Digital Markets, Competition and Consumers Bill

Written evidence submitted by Amazon (DMCCB45)

(1) Executive summary

Amazon is proud of our investment in the UK and wants to offer constructive proposals for how the Bill as drafted could be improved.

The Digital Markets, Competition and Consumer Bill creates a new competition regime, focussed on "digital activities". We agree that competition is an essential driver of consumer welfare and innovation. While Amazon welcomes the Government’s intention to adopt a flexible, targeted, and pro-innovation approach, as drafted this is not hard-wired into the Bill. The Bill would give the Competition and Markets Authority (CMA) largely unchecked discretion to decide which areas of the economy to inspect, which firms to designate and what rules to impose. It can then vary the rules over time. In theory, this allows for the regime to be bespoke, targeted, flexible and future proof. In practice, however, there are no meaningful thresholds for designation, and the Bill doesn’t require the CMA to consider the impact of any proposed interventions on customers, competition or innovation. The lack of safeguards on the exercise of this broad discretion risks creating significant uncertainty for firms. This is compounded by the lack of an effective appeals mechanism that allows Courts to review the substance of CMA decisions.

As drafted, the Bill has features which could make the UK a harder, less attractive place to invest and innovate, or launch new products and services. Rather than hoping the CMA applies the powers in the Bill proportionately, safeguards should be added to the Bill requiring it.

This note includes four practical proposals for how this could be achieved, including balancing the CMA’s wide discretion with appropriate checks and balances, ensuring appeals can address regulatory error, allowing firms to innovate without having to seek permission from a regulator, and reconsidering powers giving the CMA a ‘licence to experiment’ on customers. These changes could significantly reduce the uncertainty created by the new regime, ensuring the UK retains its world class reputation as an evidence based, proportionate and predictable regulatory environment.

(2) Amazon’s footprint in the UK

Amazon has been in the UK for more than 25 years and the UK is our third largest market globally. We have invested over £12bn in the UK in 2022 alone, and more than £56bn since 2010. Today, there are 100,000 UK SMEs selling in our Store, which make up more than half of our retail sales. In 2022 our SME partners achieved more than £3bn in export sales.

Our total UK headcount of 75,000 means we are one of the country’s largest private sector employers, and we have created 40,000 of these roles within the last three years, with the top three regions of growth being Yorkshire and Humber, the North East, and the South East. In 2022, we created more than 1,500 apprenticeships, and so far this year more than 2,300 employees are enrolled on our Career Choice skills development and training programme. Amazon enables indirect employment for more than 410,000 people across the UK, including 250,000 in the SMEs which sell on Amazon, and a further 160,000 across our supply chain.

The UK is the European centre for many of our most important businesses, such as Prime Video, and has historically been one of the countries where we launch new products and services first, such as Amazon Fresh. Last year, we invested more than £1.6 billion in infrastructure across the UK, as well as continuing to fit our facilities with the latest robotics technology. We also opened our first development centre in Swansea, Wales, which will be the new home to Veeqo, which provides software for sellers on Amazon. This is in addition to our three pre-existing R&D centres in Cambridge, Edinburgh and London, where we are working on projects such as Alexa, Prime Video, and the shopping experience on our online store.

Our UK activity also spans Amazon Web Services (AWS). In 2020, research firm Public First found that use of AWS technology added £8.7bn to the UK economy, realised by lowering costs for business and helping them innovate faster. The UK’s regulatory landscape has been a core reason why Amazon has invested disproportionately in the UK. It has been predictable, evidence-based and pro-innovation. Our hope is that this continues, and the UK remains a very attractive place for global firms to base their investment and innovation.

(3) Amazon’s position on the DMCCB

We agree with the Government that competition is a key driver of innovation and better services for consumers. We welcome the intention to ensure this Bill is targeted, proportionate and pro-innovation.

However, we do not believe that this vision is hard-wired into the Bill. This is for four reasons: First, the Bill provides a single regulator, the CMA, with very broad powers and wide discretion on how they are used. Yet there is little in the way of corresponding constraints or guardrails. Second, the lack of guardrails around the CMA’s wide discretion is compounded by the very limited appeals rights proposed in the Bill. Third, the Bill as drafted could require firms to seek ‘permission to innovate’ from the CMA. This would see the CMA deciding what technology or products are allowed to be tested and launched in the UK. Fourth, the Bill would enable the CMA to compel firms to run live trials on customers – such as changing prices or withdrawing products from sale - before the CMA has decided that there is a problem.

In order to mitigate the risks outlined above, we propose that guardrails are added to the Bill to guide the CMA’s work and ensure it is focussed on consumer welfare. These are set out in detail below.

(4) Proposals for improvement

a. Addressing the accountability gap – inserting safeguards to ensure the CMA uses broad powers and discretion in a proportionate way

The Bill provides the CMA with very broad powers and wide discretion over how they are used. Yet there is little in the way of corresponding constraints or guardrails. The CMA will decide which sectors to look at, which firms are in scope within those sectors and what rules apply to them. There is no obligation on the CMA to identify any actual or potential consumer harm either before designating a firm as having Strategic Market Status, or before intervening (through conduct requirements), neither is there a requirement to show that such interventions will address consumer harm in the future.

This is not simply a case of allowing a new regime to ‘bed in’ over time. The CMA will be able to vary the rules that apply to a regulated firm at any point. This lack of certainty over the rules framework in the UK risks undermining the confidence of companies to invest in the UK. Whilst the regime is often framed as applying narrowly to a small number of large ‘tech’ firms, the Bill would enable the CMA to regulate almost all sectors of the economy. It is critical that proportionality is embedded from the outset.

Proposed solution: Parliament could insert guardrails into the Bill to ensure the CMA uses its powers in a proportionate way, and on the basis of clear evidence of consumer welfare. This might include:

· Requiring that the CMA only imposes conduct requirements where there is an identified actual or potential customer harm to be addressed

· Ensuring conduct requirements imposed on firms can only be varied where there is a material change in market conditions, rather than at the CMA’s complete discretion; ensuring that the CMA has to take account of countervailing consumer benefits when designing (not just enforcing) conduct requirements, and only takes forward proposals where they have net consumer benefits

· Allowing firms to ask the CMA to review its designation and conduct requirements where there is a material change in market conditions

· Requiring the CMA to carry out impact assessments

· Requiring the CMA to report on how the regime is supporting innovation and investment.

b. The need for an effective appeal process that provides a safety net against regulatory error

The lack of guardrails around the CMA’s wide discretion is compounded by the very limited appeal rights proposed in the Bill. Appeals are a critical pillar in ensuring regulatory regimes work for consumers, protecting against regulatory mistakes. Such error is almost unavoidable in this highly novel regime. The CMA will be using new powers and enforcing entirely new legal concepts, some of which are not defined in the Bill. In contrast to other regulators, which can develop deep expertise in narrow areas this regime would require the CMA to be an expert regulator across potentially the entire economy.

The proposed appeal standard, however, does not allow for that. Firms subject to the CMA’s decisions will be able to be appealed only on a judicial review standard (which broadly considers whether the regulator followed the correct process), rather than on a ‘merits’ standard (which allows an appeal to consider the substance, i.e. whether the outcome is the right one for competition and consumers). This creates an accountability deficit.

Most similar regulatory regimes in the UK (e.g. water, rail, telecoms and energy) allow the ‘merits’ of many of the most important decisions to be considered at appeal. If there is a concern that proper appeal rights might result in longer processes ) or high costs, this can easily be addressed by creating a fast-track appeal process that is time limited, either by imposing an overall time limit for a decision, or by limiting evidential burdens by specifying a page limit for submissions and/or specifying the maximum number of witnesses that can be called. Concerns around costs in particular can be addressed through time limits (as the cost of an appeal is closely linked to the duration of the process) as well as setting clear procedural rules governing CAT practices (as, for instance, changing practices around oral witness evidence and cross-examination are significant contributors to rising costs and delays). These relatively simple procedural steps would enable merits reviews to act as a safety net against regulatory error, whilst addressing concerns about time or cost. There need not be a trade-off between the speed of appeals and the accuracy of decisions.

Proposed solution: Changing the standard of appeal from judicial review to full merits, but making it a fast-track system, with a time limit on the process to guarantee quick outcomes and/or alternatively introduce a review of the appeal standard after a fixed period of time of the regime being implemented.

c. Ensuring firms can innovate without having to seek permission from the CMA

As drafted, the Bill could require firms to seek ‘permission to innovate’ from the CMA. That would see the CMA deciding what technology or products are allowed to be tested and launched in the UK. Clause 20(3)(c) could mean that if a firm was found to have ‘Strategic Market Status’ in one activity (e.g. digital advertising), it could be blocked from launching a new product in another area in relation to which no competition concerns have been identified (e.g. launching a chain of high-street coffee shops) unless the CMA approves the launch in advance.

One example of how the proposals could have impacted Amazon had they been in place and Amazon had been designated as a SMS firm is related to the testing capability we developed during the pandemic. We developed in-house testing capacity so that employees, who couldn’t work from home, were able to access free PCR tests. The UK was one of only two countries where we built a lab, partly as the regulatory framework allowed us to do this quicker than we could elsewhere. In similar circumstances, where the Bill was in force, the CMA would have the discretion to require Amazon to seek the regulator’s approval to proceed. Given the uncertainty that this would add to our business planning processes – for a project where speed was of the essence - it is very possible we would have decided to build the lab in another country, where we had a quicker path to launching a testing facility.

Requiring firms to seek ‘permission to innovate’ could significantly undermine companies’ willingness to try new things in the UK (as opposed to elsewhere) for fear of being slowed down or blocked. Whilst we appreciate a ‘permission to innovate’ model is not the Government’s intention, as drafted, it risks being the unintended reality.

Proposed solution: Amending 20(3)(c) to ensure that it is applicable only to the limited circumstances in which "reverse leveraging" could harm competition and consumers. In particular, use of this power should be limited to circumstances when the firm is found to have market power in the non-designated activity which is impacted by the innovation and where there is a material link between the designated activity and the relevant non-designated activity. This would be in line with economic literature identifying where reverse leveraging could raise potential concerns. Importantly, this would safeguard the ability of firms to launch new products and services in markets where they have no market power, and ensure the UK remains open to innovation.

d. Reviewing CMA’s ‘licence to experiment on customers’ even where there is no consumer harm

The CMA can already scrutinise every aspect of a firm’s algorithms. It can compel firms to hand over the underlying source code, as well as the details and results of any algorithmic tests a firm has conducted. Clause 67 of the Bill reproduces these powers but also goes further, enabling the CMA to compel firms to run live tests on the CMA’s behalf in order to generate information. This could mean making live changes to the firm’s products and business models, such as removing products from sale or changing pricing, affecting the consumer experience of some or all of our customers, before the CMA has decided that there is a problem affecting competition. This could be done even before a firm is found to have Strategic Market Status. As drafted, this is purely for information gathering – the CMA can use these powers before they have any concerns about consumer harm. This ‘licence to experiment’ means that the CMA would be able to run tests simply to see the impact.

Proposed solution: This is a radical power, that goes beyond international precedent and has not been consulted on. No evidence has been offered as to why the CMA’s existing powers are inadequate. Amazon proposes that clauses 67 and 69 are amended to remove references to the CMA’s power to mandate the generation of new information.

(5) Conclusions

As a global business, Amazon has experience of a range of regulatory approaches. The model the UK is proposing is entirely new and novel. The Bill would grant the CMA with powers to decide which market segments to investigate, which firms to regulate and to vary the rules over time. The challenge for business is grappling with the uncertainty around the UK regulatory climate, with ramifications for investment, innovation and ultimately consumers.

The UK has historically been a global leader in the design of effective regulation that is predictable, evidence-based and proportionate. This approach is one of the reasons why Amazon has invested here. We understand the role effective regulation plays in driving good economic and consumer outcomes. However, the Bill as drafted risks undermining the intention for a targeted and pro-innovation regime. We want to work constructively with Parliament to ensure the legislation protects the UK’s hard-won status as a great place to invest and innovate. In doing so, we have proposed four practical improvements to the Bill – while seeming small, if taken forward we are confident they will have a big impact in helping the UK achieve the right balance between flexibility and regulatory certainty, and deliver on the UK’s ambition of creating an environment where all companies can compete, innovate and succeed.

June 2023


Prepared 29th June 2023