28 June 2002

Contents

Chapter 1 What is a regulatory option appraisal?

Annex A Guidelines on the regulatory option appraisal (ROA) process

Annex B Background: assessment criteria and previous Oftel analyses

Annex C Determining the appropriate level of regulation

Annex D Model templates for summarising regulatory options

1.1 It is basically a structured way to assess the advantages and disadvantages of different decision options facing Oftel. It codifies existing good practice in Oftel’s decision-making, and provides a mechanism to deliver its strategy of appropriate regulation. The regulatory option appraisal (ROA) process (outlined in Annex A) has several benefits:

• more transparent and rigorous decision-making;
• more structured analysis, based on explicit evidence;
• stronger focus on the viable options for Oftel action.

1.2 These guidelines build on Oftel’s April 2000 Guidelines for assessing the costs and benefits of regulatory action (see Annex B). The term ‘cost-benefit analysis’ (CBA) is often interpreted as requiring in-depth and technical analysis, and sometimes such analysis will be needed. In many cases, such detailed analysis is not so appropriate, but it is still important for Oftel to present its decisions clearly. These guidelines provide a framework for doing that, to be applied to the following:

• EU market reviews;
• consumer protection policy reviews;
• casework;
• ad hoc policy analysis.

1.3 The annexes to this document cover the following areas

• Annex A Guidelines on the regulatory option appraisal process;
• Annex B Background: assessment criteria and previous Oftel analyses;
• Annex C Determining the appropriate level of regulation;
• Annex D Model templates for summarising regulatory options.

1.4 The guidelines are not intended to be a straitjacket, but rather to ensure sound and transparent decision-making. The level of analysis, the appraisal criteria and the format of the analysis will vary between issues as appropriate. Allowing some discretion in the application of the guidelines will allow analysis to be adapted to the particular circumstances of individual issues. This will enable procedures to evolve over time, increasing the rigour of analysis across all areas of Oftel’s work.

1.5 It is also important to note that the appraisal process is designed to assist decision-making rather than being a prescriptive formula. The process is not mechanistic as:

• often Oftel is simultaneously pursuing a range of policy objectives, using multiple criteria, and it is rare for one option to come top on all criteria. For example, the impact on certain stakeholder groups may be crucial in some decisions. Oftel will always have to exercise its judgement;

• Oftel staff cannot fetter the Director-General's discretion.

1.6 The Better Regulation Task Force report on economic regulators recommended in 2001 that economic regulators should conduct more formalised and routine appraisals when making policy decisions. In producing these guidelines to build on its current processes, Oftel has taken account of various existing guidelines for assessing policies. To avoid carrying out separate assessments, under differing criteria, these guidelines form a single procedure that captures the key features of the following related techniques:

• regulatory impact assessments (as used by most government departments);
• customer impact assessments (produced by the National Consumer Council);
• Office of Fair Trading guidelines for competition assessment.

Links to these standalone guidelines are provided in Annex B.

The following paragraphs explain the circumstances in which ROAs should be conducted:

1.7 Significant effects. While Oftel staff may choose to use ROAs as a general decision-making tool, they should always be done where proposals for regulatory action are likely to have significant effects. In practice this means that they must be carried out at least in the following circumstances:

• policy issues (including major policy issues arising in casework) upon which Oftel is consulting or is likely to consult (formally or informally);

• occasions where particular stakeholder groups may be disproportionately affected.

1.8 Degree of Oftel discretion. In some circumstances Oftel may have to take action as a result of formal obligations placed upon it, for example when implementing obligations deriving from European Directives. However, as with other issues, Oftel will still normally have a choice about how it implements policies arising from external obligations. Where such a choice exists, an ROA should be conducted.

1.9 Direction of regulation. An ROA should be carried out regardless of any expected changes in the level of regulation, ie whether regulation is expected to rise, fall or stay the same.

1.10 Quantification. ROAs should be carried out where advantages and disadvantages are mostly not quantifiable, as well as where quantification is more relevant. The key point is that Oftel should conduct rigorous and transparent analysis for all issues.

1.11 Timing. ROAs should be produced as part of, and within the timescales for, standard decision-making process rather than as separate documents. Some specific points about the timing of ROAs are that they should be:

• conducted to inform both the making of decisions (including on whether to do anything at all) and the key issues in the implementation of those decisions;

• ideally produced in initial form in the early stages of analysis, for reporting to the client board or project champion, to avoid making early commitments that narrow the options too soon - although this may be difficult where significant data-gathering, multiple criteria and complex analysis are needed first;

• updated as necessary for inclusion at consultation and decision stages.

For policy issues there are three key stages in developing an option appraisal. The extent to which an appraisal can be usefully developed at an early stage, as distinct from the ROA produced for consultation, will vary between projects/reviews. The three stages are outlined below, along with issues to consider at each stage:

• Identify the issue to be assessed and the specific policy objectives.

• Identify some alternative options for achieving the objectives, guided by Oftel’s strategy principles. There may be legal or financial constraints on the choice of viable options. For example, for market reviews a range of options for addressing SMP is set out in Oftel’s Access Guidelines.

• Where a decision involves several elements, the number of potential options possible combinations may be very high. For example, a decision on providing consumer information may involve three options for each of the three elements of what amount of information to provide, how to make it available, and how to pay for it. It will be more manageable to select a shortlist of option ‘bundles’, that are both practical and sufficiently distinct to make clear what choices are available.

• Consider the time period over which to consider the impact of the decision. For example, market reviews may well have an anticipated date for the next review.

• When identifying different options, consider what level of regulation might be appropriate (see Annex C for guidance).

• Identify the relevant stakeholders involved. Issues may differ between types of consumers, or different types of producers (eg existing operators versus new entrants, network operators versus content providers). Include consideration of stakeholders that are indirectly affected by decisions (eg distributors).

• Assess the advantages and disadvantages of each policy option for each category of stakeholder, and the risks of each option, just based on the information known at this stage. Only very basic quantification, if any, would be anticipated at this stage.

• Include consideration of how options might be monitored and enforced, as this can be an important and expensive part of the final policy.

• Advantages and disadvantages should be recorded in relation to what would happen if Oftel did nothing different (ie record incremental rather than absolute effects). The ‘do nothing’ option does not necessarily mean that nothing will happen. Consider whether in the absence of Oftel action the same actions/effects will occur later, whether something different will happen, or whether something would happen on a different scale or to a different standard. Would Oftel activity actually displace other activity?

• Identify evidence needed to develop the analysis further.

• Consult on the basis of an ROA adjusted for new evidence/thinking since the initial stage, inviting stakeholder input on the methodology, evidence/estimates and analysis.

• Consult on the most likely options that are most practicable and are consistent with Oftel strategy.

• It may also save time later to describe why Oftel has already rejected other options that stakeholders are likely to suggest in consultation.

• Seek where practicable to quantify the monetary value of the costs and benefits (perhaps providing a range of estimates). If the material readily available does not support quantification, use plausible reasoning instead. It is recognised that benefits may often be harder to quantify than costs.

• At least attempt to compare the options on some scale for the different criteria, either by ranking them in order, comparing them on a numerical scale (eg ‘prices would fall by two per cent for option 1 or five per cent for option 2’) or indicating a ratio of the effects (eg ‘option 1 would create twice as much choice as option 2’)

• Ensure that confidentiality of information is maintained.

• Consider the risks of narrowing options too soon and of the potential outcomes of each option not being realised (see box below).

• What is the likelihood that the various costs and benefits will be realised (consider the average outcome, not just the most desirable or most likely outcome)?
• How much would the realised costs and benefits of each option have to change from your forecast in order to change your preference between options?
• How sensitive is the analysis to different assumptions?
• Is the priority to maximise the potential net benefits or to minimise the size of the risks/costs (eg to avoid breaching budget limits)?

• Limit irreversibility in the implementation of decisions – through pilot work, further research, or setting key points for reviewing the project.
• Delay the decision/action – whether to cut risks, because changes are expected in the market, or until more resources are available.
• For major decisions, do more in-depth scenario planning to consider the consequences of possible decisions in greater detail.

• Incorporate changes to the ROA from the consultation process.
• If the revised ROA indicates that disadvantages outweigh advantages of an option, either confirm that the option will not be pursued or seek to adjust some aspects of the policy to produce a net advantage for that option.
• When the statement/decision is issued, include with it a final ROA summarising the choice made and the reasoning involved.

The application of ROAs to casework differs in some respects from the policy processes, as set out below, but most of the considerations will be the same. The ROA procedures should apply to any type of case, including breaches of licence conditions, competition issues, and disputes over terms of interconnection or consumer contracts. However, the procedures will apply only where the processing of a case requires the consideration of new policy issues.

When considering the application of ROA processes, there are three distinct situations in casework:

• Where urgent action is taken for a clear breach of licence condition or anti-competitive behaviour, so that enforcement action needs to be considered, it would not be appropriate to consult. If the breach is ended quickly there seems no value in doing a formalised explanation of Oftel’s decision as it will be based essentially on straightforward application of the relevant legislation.

• Where some novel or complex features arise in the case which suggest a need for minor amendments or clarification to Oftel policy, the policy options will have to be considered, but not via a public consultation. As such cases will often involve information confidential to the parties to a dispute, it may be inappropriate to publish a formalised analysis of the evidence used by Oftel in reaching its decision on such issues. The parties to the dispute will of course be notified of the reasons for Oftel’s decision, and a summary of the decision will be published in Oftel’s Competition Bulletin.

• Some cases cover issues where policy has not yet been developed, or where Oftel’s decision would have a major impact on the telecoms industry or consumers. Such cases would be very much like any policy project, and so the same procedures would apply to these cases as for policy projects, including the analysis of options at consultation stage and in the final published decision.

A number of criteria are produced below upon which advantages and disadvantages of policy options may be assessed. This list is not intended to cover all possible criteria, and it is not intended that each of these criteria can or should be applied to each issue. Key criteria will vary widely between different Oftel projects, partly due to differing mixes of Oftel objectives, or legal constraints on the choice of criteria. Market reviews, for example, will use a quite standard set of criteria for determining remedies for weaknesses in competition. Also, while the criteria are described here in general terms, for specific issues the criteria, and how they are measured, may need more precise definition.

ROAs should consider the impact on the competitive process. The competitive process is more likely to be impacted where Oftel’s decision directly affects the main way(s) in which providers compete (eg price, differentiation of services, innovation), so it is valuable to understand how firms are competing.

Regulatory decisions can affect the competitive process in three ways:

• by affecting the supply and demand factors for firms;
• by specifying particular market outcomes, in turn affecting how firms compete;
• by directly impacting on the competitive process (eg changing barriers to entry).

Some potential impacts on competition are listed in the table below, categorised by the stakeholders more directly involved.

Options can also be compared in terms of their impacts on the following criteria, which for some decisions may be more important than competition-related criteria:

• access for consumers to the services they want/need;
• safety (eg emergency call services for consumers);
• equity (arbitrary or unfair discrimination or restrictions);
• redress for consumers; and
• privacy;

• network integrity (eg essential requirements);
• national security;
• environmental impact;
• e-commerce. Policy decisions should seek to not hinder E-commerce by introducing barriers that make it harder to operate electronically. Effects on access, conveyance and storage of information should be considered. In comparing options, consider how implementation might be done electronically; and
• the costs of implementing the policy (eg monitoring, keeping records), to Oftel and stakeholders. Both initial and ongoing costs should be considered. Implementation costs are different to the effect of the policy itself.

There is no set format for the presentation of ROAs. The key thing is that the arguments for each option are explicitly and appropriately assessed, not the form in which that is done. Tailoring the format of the ROA to the circumstances should in fact promote better analysis.

Tabular formats may be an explicit way to summarise the issues in an ROA arguments, but they are no substitute for a full examination of the issues. Annex D below gives two examples of how to compare options in tabular form.

B.1 You may already be familiar with existing types of assessment to which Oftel has referred in drawing up these guidelines. The following are links to those documents:

Regulatory Impact Assessments

Customer Impact Assessments

Office of Fair Trading guidelines for competition assessment (February 2002)

Oftel CBA guidance April 2000:

2.2 Oftel has in the past published a number of policy documents in which distinct policy options have been presented. The following is a selection of these, including some that represent a more classic cost-benefit analysis approach.

National leased lines

Charge controls on calls to mobile phones

Directory enquiries

Personal numbering

Use of automatic calling equipment

In general, Oftel should use the minimum level of regulation necessary to achieve its objectives. In practice this means considering for each issue whether to use:

• formal regulation: either sector-specific or general, ex ante or ex post;

• own-initiative informal Oftel action (eg consumer information);

• co-regulation: Oftel backs stakeholder-led initiatives through statutory fallback powers, or facilitates/participates in stakeholder groups (to either avoid formal regulation or support the implementation of formal obligations);

• self-regulation (eg codes of practice, memoranda of understanding)

Drawing on experience from previous self- and co-regulatory work (some of which was noted in Oftel’s July 2001 statement), asking the following questions may help to determine the level of regulation:

• Are the parties involved willing to work constructively towards a solution? Do they have a strong incentive to agree, in particular do they perceive a common interest? Are the benefits tangible, quick, and relatively assured? Is there a history of constructive working, or systems that promote it? If not, the prospects for successful self- or co-regulation (including monitoring its success) are weaker.

• Are the parties also able to provide the resources (money or staff) to achieve the objectives? This may constrain operators and Oftel as well as consumer groups.

• What is the priority for the kind of solution needed, in terms of the balance between quality, speed, cost, flexibility, and dependability?

• On quality, will particular stakeholder knowledge, skills or viewpoints add value?

• On cost, is Oftel able to contribute, or to facilitate agreement over funding?

• On speed, will the potential gains from self- and co-regulation (knowledge and skills) be limited by low stakeholder willingness to develop and adapt initiatives?

• On flexibility, how far can Oftel change decisions and systems (pre- and post- implementation), or monitor market developments, as well as self and co-regulatory solutions? More generally, which options offer scope to change course later?

• On dependability, will higher Oftel involvement increase the chance of delivery?

Might the level of regulation be varied, for example between commercial and practical issues, between project stages, or between parts of an overall solution?

D.1 Different options compared more directly on one summary table

• More explicitly compares different options on each of costs, benefits and risks

• May be better for decisions with several aspects, eg market data considers scope, process and publication of data

D.2 One table for each option assessing effects on each stakeholder more explicitly

Net effect on individual stakeholders is assessed more explicitly for each option