Automatically Renewable Contracts

Section 1

Summary

1.1           This statement presents our decision to amend General Condition 91 (GC 9) in order to prohibit Automatically Renewable Contracts (ARCs also referred to as ‘rollover contracts’ or ‘rollovers) to residential customers and small businesses with no more than ten employees in the fixed voice and broadband sectors.

1.2           In communications retail markets, ARCs are those that, at the end of each minimum contract period (MCP), roll forward to a new MCP by default unless the customer proactively informs their Communications Provider (CP) that they do not wish this to happen. An MCP is a fixed period of time for which a customer commits to taking services from a CP. While under an MCP, a customer is usually subject to an Early Termination Charge (ETC) should they wish to end the contract.

1.3           Since they first became a prevalent feature of the residential fixed voice sector in 2008, we have been concerned that ARCs are damaging to consumers and competition in communications markets. We recognise that ARCs may have benefits for some consumers – for example, those who wish to remain with their CP and who value the ability to move into a new minimum contract period unless they opt out. However, we believe these benefits are relatively limited and are outweighed by the costs.

1.4           Ofcom has been monitoring ARCs in UK residential and business fixed voice markets since they emerged, and we have carried out targeted research on their effects. While we have not carried out such specific research with respect to small businesses, we are confident that the results we have identified for residential ARCs customers can be extrapolated to small business customers. BT is the largest CP currently offering ARCs in these markets. Residential ARCs are also available from several smaller fixed voice CPs such as Adept Telecom, Eze Talk, italk and Axis Telecom, while TalkTalk Business, Titan Telecoms, and Optimum Calls offer ARCs to business users. We calculate that currently around 15% of UK residential fixed voice consumers are contracted to ARC packages. BT also offers ARCs in its residential broadband propositions.

1.5           Our research, in particular the econometric analysis that we commissioned on the switching behaviour of BT customers, indicates a clear causal link between ARCs and reduced levels of consumer switching. We believe this effect stems from the opt- out nature of the process for contract renewal and that any example of such a contract is likely to be harmful to consumers and to effective competition.

1.6           Our ARCs consultation document, published on 3 March 2011, set out our analysis in detail and presented proposals to prohibit ARCs for residential customers and small businesses with no more than ten employees in the fixed voice and broadband sectors. In this statement, we present and discuss the responses to the consultation that we received. The majority of respondents were supportive of our proposals, including both consumer groups and CPs. Two CPs opposed our proposals, and a number of others opposed elements of our proposals or analysis, but were generally supportive of our main proposals.

1.7           Having carefully considered all the consultation responses, we have decided to proceed with our proposed modifications to General Condition 9 to prohibit the ‘opt- out’ mechanism in ARCs. This statement contains (at Annex 2) the formal notification of modifications to General Condition 9 to give effect to the prohibition of ARCs in the fixed voice and fixed broadband small business and residential sectors. The modifications incorporate some changes to the drafting which we have made in response to the comments of respondents, but the substance of the changes is as we proposed in the consultation document.

Implementation arrangements

1.8           We recognise that the removal of ARCs from the markets in these sectors cannot happen overnight, particularly for the significant base of customers currently contracted to ARCs. We have noted the views of respondents who urged us to mandate an accelerated withdrawal, and also taken account of the practical considerations of withdrawal which requires changes to CPs’ systems. We want to ensure that the implementation is smooth and does not itself result in distortions to the market or disruption and inconvenience for consumers and business customers. Taking account of all these considerations, we have concluded that the following framework is appropriate for the withdrawal of ARCs:

1.8.1            For fixed voice and fixed broadband residential and small business ARCs the prohibition on the sale of new ARCs2 will take effect on 31 December 2011. This means that the sale of new ARCs will be prohibited from that date.

1.8.2            To enable the orderly migration of the existing customer base, we are requiring that ARCs are completely removed from the market (i.e. all existing ARCs customers migrated) by 31 December 2012. We also expect CPs to be proactive in facilitating migration for existing ARCs customers during this period, and to take a flexible approach to ETCs for ARCs customers who wish to exit after the first MCP of their contract.

1.9           We recognise that, with respect to the prohibition of ARCs in relation to small businesses, it may, at times, be difficult for a CP to identify whether or not a customer has 10 employees or less. This is because CPs may not routinely collect or hold information about the number of employees of their business customers, and because employee numbers can fluctuate. Therefore we believe a flexible approach to compliance and enforcement is needed to ensure that the small business prohibition is not inappropriately targeted, and that the compliance burden is reasonable. We have included guidance on our approach to compliance and enforcement at Annex 3 of this statement.

Section 2

Introduction

2.1           This statement is the outcome of an extensive investigation into the impact of Automatically Renewable Contracts (ARCs, or ‘rollover contracts,’ or ‘rollovers’) on consumer switching, which began in 2008 when these types of contracts first emerged in the communications sector. Following a lengthy period of evidence gathering, we published a consultation on our proposals in relation to ARCs in March 2011.3 This statement assesses in detail the consultation responses we received and sets out our final decision to implement modifications to General Condition 9 (GC 9) to prohibit the sale of ARCs in the fixed voice and broadband sectors. This section sets out some background regarding the emergence of ARCs and our investigation into their effects.

Background

2.2           Many Communications Providers (CPs) offer fixed term contracts that require customers to commit to paying for a service for a minimum contract period (MCP) in return for an incentive, such as a price discount or an equipment subsidy – for example, a mobile handset subsidy or a set-top box. In order to exit fixed term contracts before the end of an MCP, customers usually have to pay an early termination charge (ETC).

2.3           ARCs in communications markets are contracts where, at the end of an MCP (whether this is an initial or subsequent period), the contract rolls forward to a new MCP by default, unless the customer proactively informs their CP that they do not wish this to happen. (In this statement we sometimes refer to a new MCP as a “new contract”.)

ARCs in the fixed voice sector

2.4           ARCs have been a well established feature of some calls packages available to businesses, but were not common in residential markets until they were introduced by BT into its residential call packages in February 2008.

2.5           Under BT’s original offer, residential customers could opt for a 12-month ARC featuring unlimited evening and weekend calls at a discounted monthly price of

£11.54 (including VAT). This represented a discount of 21% on the standard monthly price of £14.53.4 Termination of the contract within the MCP (initial or subsequent) would result in the customer having to pay an ETC of £7.50 per month.5 Contracts were automatically renewed for a new 12-month MCP unless customers informed BT that they did not want this to happen. Customers could provide notification of this at any time, and BT sent the customer a reminder letter around one month before the expiry of each MCP.

2.6           BT subsequently broadened its ARCs to other residential call packages – Unlimited Weekend and Unlimited Anytime.

2.7           In June 2010, following discussions between Ofcom, BT, and a number of other fixed-voice CPs, BT announced significant reductions to its ETCs, and currently the ETCs which apply for each remaining month of any MCP when a customer terminates the contract early are: £2.50 for Unlimited Weekend, £3 for Unlimited Evening and Weekend, and £5.50 for Unlimited Anytime.6 These reductions applied to all customers on existing contracts as well as new customers.

2.8           At present BT still offers ARCs, but with reduced marketing emphasis – for example, the packages advertised most prominently on its website do not feature ARCs, but variants of its Unlimited Anytime and Unlimited Evening and Weekend packages are available that do feature ARCs. There are however, significant numbers of consumers who remain on ARC contracts. We estimate that approximately 15% of UK residential fixed voice consumers are currently on ARCs.

2.9           BT is the only large CP currently offering ARCs to residential fixed voice customers. However, several smaller CPs also offer them. For example:

•       Adept Telecom offers an 18-month renewable contract on its residential package providing free evening and weekend calls. Adept requires 30 days’ written notice of contract termination prior to the ending of the MCP or the contract will automatically renew to a new MCP. An ETC of £5 for every month remaining on the contract applies for customers wishing to cancel their contract within an MCP.7

•       Eze Talk offers a 24-month residential contract which automatically renews to a new MCP unless customers give notice that they do not want to renew 28 days before the contract expiry date. For customers who wish to cancel their contract during an MCP, ETCs are £8 per remaining month of the MCP.8

•       italk offers a 24-month renewable contract which can be cancelled by the customer in writing at least one month before the anniversary of the contract.9 Cancellation by the customer during any MCP (initial or following automatic renewal) will incur an ETC based on the outstanding period of the MCP. Following recent engagement with Ofcom under its Additional Charges Enforcement programme, italk has reduced its ETCs for fixed voice significantly. Its new monthly ETC is now £3 and £5.50 respectively for its italk 2 and italk 3 packages.

•       Axis Telecom offers a contract with an initial MCP of 12 months which continues from year to year unless cancelled by the customer with one month’s written notice given no earlier than one month before the end of the initial MCP or subsequent anniversaries. Cancellation by the customer at any other time results in an ETC of £149.93 if the cancelation takes place in the first month of an MCP. This figure is reduced by £6.66 for each successive month of the contract completed prior to cancellation.10

2.10        A number of CPs also provide ARCs to small business customers.11 For example:

•       TalkTalk Business offers an initial 24-month MCP which automatically rolls forward to subsequent 12-month periods unless the customer contacts the CP at least 30 days before the end of the original MCP to inform them that they do not want the contract to be renewed. Termination by the customer within the first year of the contract results in an ETC of £300, whereas termination by the customer at any other time within an MCP after the first year results in an ETC of £125.12

•       BT offers ARCs to business. The customer agrees to a minimum spend commitment, and a 1 or 2 year term which automatically renews. BT Business offer discounts based upon spend commitments. The customer spend is reviewed after 12 months (or at termination) and reconciliation charges are applied if the spend commitment is not achieved.13

•       Adept Telecom also offers ARCs to business customers. Business contracts are for either 1, 3, or 5 years. A renewable contract applies to all contracts signed after 18th November 2008. At the end of the contract period it becomes a 12- month renewable contract thereafter.14

•       Titan Telecom offers a 1 or 3-year initial term followed by subsequent periods of one year, which automatically renew at the end of each year for another year unless the customer gives at least 3 months written notice if he/she does not want the contract to be automatically renewed. Termination by the customer at any other time within an MCP results in an ETC of £150.15

•       Optimum Calls offers a 5-year initial term which automatically renews to subsequent 5-year terms unless the customer gives notice by Royal Mail recorded delivery within 30 days of the ending of an MCP. Termination by the customer at any other time within an MCP results in an ETC of £395 (excluding VAT).16

ARCs in the broadband and mobile sectors

2.11        BT has also introduced ARCs in the residential broadband sector.17 BT currently has six different broadband ARC packages, which offer a discount of between £2 and £3 per month off the standard price in exchange for a 12-month ARC.

2.12        ARCs are not currently a feature of the mobile market. We made clear in our consultation that we would have significant concerns if they were to emerge.18 However, we have not chosen to include an explicit prohibition of ARCs in the mobile sector as we have targeted our intervention at the current harm we observe in fixed markets. We would urgently revisit this issue if ARCs emerged in the mobile sector.

Legal Framework

2.13        Ofcom has a general duty under section 3 of the Communications Act 2003 (‘the Act’) to further the interests of citizens in relation to communications matters and to further the interests of consumers in relevant markets. In performing these duties Ofcom must have regard, in all cases, to the principles under which its regulatory activities should be transparent, accountable, proportionate, consistent, and targeted only at cases in which action is needed.

2.14        Further, Ofcom has powers under sections 45 – 48C19 of the Act to set (and to modify) General Conditions (GCs) for the purposes of governing the way in which CPs conduct their operations in the UK. Such conditions include those which Ofcom considers are appropriate for protecting the interests of end users of public electronic communications services (section 51(1) (a)20).

2.15        Ofcom must ensure that, when modifying a GC, this is objectively justifiable, not unduly discriminatory in relation to particular persons, proportionate to the aim sought to be achieved, and transparent (section 47(2)21).

The revised European Common Regulatory Framework

2.16        In 2009, the European Parliament and Council of Ministers agreed a package of reforms to the common framework for communications regulation which applies across all EU Member States.22 The new package was required to be transposed into the national laws and regulatory systems of Member States, including the UK, by May 2011.23 As a result, Ofcom has recently published a statement on changes to the General Conditions and Universal Service Conditions.24

2.17        The revised Framework includes amendments to the “Universal Service Directive” (“the USD”).25 These amendments include a new Article 30 which deals specifically with requirements for consumer switching, including a requirement that Member States ensure that “without prejudice to any minimum contractual period…conditions and procedures for contract termination do not act as a disincentive against changing service provider”.26 In the recitals to the Directive, it is also clear that consumers should be able to change providers when it is in their interests to do so and without hindrance by (inter-alia) “legal obstacles”, including contractual conditions.27 This means that, Article 30(6) also directly mandates Member States to ensure that arrangements at the end of contracts do not raise barriers to switching.

2.18        The requirements of Article 30(6) of the USD are reflected in the creation of a new GC 9.3, as set out in our statement on changes to the General Conditions and Universal Service Conditions.28 These requirements are relevant to our work on ARCs, where we believe that automatic renewal mechanisms which require proactive opt-out by subscribers (as in ARCs) create ambiguity around the ending of contracts, and this raises barriers to switching for some subscribers.

Ofcom’s investigation into ARCs

2.19        Since early 2008, when ARCs first emerged in UK residential communications markets, Ofcom has had serious concerns about the potential harm that ARCs may cause, particularly if they become a widespread feature of this market. This reflected our wider concern with potential barriers to switching in the communications sector, and the project has been conducted alongside our Strategic Review of Consumer Switching29 and our Additional Charges enforcement programme under the Unfair Terms in Consumer Contract Regulations 1999 (UTCCRs).30

2.20        Our analysis of the likely effects of ARCs was set out in detail in Section 3 of the consultation (and summarised in section 3 of this statement). Our main concern is that ARCs lead to the unintentional renewal of MCPs, thereby increasing switching costs in the market to the detriment of consumers and competition. We are also concerned that they induce customers looking to switch to systematically contact their existing CP thereby potentially introducing targeted save opportunities for losing providers. As explained in our consultation document on consumer switching, we are concerned that targeted save activities may result in consumer harm and be damaging to competition.31

2.21        We initially looked at ARCs in the context of our Review of Additional Charges Statement published in December 2008.32 The Review set out our guidance on additional charges, including how we would enforce the UTCCRs in the communications sector. Our current Additional Charges guidance, which was updated in November 2010 (following the Supreme Court’s judgment in the Office of Fair Trading’s Bank Charges case) states the conditions under which we believe ARC terms are more likely to be judged as ‘fair’ under the UTCCRs.33 This includes where the ARC term is transparent and a clear reminder is sent to the customer.34

2.22        However, the test of fairness under the UTCCRs is a legal test specific to those regulations and does not necessarily capture the full economic effects of a contract term. Consequently, our concerns about the impact of ARCs remained and we commissioned some market research to better understand their effects and determine whether some form of intervention was appropriate.

2.23        Our initial market research, conducted in 2009, focussed on transparency and customer awareness in relation to BT’s ARCs propositions, and included a mystery shopping exercise, and a customer survey. This focus reflected the fact that, at the time, only a relatively small proportion of BT’s ARC customers had rolled forward to a new contract (most contracts were sold in the second half of 2008) and the impact of ARCs was not yet clear.

2.24        As we discussed in the consultation document,35 our transparency research found mixed results, with some shortcomings in BT’s marketing at the point of sale. However, we took the view that targeting this issue with transparency remedies would not remove our concerns. This was, in part, because BT sends customers, in our view, a clear reminder letter before the date of the MCP renewal.

2.25        Towards the end of 2009 we commissioned Gregory S Crawford, Professor of Economics at the University of Warwick, and ESMT Competition Analysis36 to conduct an econometric analysis of BT customer data in order to identify whether BT’s ARC term had an impact on customer switching. An econometric approach was necessary in order to isolate the impact of the ARC term itself, as separate from other factors such as the price discount associated with the offer, and changes in the competitive dynamics in the market and the reductions in ETCs. The approach also allowed us to abstract away from specific features of BT’s offer and customer base and enabled us to extrapolate the findings with more confidence to other offers that could emerge in the future.

2.26        We shared the analysis with BT and took into account its criticisms of the research, as contained in two letters sent to us in October 2010.37 We subsequently asked Professor Crawford to conduct some supplementary analysis to address BT’s criticism that we had not adequately identified the impact of ARC terms separately from the associated price discount that its ARC customers receive.38

Our consultation proposals

2.27        Our consultation was published on 3 March 2011 and the consultation period lasted for 7 weeks following requests from CPs to extend the initial 6 week period. The analysis was based in large part on the econometric research conducted by Professor Crawford. We concluded that ARCs have a significant negative effect on consumer switching, despite only changing the process for contract renewal and not the length of commitment made by the customer. We further concluded that this stemmed from the opt-out nature of the renewal process and that the effect would be present for any example of an ARC that is likely to emerge in the fixed voice and broadband sectors. We also argued that the benefits of ARCs for consumers are limited. Finally, we explained that we believed our evidence applied to those small business users who make purchasing decisions in a similar way to residential consumers.

2.28        Based on these conclusions we proposed modifications to GC 9 to prohibit the sale of ARCs to residential and small business customers in the fixed voice and broadband sectors. Our proposals reflected a particular concern that ARCs may spread rapidly throughout the communications sector and become an established feature. While BT remains the only large provider to offer ARCs, because of the significant retention benefits that ARCs provide, we believe they would remain or will become attractive to CPs, and are likely to become more so as the customer bases and market shares of CPs other than BT grow.

ARCs in other sectors

2.29        Whilst we are mindful that many of the effects discussed in this statement are specific to ARCs in communications markets, we are aware that similar issues have been or are being addressed in other industry sectors.

2.30        In the energy sector, Ofgem initially considered the use of ARCs in non-residential supply and took action in 2009 to limit the length of MCPs after the initial period to 12 months.39 In March 2010, Ofgem wrote an open letter to residential energy suppliers reminding those offering ARCs of their obligations under energy supply licences and the UTCCRs.40 In January 2011 Ofgem published a consultation, which expressed concerns over the transparency of ARC terms and conditions in the energy sector and presented proposals to modify the Standard Licence Conditions in the sector to address this.41 More recently, Ofgem’s Retail Market Review consultation, which was published in March 2011 and concluded in June 2011, goes further and contains proposals to prohibit altogether the sale of ARCs to residential consumers.42

2.31        The Office of Fair Trading (OFT) has undertaken a comprehensive study into consumer contracts focussed on customers’ understanding of contract terms and on how firms approach contracts, including any practices which intentionally or unintentionally disadvantage customers.43 The study, published in February 2011, found that consumer harm can arise when terms in the small print of a contract alter the deal from what consumers understand it to be, and the OFT considers that a detrimental term includes one that poses any material obstructions to consumer switching. We think the OFT’s work is significant in understanding the effects of ARCs since the harm we have identified is partly a consequence of customers’ understanding of the effects of ARCs as well as other behavioural factors.

2.32        We have not conducted an exhaustive survey of ARCs (and similar terms) in other sectors since experiences elsewhere are likely to be shaped by the specific features of the markets concerned and may not generalise more widely. However, we noted in the consultation that many examples of automatically renewing contracts in other sectors do not subject a consumer to an ETC after the rollover and are primarily aimed at ensuring the continuation of a service that would otherwise stop.

2.33        For example, ARCs are a common feature in the insurance industry but are deployed to ensure that there is no gap in critical cover where customers have not renewed. There is no penalty for cancelling cover in these circumstances. Therefore ARCs in the insurance industry are fundamentally different to those in the communications sector where ETCs are levied for cancellation before the end of an MCP. Similarly, the automatic renewal of annual subscriptions to newspapers and magazines does not typically subject the consumer to an ETC.

Treatment of ARCs in overseas communications markets

2.34        We are aware that other regulators and some international institutions are working on contract renewal arrangements and barriers to switching generally, and in some cases on ARCs specifically. Moreover, a number of EU countries have passed primary or secondary legislation that makes ARCs, and even consecutive opt-in MCPs after the initial contract period, unlawful. For example,

•       In the Netherlands, under legislation enacted in 2009, Chapter 7 of the Telecommunications Act provides that contracts must move to a monthly basis after an initial MCP – effectively prohibiting ARCs.44

•       In France, the “Chatel Law” contains a number of relevant provisions, including requirements that CPs cannot impose contracts longer than 12 months unless the same offer is available for a period of less than 12 months, and that ETCs levied in any period after the initial 12 months of a contract cannot be greater than one quarter of the value of monthly subscriptions remaining under the contract term. The Chatel Law also prohibits any MCP greater than 24 months.45

•       In Italy, the “Bersani Law” (2007) allows consumers in the communications industry to terminate contracts with 30 days’ notice while capping ETCs to the direct costs incurred by the CPs when the service is terminated. In practice this has meant that CPs have found it very difficult to justify ETCs after the initial MCP when a subsidy which has been recovered through monthly charges ends. This effectively eliminates any commercial incentive to introduce the type of ARCs we are discussing in this consultation.

•       In its paper on Broadband Bundling, the OECD comments specifically on ARCs. The OECD notes that ARCs can “severely hamper competitive choice” and recommends that, to assist switching, governments should “prohibit any automatic initial contract renewals”.46

Impact assessment

2.35        The analysis presented in this document represents an impact assessment, as defined in section 7 of the Act. In Sections 3, 4 and 5 we discuss all of the relevant factors and options that we have considered, including their impact on stakeholders including both consumers and CPs. It is for the reasons set out in those sections that Ofcom has concluded that its decision will support the competitive process, thereby furthering the interests of citizens and consumers in fixed line telephony and broadband markets. For further information about Ofcom’s approach to impact assessments, see the guidelines, Better policy-making: Ofcom’s approach to impact assessment, which are available on Ofcom’s website.47

Equality Impact Assessment

2.36        Ofcom is also required to assess the potential impact of all our functions, policies, projects and practices on the equality of individuals to whom those policies will apply48. Equality impact assessments (EIAs) assist us in making sure that we are meeting our principal duty of furthering the interests of citizens and consumers regardless of their background or identity. We have given careful consideration to whether or not our proposal to amend GC 9 will have a particular impact on race, age, disability, gender49, pregnancy and maternity, religion or sex equality50. We do not, however, envisage that the decisions contained in this statement will have a detrimental impact on any particular group of people.

2.37        Nor do we envisage any need to carry out separate EIAs in relation to race or gender equality or equality schemes under the Northern Ireland and Disability Equality Schemes. This is because we anticipate that our regulatory intervention will affect all stakeholders equally and will not have a differential impact in relation to people of different gender or ethnicity, on consumers in Northern Ireland or on disabled consumers compared to consumers in general. Similarly, we are not making a distinction between consumers in different parts of the UK or between consumers according to income. Again, we believe that our intervention will not have a particular effect on one group of consumers over another.

Structure of the document

2.38        The structure of this statement follows that of the consultation document. Section 2 has set out some background regarding the emergence of ARCs and our investigation into their effects. In each section we have included a summary of the view of respondents on the points covered by that section together with Ofcom’s view on the points raised in consultation and our conclusions. Hence, the sections run as follows:

•         Section 3 covers the harm from ARCs, and the responses to questions 1, 2, 3, and 4 of the consultation document.

•         Section 4 explains the evidence we have gathered, and covers the responses to questions 5, 6, 7, and 8 of the consultation document.

•         Section 5 explains our conclusions and covers the responses to questions 9, 10, and 11 of the consultation document.

Section 3

The harm caused by ARCs

Introduction

3.1           This section explains our concerns about the potential impacts of ARCs. In our consultation, we used this analysis to explain in detail our concerns about ARCs, to help interpret our evidence and select the most appropriate policy options. We summarise the analysis set out in the consultation, and present and discuss the views of respondents.

3.2           This section covers the responses to consultation questions1-4. In discussing the responses, we separate out those provided by BT and Sky into their own sections since they wrote longer submissions that were critical of a number of aspects of our analysis and which contain a number of general comments that cut across the individual consultation questions. We provide a discussion of the other responses under each separate consultation question.

Summary of the consultation

3.3           In Section 3 of our consultation, we discussed the potential costs and benefits of ARC terms for consumers and set out why we had a priori concerns about their effects. Our concerns centred on the potential for ARCs to increase switching costs in the market in the form of ETCs linked to the MCP. We noted that these switching costs could have both a direct effect on consumers, by preventing them from taking advantage of better offers in the market, and an indirect one, by chilling competition amongst firms.

3.4           We recognised that ARCs may have benefits for some consumers, largely in the form of greater convenience for consumers who expect to renew their MCP, but argued there are reasons to think these benefits are limited.

The benefits of ARCs

3.5           Our consultation acknowledged that contractual innovations can create benefits for consumers and, since intervention therefore carries clear risks of regulatory failure, a thorough assessment of the likely benefits of ARC terms for consumers was warranted.

3.6           We began by noting that since ARCs only alter the process for contract renewal, and do not involve any extra contractual commitment on the part of the consumer, we should separate out the benefits stemming from the MCP from those generated by the ARC term itself.

3.7           MCPs are common in the communications sector and can provide benefits for consumers, in particular where they allow up-front costs to be spread over a period of time (such as equipment subsidies). The size of these benefits may vary with the length of the MCP, with greater price discounts being offered in exchange for longer commitment on the part of the consumer. However, since ARC terms do not increase the length of the contractual commitment, this is not a source of benefits that can be attributed to them.51

3.8           We noted that ARCs may generate savings on customer service costs since contract renewals can be processed automatically without the need for staff involvement. These savings can be passed onto consumers in the form of a lower monthly price. Finally, we noted that ARCs may provide convenience for consumers who expect to renew their contract, since they no longer have to pro-actively contact their CP.

3.9           We discussed a number of reasons why the benefits of convenience might be limited:

•       In the absence of ARCs, CPs should have an incentive to make an opt-in process as convenient as possible;

•       A consumer confident of staying with a given CP would be prepared to sign up for a longer MCP. If ARCs are preferred to this it means the consumer values the option to leave to at least some extent (and will therefore care about the ease of opting-out of renewal as well);

•       The costs of forgetting to renew an MCP must be balanced with the risk of forgetting to opt-out of an ARC. Furthermore, the risks of failing to opt-in are often not paramount since the expiry of an MCP, at least in the residential communications sector, typically means the contract reverts to a month-by-month relationship, and not the cessation of supply. Also, failure to opt-in to a new contract can be remedied straightforwardly by consumers by signing up to a new MPC, whereas failure to opt-out results in a new MCP and ETC.

The harm which results from ARCs

3.10        The main concern we identified in our consultation is that ARC terms may lead to unintentional contract renewal by customers, causing them to continue to be subject to switching costs in the form of ETC. Our argument comprised three elements: reasons why ARCs increase the likelihood of contract renewal, reasons why increased switching costs can harm consumers and competition, and reasons why ARCs might survive in the market despite the harm they cause.

3.11        We identified three mechanisms through which ARCs could increase the likelihood of unintentional contract renewal. Firstly, an ARC reduces flexibility in the timing of the renewal decision. If the renewal date corresponds to a period where the consumer has a lot of competing pressures on their time they do not have the option of putting the renewal decision back. Second, research in the field of behavioural economics identifies a strong ‘default bias’ – that is, a tendency to select default options even when the effort of making a different decision is low–which stems from consumer inertia. Finally, we noted that ARCs allow CPs to offer targeted retention offers in markets where Gaining Provider Led (GPL) switching processes otherwise make this difficult.

3.12        Our analysis of the impact of switching costs built in part on that contained in Ofcom’s recent Strategic Review of Consumer Switching52. We identified a direct effect of switching costs (the inability of individual consumers to change suppliers as better offers emerge or their personal circumstances change) and an indirect effect (the chilling effect on competition among CPs). The impact on competition stems from the reduced number of ‘free’ consumers in the market, which reduces a firm’s incentive to invest in any activity with economies of scale, including marketing campaigns, innovation and new entry.53

3.13        We noted a number of reasons why ARCs may survive in the market despite causing harm. First, consumers may not be aware of the ARC term if terms and conditions are not transparent and, while this is not necessarily an inherent feature of ARCs, the switching costs associated with them prevent consumers from changing contract if they become aware of the term at a later date. Second, drawing again on behavioural economics research, we noted that consumers may suffer from a degree of myopia and focus on the benefits of the price discount at the expense of the future effects of the ARC term. They may also be overconfident about their ability to avoid being influenced by the automatic renewal.

3.14        Third, we noted that the negative effects of ARCs on competition will fall on all consumers, whether they sign up to an ARC or not. This means each individual consumer who agrees to an ARC imposes a negative ‘externality’ on others that they will not factor into their decision. We likened this effect to traffic congestion; each extra driver on congested roads increases the average journey time for everyone and although this effect is small for any one driver, it causes significant delays when aggregated across all drivers.

3.15        We discussed the incentive of firms to introduce ARCs, given the effects we identified. We noted that the reduced customer churn would make them attractive to firms and that this is likely to increase as market shares stabilize and more firms focus their attention on customer retention strategies. We argued that adoption throughout the market could be extremely rapid since the reputational risks to any one firm from selling ARCs would diminish as they become more widespread in the market.

Weighing up the risks and the benefits

3.16        We noted in our consultation that the harm from ARCs could only be established though empirical evidence. However, based on our discussion of the likely costs and benefits of ARCs, we took the view in our consultation that there was a significant risk of harm compared to the possible benefits to consumers. Whereas there are, in our view, compelling arguments that ARCs may cause harm to consumers and competition, the benefits to consumers are much less clear.

Responses to the consultation

3.17        We received 32 responses to our consultation: 8 from CPs, 6 from other organisations and 18 from individuals. We have published the non-confidential responses on our website.54

3.18        In general, the responses were supportive of our proposals, the only two exceptions being BT and Sky. We separate out these responses and provide a detailed discussion of the points they raised in this section and in Sections 4 and 5 below.

3.19        There were also two specific aspects of our proposals that were questioned by a larger number of respondents: the precise wording of the GC and whether it has wider ranging application than we intended, and whether the GC should apply to small businesses. A number of respondents also commented on the proposed timeline for compliance, arguing that that it should be shorter. We discuss these issues in Section 5 below.

3.20        Stakeholders were broadly supportive of our proposals to prohibit ARCs in the fixed voice and broadband markets through appropriate amendments to GC 9:

•       A CP in the market who wished to remain confidential stated:

“In our experience, customers are not aware of the implications of BT’s ARCs until they decide that they might want to switch and have started to look at other offers on the market. Our sales teams report that, very frequently, potential customers have a conversation with us, want to switch to us, but can’t execute their decision to switch because they subsequently realise/remember that their contract with BT has been renewed automatically. The impact is further exacerbated when, for example, the customer’s broadband and line rental commitment periods are misaligned (e.g. 18 months on the former and 12 months on the latter). The ARC therefore leaves customers in perpetual confusion.”

And,

“To be clear, we do not have a problem with having to compete with BT for its customers when those customers reach the end of their existing contracts. Rather, what we are concerned about is that competition does not actually take place where ARCs exist because consumers have little choice but to remain with BT. BT takes advantage of the fact that customers, whether due to lack of understanding, inertia or poor timing (e.g. because they were too busy to switch within the “opt out” window), are unable to switch to us, even if they want to. Where ARCs apply, BT has no incentive to contact the customer with a more compelling deal; rather they rely on contractual terms to tie customers to its network.”

•       Another CP who wished to remain anonymous stated:

“TalkTalk Group welcomes the opportunity to comment on Ofcom’s proposals to ban automatically renewable contracts (ARCs) for residential and small business customers. We raised concerns about ARCs immediately after BT introduced them for residential customers over three years ago. We believe ARCs are harmful in that they raise switching barriers to the detriment of consumer welfare. Essentially, ARCs offer unfair protection against competition in markets with an incumbent (and effectively dominant) provider.”

And, in response to Question 3 (regarding the impact of switching costs on competition):

“We agree with Ofcom’s assessment in so far as residential consumers are concerned. ARCs will lead to increased sales and marketing costs for new entrants because consumers are tied into renewed MCPs which make them more reluctant to switch.”

And, in response to Question 4 (on the likelihood of ARCs spreading in the market):

“We believe that the current use of ARCs by BT as the incumbent provider is symptomatic of the significant amount of switching inertia among residential consumers. Although the retail telephony market is deemed to be competitive by Ofcom, it is clear that ARCs have survived in so far as BT’s product offering is concerned. We agree that there is a risk of ARCs becoming more wide-spread as the market matures and market shares stabilise. This would indeed be in line with a general temptation among larger providers to raise barriers to switching in a mature market.”

•       Scottish and Southern Energy (SSE) stated in response to Question 2 (on whether ARCs increase the likelihood of contract renewal):

“We agree with Ofcom’s points about the reduction in flexibility (paragraph 3.29) for a customer who is likely to have many calls on their time at contract renewal time, the powerful effect of the “default bias” (paragraph 3.34) on customer behaviour and also on the opportunity presented for “save” activity if a customer decides to use his ability to “opt-out” of the automatic roll-over in the ARC. As Ofcom notes, the underlying switching process widely used in the fixed line telephony market is gaining provider led (GPL) while broadband switching is predominantly losing provider led (LPL) which, from our own experience in these markets, results in a lower proportion of successful switching due, amongst other effects, to the “targeted save” opportunity afforded to the LP. In introducing the ARCs for the fixed line market, it appears that CPs are thus also importing aspects of the LPL process into the predominantly GPL switching environment and this in itself is also likely to increase the probability that the customer renews his contract.”

And, in response to Question 3:

“Our experience since entering the market is that, over time, customers have increasingly been affected by being tied into fixed-term contracts. This was not prevalent in the fixed-line telephony residential market when we entered that market – the introduction of contracts could even be observed as a defensive strategy responding to the emergence of competition. Over time, we have observed an increasing tendency for potential customers we speak to being unwilling to move to our talk service due to facing contractual penalties if they do so. Sometimes they are unaware of being in a contract and have proceeded with a decision to switch that they then wish to cancel.

As Ofcom observes, entry into a market at the retail level involves some fixed costs such as investment in systems, processes and marketing.

The ability of customers to readily switch to new offerings is crucial to support a pro-investment climate for market entry and development of new product offerings that bring competition and innovation into the established market. The introduction of contracts in the fixed-line telephony market has increased switching costs generally and the further effect of ARCs analysed by Ofcom increases these costs further – translating into a greater proportion of customers in the market being unable to readily switch to new product offering. We firmly believe that this damages competition.”

•       KCOM Group Plc (KCOM) stated:

“KCOM is supportive of moves to stop the use of automatically renewable contracts in the market. Our experience has been that customer inertia generally results in customers being entered into a new contract term without them necessarily understanding the implications in terms of their future switching options. We have experienced the negative impacts of this in areas where we compete directly with BT. Customers whose contracts have been automatically renewed are more reluctant to switch provider as a result of having to incur early termination charges when cancelling their contract with their existing provider.”

•       Citizens Advice stated, in response to Question 1 (regarding the benefits of ARCs):

“Citizens Advice recognises that ARCs may have some benefits for some customers, namely a price discount and the convenience of not having to actively renew a contract. However, we agree that these benefits are limited – both to the individual consumer and to certain groups of consumer who are well informed and aware of their rights – and we believe that they are outweighed by the harm that is caused to customers who fail to opt out. Evidence from Citizens Advice Bureaux indicates that this harm can be significant and we often hear about people who have found that their contract has been renewed without their knowledge, because they did not fully understand the terms and conditions, or in some cases against their express instructions.

We are by no means convinced that ARCs are developed and sold simply in order to benefit consumers, and we agree with Ofcom’s conclusion that they are, in fact, a mechanism to deter switching by exploiting customer inertia. In the worst cases communications providers appear to be deliberately exploiting the permissibility of automatically renewable contracts in order to prevent customers from switching.”

Sky’s response

3.21        In its response to our ARCs consultation, Sky disagreed strongly with our analysis and our proposals to modify GC 9. Acknowledging that it had previously expressed significant concerns to us, Sky argued that the now lower ETCs55 mean the situation has changed and also noted that no other major CP has introduced ARCs since 2008. Sky took the view that our proposed intervention is ‘paternalistic’ and that we had not fully explored the use of our existing powers, specifically the UTCCRs56 and Consumer Protection from Unfair Trading Regulations 2008 (the CPRs),57 to tackle transparency problems surrounding BT’s contract before proposing to modify the GC.

3.22        Sky commissioned a review by Plum Consulting of our economics analysis, in particular our use of behavioural economics, which contains a number of criticisms. Drawing on this report, Sky argued that we had relied too much on behavioural economics and had not demonstrated sufficiently that behavioural factors, rather than a lack of transparency of terms and conditions, lay behind the observed effects of BT’s ARCs. Sky referred to a number of the findings of the transparency research conducted in summer 2009 and argued that deficiencies in BT’s marketing practices could explain the effects of BT’s ARCs. Sky also noted that the research shows most BT consumers seemed happy with their contract even once fully informed of their terms and conditions and that we had not published numbers on consumer complaints.

3.23        We separate out our assessment of Sky’s response as follows. This section addresses the arguments contained in the Plum Consulting report regarding our economic analysis. In Section 4 we address the argument that low ETCs now mitigate the harm from ARCs, that a lack of transparency of terms and conditions explains the results of the econometrics and that there is no evidence of consumer harm. In Section 5 we address the argument that a GC is inappropriate and that we have not fully explored other avenues of enforcement that would be preferable to a change in the GCs.58

Sky’s response to our economic analysis

3.24        As noted above, Sky’s response to the economic analysis we presented in Section 3 of our consultation is contained in a report by Plum Consulting. The report focused on our use of behavioural economics, our argument that ARCs create externalities amongst consumers (that is, consumers accepting ARCs indirectly reduce the intensity of competition thereby creating negative ‘spillovers’ between consumers59), and our claim that ARCs could spread rapidly in the market in the future.

3.25        Regarding behavioural economics, the report argues that caution is needed when moving from the academic research to real world applications. Apparently irrational behaviour may have perfectly rational explanations and individuals may learn over time to improve their decisions.

3.26        We agree there are challenges in using behavioural economics research to identify consumer error in real world markets. However, the ‘default bias’ that we referred to is a very specific prediction from the research and has been identified as having powerful effects across numerous real world situations including in very important decisions such as pensions.60

3.27        The result is striking because the direction of a default option ought to have little effect on decisions if the cost of making a choice is low. It is difficult to identify clear rational explanations for this effect. The Plum Consulting report argues that apparent inertia may in fact be a rational decision once we account for the fact that decisions are often difficult to reverse and are made in an environment of uncertainty. The report makes reference to the economic literature on investment decisions by firms which shows that what looks like inertia in investment decisions may be a rational ‘wait and see’ strategy, in which the option to invest is kept open until more information is available. However, we do not think this offers a good explanation for why inertia in contract renewal decisions might be rational since doing nothing in the context of an ARC means being locked in to another MCP, thereby making the option to switch CP in the future more costly.

3.28        We also considered in the consultation the possibility of learning over time and argued that this was unlikely to have a strong effect since the MCP means it is costly to reverse decisions once the consumer is aware of the mistake.61 It is also important to consider that renewal decisions are relatively infrequent and consumers do not receive instant feedback about the impact of their choices, but instead at a future date when they decide to switch supplier. Both of these factors are likely to undermine the ability of learning to mitigate the effect of ARCs.

3.29        Plum Consulting also notes that individuals often resort to commitment devices to help compensate for possible biases in their choices. They provide the example of gym membership and note that paying upfront could be a rational strategy to encourage yourself to visit more often.62 Plum Consulting does not state how this might apply in the case of ARCs but one possibility would be that individuals sign on to ARCs because they are concerned about forgetting to renew their MCP. However, as we noted in the consultation the risk of failing to renew must be balanced with the risk of failing to opt-out of an ARC and it is not clear that the former would outweigh the latter, particularly in dynamic markets where the option to switch at a future date is very valuable,63 and where failure to opt-in can easily be remedied without charge, whereas failure to opt-out cannot.

3.30        We also note that inertia and a ‘default bias’ was not the only mechanism through which ARCs can encourage unintentional contract renewal. As we discussed in the consultation,64 ARCs remove the ability to delay a renewal decision and mean a consumer may end up renewing their contract if the renewal date coincides with a period where they are subject to a lot of demands on their time, for example at work. In reality, the boundary between the behavioural explanation and the rational one is likely to be blurry with most consumers subject to a mixture of the two forces.

3.31        Plum Consulting also questions our analysis of a possible ‘negative externality’ between consumers resulting from ARCs. Our argument was that consumers who sign up to an ARC indirectly harm other consumers since, with fewer ”free” consumers available in the market (i.e. those not subject to higher switching costs), there is less incentive for firms to invest in customer acquisition. We likened this indirect harm to others (or ‘externality’) to traffic congestion, emphasizing that the effect per individual customer is small but may add up to significant effects. Plum Consulting notes that the economic research we cited that explores this effect relates to a monopolistic market not a competitive one, and that externalities created through market transactions should not be regarded as genuine externalities (instead they are ‘pseudo’ or ‘pecuniary’ externalities). Plum Consulting also noted that long term contracts can have countervailing benefits.

3.32        We cited the paper by Rasmusen et al65 because it was one of the first to show that consumers might agree to harmful contract terms because their individual choices are not coordinated. The combination of switching costs created by contract terms (or in the Rasmusen paper, exclusive dealing clauses) and economies of scale mean individual decisions impact on other consumers. This insight is valid on its own, and stands apart from the rest of the paper (which concerns how a monopoly can exploit this effect to prevent entry). Furthermore, this effect can be legitimately regarded as an externality since, if consumers could coordinate their decisions, they could be collectively better off by agreeing on a different outcome (i.e. not signing up for ARCs).66

3.33        We agree that any potential harm from these negative externalities needs to be balanced against the benefits of long term contracts. As we recognised in the consultation during our discussion of the option of prohibiting subsequent MCPs, long term contracts may create benefits for consumers.67 However, we argued in the consultation that the benefits of ARCs (as opposed to the underlying MCP) are unclear.68

3.34        Plum Consulting’s report also questions our conclusion that ARCs are likely to spread in the market. We argued that adoption was more likely going forward as the relevant markets mature. Plum Consulting argued that the market was in a state of dynamic transition associated with the upgrade to Next Generation Access networks and smartphones. Plum Consulting further argued that widespread adoption of ARCs is not certain since individual firms may find benefits from transparently offering non- ARC contracts.

3.35        Our argument was based on the observation that the retail market for fixed voice is starting to settle after the rapid entry and expansion over the past few years based on access to BT’s network, in particular Local Loop Unbundling (LLU). BT’s market share has steadily declined in recent years as entrants have rolled out their own networks across the country, but this process will inevitably settle. BT’s share of residential exchange lines has fallen from around 71.2% in 2006 to 50.7% in the fourth quarter of 2010.69 Meanwhile, the growth of LLU will level off since the remaining unbundled exchanges will be in areas with lower levels of population density, making unbundling a less profitable investment.70

3.36        It is in this sense that the market is maturing; it will of course remain technologically dynamic, particularly at the wholesale level, but now that entrants such as Sky and TalkTalk have built large customer bases we think they will turn their attention more to customer retention. Given the significant retention benefits of ARCs that we have identified we think they will be an increasingly attractive option for CPs and we do not take comfort from the fact that they have not been introduced as yet. We also highlight a point made by SSE in its response that the spread of ARCs may have been delayed owing to uncertainty about Ofcom’s position regarding them and that they could spread quickly if we took a lenient approach:

“In relation to the potential spread of ARCs, we believe that up to now, there has been some uncertainty over the status of ARCs due to the comments Ofcom has made in various consultations about its investigation into the effects of the practice. If, following consultation, Ofcom decided that in some circumstances, it would be acceptable for CPs to offer ARCs, that in itself could result in more CPs quickly adopting them.”71

3.37        Regarding reputation, it is inevitably less of a reputational risk for an additional firm to introduce ARCs if more of its competitors have already done so and because of this we suggested the introduction of ARCs could be rapid. Plum Consulting notes that a firm might have the opposite incentive and seek to win customers by transparently offering non-ARCs terms. However, such a firm would be foregoing significant retention benefits and it is not clear that many customers would switch CP on the basis of the presence or not of the ARC term, as opposed to factors such as headline prices.

3.38        We accept that it is difficult to forecast developments in these markets. The relevant question for us is whether the current harm from ARCs and the risk of ARCs spreading is sufficiently high that ex ante regulation is warranted. For the reasons set out in Section 5 of our consultation, we believe this is the case.

3.39        Plum Consulting also discussed the position of other regulators on behavioural economics and ARCs and Ofcom’s own position on net neutrality. We deal with these in Section 5 below.

BT’s response

3.40        BT reiterated its view that ARCs are beneficial for consumers and, referring to the steps it has taken to ensure consumers are informed of the terms and conditions, stated that it did not believe there is justification for intervention from a consumer protection perspective. Regarding the impact of ARCs on competition, BT criticised our economic analysis, questioning in particular our view that there is a clear link between switching costs and competition. BT also noted that many of our arguments regarding contractual switching costs could apply to MCPs in general, which we are not proposing to ban. We address these points in this section. BT also provided further criticisms of the econometrics evidence and we discuss these in Section 4 below.

3.41        BT further argued that, if we did go ahead with our proposed GC, it should not apply to small business customers and that our timeline for implementation should be extended by six months. We address BT’s arguments concerning business customers and implementation in Section 5 below.

BT’s response to our economic analysis (residential consumers)

3.42        BT agreed with our assessment of the source of benefits from ARCs but argued we had underestimated their magnitude. BT argued it is very convenient for consumers to continue receiving benefits from renewal without having to opt-in. BT did not regard a longer MCP as an appropriate alternative for consumers.

3.43        In response, we note the convenience provided by ARCs depends on the amount of time and effort that would be required for renewal under an opt-in system. BT did not provide any arguments as to why an opt-in process would be particularly burdensome or inconvenient in the absence of ARCs. We accept that some extra effort would be required, but think this is low because CPs will have an incentive to make the opt-in renewal process as hassle-free as possible. We also note that we did not claim that consumers would regard a longer MCP as an equal substitute for an ARC; our point was that a consumer that does not want a longer MCP clearly cares to some degree about the ease of opting out of renewal. The convenience of not having to opt-in therefore must be balanced against the extra effort of preventing contract renewal.

3.44        BT argued that we have not demonstrated that higher switching costs and lower switching levels are damaging to the competitive process. Citing recent economic research,72 BT said the impact of switching costs is an empirical matter and we have not demonstrated an effect in the fixed voice or broadband markets. BT also noted that many of the problems of ARCs that we identify also apply to MCPs in general, which we are not proposing to ban. In particular, many of our arguments could apply to MCPs of 18 or 24 months.

3.45        Regarding the impact of switching costs, BT cites recent empirical and theoretical research which purports to show that increased switching costs can increase price competition and benefit consumers. The cited research considers the question of how firms weigh up the incentives to “harvest” existing locked-in customers by increasing prices with the incentives to “invest” in new customers by cutting prices. The research shows that, where switching costs are relatively low such that consumers are still prepared to change CP, an increase in switching costs can, in theory, increase the incentive of firms to invest by more than the incentive to harvest, thereby increasing competition.

3.46        This research is very recent compared to the literature on switching costs as a whole and it is as yet uncertain how robust the results are to different assumptions.73 We also note that the theoretical models’ predicted effects have not been clearly demonstrated empirically.74

3.47        Our main point, however, is that the issue of how switching costs affect incentives to “harvest” or “invest” did not play a central role in our analysis of the switching costs associated with MCPs. As we said in the consultation, contracts in communications markets involve a fixed price for the duration of the MCP, which removes the ability to harvest consumers who are locked-in by an ETC.75 Instead, we focussed on arguments of how switching costs limit the ability of firms to exploit economies of scale and scope by reducing the pool of “free” consumers in the market. We said this could harm price competition by reducing marketing activity, but we were also concerned with incentives for entry and innovation.76 None of the papers cited by BT consider these effects of switching costs.

3.48        BT argues that the question of whether switching costs reduce competition is an empirical one. However, in our view a legitimate presumption can be made that switching costs that do not generate clear countervailing benefits are harmful in a market where entry and innovation play a key role, particularly where an incumbent firm has a large established customer base. The ability of consumers to change CP in order to take advantage of new offers is a fundamental driver of competition, and the burden is on those who wish to demonstrate otherwise. BT has not provided evidence that switching costs benefit competition in communications markets.77

3.49        BT presents trends in Average Revenue Per User (ARPU) over time and argues that there is no indication that the introduction of ARCs has led to an increase in prices. We find this evidence of limited relevance as the real question is whether prices are higher than they would otherwise have been, not whether they have risen. ARCs are only one of a number of factors affecting competition and ultimately prices in the market. All else being equal, everything that has a material detrimental effect on competition over time will increase prices compared to the prices we would observe absent that particular set of practices (or contracts as in this case).

3.50        BT claims that many of our arguments could also apply to MCPs of a longer duration than 12 months, which we are not proposing to prohibit. The relevant arguments in the consultation that BT is referring to are those regarding the effects of switching costs78 and those made in our discussion of how harmful contracts may exist in a competitive market.79 These arguments do have broad application, since they relate to any switching costs created by contract terms, and we would agree they could in principle be applied to longer MCPs. Consequently, if MCPs did not carry clear countervailing benefits, we might have similar concerns.

3.51        However, as we acknowledged in the consultation,80 MCPs can create benefits for consumers (though this is less clear in the absence of equipment subsidies) and it could therefore be disproportionate to ban them or overly limit their length. Our concern with ARCs is the presence of switching costs that generate limited countervailing benefits for consumers. We also note that recent changes to the USD (discussed in Section 2 above, starting at paragraph 2.17) mean that MCPs are limited to a maximum of 24 months (for residential consumers) and CPs must also offer 12 month contracts.81 These changes to the Directive acknowledge the potential harm from contractual switching costs associated with lengthy contracts.

3.52        BT also argued that we had not provided empirical evidence suggesting ARCs would spread rapidly in the market. BT noted that no other CP had introduced ARCs even though most large CPs are very much focussed on customer retention. We discuss this issue above in response to Sky who also challenged our analysis. We note here that our argument was that ARCs would look increasingly attractive to CPs as the potential to capture further market share from BT is reduced. The years since 2008 have seen an increasing erosion of BT’s market share at the retail level, largely as entry and expansion based on LLU has developed. As the impacts of this start to settle we believe the relative focus on retention strategies as opposed to acquisition will increase.

Other responses

3.53        The other responses we received were generally supportive of our analysis. Stakeholders broadly agreed that ARCs can cause unintentional contract renewal and that the resulting switching costs can harm competition. There was also agreement that ARCs could spread if Ofcom did nothing following our consultation.

3.54        Finally, no respondent suggested sources of benefits for consumers from ARCs that we did not consider in the consultation, or provide evidence or new arguments against our preliminary conclusions that the benefits were relatively low.

3.55        A few respondents offered broad comments without answering our specific consultation questions. Virgin Media (Virgin) stated general concerns that, as with our recent switching consultation, we rely heavily on “behavioural economics and theoretical econometric evidence that lacks robustness to reach a conclusion that ARCs produce harmful effects on switching”. However, Virgin offered no specific criticisms of the analysis contained in Sections 3 or 4 of our consultation. Virgin’s main concern was that the wording of our GC would have too far reaching effects. We address this in Section 5 below.

3.56        Which? gave its support to our proposals and stated that ARCs are based on an unreasonable expectation for consumers to actively monitor the status of their contract and to pre-empt any future desire to switch contract. Which? referred to the results of some consumer research it carried out in March 2010, which found a general lack of awareness of terms and conditions by customers on ARCs. We explain the relevance of the transparency of ARCs arrangements in our discussion and conclusions below on responses to question 6.

3.57        A CP who wished to remain confidential agreed that ARCs can cause consumer detriment and that prohibiting opt-out processes should minimise this. The CP agreed that ARCs could result in a hindrance to the switching process. KCOM was also supportive of our proposals, commenting that in its experience inertia causes consumers to be entered into a new contract without necessarily understanding the implications. KCOM said that it had direct experience of this in those areas where it competes directly with BT. KCOM had some further comments on the precise wording of the GC which we address below.

3.58        A CP who wished to remain confidential said that its view based on “on the ground” observation was that BT’s ARCs are having a negative material effect on consumer switching. The CP said that sales teams often report that customers abandon decisions to switch when they realise or remember their contract with BT has been renewed. The CP noted that the problem is exacerbated where commitment periods for broadband and line rental are unaligned. The CP did, however, disagree with our analysis regarding the potential of ARCs to create customer save activity, saying that ARCs would not generate extra contact by the customer. We disagree since the act of opting out (by those customers who remember to do so) signals an intention to leave and provides an opportunity for BT to engage in retention activity. However, we note our proposals did not rest on this aspect of our analysis.

3.59        Almost all responses from individual consumers supported a prohibition of ARCs. Consumers mentioned the difficulty with bundles and the fact that ARCs diminish incentives for CPs to provide quality customer service to retain customers as they rely on the automatic renewal as a retention tool. Most individual responses expressed a wish for immediate compliance and many believed that mobiles should be included in the scope of the prohibition.

Question 1

3.60         Question one asked:

“Do you agree with our assessment of the potential benefits of ARCs for consumers, including the source of benefits and their magnitude? Please provide evidence to support your answer”

3.61        TalkTalk agreed that there are very few benefits of ARCs for residential customers and said they are outweighed “by some distance” by the harmful aspects of ARCs. TalkTalk noted that legitimate uses of ARCs are associated with services that residential consumers are legally obliged to have, such as car or house insurance. TalkTalk did, however, argue that there were benefits for small businesses on the basis that, in TalkTalk’s experience, small business customers appreciate the convenience of remaining with their CP at a competitive price, and of avoiding the hassle of proactively renewing their contract.

3.62        SSE stated that it was dubious about the alleged benefits of ARCs and agreed that, in highly dynamic communications markets, the option to switch supplier has significant value, making consumers on ARCs worse off than those on contracts with opt-in renewal processes. Federation of Communication Services (FCS), which represents over 300 businesses, over half of which are smaller CPs and new entrants delivering telephony into residential and business markets, also agreed that the potential benefits are negligible, noting that price discounts are more likely to derive from the MCP rather than the ARC. Citizen’s Advice agreed that the benefits of ARCs are limited and said that the primary function of these contracts is to deter switching. It provided a number of examples of cases it had dealt with where consumers had suffered harm as a result of ARC terms.

3.63        Ombudsman Services: Communications made no comment on the potential benefits of ARCs, but did question whether a comment we made about the possibility of ARC terms being optional at the point of sale82 meant we considered it acceptable for CPs to offer ARCs in this way. For the reasons set out in the consultation, we do not think extra transparency at the point of sale (which is essentially what this would entail) would mitigate the harm we identified.83

Question 2

3.64         Question two asked:

“Do you agree with our assessment that ARCs are likely to increase the probability of contract renewal (despite only altering the process for renewal)? Please explain your answer.”

3.65        TalkTalk agreed with our assessment as applied to residential consumers, and stated that consumer inertia is a significant factor for CPs seeking to take market share from incumbent CPs. SSE agreed with our analysis, in particular regarding how ARCs allow a CP to import some aspects of a Losing Provider Led (LPL) switching process into what is largely a Gaining Provider Led (GPL) market. FCS agreed with our assessment and said that inertia is undoubtedly the major factor in customers allowing a contract to roll over to a new MCP. The Centre for Competition Policy (CCP), at University East Anglia also agreed with our assessment.

3.66        Ombudsman Services noted that, based on its experiences with complaints, the process for opting out can be too onerous (for example it must be in writing) and that a reminder letter is often never sent. These concerns appear to relate to smaller CPs rather than BT. It also noted that for small business contracts there is often only the option either to roll over or terminate the service completely, without the ability to move to a monthly contract. However, the issue of whether the ability to move to a monthly contract should be mandated is outside the scope of this statement since it is separate from the issue of ARCs per se.

3.67        Lexgreen Services Ltd agreed in general with our assessment, noting that one aspect we had not accounted for is that BT’s small business ARCs only allow the opt-out decision to be made at a narrow time window at the end of the MCP. We agree that where contracts include this restriction it would exacerbate the effects of ARCs.

Question 3

3.68         Question three asked:

“Do you agree with our assessment that increased switching costs resulting from ARCs are likely to harm competition, in particular the incentives for new entry and innovation? Please explain your answer.”

3.69        TalkTalk agreed with our analysis with respect to residential consumers. SSE agreed strongly with our assessment, stating that in its experience since entering the market consumers have been increasingly affected by being tied into long term contracts. SSE argued this is a recent development representing a defensive reaction by incumbents to entry and that it firmly believes this damages competition. FCS agreed that increased switching costs are harmful to competition. Lexgreen Ltd agreed with our assessment.

CCP also agreed with our assessment and, referring to some examples, said that the theoretical literature supports this view.

Question 4

3.70        Question four asked:

“Do you agree with our analysis that ARCs could survive in a competitive market, despite causing harm to competition. Do you agree that there is a risk of ARCs spreading in the market and that this could be rapid? Please explain your answer.”

3.71        TalkTalk agreed that there is a risk of ARCs becoming more widespread as market shares stabilise. This would be in line with the temptation among larger CPs to raise barriers to switching in a mature market. SSE noted that up to now there has been some uncertainty about the status of ARCs following references made by Ofcom about its investigation into their effects, and that ARCs could spread quickly if our consultation process resulted in them being allowed. Citizen’s Advice agreed with our analysis and said that ARCs can also survive because some groups of consumers are less able to shop around, for example because they do not have internet access. It agreed that there is a risk of ARCs spreading in the market. Lexgreen Ltd agreed with our analysis, commenting that CPs would copy competitors who introduced them in a “race to the bottom”.

3.72        CCP agreed ARCs could survive in a competitive market. On the risk of ARCs spreading it said that it would be worth looking at energy markets to help understand this issue. While this could offer some insights, our view is that there are too many differences between the sectors for us to be confident in making extrapolations between them on the likelihood that ARCs will become more prominent in the market. As such, we have not based our conclusions on extrapolations from the energy market. Our reasoning as to why there is a risk of ARCs spreading was based on market dynamics specific to the communications sector, in particular the maturing of retail shares following LLU based entry. CCP also noted that if the regulator does nothing, this will send out a signal that CPs will respond to.

Conclusions

3.73        Most stakeholder responses, with the exception of BT, agreed with our assessment of the potential benefits of ARCs for residential consumers – namely that relative to the risk of harm, it is not clear that ARCs offer significant benefits for consumers. For small businesses however, BT and TalkTalk considered that ARCs have much greater benefits and should not be prohibited in this market.

3.74        Most stakeholders also agreed that ARCs increase the likelihood that a consumer’s contract will be renewed even though these terms do not affect the available options or the level of commitment on the part of the consumer (merely the process for making a choice). ARCs increase the (aggregate) level of the switching costs in the market, harming both the individual consumer and competition. This is a particular concern if ARCs become an established feature of the market, offered as standard by many CPs. Most respondents apart from Sky and BT agreed that, despite the harm they cause, ARCs may nonetheless survive in a competitive market and may become more widespread over time, as more firms turn to a customer retention rather than acquisition strategy.

Section 4

Review of evidence

Introduction

4.1           Section 4 of our consultation presented and assessed the evidence we gathered. We made use of the analysis presented in Section 3 to interpret the evidence in the light of our concerns regarding the impact of ARCs. We discussed in detail the robustness of the evidence and paid particular attention to the issue of whether the evidence we gathered on BT’s fixed voice customers could be extrapolated to other CPs and other sectors.

4.2           This section covers consultation questions 5-8. As in Section 3 of this document, when discussing the responses we separate out those provided by BT and Sky into their own sections. We provide a discussion of the other responses under each separate consultation question.

Summary of the consultation

4.3           Our primary evidence was a piece of econometric (i.e. statistical) research into the switching behaviour of BT customers that we commissioned from Professor Gregory Crawford and ESMT Competition Analysis. This was supplemented by three pieces of research focussing on the transparency of terms and conditions and consumers’ understanding of ARC terms.84

Economtric analysis

4.4           The econometric study aimed to identify the impact of ARC terms by analysing patterns of switching among BT customers. The research used data on the tariff choices and switching behaviour of 140,000 BT fixed voice customers85 during a period of 15 months ending in March 2010, along with information on prices in the market and the general state of the economy. The study employed statistical techniques to control for a number of factors that can be expected to influence switching, including the price discount offered as part of the contract and the length of time a customer had been with BT. This allowed us to isolate the incremental effect of the ARC term in a way that a simple comparison of the switching rates of ARC and non-ARC customers would not.

4.5           The study also controlled for ‘self-selection’ by BT customers.86 This refers to the possibility that customers who choose ARCs are less likely to switch than those who do not owing to unobserved factors that we do not measure and therefore cannot control for. This leads to the risk that ARCs are falsely identified as causing reduced switching.

4.6           The econometrics found that ARC terms reduce the rate of customer switching after the initial 12-month MCP by 0.33 percentage points relative to the baseline switching rate in the sample of 0.95% per month (i.e., from 0.95% per month to 0.62%). This was also 0.75 percentage points lower than the switching rate for comparable customers on fixed term contracts after the MCP ended. This is illustrated in Figure I, below.

Figure I: Estimated monthly switching rate after first MCP, by contract type, controlling for price discount and self selection

 

Source: Empirical analysis of BT’s Automatically Renewable Contracts, Professor Gregory Crawford and ESMT CA, August 2010.

 

4.7           As we acknowledged in our consultation, the data used in deriving these estimates relate to a time period in which BT charged higher ETCs than it does today.87 We asked Professor Crawford to use the available data to estimate the effect of lower ETCs on switching rates and attempt to predict the effects of the lower ETCs. Although this is only a prediction, since customers in the database are subject to a wide range of ETC levels, depending on their specific contract and the months remaining on their MCP, we were able to get a good idea of how they would respond to different ETCs.

 

4.8           Professor Crawford’s analysis is contained in an additional piece of analysis presented in Appendix 2 of his main report. The average ETC faced by a rollover customer in the sample during their second MCP was £54.80. At this ETC level the rollover term in the contract was estimated to reduce switching in the second MCP by 0.5 percentage points (from 1.92% per month to 1.42%). This estimate is broadly consistent with the main findings presented in the report, though slightly lower (a 26% reduction in switching rates rather than 34.6%).88 The estimated effect of reducing this ETC charge to £19.20 was to increase switching rates for customers on ARCs to 1.56% per month, reducing the impact of ARC from 0.5 to 0.36 percentage points. Though lower, this remains a material effect.

4.9           Consequently, we concluded in our consultation that even with lower ETCs, ARCs can be expected to have a material effect on consumer switching.

Robustness and extrapolation

4.10        Our consultation considered in detail a number of issues relating to the study including the robustness of the results, the likely underlying causes and whether the results can be extrapolated to other circumstances (for example, to CPs other than BT).

4.11        Our discussion on the robustness of the results built on the comments we received from BT in two letters89 after sharing the report with them and centred on two issues: whether the study adequately addressed the issue of self-selection, and whether the impact of the price discount associated with ARCs had been correctly identified.

4.12        The study found evidence for a small amount of self-selection in the sample and the qualitative conclusions were not altered once this was controlled for. The report is clear that the statistical techniques used to control for self-selection cannot guarantee to address for the entire problem and some self-selection may remain. However Professor Crawford notes that any residual selection bias would affect multiple parameter estimates, in particular the estimate for the effects of tenure with BT. In this context it is not possible to predict in which direction this would affect the rollover estimate.

4.13        Adequately controlling for the price effect is crucial for separating the impact of the ARC term from the price discount associated with BT’s rollover offers. The main study used a number of sources of price variation to estimate how consumers respond to lower prices, including changes in competitors’ prices and differences in prices between BT’s contracts. BT questioned the usefulness of some of these sources of price variation, including the loss of temporary discounts by customers within their initial MCP.

4.14        We asked Professor Crawford to look in more detail at the sources of price variation used in the study. His supplementary report90 confirmed that BT was right to be concerned about price variation generated by the loss of temporary discounts, since this does not appear to generate a consumer response, but that this was not an important source of information for the estimates. The primary source was a comparison between two rollover offers that were identical except for the size of the price discount. As Professor Crawford notes, this is as close as we can come to a ‘natural experiment’ and if we rely on this data alone, we continue to estimate a similar price effect and therefore similar effects of rollover terms.

4.15        BT also argued that we had not taken into account customer’s expectations about price increases or ‘loss aversion.’91 BT argued that much of the price variation in the data represents anticipated price increases that would not generate switching by consumers. The effects of these price changes could not be used to infer the effect of the ARC discount. BT further argued that the discount that rollover customers receive generates a loyalty effect that we do not account for. We did not find its arguments persuasive, since price increases will affect the benefits of switching CP even if they are fully anticipated and also there is no evidence that loss aversion is either present or would have the effects BT suggests.

4.16        As regards underlying causes, we took the view that the impact of ARCs that we identified mainly stems from core aspects of consumer behaviour, such as the ‘default bias’, rather than specific features of BT’s offer.92 On this basis we concluded that the results could be extrapolated to other CPs in the fixed voice and broadband sectors. That is, we would expect qualitatively similar effects for any version of an ARC introduced by any supplier in these sectors. Because ARCs have not emerged in the mobile sector, we did not take a firm view on whether the results could be extrapolated to mobiles. We noted, however, that we would have concerns if ARCs did emerge in the mobile sector and would urgently revisit the issue.

4.17        Finally, we argued that the conclusions could be applied to those small businesses which can be expected to behave in a similar manner to residential consumers.

Transparency research

4.18        During 2009 we commissioned three pieces of research aimed at understanding whether consumers understood ARC terms and whether BT’s marketing was transparent. Our consultation used the research to interpret the results of the econometrics. In particular, we drew from it our conclusion that, despite some evidence of shortcomings, a lack of transparency was not the main driver of the results of the econometrics.

4.19        The three pieces of research were:

•       Telephone interviews with 600 customers on ARCs (Spring Research);93

•       A mystery shopping exercise (eDigital Research);94

•       An analysis of call recordings taken from two BT call centres (Mott MacDonald Research).95

4.20        The results were somewhat mixed. Customers generally had a good understanding of the terms and conditions of their contract. However, 59% of ARC customers in their second MCP did not realise they had been rolled over and, in the mystery shopping exercise, just over half of callers were not told that the ARC contract automatically renewed at the end of each contract period.

4.21        Of most relevance for understanding the results of the econometric study is the transparency of terms and conditions at the rollover date. Although 59% of ARC customers in their second MCP did not realise they had been rolled over, since BT does send a reminder letter this is likely to be because consumers did not absorb the information that was provided rather than that BT did not make the information available. Overall, we concluded that a lack of transparency of terms and conditions was not the prime driver of the results we identified in our econometric analysis.

Responses to the consultation

Overview of responses

4.22        In general, the responses from BT and Sky reflected the most opposition to our proposals. Most respondents agreed that our findings were intuitive and reasonable, and that we had drawn the correct conclusions from the research. BT was the only respondent to offer technical comments on the econometrics, and mainly took issue with this analysis, while Sky expressed the view that our findings could be explained by transparency issues. Other than Sky, there was widespread agreement that transparency issues alone could not explain our findings, though some respondents did express concerns about the transparency of BT’s offer. There was also some disagreement over whether the findings could be extrapolated to small businesses.

Sky’s response

4.23        Sky disagreed with our interpretation of the evidence, in particular our conclusion that the results of the econometrics could not be explained by a lack of transparency of terms and conditions. This is an important issue since it was our conclusion that the effects stemmed from the opt-out nature of the renewal decision that led us to select a GC as the most appropriate means of tackling the problem. Sky also argued that we had not shown the effect would persist with the relatively low ETCs that have been in effect in the fixed voice market since around the middle of 2010. Finally, Sky noted that we had not published our own complaints data and questioned whether this was because the data does not support our case.

4.24        Sky cites a number of results from our transparency research that was carried out in summer 2009. It noted the finding in the eDigital mystery shopping exercise that between 54% and 57% of callers were not told that ARCs automatically renew at the end of each contract period. It also referred to conclusions by Mott MacDonald that BT’s call centres had shortcomings in their procedures and practices.

4.25        As we said in our ARCs consultation the results of the transparency research were mixed. However, we disagree that the shortfalls that were identified can explain the findings of the econometrics. What is important for interpreting the results of the econometrics is whether consumers have the information they need regarding their contract around the renewal date. Most of the issues Sky highlights are around transparency at the initial point of contact with a sales representative, but this is not the only source of information consumers have about their contract by the time they come to the renewal decision. After signing up, consumers would have received detailed terms and conditions in the post (and these were available online) and we believe BT made these accessible and clear. Furthermore, and of critical importance, BT sent customers a reminder letter before the date of the renewal. In the examples we have seen, this letter was written in a clear manner and should be easily understood by consumers. Sky offered no arguments as to why the existence of the reminder letter would not create transparency at the point of renewal.

4.26        Sky claims that we have made our conclusions without evidence and at one point even alleges that we concede this point ourselves.96 Sky also claims that we rely on theoretical rather than actual consumer harm. The paragraphs of the consultation that Sky references97 concern the underlying cause of the results identified by the econometrics, not whether ARCs have a negative effect on switching and are therefore harmful. At this point of the consultation we had already presented and discussed in some detail a significant amount of econometric evidence that ARC terms reduce switching in the market, despite only affecting the process for contract renewal. Sky therefore seems to have misunderstood our assessment. Sky disagrees with our judgement about the underlying causes for the effect we observe, but provides no additional evidence or arguments that we did not consider in our consultation to come to our preliminary conclusions.

4.27        Sky also referred to the results of our consumer survey which shows that many ARC customers were happy with their contract and did not intend to leave BT. Sky suggests Ofcom is focussing too much on switching levels and should accept that many consumers are actively choosing to remain with their CPs.

4.28        We would expect BT to have many long term loyal customers who have little intention of switching provider and we agree this may be the right decision for them. However, our econometric analysis demonstrated a significant incremental reduction in switching caused by ARC terms, even though the term only affects the process for contract renewal not the level of commitment by the customer. Our position is not that more consumers ought to switch supplier but that unnecessary barriers to switching should be removed. We made this point clearly in our consultation,98 so we do not believe that Sky’s argument is well-founded.

4.29        Sky also highlights that ETCs are now lower in the market than when our econometric research was conducted and states that we should do more work to understand the effects of lower ETCs. Our consultation explained in some detail why we concluded that low ETCs would still have an effect.99 In the data used in the econometrics, BT’s customers are subject to a wide range of ETCs depending on their precise tariff and the number of months remaining in their MCP, and this allows us to determine how they respond to different ETC levels. The statistical analysis suggests that even low ETCs still deter consumers from switching (and we noted that this can also be seen graphically, since switching rates jump significantly in the final month of an MCP – see Figure 4.1 in the consultation). We were clear that this is only a prediction of the effect of the new ETC levels, however we think it constitutes strong evidence since it is based on the actual switching decisions of BT’s customers. Sky offered no reasons as to why our econometric analysis would not provide a good prediction.

4.30        We also note that the UTCCRs cannot guarantee low ETCs in all cases. Generally, ‘fair’ ETCs will be low where the costs that can be mitigated after termination are a significant proportion of the headline price.100 This is not the case in broadband to the same extent as fixed voice, particularly where LLU is employed.

4.31        There are a number of points to make regarding complaints data. First, a significant element of our theory of harm relates to the wider effects of ARC terms on competition, which would not show up in complaints by individual customers. Second, we have also been looking ahead to future harm caused by the spread of ARCs which clearly would not be reflected in current complaints. Finally, it is not easy to isolate ARC complaints as they may end up being recorded under different categories. In particular, they would often show up as general complaints about MCPs and ETCs even though the underlying issue is the presence of an ARC. For these reasons complaint data was mainly used initially to guide areas of concern for Ofcom, and we have placed little weight on this source of evidence for our assessment of the impact of ARCs.

BT’s response

4.32        In its consultation response, BT repeated the argument it previously made in a letter sent to us in October 2010 that we have not adequately accounted for “loss aversion” among ARC customers when trying to isolate the impact of the price discount, and therefore took the view that we are overestimating the effect of the price discount. BT also argued that we had failed to separate out the impact of ARC terms from that of ETCs and that, once this is done, the impact of ARCs is much lower.

4.33        BT argues that because of loss aversion the price discount creates a loyalty effect among ARC customers that the econometric analysis cannot identify.101 As noted in our consultation we do not think that “loss aversion”, even if present, would necessarily have the effects BT claims.102 BT has not provided any further arguments in its response to challenge our views. We also highlight that the results in Professor Crawford’s supplementary report (at Annex 8 of the consultation) are based on a comparison of switching rates for customers on two automatically renewable contracts that differ only in their headline prices. Since this involves a comparison of two groups of ARC customers, it given a very strong indication of the impact of prices on the switching decision for ARC customers. We are confident that this work is accurately identifying price effects.

4.34        BT also noted elsewhere in its response that different specifications of the model yield different results, referring in particular to the additional analysis contained in Appendix 2 of Professor Crawford’s main report. BT noted that when the impact of the ARC term is separated out from the impact of ETCs, the ARC term is found to reduce switching by only 14.7%.

4.35        The work presented by Professor Crawford in Appendix 2 of his report is additional analysis aimed solely at the issue of predicting the impact of lower ETC levels. It does not constitute a ‘richer’ piece of analysis, as BT states, but was instead addressing a different question to the main report.103 Even within this specification we are uncertain as to why BT thinks it is more appropriate to separate out the incremental effect of the ARC term from the ETCs it causes; our main hypothesis is that the effects of ARCs are a result of the unintentional renewal of the MCP and therefore the continuing presence of ETCs. It is in fact notable that there is a residual impact even once ETCs are controlled for and this merely emphasizes the impact of ARC terms.

4.36        It may be that BT is arguing that, in the absence of 12-month ARCs, most consumers would be on 18 or 24-month contracts. However, if so, BT does not explain why this would be the case. In our view, consumers need large inducements to sign up to contacts longer than 12 months, such as expensive handsets in the mobile market. And in any case the effects of a 24-month contract would eventually end whereas ARCs renew indefinitely.

4.37        Regarding the extrapolation of our results, BT stated that if the results of our econometrics were accepted (and it did not accept them), there is no reason why they could not be extrapolated to other providers of residential fixed voice or to residential broadband.

Other responses

4.38        This section covers the remainder of the responses with respect to questions 5-8 in our consultation.

Question 5

4.39         Question five asked:

“Do you agree with our analysis regarding the robustness of the econometric results? In particular, do you agree we have adequately accounted for the effect of the price discount and selection-bias?

Please explain your answer.”

4.40        We have discussed BT’s and Sky’s comments on the econometrics in the preceding sections. Only BT offered detailed views on the econometric analysis; most respondents did not comment. TalkTalk agreed that the econometric analysis and the results seem intuitive and reasonable. SSE stated that the research seems robust and Lexgreen Services Ltd agreed with our analysis.

Question 6

4.41         Question six asked:

“Do you agree that we have correctly identified the underlying causes of the results of the econometrics? Please explain your answer.”

4.42        TalkTalk agreed that we had identified the underlying causes correctly. SSE, FCS and Lexgreen Ltd also agreed with our analysis.

4.43        As part of its response, Which? referred to research it carried out in March 2010 which suggests a lack of awareness of contract terms in the sector. The research showed that 33% of fixed voice customers who were on a contract did not know whether it was on a ‘rolling basis’ and 47% disagreed with the statement “my home phone provider has kept me informed of opportunities to stop my contract renewing”. We discuss this research here since, although Which? was supportive of our proposals and did not reach this conclusion, the research could be taken as indication that a lack of transparency is the cause of our findings. Our view is that the research points to a general lack of understanding across the industry as a whole and not specific shortcomings in the transparency of BT’s ARC. As we explained in paragraph 4.25 in response to Sky’s comments, we think the existence of the reminder letter sent prior to the renewal date means that a lack of transparency is not the main factor behind the results of our econometrics.

Question 7

4.44        Question seven asked:

“Do you agree with our assessment that the results of the econometrics can be extrapolated to: a) all providers of fixed voice services to residential consumers and small business users (with ten employees or less); b) the broadband sector? Do you agree that it is inappropriate to extrapolate the results to medium and large businesses? Please explain your answer.”

4.45        TalkTalk agreed with our extrapolation of the results of the research to all providers of residential fixed voice services and possibly to the residential broadband sector (on the basis that BT may engage in voice bundling practices). However, it did not agree that the results could be extrapolated to small businesses.

4.46        SSE agreed with our extrapolation to all providers of residential fixed voice and broadband services. It also agreed that many businesses with fewer than ten employees will behave in a similar way to domestic customers and should have the same protections. FCS also agreed that our remedies should be extended to cover voice and broadband and agreed that the switching behaviour of domestic and small businesses will generally be similar.

4.47        Ombudsman Services agreed that our findings could be extrapolated to small businesses and to broadband. With regard to the former it stated that it did not see any marked differences between the way residential and small businesses behave, since small businesses do not have staff with specific responsibility for contract scrutiny and procurement, and frequently rely on assurances from the salesman.

4.48        Lexgreen Services Ltd generally agreed with our assessment but thought that a limit of 100 employees may be a better threshold for categorising small businesses instead of ten. We have considered this comment more fully in the context of question 9.

4.49        BT did not believe that it is appropriate to extrapolate the results to ARCs with businesses of less than 10 employees, and considered that the proposed action is disproportionate to any alleged harm perceived by Ofcom.

4.50        We maintain however that the opt-out nature of the MCP renewal decision reduces consumers’ flexibility at the rollover date and may exploit consumer inertia. Therefore, the results can be extrapolated quite widely and are likely to be exhibited to a large extent by any version of an ARC that may emerge in the fixed-voice and broadband sector for both residential and small business customers. In addition, as we set out in the consultation, we do believe that in general, small business customers may behave in a similar way to residential customers. Therefore, we have concluded that the harm identified from ARCs is also applicable to this group of end users and the prohibition should extend to small businesses. This is further explained in Section 5.

Question 8

4.51        Question eight asked:

“Do you agree that the transparency of terms and conditions in BT’s ARC contracts is not likely to be the prime driver of the results identified in the econometric research?”

4.52        TalkTalk argued that it is extremely difficult to show that residential consumers have understood the concept of an ARC however clearly it is communicated to them. Consumers generally assume contracts do not contain anything “out of the ordinary” and are unlikely to read small print. SSE agreed that greater transparency in the sales process was unlikely to improve the issues we identified with ARCs, although it stated it was surprised with the results of our mystery shopping exercise. FCS agreed with our analysis, saying that in practice consumers do not recognise the significance of information provided.

4.53        Ombudsman Services agreed, saying that customers often appear not to have fully absorbed information that is provided. Copies of call recordings it had been provided with in relation to some complaints often show a large gap between what a customer was told and what they recollect. It said it believed customers generally only focus on correspondence that is clearly a bill, assuming everything else is marketing literature.

4.54        Citizen’s Advice agreed that transparency is not likely to be the main driver of the results we observe, and that the ‘default bias’ explains most of the results. However, it said it was not convinced that the evidence we presented indicates there is not a transparency problem at all, noting the finding in our mystery shopping exercise that around half of callers were not told the contract renewed automatically. Despite this, Citizen’s Advice agreed transparency on its own would not be sufficient and that ARCs should be prohibited.

4.55        Lexgreen Services Ltd did not agree that transparency issues were not the prime driver of our findings, noting that many consumers do not read small print and that sales people have little incentive to explain how to terminate a service they are trying to sell. However, Lexgreen Services Ltd fully agreed with our proposals and did not suggest transparency remedies were more appropriate.

Conclusions

4.56        In the course of our research we have reviewed a number of sources of evidence into the impact of ARCs on consumer switching. Our core evidence is the detailed econometric analysis of the switching patterns of BT’s customers that we commissioned from Professor Crawford and ESMT Competition Analysis. The work shows that, even after allowing for the price discount offered with the ARC and the possibility of self-selection by BT’s customers, ARCs are associated with a significantly lower level of switching after the initial MCP. We also find evidence that switching is deterred even at relatively low levels of ETCs. Of those respondents that commented on our econometric analysis, apart from BT and Sky, most agreed with our preliminary conclusions that the effect we observe stems from the opt-out nature of the MCP renewal decision rather than transparency issues.

4.57        We have concluded that the opt-out nature of the MCP renewal decision reduces consumers’ flexibility at the rollover date and may exploit consumer inertia. Therefore, the results can be extrapolated quite widely and are likely to be exhibited to a large extent by any version of an ARC that may emerge in the fixed-voice and broadband sector for both residential and small business customers. Apart from BT and TalkTalk who do not believe that the results can be extrapolated to small businesses, most respondents agreed with our conclusions to include small businesses in the scope of our proposed GC.

4.58        We set out in the consultation that we believe small business customers behave in a similar way to residential customers and that the harm identified from ARCs was also applicable to this group of end users. We have therefore concluded that it is warranted and justified to apply the modified GC to small businesses.

Section 5

Ofcom’s proposals for addressing the harm caused by ARCs

Introduction

5.1           Section 5 of our consultation discussed the options for addressing the harm from ARCs that we had identified. Our discussion was based on our preliminary conclusions that ARCs are harmful for consumers and competition, that the effect stems from the opt-out nature of the renewal decision and would be created by any version of an ARC, and that there is a risk of rapid adoption throughout the market in the future.

5.2           This section covers consultation questions 9-11. We start by recapping the proposals contained in our consultation, before setting out a general overview of the responses received, and the responses received in relation to each specific consultation question. We then set out the conclusions we have drawn in light of those responses.

Summary of the consultation

5.3           We identified four options in our consultation:

•       Option 1: do nothing;

•       Option 2: introduce a General Condition (GC) aimed at improving consumer information;

•       Option 3: introduce a GC preventing subsequent MCPs, once the initial one has ended;

•       Option 4: introduce a GC requiring MCP renewal to be an opt-in process.

5.4           We took the view that doing nothing would not address the current or potential future harm in the market and would mean our regulatory response would lag significantly behind events if ARCs spread more widely throughout the market. Furthermore, we felt the UTCCRs, while helpful in tackling certain aspects of consumer harm, were not sufficient to tackle our core concerns.

5.5           After reviewing the options for intervention, in particular assessing whether they met the tests set out in section 47(2) the Act of being objectively justified, not unduly discriminatory, proportionate and transparent in relation to what it is intended to achieve, we provisionally concluded that Option 4 would be preferable.

5.6           We felt that Option 2, although lighter touch, would not be consistent with our conclusion that the harmful effect of ARCs are a result of the opt-out nature of the renewal decision, not a lack of consumer information. We felt that Option 3, which has been adopted via legislation in other EU countries, might be disproportionate because it removes from consumers the option of benefiting from price discounts and other benefits firms are willing to offer in exchange for an MCP.

5.7           We provisionally concluded that Option 4 would be the most appropriate form of intervention and that it would meet the criteria set out in the Act. This option requires consumers to make active opt-in decisions at the end of each MCP if they wish to renew their contract. This targets the root cause of the problem (the opt-out nature of automatic renewals) and applies widely enough to address our concerns that any version of an ARC will be harmful. We argued that there is a risk of ARCs spreading rapidly in the market going forward and that retrospective intervention would be slow to deal with the resulting harm. We noted that a ban on opt-out renewals still preserves for consumers the option to renew their contracts proactively and that CPs would be free to make this as easy as possible. We therefore felt the proposed intervention would be proportionate.

5.8           Regarding the timeframe for compliance, we noted that CPs with existing customers on ARCs would have to make any necessary systems changes and develop a replacement product and this is likely to be more burdensome with a shorter compliance timeframe. We took the view that many of the costs of implementation would be those that a CP is used to incurring on a regular basis, such as writing to consumers and informing sales staff of changes to terms and conditions, and should be able to manage efficiently. Given this, and since the vast majority of current ARC customers are on 12-month contracts, we proposed that a compliance period of 12 months was most appropriate.

Responses to the consultation

General overview of responses

5.9           Most of the responses welcomed our proposals for a modified GC 9, prohibiting the sale of ARCs. The two main exceptions were BT and Sky (Virgin Media also expressed some reservations about Ofcom’s reliance on behavioural economics but mainly focussed on practical issues surrounding the wording of the GC). Other than these respondents there was general agreement that a GC was appropriate and that this should cover both fixed voice and broadband services.

5.10        There was disagreement among respondents on the issue of whether the GC should cover small businesses and on the appropriate implementation timelines for those CPs with existing customers on ARCs. A number of respondents also disagreed with our proposal not to include mobile services.

5.11        We now consider the responses received to each specific question below.

Question 9

5.12        Question nine asked:

“Do you agree that intervention regarding ARCs is warranted? Do you agree that a General Condition restricting opt-out renewal processes (i.e., option 4) is the most appropriate form of intervention? Do you agree that a GC should cover both residential and small business customers, but not medium or large businesses? Do you agree that mobile communications should be out of scope at this stage? Please explain your answer, and provide details of any alternative intervention you think more appropriate.”

5.13        Sky’s response strongly objected to our proposals. We have dealt with a number of its arguments in previous sections and set out below its specific objections in relation to question 9.

5.14        In Sky’s view, the introduction of the proposed modified GC would not constitute a targeted or proportionate approach to regulation. Sky maintains that Ofcom has not fully explored the use of existing consumer protection tools – specifically the UTCCRs and the CPRs – and that intervention through the creation of a modified GC is, at this stage, unwarranted. Sky also suggests that Ofcom gain a more thorough understanding of the impact of lower ETCs in order to properly inform our proposals. We discuss this at paragraph 5.33 below.

5.15        Sky also makes the point that the policy options discussed in section 5 of our consultation do not include a middle option in which existing powers would be used to tackle transparency problems. Sky suggests that we dismiss too rapidly the option of requiring CPs to supply consumers with more information regarding their contract, and note that quality of information may be more important than quantity.

5.16        Sky questions the robustness of behavioural economics and states that the existence of consumer bias (as identified by behavioural economics) is not a legitimate basis for intervention, labelling this ‘paternalistic’.104 Sky refers, in particular, to its report commissioned from Plum Consulting which notes that regulators too can be subject to bias. Sky states that ‘paternalistic’ intervention sets a very dangerous precedent of regulating on the basis that consumers are not capable of making informed decisions, even where terms and conditions are transparent.

5.17        BT also provided a comprehensive response, setting out the steps it has taken to protect consumers on ARCs, including transparency of terms and conditions, the reminder letter that is sent prior to the roll over date, and the ability of the consumers to opt-out of their next rollover at any point during their current MCP. BT also noted we had concluded BT’s contracts did not appear to breach the UTCCRs. BT argued that this means there is no consumer protection justification for intervention.105

5.18        BT also opposes the creation of a GC more generally, principally because it disagrees with our econometric analysis and our arguments that switching costs harm competition, as set out in Section 3. In Section 4 we addressed those concerns and we now consider the further arguments that BT makes with respect to question

5.19        BT disagreed with our proposal to include within the scope of the GC businesses with 10 or fewer employees. This is on the basis that our proposal to cover small businesses appears to have been an afterthought and was not fully explained. It noted that it was not asked to provide any information on small businesses and these contracts were not part of our discussions. Furthermore, all the data BT provided for use in our econometric analysis related to residential consumers not businesses.

5.20        BT also argued that its business contracts that include automatic rollover terms should not fall within the scope of the proposed GC, since they involve a spending commitment that must be met within a particular time period rather than an MCP per se. Business customers have a choice of the level of spend they can commit to and receive price discounts in return. While customers must pay a reconciliation charge if they terminate without meeting their spending commitment, this is not linked to the time at which they terminate.

5.21        BT also expressed concerns with our proposed threshold of 10 employees or fewer for distinguishing small firms. BT said that businesses with few employees may still be large in terms of their spending on BT services and may therefore wish to receive the benefits that are offered in return for spending commitments. It also argued that it does not keep records on the size of business customers and could not implement our proposed distinctions without incurring extensive costs, since systems changes would be needed to identify, record and monitor customer size. BT estimated that changes would cost £3 million per annum resulting from the additional need to contact customers, and involve a cost of £1.3 million to develop duplicate non-ARC pricing options. These costs would be passed on to consumers.

5.22        As noted above TalkTalk agreed with our proposals with respect to the residential fixed voice and broadband sectors, but did not think the GC should extend to small businesses.

5.23        SSE agreed with our proposals, adding that a case could be made for applying them to the mobile sector as well since there would be no compliance costs for any CP and we make clear that we would have concerns if ARCs emerged in the sector.

5.24        FCS agreed with our proposals to include small businesses but also argued for a prohibition in the mobile sector as a preventative measure.

5.25        Ombudsmen Services supported our proposal for intervention but noted that we had not made clear whether CPs would be required to provide month-by-month service at the end of each MCP. In relation to both residential and small businesses it was outside the remit of our consultation to consider mandating a month to month option. However, we are unaware of any residential contracts that either disconnect the customer at the end of the MCP or that do not revert to monthly supply at the end of the MCP. We think it is unlikely they would emerge, since their introduction would almost certainly create a lot of unintended disconnections and prompt consumers to switch suppliers. Furthermore, application of a new MCP to which customers are not required to proactively opt-in would be likely to be regarded as an ARC, and hence prohibited under the amendments to GC 9. Regarding business contracts, the issue of whether contracts should be required to revert to a month-by-month basis in the event that the customer does not renew the MCP is separate from that of ARCs and is outside the scope of this statement.

5.26        Ombudsmen Services also noted that problems are frequently most severe in business contracts and welcomed our inclusion of small businesses in the GC.

5.27        Citizen’s Advice stated it “strongly supports our proposal to prohibit opt-out renewal processes”. It provided examples of cases it had dealt with where consumers had suffered harm as a result of ARCs. These related to a number of (unnamed) CPs and some appear to involve deliberate attempts to prevent opting out. Citizen’s Advice disagreed, however, that mobiles should be excluded from the GC, pointing to the OECD’s recommendations that ARCs be prohibited and Ofgem’s recent proposals to prohibit ARCs in energy markets as demonstrating that ARCs are widely recognised as harmful. As we said in our consultation,106 we share concerns that ARCs could cause detriment in the mobile market if they emerged. Although we have chosen to target sectors where ARCs currently exist, we said we would urgently revisit the issue if ARCs emerged in the mobile sector.

5.28        Lexgreen Services Ltd supported our proposals but argued that the threshold for it to apply to business customers should be 100 employees and also disagreed that mobiles (and VOIP) should be outside the scope of the GC.107

5.29        CCP agreed that intervention is required but stated that medium sized businesses should also be covered (it did not define what it meant by medium businesses, however). CCP also disagreed that mobile services should be excluded from the scope of the GC.

Our response

5.30        As set out in our consultation, Ofcom has given careful consideration as to whether or not the UTCCRs might address the specific economic harms that we have identified in relation to ARCs. We concluded that, while these regulations may be helpful in tackling certain aspects of consumer harm, particularly in relation to the overall fairness and transparency of contractual terms used in ARCs, they are unlikely to target our core concern, which is the adverse effect on consumers resulting from higher switching costs and reduced competition. Accordingly, we do not accept this aspect of Sky’s response.

5.31        Ofcom has also given careful consideration to Sky’s view that the CPRs can be used to address the harm caused by ARCs. The CPRs are designed to prohibit “unfair” commercial practices, namely those which contravene the requirements of professional diligence and which materially distort the economic behaviour of the average consumer with regard to products. Such unfair behaviour includes misleading actions and omissions (for example, providing false information or failing to include material information about a product) and aggressive commercial practices. Again, while there may be scope for applying the CPRs in specific cases where ARCs arise, this legislation does not address Ofcom’s core concern which is that ARCs create a barrier to consumer switching and competition in the market.

5.32        As we made clear in our consultation document, the harmful effects we observed from ARCs are most likely to stem from the opt-out nature of the renewal process. We do not think providing consumers with more (or different) information would significantly mitigate this harm since reminder letters may be clear and sent to the consumer near to the renewal date, as is currently the case with BT’s contracts. Sky does not provide reasons or evidence as to deficiencies it has observed in reminder letters or the forms of information that it believes would be more effective. This is also in a context where most CPs and consumer organisations agree with our preliminary conclusions that the potential benefits of ARCs on consumer welfare are limited.

5.33        We do not agree with Sky that more research is needed into the effects of lower ETCs. As we set out very clearly in our consultation,108 the data we obtained from BT contains sufficient information to show that even low ETCs deter consumer switching. We note that Sky did not offer any specific comments on why this analysis was flawed.

5.34        Our consultation was also very clear about the possibility of regulatory failure. Section 3 of our consultation begins with a detailed discussion of the possible benefits of ARC terms for consumers, recognizing the risk that some consumers may be made worse off by a ban. Our arguments in favour of a GC (including the fact that ARCs have the effect of increasing switching costs and our view that that the benefit to consumers from ARCs is unclear), acknowledge clearly the potential for regulatory failure109 and set out why we think the risks are low in this case.110 As we explained, prohibiting ARCs only changes the process for contract renewal, not the options available for consumers. We note Sky did not provide reasons why we have underestimated the benefits of ARCs for consumers, nor suggest any other mechanism through which our GC would create regulatory failure.

5.35        We also do not accept that our proposals are ‘paternalistic’. Our interest is on the effect that ARCs have on consumers and competition, and using the insights of behavioural economics, in combination with the econometric evidence, allows us to gain a much more realistic picture of how consumers actually behave. This does not involve a judgement about what consumers are ‘capable’ or ‘incapable’ of, rather how they actually make decisions. As we have made clear, we do not make any judgements about whether or not consumers ought continuously to renew their MCP; we are only concerned that renewal should be an active decision in order to ensure that consumers are fully aware of their right to switch and can exercise those rights effectively.

5.36        The Plum Consulting report submitted by Sky, (also discussed in Section 3) compares our intervention with proposals by Ofgem and recent research by the OFT,111 and argues that they have taken a more pragmatic view that emphasizes the use of existing mechanisms and improved consumer information. The OFT study took a broad look at the issue of consumer contracts in general and did not consider the specific issue of ARCs in communications markets. We agree with the general conclusions of the OFT report and do not see any conflict between its overall recommendations and our proposals. The OFT report presents a framework for assessing consumer harm which involves questions such as whether consumers can learn over time, whether contract terms have countervailing benefits and whether the contract term has wider implications for competition in the market. Our consultation considered all of these elements. Furthermore, our findings were based on an extensive analysis of switching behaviour in the fixed voice sector as well as other sources of evidence, and we are therefore able to identify and justify more specific interventions than those discussed in the OFT’s high-level report. The OFT report discussed intervention in the context of the tools it has available and did not compare the UTCCRs or CPRs to the interventions that other regulators have available in their sectors. Regarding Ofgem, we note that their current proposals in the residential energy sector involve a prohibition on the sale of ARCs.112

5.37        While many of our concerns centred around the impact on competition, we disagree that there is no consumer protection justification for our GC. What is important is the impact that the contract terms have on consumers and, since we believe they induce unintentional contract renewal that subjects a consumer to ETCs, this generates a consumer protection issue that is separate from the follow-on effects on competition.

5.38        In relation to the inclusion of small businesses, we maintain that the opt-out nature of the MCP renewal decision reduces consumers’ flexibility at the rollover date and may exploit consumer inertia. Therefore, the results can be extrapolated quite widely and are likely to be exhibited to a large extent by any version of an ARC that may emerge in the fixed-voice and broadband sector or residential and small business customers.

5.39        We acknowledge that businesses with 1 – 10 employees are not a homogeneous group and that it is possible that some users in this group may make purchasing decisions in a way which is more consistent with larger businesses than residential consumers. However, as we set out in the consultation, we believe that a significant proportion of small business customers behave in a similar way to residential customers, especially in relation to switching. Smaller businesses typically have less resource than larger businesses to devote to researching the best deal and may inadvertently renew their contracts at the end of the MCP. Therefore, we have concluded that the harm identified from ARCs is also applicable to this group of end users and the prohibition should extend to small businesses. We have included (at Annex 3 of this statement) guidance on our approach to compliance and enforcement to assist CPs.

5.40        In relation to Lexgreen Services Ltd’s view that the threshold for small businesses should be set at 100 employees, and CCP’s view that the GC should apply to “medium” sized businesses, we do not believe that this is appropriate. This is principally because firms this size typically have dedicated resources for procurement and for dealing with contractual arrangements with suppliers, and are less likely to be affected by the default bias than small businesses who, in our view, are much more akin to residential consumers. As the threshold for the definition of small businesses is raised there is an increasing risk of regulatory failure and of preventing the sale of ARCs to firms that would not be affected by the renewal term. Furthermore, the 10 employee threshold is consistent with the way in which “small business customer” has been defined in the other GCs.

5.41        In relation to the inclusion of mobile, we stated in the consultation that because ARCs had not emerged in the mobile sector, we did not take a firm view on whether the results could be extrapolated to mobiles. We noted, however, that we would have concerns if ARCs did emerge in the mobile sector and would urgently revisit the issue. Since publishing the consultation, ARCs have not emerged as a feature of the mobile market and although responses from some stakeholders indicated that we should include mobiles in the scope of our prohibition as a preventative measure, we believe it would be proportionate to target our intervention at the current harm we observe in fixed markets. Therefore, we have not chosen to include an explicit prohibition of ARCs in the mobile sector, although our position remains that we would urgently revisit this issue if ARCs emerged in this sector.

Question 10

5.42        Question ten asked:

“Do you agree that we should require compliance with the proposed GC amendment immediately it comes into effect? Do you agree that a period of 12 months for compliance with our proposed GC for CPs with existing customers subject to ARC terms is sufficient to remove the harm associated with ARCs without imposing disproportionate costs on CPs? If you disagree, please provide evidence to support your view and suggestions for a more appropriate time frame.”

5.43        TalkTalk agreed that the proposed GC should apply immediately when it enters into force, but said that a 12-month implementation period for those CPs with customers currently on ARCs was overly generous and that a three month period would be sufficient.

5.44        SSE also argued that our proposed implementation period for CPs with customers already on ARCs was too long. Given that renewal results in a further 12-month MCP, our proposals mean that it would take two years for all the effects to be removed. SSE argued instead that all contract renewals should be opt-in after “a few months”.

5.45        A confidential respondent argued that 12 months was too long and that 6 months should be sufficient.

5.46        Ombudsmen Services also considered our proposed timeline to be too long and suggested a period for implementation of between three and six months.

5.47        Lexgreen Services Ltd said that compliance should be immediate for all CPs including those with existing customers on ARCs.

5.48        KCOM was concerned that the wording of the GC would allow a CP to automatically renew existing customer contracts within the 12-month implementation period, and suggested that the wording of the GC be amended to prevent this.

5.49        FCS and Citizen’s Advice agreed with our proposals on implementation timelines.

5.50        BT said that our proposed timeline for compliance presents them with serious challenges. While it agreed that migration of residential customers would take 12 months it said that it could not start the migration of existing customers until the end of December. The reasons given for this are that BT would need to have alternative products in place, change its system-generated customer communications, ensure that all channels were fully briefed and trained, and a change of this size would involve large system changes which would impact a number of other systems and platforms. BT also said that it could only meet these deadlines if Ofcom published a final statement in mid-July. For business contracts, BT said that it would be extremely difficult to start work on compliance before the quarter starting January 2012.

Our response

5.51        We recognise that the removal of ARCs from the markets in these sectors cannot happen overnight, particularly for the significant base of customers currently contracted to ARCs. We have noted the views of respondents who urged us to mandate an accelerated withdrawal, and also taken account of the practical considerations of withdrawal which requires changes to CP systems. We want to ensure that the implementation is smooth and does not itself result in distortions to the market or disruption and inconvenience for consumers and users. Taking account of all these considerations, we have concluded that the following framework is appropriate for the withdrawal of ARCs:

•        For fixed voice and fixed broadband residential and small businesses ARCs the prohibition on the sale of new ARCs113 will take effect on 31 December 2011. This means that the sale of new ARCs will be prohibited from that date.

•        In relation to all other residential and small business ARCs, in order to enable the orderly migration of the existing customer base, we are requiring that these are completely removed from the market (i.e. all existing customers migrated) by 31 December 2012. We also expect CPs to be proactive in facilitating migration for existing ARCs customers during this period, and to take a flexible approach to ETCs for ARCs customers after the first MCP of their contract who wish to exit.

5.52        We have included a footnote in the GC (footnote 3 of the notification at Annex 2 of this statement) to clarify the implementation period for bringing GC 9.3(a) and (b) into force.

Question 11

5.53        Question eleven asked:

“Do you agree with the General Condition we propose? If not, what alternative form of GC do you consider more appropriate? Please explain your answer.”

5.54        A number of respondents agreed with our proposed wording of the GC, while others disagreed and made suggestions for clarifying or improving it.

5.55        TalkTalk agreed with the wording of our proposed GC (subject to its reservations about our inclusion of small businesses). SSE also agreed with our wording but argued that the GC should be widened to include mobile services and should have a shorter implementation timeline. Citizen’s Advice agreed but urged us to include mobiles.

5.56        Ombudsmen Services raised concerns that our wording would allow CPs to place contractual time limits on opt-in renewal decisions, such that a decision to renew could only be made far in advance of the end of the current MCP. We agree that such practices, if they emerged, could potentially be detrimental to consumers and competition since they force a customer to make a decision far in advance of knowing their future circumstances and the offers available in the market and will not have a clear view of any countervailing benefits. If these practices did emerge, we would monitor developments closely and consider what action, if any, is appropriate.

5.57        KCOM said that the use of the term “End-User” included in GC 9.3 could cause confusion and that it would be preferable to use the term “Domestic and Small Business Customers” as used in GC 14. Sky also queried the use of this term, noting that GC 9 already applies to a range of different consumers and that it is not clear why we are seeking to introduce a new definition of “End User”.

5.58        Virgin also expressed a number of concerns about the wording of the proposed GC. Virgin notes that the terms “minimum contract period” and “further minimum contract period” are undefined and that the condition goes well beyond simply addressing the harm that Ofcom has identified. In particular, Virgin states that the GC as drafted would capture all situations where a service may continue – for example, on a month to month basis, terminable on notice without the payment of an ETC – once an MCP has come to an end. In Virgin’s view, it might be possible to interpret the notice period in the “month to month” scenario as a new MCP that automatically renews if the customer chooses not to terminate the contract within that time.  This would mean that at the end of each notice period (which may, for argument’s sake, be 30 days) a CP would need to contact the customer in order to gain their Express Consent to continue receiving the service. In this situation the effect of the GC would be much wider-ranging than Ofcom’s stated policy intention of preventing the automatic renewal of contracts for periods of twelve months or more. Sky has similarly echoed this concern in its response.

5.59        Further, in relation to Ofcom’s proposals for securing consumers’ Express Consent to enter into a further MCP, Virgin queries why CPs are restricted to contacting customers no earlier than half way through their existing MCP and no later than two weeks prior to the expiry of that period. In Virgin’s view, provided that customers are informed in a sufficiently transparent manner, they should be able to give their Express Consent at any time throughout the duration of an MCP. KCOM and Sky have also raised similar concerns, stating that CPs may be prevented from making offers to upgrade their customers in return for a new MCP outside the proposed timescales. KCOM has stated that CPs should be able to contact their customers within the first half of the year for the purposes of offering service upgrades.

5.60        Finally, Sky has raised the point that the proposed modifications appear to narrow the overall scope of GC 9.3 by applying only to Consumers and Small Businesses, rather than to “End Users” more generally as provided for in Article 30(6) of the USD. Sky has suggested that Ofcom may wish to clarify its position, insofar as the modifications to the GC appear to conflict with the intention of the USD.

Our response

5.61        We set out in our consultation that Ofcom was still awaiting the outcome of a concurrent consultation, the “Changes to the General Conditions and Universal Services Conditions,”114 which set out our proposals for amending GC 9 in light of changes that had been made to the USD. This included the insertion of a new GC 9.3, ensuring that conditions or procedures for terminating contracts do not act as a disincentive against switching providers.115 This proposal, if implemented, would form the basis upon which Ofcom would make its further GC with respect to ARCs.116

5.62        Our consultation on the Changes to the General Conditions and Universal Services Conditions ended on 7 April 2011 and our statement was published on 25 May 2011.117 It was decided that GC 9 would be amended by inserting a high level requirement ensuring that CPs’ conditions or procedures for terminating contracts do not act as a disincentive for all End Users against switching providers. It was decided that “End-Users” would include all customers and all businesses. Ofcom acknowledged that stakeholders had raised concerns about the number of different definitions contained in GC 9, such as “Consumer”, Subscriber”, “End-User” and “User”, but that these terms were clearly expressed in the Framework Directive and so should be readily understood by stakeholders.118 As proposed in our consultation addressing ARCs, the new GC 9.3 will now form the basis for making our GC.

5.63        The new GC 9.3 states:

“Without prejudice to any initial commitment period, Communications Providers shall ensure that conditions or procedures for contract termination do not act as disincentives for End-Users against changing their Communications Provider”.

5.64        This wording differs slightly from that contained in our consultation on ARCs in that, following stakeholder responses to the Changes to the General Conditions and Universal Services Conditions consultation, the phrase “minimum contract period” was replaced with the phrase “initial commitment period.” An “initial commitment period” has been defined in GC 9.4 as being:

“the period beginning on the day on which the Communications Provider and Consumer have agreed that the contract shall begin and ending on a day falling no more than 24 months thereafter”.

5.65        As stated in our statement on the General Conditions and Universal Services Conditions,119 Ofcom interprets “initial commitment period” as being the period of the contract during which the consumer cannot leave without being liable to pay compensation to the CP (i.e. by way of an ETC). After the expiry of that period, the same contract may still be in (continuous) operation, but without a requirement for the consumer to compensate the CP in relation to termination. Further, the word “initial” does not refer only to the first contract entered into between a CP and a consumer. Ofcom considers that each and every contract between a CP and a consumer will have an initial commitment period. This means that second and subsequent contracts that are entered into between a CP and a consumer will also have an initial commitment period (which, because of GC 9.4, cannot be longer than 24 months).120

5.66        We are satisfied that, for the purposes of modifying the GC as proposed in our ARCs consultation, the substitution of “initial commitment period” in GC 9.3 is wholly consistent with those proposals. This is because the concept of an “initial commitment period” is synonymous with the concept of an “MCP” as set out in that document. Further, where stakeholders such as Virgin have expressed concern at the lack of definitional clarity with respect to “MCP”, we believe this issue is now resolved by the introduction of the new wording.

5.67        We note the concerns raised by stakeholders (in particular, Virgin and Sky) that the wording of the proposed GC could be interpreted as extending to contractual arrangements between a CP and a customer which continue beyond the expiry of an MCP (or, under the new terminology, an initial commitment period). This may include, for example, where services continue to be provided on a month to month basis, terminable on notice without the payment of an ETC. Ofcom confirms that it was not (nor is) its intention that such arrangements should be caught by the modifications to GC 9.3. Rather, we are concerned with targeting the way that ARCs lock a customer into contractual arrangements that may only be broken by paying an ETC to their CP. We have sought to clarify this in the GC by including a footnote.

5.68        We also note stakeholders’ concerns with the proposed timeframes by which Express Consent must be sought. To clarify, the reason why we proposed a window of “no earlier than half way through the existing minimum contract period and no later than two weeks prior to the expiry of that period” is first to ensure that CPs obtain consent for each and every MCP (rather than simply asking the customer to provide one consent at the outset for all subsequent MCPs that they may enter into) and secondly to ensure that CPs do not wait until the last minute to contact their customers, which could result in customers panicking and simply signing up to a further MCP by default.

5.69        However, having considered stakeholders’ comments, we agree that it is preferable to keep the wording of GC 9.3(b)(iv) as simple as possible. On that basis we have removed the references to specific time-frames. However, bearing in mind our initial concerns, we have added the following wording “…separately for each initial commitment period and in a manner which enables the Consumer or Small Business Customer to make an informed choice”.

5.70        Finally, with respect to Sky’s concern that we may be narrowing the overall scope of GC 9.3 (and, by extension, Article 30(6) of the USD) by applying the GC only to consumers and small business customers, rather than to “End Users” more generally, we do not accept that this is the case. This is because the modified GC is prefaced by the words, “In particular, but without limiting the extent of this paragraph” which makes it clear that these amendments are just one example of how disincentives to switching may arise in relation to a particular group of End-Users.

5.71        We do, however, acknowledge the comments from stakeholders (such as Sky and KCOM), querying why we are introducing a new definition of “End User” for the purposes of our amendments and suggesting that, in the interests of minimising confusion, Ofcom should opt for the definition of “Domestic and Small Business Customers” that has been used in GC 14.

5.72        The definition of “Domestic and Small Business Customer” that is used in GC 14 has been taken from section 52(6) of the Act. The term refers to a customer of a CP who is neither:

“(a) himself a communications provider; nor

(b) a person who is such a customer in respect of an undertaking carried on by him for which more than ten individuals work (whether as employees or volunteers or otherwise)”.

5.73        The reason why Ofcom did not initially propose referring to this definition was to minimise both duplication and potential confusion in defining residential customers, in circumstances where GC 9 already contains a general definition of “Consumer”. However, in setting the threshold for small businesses of no more than 10 employees, we relied upon the guidance provided in section 52(6) of the Act (which  is also the basis on which “small business customer” has been defined elsewhere in the GCs). Having given careful consideration to the points raised, we agree that in the interests of consistency with the GCs overall, the definition of “Small Business Customer”, could be usefully clarified by using the express language contained in that section.

Conclusions

5.74        Having given careful consideration to the consultation responses, together with the analysis set out in our consultation document, we have come to the conclusion that ARCs do cause harm to consumers and that the impact of this harm stems from the opt-out nature of these contracts. It is on that basis that we have decided to proceed with the modifications to GC 9.3. As we explained in the consultation document (and in this statement) we have fully considered our options for addressing the harm caused by ARCs and concluded that this can only really be properly addressed through the creation of a bespoke modification to the GCs.121 Accordingly, for the reasons set out above, we have decided to proceed with the amendments to GC 9.3 as set out in Annex 2.

5.75        In summary, the changes we are making to our proposed amendments to GC 9.3 following the publication of our Changes to the General Conditions and Universal Service Conditions122 and the responses we have received from stakeholders in relation to our consultation on ARCs are:

(a)    to substitute “minimum contractual period” for “initial commitment period”, relying on the definition of “initial commitment period” contained in GC 9.4;

(b)    to further clarify the scope of the amendments to GC 9.3 by:

i)    reformatting the amendments to GC 9.3 and further clarifying the definition of “Small Business Customer”;

ii)   making it clear that where a Consumer or Small Business Customer has come to the end of an initial commitment period, but continues to receive services from a Communications Provider on a periodic basis (for example, month to month), terminable on notice by that Consumer or Small Business Customer without charge or penalty, this will not be considered to be a renewal of the contract for a further initial commitment period;

(c)     to simplify the requirements on gaining Express Consent;

(d)    to clarify the implementation period for the modifications to GC 9.3 coming into force.

5.76        We consider that our modifications to GC 9.3 in relation to the requirement to offer contracts with minimum terms meet the criteria set out in section 47(2) of the Act being:

•         objectively justifiable, on the basis of the harm that Ofcom has identified, as set out both in its consultation and in this statement;

•         not unduly discriminatory, as the revised requirement will apply to all CPs;

•         proportionate, as the change is the minimum necessary to address the harm that Ofcom has identified, while also preserving consumer choice and the ability of CPs to develop competitive offerings; and

•         transparent, as both the purpose and requirements of the proposed modifications to the GC are clear and the GC itself will be readily accessible on Ofcom’s website.

Annex 1

List of respondents

We received 32 responses to our consultation in total.

The following stakeholders submitted non-confidential responses to our consultation. All non-confidential responses (including those where the respondent wished to withhold their name, and therefore their name is not listed here) can be found at: http://stakeholders.ofcom.org.uk/consultations/arcs/?showResponses=true

•       Bromley, Dr M

•       BT

•       Casey, Mrs R

•       Centre for Competition Policy, University of East Anglia (CCP)

•       Citizen’s Advice

•       Fawcett, Mr G

•       Federation of Communication Services (FCS)

•       Hamill, Mr M

•       KCOM Group Plc

•       King, Dr C

•       Knight, Mr G

•       Lambert, Mr B

•       Lexgreen services

•       Ombudsman Services: Communications

•       Ridgley, Mrs

•       Scott, Mr G

•       Scottish and Southern Energy plc (SSE)

•       Sky

•       TalkTalk Group

•       Tan, Dr K

•       Virgin Media

•       Which?

Annex 2

Notification of modifications to the General Conditions

NOTIFICATION MODIFYING THE GENERAL CONDITIONS UNDER SECTION 48(1) OF THE COMMUNICATIONS ACT 2003 MODIFICATION OF GENERAL CONDITION 9 OF PART 2 OF THE GENERAL NOTIFICATION REGARDING THE REQUIREMENT TO OFFER CONTRACTS WITH MINIMUM TERMS

BACKGROUND

A.               The Director General of Telecommunications published on 22 July 2003 a notification setting general conditions under section 45 of the Act which took effect on 25 July 2003 (the “2003 Notification”). Since July 2003, the general conditions so set have been modified on several occasions and new general conditions have been set by Ofcom (collectively, the “General Conditions”).

B.               On 3 March 2011, Ofcom published a Notification under section 48(2) of the Act setting out proposals to modify the General Conditions (the “Notification”).

C.               A copy of the Notification was sent to the Secretary of State in accordance with section 50(1)(a) of the Act.

D.               In the Notification (and accompanying explanatory statement), Ofcom invited representations on the proposals by 12 May 2011.

E.               By virtue of section 48(5) of the Act, Ofcom may give effect to the proposals with respect to which it has published a Notification, with any modifications that appear to Ofcom to be appropriate, where Ofcom has:

(i)  considered every representation about the proposals made to Ofcom within the period specified in the notification; and

(ii) had regard to every international obligation of the United Kingdom (if any) which has been notified to Ofcom for this purpose by the Secretary of State.

F.               Ofcom received responses to the Notification and has considered every such representation made in respect of the proposals set out in the Notification (and the accompanying explanatory statement); and the Secretary of State has not notified Ofcom of any international obligation of the United Kingdom for this purpose.

DECISION

1.              In accordance with section 48(1) of the Act Ofcom sets out the modifications to the General Conditions made in the Schedule to this Notification.

2.              The effect of, and Ofcom’s reasons for making, these modifications are set out in the accompanying explanatory statement.

3.             General Conditions 9.3(a) and 9.3(b) come into force on 31 December 2011 in relation to all Consumers and Small Business Customers who are not, as at that date, in a contract with a Communications Provider where, at the end of an initial commitment period, that contract will automatically renew for a further initial commitment period without that Consumer’s or Small Business Customer’s Express Consent. For all other Consumers and Small Business Customers, GCs 9.3(a) and 9.3(b) come into force on 31 December 2012.

4.              Ofcom is satisfied that these modifications satisfy the test in section 47 of the Act.

5.              In making these modifications, Ofcom has considered and acted in accordance with its general duties in section 3 of the Act and the six Community requirements in section 4 of the Act.

6.              In this Notification:

(i)  “2003 Notification” has the meaning ascribed in recital A above;

(ii)  “Act” means the Communications Act 2003;

(iii) ”Consumers” has the meaning ascribed in paragraph 9(7)(b) of the Schedule to this Notification;

(iv)  “Initial Commitment Period” has the meaning ascribed in paragraph 9.4 of the Schedule to this Notification;

(v)  “Notification” has the meaning ascribed in recital B above;

(vi)  “General Conditions” have the meaning ascribed in recital A above;

(vii)  “Ofcom” means the Office of Communications; and

(viii)  “Small Business Customer” has the meaning ascribed in paragraph 9(b)(v) of the Schedule to this Notification.

7.              Except insofar as the context otherwise requires, words or expressions shall have the meaning assigned to them in this Notification and otherwise any word or expression shall have the same meaning as it has in the Act.

8.              For the purpose of interpreting this Notification:

(i)  headings and titles shall be disregarded; and

(ii)  the Interpretation Act 1978 shall apply as if this Notification were an Act of Parliament.

9.              The Schedule to this Notification shall form part of this Notification.

Signed by Claudio Pollack

Group Director, Consumer Group

A person authorised by Ofcom under paragraph 18 of the Schedule to the Office of Communications Act 2002

13 September 2011

SCHEDULE

MODIFICATION TO GENERAL CONDITION 9

REQUIREMENT TO OFFER CONTRACTS WITH MINIMUM TERMS

This Schedule shows the modification to General Condition 9. Additions are in italics.

9.  REQUIREMENT TO OFFER CONTRACTS WITH MINIMUM TERMS1

9.1              Communications Providers shall, in offering to provide, or providing, connection to a Public Communications Network and/or Public Electronic Communications Services to a Consumer, and other End-Users on request, offer to enter into a contract or vary an existing contract with that Consumer, or other End-User, which complies with the following paragraphs. 9.2 Any contract concluded after 25 May 2011 between the Communications Provider and a Consumer, and other End-Users on request, shall specify at least the following minimum requirements in a clear, comprehensive and easily accessible form:

(a)    the identity and address of the Communications Provider;

(b)    the services provided, including in particular whether or not access to Emergency Services and Caller Location Information is being provided, and any limitations on the provision of access to Emergency Services;

(c)     information on any other conditions limiting access to and/or use of services and applications (where such conditions are permitted under national law);

(d)  details of the minimum service quality levels offered, namely the time for initial connection and any other quality of service parameters as directed by Ofcom;

(e)    information on any procedures put in place by the undertaking to measure and shape traffic so as to avoid filling or overfilling a network link, and information on how those procedures could impact on service quality;

(f)      the types of maintenance services and customer support services offered, as well as the means of contacting these services;

(g)    any restrictions imposed by the provider on the use of terminal equipment supplied;

(h)    the Subscriber’s options as to whether or not to include his or her personal data in a directory, and the data concerned;

(i)      details of prices and tariffs, the means by which up-to-date information on all applicable tariffs and maintenance charges may be obtained, payment methods offered and any difference in costs due to payment method;

(j)      the duration of the contract, and the conditions for renewal and termination of services and of the contract, including:

(i)      any minimum usage or duration required to benefit from promotional terms,

(ii)     any charges related to portability of numbers and other identifiers, and

(iii)    any charges due on termination of the contract, including any cost recovery with respect to terminal equipment;

(k)     any applicable compensation and/or refund arrangements which will apply if contracted quality service levels are not met;

(l)      the means of initiating procedures for the settlement of disputes in respect of the contract; and

(m)   the type of action that might be taken by the Communications Provider in reaction to security or integrity incidents or threats and vulnerabilities.

For the avoidance of doubt, any contract between the Communications Provider and a Consumer concluded before 26 May 2011 shall specify the following minimum requirements prescribed by paragraph 9.2 as it applied prior to 26 May 2011.

9.3              Without prejudice to any initial commitment period, Communications Providers shall ensure that conditions or procedures for contract termination do not act as disincentives for End-Users against changing their Communications Provider. In particular, but without limiting the extent of this paragraph:

(a)    Communications Providers who:

(i)      are providing Fixed-Line Telecommunications Services and Broadband Services to Consumers must not, at the end of those Consumers’ initial commitment period, renew those Consumers’ contracts for a further initial commitment period2 unless that Communications Provider has first obtained those Consumers’ Express Consent;3

(ii)     are providing Fixed-Line Telecommunications Services and Broadband Services to Small Business Customers must not, at the end of those Small Business Customers’ initial commitment period, renew those Small Business Customers’ contracts for a further initial commitment period unless that Communications Provider has first obtained those Small Business Customers’ Express Consent.

(b)    For the purposes of Condition 9.3(a):

(i)      “Fixed-Line Telecommunications Services” means Narrowband call and/or line rental services;

(ii)     “Narrowband” means services provided over a Public Telephone Network;

(iii)    “Broadband Services” means services that allow for the transfer of high volumes of data at high speeds;

(iv)   “Express Consent” means the express agreement of the Consumer or Small Business Customer to contract with the Communications Provider in relation to each initial commitment period, where the Communications Provider has obtained such consent separately for each initial commitment period in a manner which has enabled the Consumer or Small Business Customer to make an informed choice;

(v)     “Small Business Customer”, in relation to a public communications provider, means a customer of that provider who is neither himself a communications provider, nor a person who is such a customer in respect of an undertaking carried on by him for which more than ten individuals work (whether as employees or volunteers or otherwise).

9.4              Communications Providers shall not include a term in any contract with a Consumer for the provision of Electronic Communications Services concluded after 25 May 2011 preventing the Consumer from terminating the contract before the end of the agreed contractual period without compensating the Communications Provider for so doing, unless such compensation relates to no more than the initial commitment period (being the period beginning on the day on which the Communications Provider and Consumer have agreed that the contract shall begin and ending on a day falling no more than 24 months thereafter).

9.5              Communications Providers shall ensure that Users are able to subscribe to a contract with a maximum duration of 12 months.

9.6              The Communications Provider shall:

(a)    give its Subscribers adequate notice not shorter than one month of any modifications likely to be of material detriment to that Subscriber;

(b)    allow its Subscribers to withdraw from their contract without penalty upon such notice; and

(c)     at the same time as giving the notice in condition 9.6 (a) above, shall inform the Subscriber of its ability to terminate the contract without penalty if the proposed modification is not acceptable to the Subscriber.

9.7              For the purposes of this Condition:

(a)    “Communications Provider” means a person who provides Public Communications Networks and/or Public Electronic Communications Services;

(b)    “Consumer” means any natural person who uses or requests a Public Electronic Communications Service for purposes which are outside his or her trade, business or profession;

(c)     “User” means a legal entity or natural person using or requesting a Public Electronic Communications Service

Annex 3

Guidance on Compliance with the Prohibition of ARCs under General Condition 9.3

Purpose of this Guidance

These Guidance notes are produced by Ofcom and do not form part of General Condition

9.3  (GC 9.3). They are intended to assist Communications Providers (CPs) by outlining Ofcom’s expectations and providing some guidance as to Ofcom’s likely approach to investigating compliance with GC 9.3. The Guidance is not binding on Ofcom and Ofcom will at all times determine compliance on the basis of individual circumstances whilst having regard to these guidelines. However, where Ofcom decides to depart from the Guidance it expects to give reasons for doing so. Words and expressions used in GC 9.3 shall have the same meaning when used in this Guidance.

GC 9.3 has been modified to prohibit the use of Automatically Renewable Contracts (ARCs) in the provision of fixed voice and fixed broadband services to residential Consumers and Small business Customers (together ‘customers’ for the purposes of this Guidance). An initial commitment period is the period beginning on the first day a contract takes effect and ending on a day falling no more than 24 months thereafter. The modifications to GC 9.3 mean that CPs cannot roll forward (or automatically renew) a customer contract to a new initial commitment period following the expiry of an initial or subsequent initial commitment period without having obtained the Express Consent of the customer.

The Guidance covers three main areas:

1.     The implementation period by which the requirements of GC 9.3(a) (which relate specifically to ARCs) will come into effect.

2.     The requirements of ‘Express Consent’ as it is defined in GC 9.3(b)(iv), and in particular ‘informed choice’.

3.     Particular issues raised by GC 9.3(a)(ii) which prohibits the provision of ARCs to small business customers.

1.     Implementation period

Ofcom’s objective is to ensure that the harm from ARCs is ended as soon as possible and is not extended in the meantime. However, we recognise that ARCs cannot be withdrawn from the market overnight. Therefore the modification to GC 9.3 includes an implementation period to allow time for an orderly withdrawal of ARCs. The modification to GC 9.3 (footnote 3 of the notification at Annex 2 of the ARCs statement) states both the period by which CPs must stop the sale of new ARCs, and migrate existing ARCs customers off ARCs. These periods are set out below.

Stop-sell

The prohibition against selling new ARCs to customers who are not currently on an ARC will come into force on 31 December 2011. This means that it will be unlawful to sell new ARCs after this date. This period was set to allow CPs selling ARCs to stop doing so, taking account of the need to adjust their systems as necessary, and to avoid disruption to consumers or markets as a result of rushed implementation.

Before 31 December 2011, we expect all CPs who sell ARCs to take reasonable steps to limit additional sales. Also, we expect CPs to take a reasonable approach to new or existing ARCs customers who wish to exit their contract before the end of any subsequent (or further) initial commitment period. For example, in such circumstances we think it would be reasonable for CPs to waive any ETCs or exit penalties over and above sums required to recover any outstanding equipment subsidy and other costs directly incurred as a result of the termination.

Migration of existing ARCs customers

For customers who are on an ARC as at 31 December 2011, the prohibition against selling ARCs will come into force on 31 December 2012. This means that any customer who is on an ARC as at 31 December 2011 must be migrated off this ARC by 31 December 2012. This period recognises that the most efficient method of migration may be to transfer each customer at the expiry of the initial commitment period that they are on, to an alternative deal.

However, we recognise that other migration methods may be used, and that ARCs providers may proactively migrate customers away from ARCs at a time other than the ending of an initial commitment period. This will be necessary for ARCs which involve contract terms greater than 12 months where the expiry date does not fall until after 31 December 2012. In these cases, ARCs providers will need to ensure that customers are migrated away from ARCs before the end of the current initial commitment period of their ARC.

As with stop-sell, we expect CPs to take a reasonable approach to ARCs customers who wish to exit their contract before the end of any subsequent (or further) initial commitment period. For example, in such circumstances we think it would be reasonable for CPs to waive any ETCs or exit penalties over and above sums required to recover any outstanding equipment subsidy and other costs directly incurred as a result of the termination.

2.     ‘Informed choice’

General Condition 9.3(iv) defines ‘Express Consent’ as follows:

“Express Consent” means the express agreement of the Consumer or Small Business Customer to contract with the Communications Provider in relation to each initial commitment period, where the Communications Provider has obtained such consent separately for each initial commitment period in a manner which has enabled the Consumer or Small Business Customer to make an informed choice ;

The requirements of this definition are clear. However, we think it is important to clarify that the timing of Express Consent and method by which it is obtained are important in order for customers to be able to make an informed choice.

Method and timing for obtaining Express Consent

Where Express Consent is initiated by a customer, we think it is likely to be reasonable for it to be given at any time.

In all other circumstances, CPs should ensure that customers have sufficient time to properly consider the deal they are being offered (including, for example, allowing them time to,consider the market more generally) before setting deadlines requiring them to opt in to a further initial commitment period. For the avoidance of doubt, the guidelines on the timing of obtaining Express Consent are in relation to CPs contacting customers for the purposes of renewing an initial commitment period rather than, for example, for offering an upgrade or a different deal.

Ofcom has not prescribed specific time frames with which CPs must comply, however there are certain types of behaviour that are unlikely to satisfy the requirements of GC 9.3(iv).

These include (but are not limited to) the following examples where:

•       A CP has asked a customer to provide a “one off” consent which purports to cover all initial commitment periods that that Consumer or Small Business Customer may subsequently enter into (“stacking”);

•       Consent is sought at a time which is too far in advance of the ending of the initial commitment period for a customer to reasonably know what other offers may be available at that time;

•       A CP contacts a customer either on the day that their initial commitment period is due to expire, or very shortly before that day, and requests their consent to enter into a further initial commitment period in circumstances where that customer has not been given an opportunity to consider what other offers may be available,

Ofcom expects CPs to have reasonable steps in place to prevent stacking and to ensure reasonable and appropriate timing for obtaining Express Consent. Therefore, other things being equal, it is generally likely to be reasonable for Express Consent to be obtained by CPs no sooner than six months before the end of each initial commitment period.

3. GC 9.3(a)(ii)
The definition of ‘small business’

The definition of a small business customer for the purposes of General Condition 9.3(a)(ii) is consistent with the definition in Section 52(6) of the Communications Act 2003 which is reproduced here for ease of reference.

In this section “domestic and small business customer”, in relation to a public communications provider, means a customer of that provider who is neither—

(a)  himself a communications provider; nor

(b )a person who is such a customer in respect of an undertaking carried on by him for which more than ten individuals work (whether as employees or volunteers or otherwise).

This means that the prohibition in General Condition 9.3(a)(ii) applies to small business customers with 10 or less employees (the ’10 employee threshold’).

We recognise that it may, at times, be difficult to identify whether or not a customer has 10 employees or less. CPs have informed us that they do not routinely collect or hold information about the number of employees of their business customers. Furthermore, employee numbers can fluctuate over short timescales. We accept, therefore, that estimates as to whether or not a customer falls within the 10 employee threshold may not be precise.

Approach to Enforcement

We note that other regulatory requirements also apply to small business customers as defined by Section 52(6) of the Act. For example, General Condition 14 (GC14) requires that an alternative dispute resolution (ADR) scheme be made available to small business customers. We note also that Ofcom has indicated that it will take a pragmatic and flexible approach to compliance monitoring and enforcement for these requirements. For example, in guidance on GC 14 we said “…we would be satisfied if, rather than having to contact the Complainant to determine whether they have ten or fewer employees (and is therefore potentially ‘eligible’ to take a case to ADR), a CP instead had reasonable processes in place for determining whether business customers are likely to be small businesses for the purpose of this obligation (for example, making an assessment based on annual communications expenditure of that customer).”126

We will take an approach consistent with this to enforcement of GC9.3(a)(ii). In assessing compliance, we will consider whether CPs have taken reasonable steps to identify business customers to whom the prohibition applies. For example, they may (but not be limited to):

•       Identify the size of the business by the annual communications spend and ensure that packages without ARCs are targeted to low spending small business customers.

•       Identify the size of the business by the number of lines it has, and ensure that packages without ARCs are targeted to small business customers with few lines.

•       Ensure that where customers self select an ARC (for example, by purchasing online), they may easily identify themselves as being ‘eligible’ for an ARC.

In addition, in assessing compliance, we will expect CPs to take reasonable steps to inform staff and existing ARCs customers affected by the amendments to GC 9.3 of these new regulations. For example, CPs may:

•       Ensure that sales staff are comprehensively briefed on the regulations.

•       Provide clear information to customers about the regulations on ARCs.

•       Ensure that sales scripts and contract negotiations include necessary information about, for example, migration process, key dates and charges, and any exit procedures.

We also expect that CPs will take a reasonable approach to redress in cases where a small business customer has been sold an ARC inadvertently. Generally, we would expect the CP to enable the customer to exit the contract or move to another package penalty free in such cases (after the ending of any initial MCP).

We believe this flexible approach is an appropriate way to monitor compliance and enforce the small business prohibition for businesses with 10 or less employees.

Glossary

Act: means the Communications Act 2003

Automatically renewable contracts (ARCs): in communications markets, ARCs are contracts where, at the end of a minimum contract period (an MCP – a fixed period of time for which a customer commits to taking services from a Communications Provider), whether this is an initial or subsequent period, the contract is automatically renewed to a new MCP by default unless the customer proactively informs their Communications Provider that they not wish this to happen.

Behavioural Economics: a field of economics which explores the ways in which individual decision-making is influenced by factors such as emotions, habits and cognitive limitations.

BT: British Telecommunications plc.

Bundle: where a customer purchases two or more services from the same CP and receives only one bill from the CP. The customer may or may not receive a discount.

Communications Provider (CP): a person who provides an Electronic Communications Network (ECN) or provides an Electronic Communications Service (ECS)

Consumer: any natural person who uses or requests a publicly available electronic communications service for purposes which are outside his or her trade, business or profession.

Customer churn: also known as customer attrition or customer turnover, the rate at which a business loses its customers.

Default Bias: derived from research into behavioural economics, the default bias is a type of behaviour based on the common observation that regardless the information available or the ease of making a choice, individuals are significantly influenced by default options (and furthermore that they are unlikely to anticipate these effects in advance).

Early Termination Charge (ETC): a charge levied on consumers who terminate their contract before the end of any Minimum Contract Period (or Subsequent Minimum Contract Period).

Fixed-line: means Narrowband call and/or line rental services provided to consumers and small business consumers.

Gaining Provider: the CP to whom the customer is transferring.

Gaining Provider Led (GPL) process: switching process where the consumer only needs to contact the CP they are transferring to in order to switch.

General Conditions (GCs): a set of regulations that apply to anyone who provides an Electronic Communication Service or an Electronic Communications Network as defined in the Act.

Initial Commitment Period: defined in GC 9.4 as “the period beginning on the day on which the Communications Provider and Consumer have agreed that the contract shall begin and ending on a day falling no more than 24 months thereafter”. It is the period during which the consumer cannot leave a contract without being liable to pay a charge to the CP (i.e. by way of an ETC).

Losing Provider: the CP from whom the customer is transferring.

Losing Provider Led (LPL) process: switching process where the consumer needs to contact the CP they are transferring away from as well as the CP they are transferring to, in order to switch.

Minimum Contract Period (MCP): a fixed period of time for which a customer commits to taking services from a CP.

Narrowband: services provided over a traditional Public Telephone Network, excluding services provided over a Cable Network (as defined in General Condition 24.19).

Ofcom: Office of Communications. The regulator for the communications industries, created by the Office of Communications Act 2002.

Rollover date: the date on which an ARC will automatically renew into a new, subsequent MCP, unless the customer contacts the CP to prevent this from happening.

Small businesses: businesses with up to ten employees, as set out in the Act.

Subsequent Minimum Contract Period: a consecutive MCP, starting at the conclusion of the initial MCP.

Switching costs: costs incurred by changing supplier that are not incurred by remaining with the current CP. There are several types of switching costs including transaction costs, compatibility costs, learning costs, contractual costs, equipment costs, uncertainty costs, psychological costs, shopping costs and search costs.