Transcript
Check against delivery.
Introduction
Competition law supports the integrity of markets, so businesses have the incentive to operate more efficiently, price competitively and offer better products and services to their customers. This benefits consumers, the economy, and the community more broadly, with lower prices, greater innovation, and higher quality goods and services.
As the national regulator responsible for enforcing compliance with the Competition and Consumer Act 2010 (CCA), in the 2021-22 financial year alone, we took action to address exclusive dealing in ice-creams, roof tiling bid-rigging, cartels in the pharmaceutical sector, alleged misuse of market power in card payments, in addition to actions in the power tools, money exchange, building supplies markets, and we resolved a misuse of market power investigation against Telstra with a court enforceable undertaking. Our focus is deterring anti-competitive and cartel conduct in Australia – regardless of the industry.
Our competition enforcement is supported, in no minor way, by our market inquiry functions. We started examining digital platform markets in 2017. The strength of our market inquiries is in their open and consultative processes in parallel with the use of compulsory information gathering powers.
Throughout three separate inquiries, we have looked at:
- the impact of digital platforms on news and journalism content and advertising services markets in the Digital Platforms Inquiry, published in 2019
- the supply of digital advertising technology services through the Ad Tech Inquiry, published in 2021
- a range of other digital platform services including search, app stores, online private messaging services, online retail marketplaces, social media services, as well as our current report into the expanding ecosystems of digital platforms. All of this work has been under our ongoing Digital Platform Services Inquiry which ends in 2025. As part of this Inquiry, we also recommended regulatory reform to address the array of harms we have identified, in our September 2022 report.
I joined the ACCC as the Enforcement Commissioner 14 months ago after 24 years as a partner in leading commercial law firms. Over the last 14 months I have had the benefit of the considerable breadth and depth of expertise built up over 6 years through our digital platforms work and I have developed a direct and deep appreciation of the challenges of using our existing Part IV provisions to address competition harms in this space.
In the year of the 30th Anniversary of the Hilmer Inquiry, I have come to appreciate the significant limitations of our economy wide regime of ex post competition prohibitions enforced through Federal Court proceedings in addressing harms in digital platform markets.
This is the journey I want to take you on today.
Starting position
Firstly, I will out myself as a competition purist.
I have significant faith in our competition law regime in Australia and consider our track record speaks for itself in the ability of an agnostic competition law to address most harms we have seen across the economy.
The elegance of the general competition rules advocated for by the Hilmer Committee in the 1993 report is that they are flexible, and allow for a consistent approach to competition policy across industries. As the Hilmer Report states: there are compelling efficiency and equity arguments for ensuring that competitive conduct rules of the kind proposed in this Part are applied uniformly and universally throughout the economy. The expectation was that the Hilmer reforms would future-proof our competition regime so that the laws were well placed and generalist enough that they could be applied to existing and future industries without the need for significant changes.
However, we’re not in 1993 anymore. Three years after the Hilmer Report, Larry Page and Sergey Brin, who were students at Stanford at the time, launched a search algorithm first known as BackRub. In 1998, it was officially launched as Google.
Our generalist competition law regime in Part IV of the CCA has demonstrated an ability to cross industry lines. However, the rise of digital platforms and a global digital economy have brought with them significant change. Many industries have changed beyond recognition in the 30 years since the release of the report.
Disruptive technological advances have changed not only how we conduct competition investigations, but also introduced new theories of harm and questions about the tools in our analytical toolkit – all in contexts where the markets directly interact with consumers on a global scale.
Digital platforms
I don’t think anyone could argue with the notion that digital platforms have significantly changed our society, bringing huge benefits to Australian consumers and businesses. They have changed how we work, study, communicate, shop, entertain ourselves and do business, never more so than during the pandemic.
The benefits of the digitalisation of the economy are enormous and have allowed for significant benefits to consumers and businesses in terms of reducing search costs, the fostering of social relationships and increased the ability for small businesses to locate customers.
However, the benefits of the digital revolution will only promote efficiency and consumer welfare if digital platform markets are effectively competitive. The conditions for effective competition are well understood. As the 1955 report of the US Attorney-General’s National Committee to Study the Antitrust Laws observed:
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 or charging more. Where there is workable competition, rival sellers, whether existing competitors or potential entrants... would keep this power in check.
We have found that global digital platforms are engaging in conduct that may have the effect of limiting competition, distorting innovation, and entrenching market power. This conduct includes potentially anti-competitive self-preferencing, tying, exclusivity agreements, erecting barriers to switching, denying interoperability, and withholding access to important hardware, software, and data inputs.
We have also identified significant issues related to a lack of transparency, and the ability of digital platforms with market power to unilaterally impose and degrade the terms on which services are provided to consumers and business users.
Not all digital platforms have the power to ‘give less or charge more’. The characteristics that some global digital platforms have in combination, that create the significant risk of substantial competition harms, are:
- Digital platform services are ubiquitous for everyday consumers, with one recent study stating that Australians spend an average of 5 hours and 51 minutes daily online. From checking the weather, public transport timetables, responding to group chats, banking, searching for local restaurants, and accessing local and community information in times of crisis and disasters, digital platform services constantly surround us.
- The dependence of downstream markets. Digital platforms often act as distribution channels and vital intermediaries in value chains. This ranges across everything from the criticality of social media and search advertising for small business to the ad-tech supply chain where Google has a 90-100% share of impressions for publisher ad-server services and 80-90% share of advertiser ad server impressions. Its publisher ad-server is considered a ‘must have’ service by those in the advertising industry.
- Business model of ecosystems. Digital platform firms each have slightly different strategies, but a key trend is them broadening their ecosystems and moving into new areas and technologies. We are currently considering potential competition and consumer issues from the expanding ecosystems of digital platform providers in Australia with a report due to the Treasurer in September. We have observed digital platform service providers leveraging existing positions in certain products and services, such as unique access to valuable user data and significant user bases, to offer new and improved products and services in related markets. For example, Meta using its social media user base to expand into online shopping and dating. While ‘conglomerate’ business strategies and competition agency concerns with the leveraging of power across markets are not new, ‘ecosystem’ business models are generally characterised by the use of user data in ways that are not transparent and are difficult to replicate by potential competitors.
- Finally, while the services offered by different types of digital platform businesses may differ, many digital platform services share particular economic characteristics which tend towards concentration, making it difficult for smaller or new platforms to develop services to compete. These include network effects, extreme economies of scale and scope, consumer inertia, switching costs and, access to, and use of, vast amounts of data. For example, Google’s Search index contains hundreds of billions of web pages and is over 100,000,000 gigabytes in size. Not only do these characteristics tend to lead to concentration, but they also enhance the ability of and incentive for incumbent platforms to raise barriers to entry and expansion for rivals or would be rivals, potentially through anti-competitive conduct, in a race to entrench their market power.
As a result of these characteristics, what I will call ‘significant’ digital platforms are likely to be unconstrained when engaging in conduct that is likely to lessen innovation and reduce efficiency. The competitive harm from such conduct may only be obvious, in retrospect. Three cases demonstrate this point, one determined and two pending:
- On 27 June 2017, the European Commission (EC) fined Google €2.42 billion for abusing its market dominance as a search engine by giving illegal advantage to its own comparison shopping service, Google Shopping. The EC found that from 2008, Google changed its strategy to systematically give prominent placement to its own comparison shopping service, and demote rival comparison shopping services in its search results. The EC found these practices amounted to an abuse of Google’s dominance in the general internet search market by restricting competition in comparison shopping markets. The case took more than 7 years after the EC opened a formal investigation. In November 2021, the General Court of the European Union largely dismissed Google’s appeal.
- In January 2023, the US Department of Justice (DoJ) brought proceedings against Google, alleging that Google monopolised digital advertising technologies through a course of anticompetitive and exclusionary conduct spanning 15 years. The DoJ alleges that Google’s conduct included acquiring competitors to obtain control over key digital advertising tools, forcing the adoption of Google’s tools across services by users, distorting auction competition and auction manipulation.
- In December 2020, the US Federal Trade Commission (US FTC) brought proceedings against Facebook (now Meta), alleging that the company is illegally maintaining its personal social networking monopoly through a years-long course of anti-competitive conduct. The court documents, filed by the US FTC refer to Meta’s alleged systematic long term strategy to illegally maintain its monopoly position.
Enforcement challenges – the practicality of enforcement
I have significant faith in our robust competition law regime to address conduct across industries. Enforcement of these laws can address a range of conduct through broad, flexible provisions, and allows for nuanced assessment of the actual effects of discrete conduct based on the specific facts of a case brought before the court.
Whether we seek to enforce sections 45, 46 or 47 of the CCA as the enforcement agency we must bring specific pleadings and facts before the court to prove the purpose or effect of substantially lessening competition in a market. For section 46, we must establish that specific unilateral conduct was engaged in with the relevant purpose or had the relevant effect, to be tested against what would likely have occurred without the conduct. The necessity to lead probative evidence means that any such proceedings are most likely brought after the event.
The strength of Hilmer’s advocacy for general, industry-agnostic competition law regime is not diminished but we consider it is not enough to address the harms from anti-competitive conduct by some digital platforms. Indeed, we are behind the European Union, the United Kingdom and Germany in recognising that a new regulatory regime is needed, and I’ll return to this later.
Why are significant digital platforms different, from an enforcement perspective?
Competition law enforcement is almost always difficult. Conducting investigations and instituting proceedings on the basis of a substantial lessening of competition analysis will often be a long and arduous journey. What we have seen internationally is that proceedings against significant digital platforms for unilateral conduct take years to complete, long after the harm caused is entrenched. This is especially true where conduct affects competition at multiple levels across a supply chain or across multiple markets, such as in the European Commission’s Google Android and Google Shopping cases. Of course, the fact that proceedings for unilateral conduct may take a long time is not unique to significant digital platform cases. As I represented the respondents in the 10 year section 46 and 45 case brought by the ACCC against Cement Australia, I can attest to that. However, I can also say that cases involving unilateral conduct against significant digital platforms bring a level of complexity, scale, and forensic challenge that makes the Cement Australia case look like a doddle.
The fact that competition investigations are long and complex is not of itself why regulatory reform is necessary. It is also that the time that such cases take is relevant to the potential efficacy of remedies in dynamic markets and technology industries that are characterised by extreme economies of scale and scope, and network effects. Firstly, when technological innovation happens at a pace, we don’t see what may have emerged, but for the conduct. Relatedly, the remedy that may be fashioned may come too late to restore competitive dynamics. The EU ‘choice screen remedy’ may prove the proof of this proposition. There are questions about the effectiveness of the EU choice screen implemented to date, with the ACCC observing that it has had a limited impact on market concentration and consumer reach. I would suggest the delay between the identification of the competitive harm and the institution of the remedy has been significant in the apparent failure of the choice screen remedy.
It is also significant that we do not have divestiture as a remedy for contravention of section 46 as the US does.
The technologies used by digital platforms also present real challenges for the detection and proof of potentially anti-competitive conduct. Where opaque algorithms and machine learning drive market outcomes, both the detection and the characterisation of conduct is more challenging, compared to, for example, the manifest and observable conduct of Cement Australia in securing exclusive access to available fly ash capacity.
Cases against global digital platforms present other unique challenges. Where global firms are involved, there are difficulties in obtaining evidence from decision makers and business documents to evidence conduct and its purpose. These difficulties are compounded by complicated corporate structures and diffuse and interacting digital systems, which make it difficult, without intimate knowledge of how these companies are structured and how platforms operate, to identify, locate and compel the production of records and algorithms.
Digital platform markets also present unique analytical challenges for enforcement agencies and courts. Traditional approaches need to be adapted to reflect the unique conditions found in digital markets including the prevalence of multi-sided markets, strong network effects, the importance of data as an input (and potential source of market power) and low or zero monetary price business models.
There are inherent theoretical, jurisprudential and forensic challenges in answering the question can this significant digital platform offer less or charge more and get away with it?
How does one undertake the hypothetical monopolist, or SSNIP, test in a multi-sided market with zero dollar prices on one side of the market? Should we pivot to a question of decreasing quality rather than an increase in price as we saw in the Google Android decision in Europe? This comes with its own difficulties as determining what a decrease in quality is not as clear as what constitutes an increase in price. To what extent is more advertising a bad thing? To what extent is the ability to control access to personal data a ‘quality characteristic’ from the perspective of the user? If it is, what is the value of the exchange that occurs between the user who trades their data with the ‘free’ content provided by the platform? How does that compare to what would occur in the counterfactual?
What are the analytical tools in our existing jurisprudence that provide a framework to answer the question whether particular conduct has harmed consumers and the competitive process where consumers face a zero dollar price for the substantive service, and instead typically pay for the service through exposure to advertising and use of personal data?
While a SSNIP focussed on quality and not a monetary price may be of some use, our consideration of market definition also needs to be adjusted to consider ecosystem driven business strategies. As noted in the recent issues paper released for our 7th interim report of the Digital Platform Services Inquiry, large digital platform service providers, like Alphabet (Google), Amazon, Apple, Meta (Facebook) and Microsoft, continue to invest heavily across different sectors and technologies, creating a web of interconnected and re-enforcing products and services. Access to and use of vast amounts of user data provides the analytical basis for significant digital platforms to both develop and innovate across multiple products and services but can also provide the conditions to ‘lock’ consumers into ecosystems. Interconnection between platforms, apps, operating systems, and digital services creates a seamless experience for consumers. However, it can lead to bottlenecks, interoperability issues and potential leveraging conduct increasing barriers for potential competitors. Since competition occurs in different ways both within and across product boundaries in respect of digital platform ecosystems, this creates significantly more difficulty in undertaking market definition for the purpose of competition analysis.
Beyond the practical analytical challenges digital markets pose, there are other difficulties associated with our competition law framework.
As an ex post judicial enforcement regime we must necessarily plead specific unilateral conduct, provisions in contracts, arrangements or understandings or acquisitions and supplies in breach of sections 46, 45 and 47 respectively. This means it is not well-suited to the more systemic concerns we are seeking to address, including a range of conduct over many years and across multiple markets and services.
The remedies we can seek in Part IV cases are necessarily (and appropriately) limited to the specific pleadings and factual finding of the court. As an enforcement agency this presents the real risk that even a successful prosecution may not address competitive harm as significant digital platforms operating in multiple interconnected and often multi-sided markets can pivot their strategies to achieve similar outcomes through different means.
Perhaps curiously, given the antecedents of our law, as enforcers we have additional challenges to the US agencies. An example of this is the Department of Justice’s ad-tech case which involves numerous alleged different instances of changes to products and services in the ad tech stack over a period of 15 years to push market participants to Google’s services and consolidate Google’s position. The filing cites a sequence of alleged practices including acquisitions of actual and potential competitors, exclusionary conduct, opaque self-serving rules that gave Google advantages over competitors, and the network bidding agreement with Meta.
Under section 46 we face the need to plead ‘conduct’ – is that in the singular or plural? – and the counterfactual. But for the conduct what would be likely to have had a real chance to occur? Can an overarching business strategy constitute ‘conduct’ or does it have to be pleaded atomistically? The US does not face these challenges with their ‘monopolisation’ prohibition or their ‘unfair competition’ prohibition. Proving in an Australian court a credible theory about how a market or many global markets would have been likely to have evolved in the absence of particular conduct is a seriously challenging evidentiary task.
In dealing with well-resourced digital platforms that operate globally the ACCC is at a profound disadvantage. Business strategies are invariably developed by executives outside of Australia. The design, development and recording of platform algorithms occurs outside of Australia. Often the competitors or potential competitors harmed by anti-competitive conduct operate outside of Australia. Others may say that one response to this forensic challenge may be why not leave the task to the DOJ, FTC and European Commission? But what about the Australian consumers and small businesses that are deprived of choice and better products and services from such conduct?
These challenges also raise issues in respect of remedies. In designing remedies, the aim is to seek to deter anti-competitive conduct but also to restore the competitive conditions as close as possible to that “but for” world. Our courts appropriately do not want to make orders that require them to be in an ongoing supervisory role and seek to make orders in which the requirements are objectively clear and self-executing. Divestiture is not an available remedy for breaches of sections 45, 46 and 47, but for conduct by significant digital platforms, bare declaratory injunctions prohibiting the repeat of pleaded and found conduct will be unlikely to restore the conditions for competition.
In summary, while there are benefits to flexible, general laws, without other complementary measures, the case-by-case enforcement of competition laws through the courts is poorly suited to addressing the issues we have identified with significant digital platforms. Namely, its inability to efficiently address a range of broad and systemic conduct, and the limited effectiveness of remedies. And digital platform services are far too important to Australian consumers and businesses to simply give up on achieving effective competition in these markets.
This is where I reached the much-debated fork in the road. Competition enforcement or regulation. However, the truth is the regulatory reforms proposed aren’t an antonym to competition enforcement, but a complementing tool.
Regulatory reform is the solution
Australia was one of the first jurisdictions to consider competition, consumer and data issues together in relation to digital platforms, through the original Digital Platforms Inquiry.
In considering whether regulatory reform is required to address the broader competition issues we have encountered in our inquiries, we have been informed by lessons out of our parallel ongoing investigations into digital platform services, as well as international developments and discussions with our overseas counterparts who have also been investigating anti-competitive conduct of these global platforms.
There is recognition from regulators around the world that enforcement of current laws is not enough on its own to deal with the competition challenges posed by significant digital platforms. Consensus is growing that urgent reform is needed.
The market failures we have identified in our inquiries and through our competition investigations are not unique to Australia and while globally our competition law systems differ, there is broad consensus of the need for action.
Many jurisdictions are already acting and implementing, or considering, reforms targeted at significant digital platforms to address competition and consumer harms.
- In Europe, Germany has introduced a potentially super-powered enforcement tool for digital platforms and the European Union has passed the Digital Markets Act and the Digital Services Act, which includes rules for gatekeeper platforms to promote contestability and fairness in the digital sector and strengthens consumer rights online.
- The United Kingdom’s pro-competition regime, which will allow for targeted obligations in platform specific codes of conduct, is expected to be passed within the year, and several bills to counter anti-competitive practices are being considered in the US.
- There is growing support for ex ante regulation in India and Japan, with both countries looking at what reforms may be required in their jurisdictions.
We cannot expect the work done by international regulators to bring about change in Australia. We’ve observed that global businesses will make changes in the jurisdictions that act to give extra protection to their consumers and businesses but will not as a rule voluntarily extend those protections in jurisdictions that don’t.
To address these issues, we have recommended new competition measures for digital platforms, including new service-specific mandatory codes of conduct for ‘designated digital platforms,’ based on principles set out in legislation. These codes would introduce targeted obligations to address specific anti-competitive conduct we have previously observed, like specific forms of self-preferencing, tying, certain exclusivity agreements, certain impediments to interoperability, and specific forms of unfair business dealings.
To be clear, we are not suggesting new provisions that say, for example, “platforms cannot anti-competitively self-preference”. This would be just ex-post competition regulation via a different means. Instead, we are proposing prescriptive obligations targeting specific forms of behaviour.
Where practical, ex ante obligations can also be designed to address systemic or structural obstacles to effective competition in some of these services (such as access to data and other barriers to entry and expansion), in addition to specific types of harmful conduct.
In this way, new ex ante obligations would allow for dynamic and flexible measures that could be targeted and tailored to relevant digital platform services and those platforms with the ability and incentive to harm competition by engaging in anti-competitive conduct. Tailoring requirements to a particular service can also assist in balancing of the risks of over- and under- enforcement in dynamic markets. Treasury has concluded its consultation on our recommendations, and we await policy decisions by Government in this area.
Conclusion
Regulatory and enforcement tools must be able to account for the rate of technological change to avoid the far-reaching consequences of anti-competitive conduct in services that are central to the economy. I believe I have made the case for why our existing generalist prohibitions must be complemented by new targeted regulation of significant digital platform services.
While the specificity and targeting under the new sector-specific rules would address many of the issues identified in the work of the ACCC, the flexibility possible through existing case-by-case enforcement under the broad general provisions of existing competition laws remains critical.
I do not want to suggest we are unduly deterred from taking enforcement action under the current provisions of the CCA. We are continuing to actively investigate competition issues in digital platforms markets and as I said before, I have seen my fair share of long competition cases. Our recent commencement of section 46 proceedings against Mastercard is evidence of our capacity to bring complex cases against very well-resourced respondents. We have also invested heavily in our working relationships with overseas regulators and have the capacity to receive substantial investigative assistance and gather information and evidence from some key jurisdictions including the United States of America.
We also welcome the amendments to the Competition and Consumer Act to clarify that section 155 notices may be served on persons (whether in, or outside Australia) and the recent amendments to the Federal Court rules regarding extraterritorial service. These provisions assist us to investigate people and companies carrying on business in Australia wherever they are physically located.
We are not shying away from investigating and taking action against anti-competitive conduct that harms consumers and interferes with the competitive process – whatever the industry. However, new measures for digital platform services, coupled with existing case-by-case enforcement, is needed to promote a more dynamic, competitive and innovative digital economy.