Competition law and the communications sector

Speech to the UCL Jevons Insitute for Competition Law and Economics Annual Colloquium

By Ed Richards

Chief Executive, Ofcom

7 July 2010

My topic today will be the complex regulatory environment in the communications sector. There are three themes I would like to cover:

First it is useful to consider the state of ex-post and ex-ante regulation in the communications sector.  Where do we find ourselves in this debate and in what direction are we likely to be heading?

Second, how can regulators use consumer policy to encourage competition, both apart from, and in close connection to ex-ante and ex-post competition powers? I want to assess the balance of regulatory powers in the sector using net neutrality as a mirror for the challenges we face.

And finally, I’d like to consider one or two of the key regulatory challenges we are facing in communications regulation, and look at the stumbling blocks in our path to making timely decisions which can keep pace with and adapt to the fast moving terrain of today’s communications sector.

Taking the first of these, I want to conisder the balance of regulatory powers in the communications sector and focus on the question:  What is the most effective way of dealing with competition concerns in the communications sector?

We think the answer is that it depends on the specific concern.  Take the case of dealing with enduring bottlenecks, for example wholesale inputs in fixed telecom markets.  I think most regulators in Europe and globally would agree ex ante regulation seems a more effective solution than ex post.  In 1987 New Zealand adopted a “no sector” regulatory experiment, purely relying on ex-post competition law and contractual undertakings, and found it could not sustain the competitive and efficient telecoms market it wished to see when it adopted its light-regulatory system. What was the result?  New Zealand eventually adopted a policy that closely mirrors our own in the UK (and in the EU generally).  A regulatory system that focuses on the close supervision of the upstream enduring bottleneck elements of the incumbent telecoms network, but allows competition downstream.

Consistent with this approach, the European Commission and a number of commentators in the discussions over the implementation of the EC Framework for Telecom regulation have frequently argued that ex ante regulation is appropriate for enduring bottlenecks but that over time telecom markets that are not enduring bottlenecks should witness a transition from ex ante to ex post regulation.

1. Between 2003-2007 the number of markets on the EC list likely to require ex-ante regulation fell from 18 to 7.
2. Ofcom has also pushed ahead with reducing ex ante regulation where appropriate. We’ve removed regulation from retail telephony markets last year and deregulated wholesale broadband access in nearly three quarters of the country in 2008 (and never imposed regulation in retail broadband markets).
3. But in common with other competition authorities we have found most of the Competition Act cases we have investigated to be lengthy, complex and resource intensive, both for us and for the industry participants.

The hurdles involved in establishing a breach of ex-post competition laws are high. And these tend to be highly complex cases that take a very long time.  Speaking as a regulator that has and does use these powers, it is possible to observe that this may be one reason why ex-post powers have been used so little by sectoral regulators more generally.

However, in telecoms there is a more significant reason.  Industry has a much swifter, guaranteed route open to it through the system of dispute resolution established by European law.  Competition issues in communications markets are frequently raised as regulatory disputes which Ofcom has a statutory obligation to resolve in 4 months, rather than years.

Most disputes are in relation to pricing issues alleging abusive/unfair pricing/refusal to supply – our powers include requiring supply, setting the price and other terms going forward and also crucially discretion to award repayment.

In contrast, a complainant in a Competition Act case needs first to persuade the competition authority to investigate and find an infringement and then will have to sue for damages in court separately (at its own litigation costs and at risk of the defendant’s costs).  We are not able to charge for handling most disputes, so they represent a comparatively cheaper option for companies when compared with litigation.

So where does this leave us?  It suggests to me that the pure distinction between ex ante and ex post is likely to be too simplistic.  Competition Act powers are important in the overall toolkit available to economic regulators, but they have to be viewed and deployed in conjunction with other powers, such as our ex-ante powers under the framework and our dispute powers.

We can all appreciate the dangers of defaulting to relying too heavily on ex ante regulation in terms of shaping the market with unintended consequences.  We can also appreciate the dangers of relying too heavily on ex post regulation as New Zealand discovered.

Instead I think recent experience here and internationally points us towards the idea that a national regulator needs to use a balance of powers to address complex competition problems, and should not overstate or seek to hold ideologically pure technical positions above the interests of competition and consumers, which is what ultimately matters.

We also need a regulatory system that can react quickly to change and impose remedies on complex fast changing communications markets where that is necessary.  I will come back to this at the end of my talk.

Important role of consumer support and protection.

In our discussion of ex-ante and ex-post regulatory powers we must also not lose sight of the complementary role consumer support and protection can play. And we should consider how this contributes to the broader debate around regulation?

Ofcom works hard to avoid poor consumer outcomes by using both competition policy and consumer policy.  Competition policy is often characterised as the set of regulatory instruments used to address supply-side market failures – typically those that arise from high market shares and barriers to entry and expansion.

Left unchecked, these could lead to one or more suppliers being in a position to exploit a dominant position.  Consumers suffer as a result of excessive or inefficient pricing, poor choice, and weaker incentives to innovate and invest in new products and services.

If competition policy seeks to address the supply side, consumer policy can be characterised as being concerned with “demand-side” market failures, such as information asymmetries and transaction costs.  These market failures prevent consumers, first, from understanding their choices, and then from exercising those choices effectively.

We want to ensure that consumers can make well-informed decisions about the services they receive. One of the best examples is our Broadband Speeds Code, where we have helped to create an industry solution to the very poor quality information consumers were receiving about the speed of their connection.

This is all part of ensuring that consumers have a range of information at their fingertips about the products they are buying. We want to enable consumers to make effective decisions on where to spend their money. And even if each individual consumer does not analyse all the information, some ‘lead consumers’ will be able to influence the prevailing environment by providing sufficient switching behaviour to discipline the market.

Consumer empowerment is also perhaps one of the strongest routes to deal with consumer issues online. And, interestingly the development of the web and online communities opens new doors for regulatory bodies to encourage openness in practices, prices and charges for those it regulates. As Commissioner Almunia said earlier today “on the internet information travels fast and users are very reactive”.

I agree with the Commissioner’s sentiment very strongly, which of course makes it an even greater irony that the one service for which you cannot find reliable and accurate information on the internet is your network access to the internet.

So, later in July we will publish the second round of our research on access speeds to provide consumers with this information to enable effective choice.

Combining better consumer information with the ability easily to switch providers, we can help move towards fully functioning competitive markets.  Changing communications provider is typically more challenging than, for example, changing supermarket, and importantly it requires co-ordination among multiple providers.  There may therefore be a need for specific rules around switching processes, and this is something we are considering in depth at present.

Net neutrality as a case study

To illustrate my point about the effectiveness of different regulatory powers, I’d like now to turn to net neutrality as an area where there is an interesting duality of ex-ante and ex-post regulation in the potential solutions to an anticipated competition issue.

Net neutrality or Internet traffic management is one of the hot topics in the communications sector. Communications regulators around the world are starting to turn their attention to this issue. The FCC calls it “preserving the open internet”. Ofcom released its own discussion paper last month and the Commission has just published its consultation.

At the heart of the net neutrality debate is the concern that traffic management could be used as a form of anti-competitive discrimination. And at an illustrative extreme, that a provider could limit the access of its subscribers to its own services, its own content, its own subset of the Internet.

Here in the UK, there have been no formal complaints about anti-competitive discrimination, although there have been a number of modest disagreements between content/service providers and ISPs/mobile operators. It is in this vein that we do not currently see a compelling reason for preventing, ex ante, all forms of discrimination using our sector-specific regulatory powers.

But if genuine problems of anti-competitive practices in relation to traffic management emerged, we would of course have the ability to intervene applying our full range of ex post competition powers as appropriate.

This allows us to take a measured approach, allowing certain practices – such as permitting operators and ISPs to set differentiated quality of service – which may prove beneficial to consumers, but which could be caught by a blanket prohibition, whilst at the same time being able to take effective action to curb any genuinely anti-competitive practices that may emerge.

Although the evidence at this stage suggests a blanket prohibition is undesirable, our initial stance in this debate is that consumer transparency must be guaranteed wherever traffic management occurs. Consumers need to have information available to them. They need to know what policy their internet provider applies, and how this affects the service they receive. In a competitive market, they can then exercise choice including this criterion.

Developing some basic principles around transparency, and ensuring that operators and ISPs comply with these principles, is consistent with our broader functions and duties as a sector regulator.  Our understanding of the sector and the challenges it presents to consumers, backed by original research and carefully gathered evidence, gives us a particularly strong insight into designing and applying appropriate transparency rules to the communications sector.

So the ideal solution to the complex competition issues involved in the net neutrality debate would be to use a combination of appropriate ex-ante and ex-post regulation, with the scope to apply tough but targeted competition powers to genuine abuse, whilst securing the high level of consumer transparency required in this complex area through our instruments of sector-specific regulation.

I would now like to turn to the regulatory challenges on the horizon, and the problems we face in ensuring good outcomes for UK citizens and consumers, and in promoting effective competition.

So what are the emerging issues and questions we face in communications regulation?

First, new bottlenecks are emerging that are different from those we are traditionally used to. In the past bottlenecks have arisen from large economies of scale and sunk costs in the provision of physical assets. Access networks are the most obvious and well known.  New bottlenecks, on the other hand, are likely to arise from large economies of scale and network effects in information goods, often as a result of tipping effects in two-sided markets.

Google is the example on many people’s lips in Europe and the US as a number of commercial rivals have started submitting complaints over alleged abusive practices. Consumers use Google to find what they want from the internet, which creates an opportunity for advertisers, which allows Google to make money and to maintain its attractiveness to search customers.

In this case the alleged bottleneck has arisen from innovation, and winning a competition for the market.  The difficult task for regulators is to avoid suppressing reward for innovation and risk balanced against the need to avoid allowing the creation of an enduring monopoly with the associated consequences.  And of course assessing whether or not such a position is or is not enduring, a particularly difficult task where very high market shares persist alongside high rates of technological innovation.

Second, we have the challenge of consumer outcomes in oligopolistic markets with a limited number of players but high barriers to entry and growth.  Communications markets have a natural tendency towards oligopoly.  There are significant scale economies and entry barriers (such as the availability of spectrum in mobile markets); therefore the market will not support a large number of players (particularly at the wholesale level).  But, there is also sufficient opportunity for innovation and product differentiation to avoid a monopoly.

This has been true for mobile networks, traditional broadband networks and probably for superfast broadband networks. We have taken the view that it is possible for a market with a small number of players to operate competitively, as with the mobile market with five, now four main players. We have also deregulated the wholesale broadband access market when there are four or more players operating in an exchange area.

However, in these markets with a limited number of players and high barriers to entry and expansion, we remain vigilant as all too often there is a danger of less innovation and higher prices for consumers.

Thirdly, there are also increasingly blurred market boundaries where products are imperfect substitutes for one another. A good example of this in the communications sector is between next generation and current generation access services. Are they in the same market? The answer is not entirely clear. While there is some constraint between the two, the substitution is likely to be imperfect.   As one generation of products gradually replaces another, our conventional approach to market definition struggles to capture the potential constraining effect of the evolving new product.

So where does all this leave us?  The regulatory environment in the communications sector is as complex as ever. Technological developments in digital media and online and emerging regulatory problems like the ones I have just described throw up new challenges for regulators. We have a lot of important decisions to make and much to be wary of.

Our stakeholders demand the ‘right’ decision, but also a ‘timely’ decision as they react to fast moving changes in technology, consumer behaviour and in competitive strategy.  So I think that a final question that we must ask is whether the overall approach is one which is likely to ensure that the regulator makes good decisions in a timely and efficient way?

Appeals and the science of appeals

One of the key benefits of our regulatory intervention is to strengthen dynamic competition.  That is, it is to allow more efficient providers to enter and grow in the markets we regulate. We have achieved this through, for example, access remedies, number portability interventions and interventions on switching processes.

One of the issues we face over and over again is the strong focus on quantitative cost-benefit assessments of regulatory interventions during a full merits appeal to which we are subject.  Assessments, as it has been put to us, which must be able “to withstand profound and rigorous scrutiny”.

Of course our decisions need to be soundly based, but the assessments we are required to carry out are often highly complex, take considerable time and inevitably involve forecast and estimation. The quest for ever more precise quasi-scientific quantification may be seen to hamper our effectiveness both in intervening where necessary, and in reducing regulation where possible.

It also makes it risky for us in cases where the direct evidence of effects on consumers may not be absolutely clear or precisely estimated. For instance, the main benefit of many of our activities, the increase in competition over time, is by its very nature difficult to quantify.  Yet I would suggest that history tells us that it is relatively hard to identify economic problems where the cause has been ‘too much’ competition, rather than too little.

While the changes we bring in through our regulatory decisions can very definitely impose costs on incumbent providers, an excessive focus on quantification of those costs during an appeal process would seem to place greater weight on the short-term costs than on the long term benefits from the increase in dynamic competition.

A number of commentators, have questioned the apparently excessive reliance on a rather static modelling of costs and benefits against a more qualitative analysis of the long-term dynamic effects of particular interventions. They have made an analogy with the over-reliance on quantitative risk models by a number of well-known banks.

It’s time perhaps to return from Immanuel Kant’s clever inversion beloved of the academic community – “That may be all right in practice, but it doesn’t work in theory”, to the more mundane alternative “it may work in theory but what damage is it doing in practice?”.

While the burden of proof might seem appropriate for quasi criminal Competition Act investigations, it might not represent the right balance for regulatory interventions in areas where regulation needs to be fine tuned over time; where regulation needs to react rapidly to changing circumstances, and where there is a well established ex-ante framework which with effective implementation demonstrably delivers benefits to consumers through more effective competition.

Now, I believe very strongly that stakeholders have a right to a meaningful and fair appeal of our decisions, and indeed those of any other regulators. It is entirely correct that regulatory decisions are appropriately scrutinised.

However, it is important also to examine whether the current system is working effectively more generally.

You could reasonably take the view that the appeals model in telecoms is right exactly as it is today.  But also note the fact that Ofcom has to deal with significantly more appeals than any other economic regulator carrying out an equivalent function, but who are not subject to the same standard of review.  The question is then perhaps better phrased as: If this standard of review is correct in telecoms, then why should all economic regulators not have the same standard of review as Ofcom?  It would certainly be a red letter day for the legal community.

An effective appeals regime is one in which there is a good balance between the competing requirements of justice and efficiency.  Justice requires that there should be a means for correcting material errors which cause harm to affected parties, while efficiency requires there to be a way of resolving issues in a timely way.

Consider a World Cup sporting analogy.   If the ball crosses the goal line by a metre (or was it just 50cm?) but it is not spotted by the referee, justice suggests that there should be a means of correcting what is a material error in the decision; most obviously by having a review of the decision involving a fresh look at the evidence. Frank Lampard and Fabio Capello would surely agree.

But should we also have a full review of every incident in the game? Every dispute over a ‘throw in’ on the half way line, involving a full review of all the evidence? This would surely be to emphasise justice excessively at the expense of efficiency.

It is now time to ask, whether the regulatory system strikes the right balance between justice and efficiency in decision making.  Or whether we have ambled somnolently into a world where regulators are expected to make timely decisions to promote competition, but find it ever increasingly difficult to do so.

Today, I’ve described a world in which the old, hard-and-fast distinction between ex-ante and ex-post is a less useful way of thinking about how to regulate the communications sector.

I’ve explained the role of consumer policy in complementing and supporting supply side intervention.

And I’ve set out where I think the challenges to timely and effective decision-making now exist, and how we might begin to think about them.

The basis for this approach are all there, but we do not yet, in my view, have the optimal mix of ingredients.