Easyjet Airline Co Ltd, R (on the application of) v Civil Aviation Authority

On 18 January, the CAA published a document entitled “CAA statement on the process for the remaining stages of the review of Heathrow and Gatwick price controls”. It stated:

“Over the last few weeks there has been a substantial amount of productive engagement between BAA, airlines and the CAA on issues associated with the setting of price controls at Heathrow and Gatwick airports from 1 April 2008. It is apparent to the CAA that much progress has been made, but there remain a number of topics on which continued engagement between BAA and its airline users may yield further benefit. Against this background, and given the commitment to the review shown so far by the various parties, the CAA wishes to ensure that all parties are aware of the process for the remaining stages of the review of the price controls at Heathrow and Gatwick airports.

On 20 November 2007 the CAA published for consultation its proposals for price controls for Heathrow and Gatwick airports for the five years from 1 April 2008 to 31 March 2013. In formulating these proposals the CAA had regard to the recommendations of the Competition Commission in its report (published by the CAA on 3 October 2007) and built on agreements reached between the airports and airlines through the process of constructive engagement at the two airports as well as over two years of consultation. The CAA invited written responses to its proposals by 21 January 2008 to be followed by oral hearings with the main stakeholders to be held at the end of January.

The CAA’s document recognised that in certain respects some of its proposals were necessarily provisional in nature. The CAA’s final decision on the price control conditions would therefore take account of the representations made in response to its proposals as well as any new or updated information that emerged during the period of consultation up to the end of January 2008. The CAA’s proposals document identified a number of specific areas where material new information either had been received from BAA shortly before the CAA published its proposals that it had not been possible to reflect in the proposals or was likely to emerge subsequently through the continuing process of constructive engagement. In addition, at the industry seminar held on 3 December 2007 the CAA described the issues that were still outstanding, how they were being taken forward and, where possible, the scale of their potential impact. (The CAA subsequently published on its website the presentation it gave at the seminar). These issues included:

(i)

Operating expenditure, including the costs of meeting security requirements and service standards;

(ii)

Capital expenditure, in particular the timing and scope of particular projects;

(iii)

Cost of capital and financeability;

(iv)

Gatwick traffic forecasts; and

(v)

The treatment of Heathrow expansion

The CAA expected many of these significant issues to be taken forward, and where possible resolved, through further airport/airline discussion during the final two months of the process of constructive engagement. In deciding the most suitable process for the remaining stages of the review the CAA has had in mind the need to ensure proper process while allowing the CAA to be in position in March to make its final price control decision on the basis of the most up to-date information and analysis. The CAA considers it most likely that this will enable it to reach a final decision that will best meet its statutory objectives in section 39 of the Airports Act 1986.

In its November consultation document the CAA invited written responses by 21 January 2008. In October 2005, the CAA set out the process it would follow for the review. The CAA said that it would not rule out responses made after the deadline it had set but could provide no general assurance that it would be able to take account of ‘late’ responses. Having reviewed its processes the CAA is proceeding as follows:

Written representations should be made to the CAA by 21 January 2008, the original date, on the basis of respondents’ latest information. This will allow the CAA to review the responses and publish them promptly, as they will be a key background to the discussion at the oral hearings that are being held by the CAA on 28 and 29 January 2008.

Where outstanding issues are the subject of continuing dialogue between airlines and airports, the CAA encourages the parties to make continuing efforts to achieve resolution by the end of January. These issues should be identified in submissions made by 21 January 2008.

The CAA will not be able to take account of any written or oral representations made after Thursday, 31 January 2008. This is therefore the final deadline for any information resulting from constructive engagement.

On or about 21 January, easyJet submitted detailed written representations to the CAA on the CAA November Proposals. It complained about BAA’s November submissions, expressing concern that BAA had claimed increased operating costs without adequate disclosure to or discussion with the airlines. Additionally, these increased costs (particularly the increased security costs) had not been subject to any degree of real scrutiny. It complained at paras 112 and 113: “The increases to operating costs proposed by BAA are unprecedented and unexplained….BAA has provided no transparency on the increased security costs detailed above, other than a brief presentation to airlines and the CAA made on 19 December 2007. An analysis of the data provided suggests that incremental costs are significantly overstated….” At para 119, it said that BAA’s estimates of security costs “are unreasonable and lack credibility”. At para 121, it said that BAA had not made “any robust case for the inclusion of these extra operating costs….the CAA cannot reasonably seek to include any element of these in its regulatory settlement”.

Meanwhile, BA had been continuing to represent the airlines in discussions with BAA on the subject of the modelling of central search and staffing requirements for additional security. As the 31 January deadline was fast approaching, BA and BAA decided to produce a joint paper in which they identified the areas of agreement and disagreement between them. On 25 January, a joint paper was submitted by BA and BAA setting out the position as they saw it. The central differences between them were summarised in their conclusion as follows:

“BAA believes that its modelling approach is robust and therefore the costs that were presented to the tripartite meeting on 19 December 2007 (ref. BAA-Q5-586) are an accurate representation of the costs that will be incurred by BAA in order to provide the agreed 95%: 5 minute service level within a “more than one bag” security regime.

BA believes that BAA’s modelling adds contingencies at all stages, resulting in an inflated forecast of the number of lanes required and the costs for central search. We believe that 3 fewer lanes are required and that at least £10m of annual opex can be reduced compared with BAA’s current estimate for central search.

Most of the work has concentrated on the central search costs because we believe that this accounted for the major part of the costs as well as the major part of the cost increase. Information provided last week by BAA shows that the other security costs are just as significant as central search, however it has been difficult to examine this and the data that we have seen does not account for the large increases proposed.”

The CAA held oral hearings on 28 and 29 January. At the outset of the meeting of 28 January, Dr Bush, head of the Economic Regulation Group (“ERG”) at the CAA said:

“…what we have set out in a note we put on the web I think last week…was that there are some continuing constructive engagement discussions going on. The last dates for submission for anything is the 31st. If at the end of these sessions parties feel there are points where they need to add some further clarification to what they have said, then they may put something further in, but everything has to be in by the 31st and that is the final deadline. ”

Mr Carrivick, chief executive of BARUK (a body representing the interests of the airlines) complained that the late production of material from BAA had made it difficult to engage in a fruitful CE. At the second meeting, the airlines were well represented. They said that BAA had not given them sufficient information to enable them to make “an informed decision” or play a full part in the CE.

On 31 January, BA prepared a document entitled “Annex 1” which was a detailed response to BAA’s modelling of security costs. It stated that BA was continuing to work with BAA. At para 3 it stated: “In summary, our main concerns at both Heathrow and Gatwick are that BAA have included excessive costings in their proposals and have not sought to deliver a reasonable level of operating cost (manning levels) to deliver the required service standard, but an overstated one”. At para 9, BA made a comment about excessive contingency figures claimed by BAA. They said: “We note in addition, we have not been able to review other security areas and consider this is an area that warrants further review by the CAA”. At para 13, BA made various recommendations to the CAA. These included:

“(ii)

The airlines repeatedly suggested to BAA that:

(a)

The airlines and BAA work collaboratively to determine the reasons for the differences between the modelling.

(b)

The airlines would welcome BAA running the CAST model to verify the findings of the BA simulation model.

(c)

The airlines would like to collaboratively devise an approach with BAA that both parties agree would accurately evaluate the additional infrastructure requirements to accommodate any increases in demand caused by potential extreme exogenous shocks.

(iii)

Until today, 31st January, BAA's position has been that there is insufficient time left. However, some of the information requested has been provided today and the airlines are willing to consider further work if this is of benefit to the CAA.”

Later in the document, BA made a number of specific recommendations of areas of potential savings in BAA’s costs. At para 54, they said:

“Finally with respect to Gatwick, we highlight, as with Heathrow, we have not been able to review non central search costs and therefore believe the CAA should review and scrutinise such cost closely as we are concerned these too may contain excessive contingency and associated cost”.

The CAA panel was due to meet again on 5 February. The report prepared for that meeting noted at para 39 that BAA maintained that its modelling was “robust” and that the security costs it had put forward were “an accurate representation of the costs that will be incurred by BAA in order to provide the agreed 95%: 5 minute service level within a “more than one bag” security regime”. The report continued:

“40.

Airlines have spent more than two months considering these issues in detail with BAA and have struggled to get to the bottom of BAA’s proposal. Nevertheless, airlines have been able to identify enough shortcomings in BAA’s modelling of central search costs to cast doubt over the rigour of BAA’s modelling of other elements e.g. the additional costs of compliance and staff/transfer search where BAA has failed to provide much support for or validation of its proposals.

41.

In order to facilitate debate at the Panel ERG has set out four potential options for dealing with this:

∙ Option 1 – accept BAA’s proposal in full. This option does not seem credible given the scale of the increases proposed, the lack of detailed explanation from BAA and the modelling issues that have been identified by airlines.

∙ Option 2 – retain the current (Nov 07) numbers in price caps with an undertaking to carry out a systematic audit and review of BAA’s projections over the spring with a view to adjusting prices in 2009/10 if BAA is able to provide robust evidence to support its projections. The advantage of this approach is that it signals publicly to BAA that we take a very dim view of BAA’s approach and that BAA must go much further in justifying and explaining cost increases of this magnitude to both the CAA and to airlines. This would also reverse the perverse situation in which the CAA and airlines are seeking to justify reducing BAA’s projections and would instead place the onus on BAA to properly substantiate and validate its proposals. The disadvantage is that this approach creates uncertainty over the settlement as a whole and is likely to have wider implications for the implementation of the service quality regime in particular.

∙ Option 3 – present our misgivings to BAA and give them the opportunity to present a more robust case for including these costs. However, it seems unlikely that BAA would be able to provide much more evidence in the timescales available than it has already exhibited.

∙ Option 4 – apply judgment to reach a view on the appropriate level of costs on the basis of the material that BAA has provided.

∙ (a) one approach would be to accept the airlines proposed reduction to central search costs (i.e. £16m p.a.) and apply an aggressive efficiency factor (e.g. 3% p.a.) to the remainder of security costs;

∙ (b) an alternative would be to base the settlement on a one-bag scenario and effectively disallow the costs BAA has identified for moving back to a more than one-bag approach.”

At the meeting, the CAA panel decided to postpone the discussion of the security costs issue until its next meeting.

Mr Storey, the CAA’s Head of the Price Controls section, emailed BAA with an agenda for a meeting on 8 February, stating that its main concern with BAA’s submissions on security costs was that it had not provided “adequate evidence to justify the proposed opex increases which are very material to the overall price cap assessment…In addition, the CAA needs to see much more evidence….to substantiate the forecasts that BAA is now putting forward…”. In order to meet these concerns, the CAA identified 18 questions which they wanted BAA to answer. These were:

“1.

BAA/Q5/608 sets out a high-level summary of the numbers behind BAA's road map. In relation to the summer 2006 base: please disaggregate these costs by central search, transfer, staff, patrolling, control post or other split as appropriate (where you have included estimates or made pro rata adjustments, please set out details of your assumptions.)

2.

Please provide an equivalent disaggregation for BAA's most recent projection of security costs (i.e. those underlying BAA/Q5/621).

3.

For each of these submissions, please also provide the same disaggregation split by man-year.

4.

Please provide a copy of the model(s) that you have used to support your projections of additional security costs.

5.

There appears to be a difference between BAA and the airlines regarding the performance standard to which BAA is planning and resourcing. The airlines’ modelling appears to suggest that BAA should plan to fail the 5 minute central search target on up to 5% of occasions (i.e. so that no rebate of bonus would be due). BAA's modelling suggests that it is resourcing to achieve a higher certainty of exceeding the 95/5 standard, which might suggest an in-built upward bias to opex.

Joint response by BAA and BA on modelling of central search

6.

Table C on page 5 of the submissions sets out details of BAA's assumptions on central search and other security costs. Please provide more details of, and evidence to support, the £700k you have allocated for cleaning/waste/equipment.

7.

Please provide evidence to support your projection of 'Blackjack' costs (e.g. contacts with suppliers).

8.

In relation to training, communications, cleaning, the CAA is concerned that provision for these costs may already be included in the opex baseline. Please provide evidence to support your view that these costs are incremental.

9.

Table 4 sets out a reconciliation to BAA's financial plan and includes a row for 'staff costs (incl pension adjustment)'. Please provide details of the adjustment that has been made.

10.

Table 4 includes a provision for '3rd Party Agency' of £2.8m. Please provide more details of, and evidence to support, these costs. How is this item accounted for in Table C?

11.

Please confirm whether the pension costs included within the additional staff costs are based on accounting (service cost) cost rather than cash.

12.

Appendix 12 summarises BAA's evidence on the impact of lifting the 1 bag restriction on images per pax. The appendix notes that the impact of laptops and bags was based on surveys examining these items. Please provide more details of these surveys.

Security Cost presentation for CAA / Gatwick Airlines / BAA

Gatwick tripartitite meeting (BAA/Q5/586).

Phase 1

13.

Slide 7 refers to a DfT review and audit of other areas which led to an increase in perimeter patrolling and external control posts, including airside control zones. Please confirm when this audit and review concluded. Please provide more details of the changes to man-year requirements and costs that were implemented as result of this audit. Please also provide a copy of the output of the review and audit (e.g. DfT report).

14.

Slide 12 refers to a '23% improvement included in base' as well as further productivity assumptions which were built into the forecasts. Please set out details of how these efficiency assumptions were incorporated into BAA's July 2007 projections of security costs.

Phase 2

15.

Slide 15 refers to a 'Review of Performance - both Central Search and Others' and notes that this process was assisted by DfT and Group Security. The slide suggests that this review led to a requirement for - 171 additional staff plus additional 'Blackjack' contractors. Please provide a report or other output which summarises the findings of this process.

16.

Slide 19 shows modelling of the trade-off between improved queuing standards and additional costs. Please provide the underlying data for this analysis (NB this may be addressed by question 4 above).

17.

Slide 21 suggests that the additional 87 man-years required for central search will cost £4.4m. Please provide details of the assumptions which underpin this calculation.

18.

Slide 22 refers to the requirement to meet new Vehicle searching requirements which came into effect from May 2007. Please provide details of the contracts and / or tender document put in place with Servisair and PACE to fulfil this requirement.”

Notwithstanding the deadline of 31 January, therefore, the CAA embarked upon a further process of consultation with BAA. On 8 February, the CAA and BAA had a day-long meeting to consider BAA’s security costs projections. A contemporaneous internal CAA email records that “BAA made a reasonably robust defence of their forecasts and we are expecting more data from them today which I will analyse and process for the panel meeting on Tuesday”. It will be noted that the airlines were not involved in these or in any further discussions with BAA or the CAA.

During this period of further consultation, BAA provided a significant quantity of further information in response to the 18 questions set out in the agenda for the meeting of 8 February. These included models used to determine the infrastructure (the number of search lanes) and resources (the number of security officers) that were required. The further information submitted by BAA included, by way of example, (a) breakdowns of security costs by function, man-year and cost; (b) multipliers for converting staff resource requirements into rostered hours and then into labour costs; and (c) information concerning security costs at other BAA airports. On 12 February, BAA also provided the CAA with a copy of its security demand model in an executable form.

An executable form of model is a complex spreadsheet which calculates the number of security lanes likely to be needed to handle predicted passenger levels to a desired standard of queuing times. The number of security lanes that are required has an obvious bearing on the level of labour costs that are incurred on security. If the model is executable (ie its formulae are active), it can be tested for its robustness and also for its sensitivity to alternative inputs and assumptions.

Having now seen the model, Mr Clayton says at para 54 of his second witness statement:

“If airlines had had access to BAA's model, they could have run a variety of alternative service options though it, or through any improvements to that model, and seen what difference it made to resource requirements. Our perspective on such data might have been very different from that of BAA, which was concerned to achieve the maximum earnings from any hours and secure the maximum cushion on its levels of staffing.”

On 15 February, the CAA panel met again. A presentation was made of the current state of the analysis of BAA’s case on security costs. The conclusion was that the accumulation of evidence pointed towards BAA being “over-cautious”. The panel agreed on a £7.5 million reduction in the 2008/09 base level and a 3% annual efficiency factor to be applied to the increment level of costs above the base level used in the CAA November Proposals.

None of the new security costs information that BAA provided during February 2008 was shared with easyJet or any other airlines at the time. Neither easyJet nor any other airline was informed of, or invited to participate in, the post-January round of consultation. Indeed, easyJet was not aware that any further consultation had taken place on the security costs issue until it received BAA’s Summary Grounds in the present proceedings.

The CAA’s conclusion was announced in the March Decision. It reflected the decision that the panel had reached on 15 February. It is not necessary to refer to the March Decision in detail. It contains a summary of the history which I have recounted above. It makes no mention of the important dialogue in February 2008 between the CAA and BAA, an omission which the judge said at [62] was “unfortunate”. At para 5.31, however, the report states:

“The CAA has listened carefully to the arguments advanced by both sides in bi-lateral and tri-lateral meetings to discuss these complex issues in some detail. In parallel, the CAA has undertaken it own detailed review of BAA’s analysis and modelling assumptions”.

At para 5.41, the CAA said that it would be appropriate to increase the 2008/09 base level of security costs by £24 million rather than the £32 million proposed by BAA. At para 5.45, it said that it considered that it was unfortunate that BAA was not able to identify more of its proposed security cost increases for Gatwick earlier in the process. The overall conclusion was that the operating costs allowed for the whole of the Q5 period was £1,381 million with a yield per passenger in the first year (2008/09) of £6.79.

The grounds of challenge

The grounds of challenge are that the March Decision was unlawful because (i) there was procedural unfairness in the manner in which the CAA consulted the airlines before making the decision and (ii) the CAA November Proposals, which purported to be a notice pursuant to regulation 12(1) of the 1986 Regulations, was not a valid notice. A further ground of challenge was that there was a breach of a legitimate expectation on the part of the CAA in conducting a “one-sided and extensive consultation with BAA” on the security issue after 31 January 2008. But Mr Béar QC sensibly agreed that, if the first ground of challenge does not succeed, this further ground will fail and if the first ground succeeds, this further ground does not add anything to easyJet’s case. In these circumstances, I propose to say no more about the legitimate expectation issue.

The first ground: the failure to consult the airlines after the 31 January 2008 deadline was unfair.

The judgment

Mr Béar has helpfully summarised the reasoning of the judge on the unfairness issue (which appears at [58] to [62] of his judgment) as follows. (1) The CAA was entitled and obliged to obtain all information it considered necessary to perform its price-setting function. The further information from BAA obtained after 31 January 2008 was information which it was correct for the CAA to obtain. (2) The CAA did not feel that it needed to seek further representations after February (sic) 2008. It gave careful consideration to all relevant material and concluded that it could not have been assisted by further representations from the airlines. The CAA as the expert was entitled to reach that conclusion and the court should exercise restraint in judicial reviews of regulatory bodies. (3) easyJet has now seen the models supplied by BAA to the CAA in February 2008. There was no specific evidence that further representations could have been made which might have affected the final result. The judge was not persuaded that easyJet was prejudiced. (4) Other airlines at Gatwick had not joined in the claim. This suggested that the figure reached was not one which could in the end be regarded as wrong. (5) BA, which had acted on behalf of the airlines in examining the information supplied by BAA up to the 31 January deadline, had been content to act on what BAA supplied, although it was not satisfied with BAA’s figures.

Summary of the parties’ submissions

In a nutshell, easyJet’s complaint is that CAA acted unfairly in failing to consult the airlines after 31 January 2008. After its own clearly-stated deadline for accepting representations from any party had passed, the CAA obtained and took into account highly material evidence from BAA explaining BAA’s underlying calculations for its estimated additional security costs. The CAA did not inform the airlines about this material and did not afford them any opportunity to scrutinise or comment on it. Moreover, while the CAA afforded BAA an opportunity to improve its case and deal with points in areas in which the CAA informed BAA that evidence was lacking (which opportunity BAA took), no such opportunity was afforded to the airlines.

Mr Béar’s submission is simple and is put most persuasively. He says that there was extensive consultation of the airlines from December 2005 through the medium of the CE and by the CAA leading a working group comprising airline and airport representatives. Between November 2007 and January 2008, there were detailed discussions between the CAA, the airlines (in particular, BA) and BAA in which BAA’s estimated additional security costs were put under close scrutiny. I have already set out the history in some detail. By the time easyJet made its detailed written submissions on 21 January, it was still complaining that BAA’s estimated increased costs were unexplained and BAA had not made a “robust” case for the inclusion of the claimed increased costs. By early February, despite the best efforts of BA, the CAA considered that BAA had not provided adequate evidence to justify the estimated costs that it was now putting forward (see the agenda for the meeting of 8 February). Crucially, the material now provided by BAA to the CAA (see [27] above), which included a copy of its security demand model in fully-executable form, was considered by the CAA to be sufficient to enable it to make a reasonable estimate of the security costs. The reference in the record of the meeting of 8 February to BAA having made a “reasonably robust defence of their forecasts” is of particular significance.

Mr Béar submits that it was unfair to involve the airlines in the consultation process to the point where they (and others) had demonstrated that BAA’s evidence did not justify its estimated security costs, but to deprive them of the opportunity of testing and commenting on the evidence that BAA put forward to make good the very shortcomings that had been previously identified and which easyJet had stated prevented it from making fully informed comments in its written submissions of 21 January.

Mr Beloff QC and Mr Sharpe QC seek to meet this case in two ways. First, they say that the duty to consult is precisely and exhaustively defined by the statutory regime. The only statutory obligation to consult is to publish notice of the conditions or modifications that the CAA intends to impose (which was done in the CAA November Proposals) and to take account of any representations received within 30 days: see regulation 12 of the 1986 Regulations. There is no duty to consult other than that imposed by regulation 12. Secondly, they say that, if and in so far as procedural fairness required consultation of the airlines beyond the scope of regulation 12, CAA discharged this further obligation on the facts of the present case. I shall consider these two submissions in turn.

Is regulation 12 sufficient to meet the requirements of fairness?

I would be inclined to accept that, if the CAA had not embarked on a process of consultation which went beyond what was required of it by regulation 12, it would not have been obliged to do so. Regulation 12 is a reasonable legislative measure intended to ensure that the CAA, as the regulatory authority charged with the duties stated in section 39(2)(a) to (d) of the 1986 Act, ascertains the views of interested parties on a proposal to impose or modify a condition in relation to an airport.

It is true that the courts are often willing to supplement the procedure prescribed in legislation where the statutory procedure is insufficient to achieve fairness and justice and where the supplementary procedure would not frustrate the purpose of the legislation. Thus, in S,T and P v London Borough of Brent and others [2002] EWCA Civ 693, [2002] ELR 556 at [14], Schiemann LJ said:

“So far there is in our judgment nothing in the scheme which blocks the application of the longstanding default principle that the common law will supplement such procedures to the extent necessary to ensure that they operate fairly”.

The question that arises, therefore, is what fairness requires in the context of the functions of the CAA under sections 39 to 45 of the 1986 Act. In Lloyd v McMahon [1987] AC 625 at p 702H, Lord Bridge said:

“To use the phrase which better expresses the underlying concept, what the requirements of fairness demand when any body, domestic, administrative or judicial, has to make a decision which will affect the rights of individuals depends on the character of the decision-making body, the kind of decision it has to make and the statutory or other framework in which it operates. In particular, it is well-established that when a statute has conferred on any body the power to make decisions affecting individuals, the courts will not only require the procedure prescribed by the statute to be followed, but will readily imply so much and no more to be introduced by way of additional procedural safeguards as will ensure the attainment of fairness. It follows that the starting-point for the examination of all the appellants’ submissions on this aspect of the case is the Act of 1982”

In R v Secretary of State for the Home Department, ex parte Doody [1994] 1 AC 531, 560D Lord Mustill explained in a celebrated passage what fairness requires, a matter which he said was “essentially an intuitive judgment”. The six propositions that he derived from the cases were:

“…(1) where an Act of Parliament confers an administrative power there is a presumption that it will be exercised in a manner which is fair in all the circumstances. (2) The standards of fairness are not immutable. They may change with the passage of time, both in the general and in their application to decisions of a particular type. (3) The principles of fairness are not to be applied by rote identically in every situation. What fairness demands is dependent on the context of the decision, and this is to be taken into account in all its aspects. (4) An essential feature of the context is the statute which creates the discretion, as regards both its language and the shape of the legal and administrative system within which the decision is taken. (5) Fairness will very often require that a person who may be adversely affected by the decision will have an opportunity to make representations on his own behalf either before the decision is taken with a view to producing a favourable result; or after it is taken, with a view to procuring its modification; or both. (6) Since the person affected usually cannot make worthwhile representations without knowing what factors may weigh against his interests fairness will very often require that he is informed of the gist of the case which he has to answer.”

I also bear in mind that, as Simon Brown LJ said in R v Secretary of State for Education and Employment and North East London Education Association, ex p M [1996] ELR 162, 206-7, one should avoid a mechanistic approach to the requirements of consultation.

I incline to the conclusion that the discharge of the regulation 12 obligation would be sufficient to meet the requirements of fairness for the following reasons. First, the fact that the Secretary of State has prescribed the scope of its obligation to consult is a relevant factor, although I accept that it is by no means decisive, since ultimately whether a process is fair is a matter for the courts to decide. Secondly, the CAA is a statutory regulator in the field of economic regulation in the general interest of all those who use Gatwick airport. The need to give the right to affected or interested parties to make full representations is less pressing in this context than it is in other contexts where the decision affects the fundamental rights of individuals such as, for example, the liberty of the subject, the right to receive publicly-funded accommodation, health treatment or education or a benefit (such as a licence) which affects a person’s livelihood. Thirdly, the CAA is not acting as a judicial or quasi-judicial body adjudicating between competing claims. It is not its role to balance the airlines’ interests against those of BAA. Rather, it must carry out its own investigation and make an objective assessment of the level of charges which, in its judgment, is most appropriate to achieve the statutory objectives. It conducts it own investigations with its specialist staff and is assisted by the work done by the CC and has to act in the manner which it considers is best calculated to achieve the goals stated in section 39(2) of the 1986 Act. Fourthly, in the light of the foregoing, the requirements of fairness are satisfied by giving all affected or interested parties the opportunity to make written representations on the conditions or modifications that the CAA intends to make (after conducting its own investigations and taking account of a CC report) within 30 days of publication of the proposals.

But I do not find it necessary to reach a firm conclusion on the question whether the discharge of the regulation 12 obligation would be sufficient to satisfy the requirements of procedural fairness. That is because it is not in dispute that, whether or not consultation is a legal requirement, if it is embarked upon, it must be carried out properly: see R v North and East Devon Health Authority, ex p Coughlan [2001] 1 QB 213 at [108]. It is plain from the facts which I have set out in some detail above that the CAA embarked on a process of consultation with the airlines (and others) which went far beyond the confines of regulation 12. Having taken this course, it is not open to the CAA to take refuge in the limited scope of regulation 12. It was incumbent on the CAA to carry out the consultation properly and fairly.

Was the consultation that was undertaken adequate to meet the requirements of fairness?

I have already summarised the reasons why Mr Béar submits that the consultation was unfair: see [36] to [38] above.

Mr Béar relies on R (Eisai) v NICE [2008] EWCA Civ 438, (2008) 101 BMLR 26 as an example of a case where the court held that, having embarked on a process of consultation, the decision-maker had not conducted the process properly. Mr Béar submits that there are striking similarities between the facts of that case and those of the present case and that the results in the two cases should be the same. In Eisai, the appellant pharmaceutical company held the UK marketing authorisation for a drug used in the treatment of Alzheimer’s disease. NICE decided that it was not cost efficient and published guidance to that effect. The claimant sought to challenge by judicial review the procedure by which NICE had reached its decision. In particular, it alleged that as a consultee of NICE, it should have had access to a fully executable version of the economic model that NICE had used to determine the cost effectiveness of the drug, rather than the only partly executable version which NICE had made available to all consultees. Its challenge succeeded on appeal to this court. It was argued on behalf of NICE that there was no duty to consult at all. At [24], Richards LJ doubted whether that was right, but said that “in any event, whether or not consultation was a legal requirement, if it is embarked upon, it must be carried out properly”.

The particular feature of Eisai to which Mr Béar draws attention is that the claimant complained that it could not comment on the model that had been provided to it by NICE since it was not in a fully executable form. This meant that it could not comment on crucial aspects of the evidence on which it wished to comment. Richards LJ (with whom Tuckey and Jacob LJJ agreed) accepted the submission that consultees were entitled to check and comment on the evidence relied on by the decision-maker, rather than having to take the decision-maker’s work on trust.

The robustness or reliability of the model was a key question. At [45], Richards LJ said:

“Whether the model has weaknesses is a matter on which consultees may properly have something to say. Indeed, they already do have things to say on the basis of the read-only version, and attention is properly paid to their representations. The carrying out of additional tests that are possible only with the fully executable version does not give rise to any difference of principle. On the face of it, to limit the extent to which consultees can engage in the legitimate task of testing such an important element in the appraisal process does seem to me to be unfair.”

I accept that there are apparent similarities between Eisai and the present case. But the principles of fairness should not be applied “by rote” (per Lord Mustill in Doody) or by adopting a “mechanistic approach” (per Simon Brown LJ in ex p M). What fairness requires depends on the facts and context of the particular case and on a consideration of all the circumstances of the case. Despite the apparent similarities between the two cases, the fact that this court held that there was procedural unfairness on the facts of Eisai does not necessarily mean that the same conclusion should be reached on the facts of the present case. In any event, in my judgment the similarities are more apparent than real.

It was an important feature of Eisai that, throughout the consultation process, the claimant was asking for a copy of the fully executable version of the model. In the present case, the airlines had made a considerable contribution to the process between December 2006 and January 2008. They had tested the evidence that had been submitted by BAA before the end of January 2008 and had exposed its weaknesses. They had exposed most effectively BAA’s failure to explain important aspects of its case. They had identified the particular areas where BAA would have to provide further material to justify its estimated additional security costs. But critically, on 31 January 2008 (the final date for consultation), BA acting on behalf of the airlines stated in Annex 1 that some of the information that it had requested had been provided by BAA “and the airlines are willing to consider further work if this is of benefit to the CAA” (see [21] above) and “ we have not been able to review non central search costs and therefore believe the CAA should review and scrutinise such cost closely as we are concerned these too may contain excessive contingency and associated cost” (see [22] above).

It can be seen, therefore, that the airlines were content that the final stage of the scrutiny of BAA’s estimates should be undertaken by the CAA without necessarily receiving any further input from themselves. Neither BA nor any of the other airlines objected to the CAA’s completion of the exercise without any further contribution from themselves. They said nothing at the time to indicate that they thought that there was anything unfair about that. It is true that they were aware of the 31 January deadline. But they could have protested that it was unfair that they were being denied the opportunity of testing any further evidence that might be submitted by BAA and they could have asked the CAA to grant a short extension of the consultation period. They did neither of these things. In fact, a short extension of time would probably have sufficed, since by 15 February the CAA had made the decision which was subsequently incorporated into the March Decision. Although it is not a decisive point, it is also of some significance that, apart from easyJet, none of the airlines has ever asserted that there was unfairness on the part of the CAA, let alone joined in these proceedings.

Since the completion of the hearing, the parties have helpfully analysed the 18 questions formulated by the CAA in the agenda for 8 February (see [25] above) and BAA’s responses to these questions. Mr Béar submits that it is immaterial that many of the questions were prompted by work done by BA on behalf of the airlines, because, he submits, nothing in BA’s representations justified the CAA asking the 18 questions at all, let alone in a way which would deprive the airlines of an opportunity to make representations on the answers to the questions.

In my judgment, it is clear that many of the questions were prompted by work that had been done by BA on behalf of the airlines. That is evident from a comparison of the 18 questions with the content of Annex 1 (part of which is set out at [21] and [22] above). They broadly covered much of the ground that had been covered in detail during the discussions between the airlines and BAA. There is disagreement between the parties as to precisely which of the questions were prompted by BA. I do not find it necessary to resolve this dispute. The airlines could not reasonably have been surprised if, following the submission of Annex 1, the CAA asked BAA the 18 questions (or at least many of them).

It was BAA’s answers to these questions that enabled the CAA to reach a final conclusion on 15 February, which led to the March Decision. I have already explained why the attitude evinced by BA on behalf of the airlines in Annex 1 is fatal to easyJet’s argument that it was unfair not to allow the airlines to make representations on BAA’s late material, which principally comprised the responses to the 18 questions. The airlines were content to allow the CAA to assess the responses without any further help from them.

The essential factual difference between Eisai and the present case, therefore, is that in Eisai the claimant made it clear that it wanted to see and comment on the fully-executable version of the model. In the present case, the airlines were content to leave the completion of the process of scrutinising and assessing BAA’s estimated security costs to the CAA without any further input from them. I accept that the airlines might well have been able to contribute something of value to the CAA’s assessment of the material submitted by BAA in February. But the mere fact that they did not have the opportunity of making a further contribution does not necessarily mean that there was procedural unfairness; and for the reasons that I have given, on the facts of this case, there was no unfairness.

A further important distinction between the present case and Eisai is that the NICE function under consideration in Eisai (deciding whether to encourage the use of a drug in the NHS) is not analogous to the function of fixing price caps at airports in pursuance of the statutory objectives identified at section 39 (2) of the 1986 Act.

The decision as to whether there was unfairness ultimately calls for what Lord Mustill described in Doody as an “intuitive judgment” on the basis of all the circumstances of the case. For the reasons that I have given (which to some extent differ from those of the judge), in my judgment, there was no unfairness here.

The second ground: the regulation 12 issue

As we have seen at [11] above, the CAA November Proposals purported to be a notice of the conditions which it intended “to impose for the purposes under regulation 12(1) of the [1986 Regulations]”. In summary, Mr Béar submits that, since it was clear on the face of the CAA November Proposals that they took no account of BAA’s November 2007 submissions on security costs, the charges set out by the CAA were not charges which it intended to impose. In other words, the purported regulation 12(1) notice was not in truth a regulation 12(1) notice at all. He further submits that this was not a procedural irregularity which was capable of being, or which was, cured by the CAA pursuant to regulation 18.

The judgment

The judge dealt with this issue at [43] of his judgment in these terms:

“I did at one stage wonder whether the notice was indeed within Regulation 12 on the basis that there were no firm proposals in a number of respects. Regulation 12 requires that the notice should contain the conditions or modifications the CAA intends to impose. The lateness of the increased security costs made it impossible to reach a firm view on this major issue and there were other investigations to be made as a result of the CC’s recommendations. It would have been open to the CAA to defer the imposition of Q5 for up to 12 months, but in the end the claimant has not submitted that that course should have been adopted. Suffice to say that there were good reasons not to do that. But, as was made clear in Paragraph 20 which I have referred to, the intention to reach a particular figure was expressed albeit it might change as a result of the further consultation or CE to be carried out. Thus I am satisfied that the notice did comply with Regulation 12.”

Discussion

Mr Béar submits that para 6.30 of the CAA November Proposals made it clear that the figures took no account of BAA’s November 2007 submissions on security costs. At para 6.31, the document stated that the CAA had adopted the CC’s projections of security costs “as a holding position”. Mr Béar submits that a “holding position” is something different from a conclusion, even a provisional one. He says that, in reality, the CAA did not even have a provisional final figure, since it had been unable to take into account BAA’s November 2007 submissions. It had not ruled out the claim based on those submissions, nor had it taken a provisional view on it. In short, it was clear that the figures put forward in the CAA November Proposals, on which it sought representations within 30 days (see regulation 12(3)), did not reflect the figures which, subject to those representations, the CAA intended to impose.

Mr Beloff and Mr Sharpe seek to uphold the judge’s reasoning. They submit that the CAA November Proposals were, as they stated on their face, a notice under regulation 12(1). They gave notice of the figure which the CAA intended to impose, viz price caps starting with £6.07 per passenger in 2008/09 increasing annually by no more than RPI inflation plus 2%. The CAA November Proposals made clear the target at which affected persons were to aim. Both the airlines and BAA were able to, and did, make meaningful representations on them. easyJet’s 26 page submission dated 21 January 2008 betrays no hint of confusion about what the CAA was proposing to do.

As regards regulation 18, Mr Béar submits that this cannot be invoked by the CAA to save the day. The deficiency in the CAA November Proposals amounted to a complete failure to comply with regulation 12(1). It was not a mere irregularity with which regulation 18 is concerned. Furthermore, there is a subjective component to regulation 18. The starting point must be a recognition by the CAA that there has been a procedural irregularity which leads to the CAA taking “such steps as it thinks fit before reaching its decision to cure the irregularity” (my emphasis). There was no such recognition here and the CAA did not take such steps as it thought fit to cure the irregularity.

I would accept Mr Béar’s submissions. Regulation 12 provides the only opportunity afforded by the statutory regime for affected or interested persons to make representations on the conditions or modifications which the CAA intends to propose. It is not sufficient to satisfy regulation 12(1) that the CAA gives notice of figures which amount to a target at which affected persons can aim or that the figures are ones on which they can make meaningful representations. In my judgment, regulation 12 contemplates the publishing of a notice of the conditions or modifications which the CAA intends to impose, subject to change in the light of any representations made in response to the notice. If it knows that it is in possession of important material which it has not yet taken into account, it cannot properly say that it has determined a figure which it intends to impose. In such circumstances, it can say that it may impose a published figure; but it cannot say that it intends to do so, because it does not yet know. In my judgment, regulation 12 assumes that the CAA has considered all the important material that has been presented to it and does not expect to receive further such material (apart from any representations it may receive in response to the notice). I refer to “important” material because there is no doubt that the BAA November submissions on its estimated security costs were of central importance to the CAA’s assessment of the price cap. It may well be that a valid regulation 12(1) notice can be served even if the CAA has not fully considered all the material that it has received and expects to receive, provided that the material that has not been considered is not of central importance. I do not need to explore this issue on the facts of the present case.

Mr Beloff and Mr Sharpe submit that the CAA November Proposals were a “provisional” decision by the CAA to impose the price caps stated in them. I would agree that a proper notice published under regulation 12 is a notice of the conditions or modifications which the CAA has provisionally decided to impose and which it provisionally intends to impose. If the consultation exercise prescribed by regulation 12 is to be meaningful, the CAA’s decision and intention must necessarily be provisional, because the CAA may change its mind in the light of any representations it receives. But subject to any response by the CAA to those representations, regulation 12(1) requires the CAA to announce that, at any rate so far as it is aware, it has reached its final decision as to the conditions or modifications that it intends to impose.

That is very different from the situation where the CAA knows that it has not completed its consideration, because it has, or is about to receive, relevant material which it has not yet considered. In that situation, in my view it cannot properly say that it has reached any conclusion, whether provisional or otherwise as to the conditions or modifications that it intends to impose.

On the facts of this case, there is no doubt that the CAA had not considered the BAA November 2007 submissions. The estimated additional security costs claimed by BAA were, and were known by the CAA to be, of central importance to the fixing of the price cap for Gatwick airport during Q5, and so it has proved to be. In these circumstances, I do not consider that the CAA November Proposals constituted a valid notice under regulation 12(1). If the document was not a valid notice, validity could not be conferred on it by the simple expedient of a statement by the CAA that it was a notice published under regulation 12(1).

I do not see why regulation 18 could not in principle have been invoked by the CAA in the circumstances which occurred in this case. In my view, the failure to publish a notice of the proposed conditions or modifications which the CAA intended to impose was a procedural irregularity within the meaning of regulation 18. The CAA should have recognised that it was a procedural irregularity and decided whether it had caused any prejudice and, if so, taken such steps as it thought fit to cure the irregularity before reaching its decision. It is not suggested on behalf of the CAA or BAA that the CAA did recognise that there had been a procedural irregularity or that it considered what steps, if any, it should take to cure the irregularity. In these circumstances, I do not see how the CAA can rely on regulation 18.

But that is not the end of the matter. The airlines were aware of the deficiencies in the CAA November Proposals. These were clear on the face of the document. For the reasons given by Mr Béar (which I have effectively accepted) it was apparent that the notice purportedly given pursuant to regulation 12(1) was not a good notice. But the airlines did not object. On the contrary, they continued to participate in the consultation process that I have described earlier in this judgment and on 21 January 2008, easyJet made detailed representations in response to the CAA November Proposals. It is true that in its representations, it complained about the adequacy of the material provided by BAA. But it did not state that the CAA November Proposals did not constitute a valid regulation 12(1) notice. Instead, through the agency of BA, it continued to participate in the consultation process until 31 January. In other words, it did not seek to take the technical point that has been advanced so skilfully by Mr Béar. Rather, it was concerned with the detailed consultation process that was taking place on the ground.

In my view, it is clear that easyJet was content to make its representations (latterly through BA) informally in the ongoing process of consultation and had no interest in the form of the CAA November Proposals. Provided that it was given the opportunity to take a full and meaningful part in the consultation process, easyJet was content. In my view, it is now too late for easyJet to challenge the CAA November Proposals or (more importantly) the March Decision on the grounds that it had not been preceded by a valid regulation 12(1) notice. The real challenge of easyJet has been to the fairness of the consultation process as a whole. That is the subject of the first ground of challenge which I have already rejected.

Overall conclusion

I would therefore dismiss this appeal on both grounds. My reasons differ to some extent from those of the judge. Standing back from the considerable detail that this case has generated, I cannot accept that the airlines in general (or easyJet in particular) have been unfairly treated by the CAA. They took a very full part in the consultation process until 31 January. Their real complaint is that they were not asked to comment on the important material that was submitted by BAA between about 31 January and 15 February. In disagreement with the judge, I accept that the airlines might have been able to make meaningful representations on this late material. But it is clear that they were content to leave it to the CAA to deal with the last stage of the process without any input from themselves. Perhaps they did not ask to be involved in the last stage because they believed that the CAA would refuse their request, since it had been made clear that the deadline for representations was 31 January. But the fact is that the CAA did receive further submissions from BAA after the deadline. Having decided not to ask the CAA to be allowed to make some further comments, and especially in the light of the comments in the BA Annex 1 document of 31 January (see [21] and [22] above), I do not see how the airlines can properly complain that there was unfairness in this case.

Lord Justice Maurice Kay :

I agree and simply add these few words as a footnote.

My starting point is that, in the words of Lord Mustill (paragraph 43, above) “what fairness demands is dependent on the context of the decision”.  The decision in the present case does not impact on personal liberty, a person’s home, the use which a property owner may make of his property or the right to conduct a business.  Its context is the regulation by a statutory body of one aspect of the prices charged by a private monopoly supplier to its customers.  The duty imposed on the CAA by section 39(2) of the 1986 Act is to act “in the manner which it considers is best calculated (a) to further the reasonable interests of users of airports … ; (b) to promote the efficient, economic and profitable operation of airports; (c) to encourage investment in new facilities at airports … ; and (d) to impose the minimum restrictions that are consistent with the performance by the CAA of its functions”.  Essentially, the CAA is carrying out a balancing exercise.  The airports need to charge the airlines for their services but the regulator has to ensure that they do not abuse their market power by overcharging.  The “users” whose interests the CAA is required to further include both the airlines and their passengers (section 82).  In the context of the present case, airlines and their passengers have an interest in the development of security measures and in protection against being overcharged for them.  The ultimate issue is not the provision or non-provision of a service.  It is simply the charge that may be levied by the airports per passenger.

This puts the decision of the CAA at the “soft” end of the spectrum.  It is no doubt why the statutory obligations imposed by Regulation 12 are light.  It is axiomatic that, having voluntarily decided to engage in a consultative process, the consultation must be carried out fairly.  However, the requisite fairness should reflect the context as I have described it.  It is for this reason that I reject Mr Bear’s submission that the present case is in all fours with Eisai, where the regulatory decision was effectively as to whether or not the company should be enabled to market their drug within the NHS.  I see that as a significantly more intrusive decision which is more likely to attract a higher level of procedural fairness.

It is by reference to the context as I have described it that it is necessary to examine the position that arose at the end of January and into February 2008.  As to that, I can add little to the analysis of Dyson LJ.  As I see it, the non-statutory part of the consultation exercise enabled the airlines, individually and through their agent BA, to highlight shortcomings in BAA’s submission to the CAA.  As at 1 February, and at least partly as a result of the airlines’ input, the CAA was unwilling without more to accede to BAA’s case.  What it did was to raise with BAA the 18 questions set out in paragraph 25, above.  To put it at its lowest, the 18 questions were conditioned to a significant extent by the points raised by BA and the other airlines by 31 January.  In these circumstances, and given the way in which BA had left the matter with the CAA on 31 January (as to which, see paragraphs 52 and 53, above), I do not consider that it was unfair of the CAA to proceed as it did, requiring BAA to satisfy it by reference to the 18 questions.

For these reasons, which do not differ from those of Dyson LJ, I, too, would dismiss easyjet’s appeal.

Lord Justice Rimer: