per Lord Simon.

[127] Our decision does not turn on this issue. However, we have formed a view on

[128] By way of introduction, the Commission attributed several specific losses of allocative efficiency to the transaction. These it quantified and deducted from quantified efficiency gains. The losses would arise in the community newspaper advertising and Sunday newspaper advertising markets, from increases in the cover and subscription price for Sunday newspapers and from the implementation of a paywall for online news.¹⁵⁵The High Court found that a paywall is not likely following the transaction and the Commission contends that it was wrong.¹⁵⁶On this issue turns between $0 and $[12–16] million in efficiency losses per year and a minor consequential change in the value assigned to wealth transfers from consumers to non-New Zealand shareholders in the merged entity.¹⁵⁷

[129] The Commission contends that a paywall is likely, highlighting a recent announcement by NZME that it intends to launch a paywall for “premium content” on its website for online news, nzherald.co.nz. It also highlights a report from Fairfax that paid digital subscriptions for three of its Australian media, The Sydney Morning Herald, The Age and The Australian Financial Review, have increased in number for four consecutive years.

[130] Competition between the appellants is presently a major impediment to a paywall, as the High Court recognised.¹⁵⁸The evidence before the Commission was

¹⁵⁵CC determination, above n 1, at [1239], [1244]–[1246] and [1266]–[1268].

].¹⁵⁹The Court nonetheless found a general paywall unlikely, because the loss of advertising revenue would exceed subscriptions.¹⁶⁰

[131] We prefer the view that some form of paywall, most likely a restrictive content paywall like that recently announced by NZME, would be implemented by the merged entity. The Commission emphasises confidential information provided by NZME that detailed a possible subscription model it could implement. [

].¹⁶¹Assuming the merged entity employed a similar model and priced subscriptions to offset the loss of advertising revenues just detailed, a restricted-content paywall is a commercially feasible option for the merged entity, and there is a real chance one could be implemented following the transaction given the significant reduction in competitive pressure that flows from the merged entity controlling both nzherald.co.nz and stuff.co.nz.

[132] Accordingly, we reinstate the Commission’s findings on quantified detriments as they relate to the paywall. This means that the quantified net benefits attributable to the transaction is approximately $[65–75] million to $200 million.

[133] We may state our conclusions shortly. We accept the range of likely quantified efficiency flowing from the transaction, while noting that they are mere estimates the reliability of which is difficult to assess and recording that the range is somewhat reduced by reinstatement of the paywall.

[134] We consider that quality and plurality detriments are very likely to result from the transaction. We have examined quality effects first because we consider them more immediate and find them easier to gauge. Counsel used journalist and editorial

_{staff numbers as a proxy for quality, as noted. We consider that the
transaction will}

cause a substantial reduction in numbers. We agree with the High Court that quality detriments are very likely and substantial. We consider that they are sufficient in themselves to outweigh the transaction’s benefits.

[135] Additional plurality losses are also very likely and substantial. We agree with the High Court and the Commission that plurality is a characteristic of media markets that is vitally important to the community. We also agree with the High Court that the loss of plurality attributable to the transaction would very likely be irreparable.

[136] These conclusions flow from the appellants’ powerful position in New Zealand markets that are already highly concentrated, the attenuation of competitive pressure to maintain quality of journalism post-transaction, and the incentives that the merged firm would experience to cut costs.

[137] In the result, we find that detriments clearly outweigh benefits, and not by a small margin. It follows that authorisation was properly declined below.

[138] As noted, this is a general appeal, not confined to the questions that the High Court identified when granting leave. It is nonetheless appropriate to answer those questions:

a) Was the High Court correct, as a matter of law, to find that the Commission had jurisdiction to take into account non-economic, unquantified detriments (in the form of plurality losses) when applying the legal test for authorisation under s 67(3) of the Commerce Act 1986?

b) Did the High Court err in law and fact when applying the statutory test of whether the unquantified detriments and plurality detriments identified by the NZCC were “likely”, and were attributable to the transaction?

The High Court erred to the extent that it assigned weight to a risk that it considered remote, namely the risk that a single owner would exploit the merged entity for political purposes. It did not otherwise err.

c) Did the High Court err in fact and law in its approach to balancing unquantifiable detriments against the net qualified benefits of the transaction?

[139] The third of these answers disposes of the appeal on the merits. The appeal is accordingly dismissed.

[140] The appellants are jointly and severally liable to pay the respondent costs for a complex appeal on a band B basis and usual disbursements. We certify for second counsel.

Court of Appeal of New Zealand

NZME Limited v Commerce Commission [2018] NZCA 389; [2018] 3 NZLR 715 (26 September 2018)