Scottish Court of Session Decisions

You are here: BAILII >> Databases >> Scottish Court of Session Decisions >> Clyde Solway Consortium, Re Application For Judicial Review [2001] ScotCS 8 (16 January 2001)
URL: http://www.bailii.org/scot/cases/ScotCS/2001/8.html
Cite as: [2001] ScotCS 8

[New search] [Help]

OUTER HOUSE, COURT OF SESSION

OPINION OF LORD PENROSE

in the petition of

CLYDE SOLWAY CONSORTIUM

Petitioners;

for

Judicial Review of procedures implemented by the Scottish Ministers in determining a contract to be entered into relating to the Management and Maintenance of the Scottish Trunk Road Network, South West Unit; and for suspension and interdict

________________

Petitioners: Glennie, Q.C., Arthurson, Simpson & Marwick, W.S.

First & Second Respondents: Anderson, Q.C., Shand, Solicitor's Office for Scottish Ministers, Solicitor's Office for Advocate General

Third Respondents: Cullen,Q.C., Macroberts

16 January 2001

[1] The petitioners are a consortium of local authorities formed with a view to tendering for a trunk roads management and maintenance contract to be let by the first respondents, the Scottish Ministers, in respect of an area identified as the South West Unit of the Scottish Trunk Roads Network.

[2] In putting the work out to tender, the first respondents required to comply with the Public Works Contracts Regulations 1991 (S.I. 1991 No 2680), as modified for application in Scotland following devolution by the Scotland Act 1998 (Consequential Modifications) (No 1) Order 1999 (S.I. 1999 No 1042), schedule 1 paragraph 20. The 1991 Regulations gave effect to the then current Directive of the Council of the European Communities concerning the award of public works contracts, now superseded by the consolidating Council Directive of 14 June 1993 (93/37/EEC).

It was a matter of agreement that there was no material difference between the provisions of the Regulations applicable in this case and the equivalent provisions of the Directives in force from time to time, and that the European jurisprudence relating to the Directive instructed the proper approach to the interpretation and application of the Regulation.

[3] In accordance with the Directive, the first respondents gave notice in December 1999, by advertisement in the Official Journal of the European Communities, of their intention to put the contract and three related contracts for the remaining trunk roads network areas of Scotland out to competitive tender. The petitioners and other contractors, including the third respondents, Amey Highways, responded to the notice.

[4] The notice published by the first respondents in the Official Journal related to all four contracts. The South West Unit was defined in terms of specified lengths of named trunk roads covering a total route length of approximately 672 km. The nature and extent of the works were described as follows:

"3 (a) Site: It is proposed to let 4 separate contracts in respect of certain defined units on the trunk road network in Scotland, United Kingdom. The contracts will apply to the following 4 units...."

The four units were then described in detail which it is unnecessary to repeat for present purposes. It was then provided:

"(b) Nature and extent of the works ...

4 separate contracts for separate geographical areas of Scotland for the provision of integrated network management and maintenance activities on the specified roads within each unit. Each contract includes, inter alia, a mobilization period, a defects liability period, and a 5-year period for the provision of integrated network management and maintenance activities, with the possibility of extension(s) to such 5-year period, for a period or periods not exceeding a total of 2 years. The approximate works values per annum (turnover) of the contracts ... are as follows:

....

South West Unit: 26 000 000 GBP..."

It was intimated that details of award criteria would be provided in the tender invitation documents.

[5] A preliminary selection procedure was carried out which identified the petitioners, the third respondents and two other contractors as appropriate bodies to be invited to submit tenders for the South West Unit. The selection procedure was not contentious in these proceedings, and details of this stage in the process were not discussed.

[6] The selected tenderers were invited to submit tenders in accordance with instructions and a form of tender issued by the first respondents. The tender form included a schedule of defined and numbered work items divided into series. There were just short of 6,000 separate items. Some of the items called for a fixed rate or fixed rates for stated periods of work of the appropriate class over the duration of the contract period. Some called for unit rates for items of work quantified in terms of volume or lineal or other dimensional extent, but in respect of which no indication was provided of the total amount of work likely to be performed under the contract. With the possible exception of the fixed rate items, the tenders were essentially for a "rates" contract, in anticipation that the quoted rates would apply to the quantities of work which required to be performed from time to time during the term of the contract.

[7] Tenderers were issued with an "information pack" containing historical information about equivalent work carried out over the four year period from1996 to 1999, and directed to an "information room" where further information could be inspected. I was informed by Mr Anderson for the first respondents that the current tendering exercise was the first carried out using four contract areas for the whole country. In the previous round, which included the period 1996-9, there had been three areas in the central belt, managed and maintained by operating companies, and five rural areas where the services were provided under management agency agreements with local roads authorities. As a result of the accounting systems operated by the different bodies involved there did not exist a comprehensive record of the quantities of work carried out. The records set out the cost of the work. The information provided to tenderers for the current round therefore presented financial information. Tenderers, and the first respondents themselves, were required by the application of professional skill and experience to deduce from the financial information the quantities of the several items of work which had generated the costs. Leaving aside for the moment the position of the first respondents, it appeared that tenderers making an assessment of the likely volumes of work in any series of items, or in any individual item, by reference to the historical data, would have required in the first place to estimate historical quantitative data by applying information about rates prevailing over the four base years to the recorded expenditure, whether that information was already in their possession or collected from published and other sources, and then extrapolating forward over the anticipated contract term to determine the total quantities likely to be involved before determining the rates to be quoted for the individual items, given the information available to them about the budget constraints which might apply. There was no requirement on any tenderer to disclose the basis of its calculation, and indeed no mechanism for so doing.

[8] The instructions for tendering were complex. For present purposes it is sufficient to take general note of some of the provisions, and to consider others in more detail. There was provision for inspection of the unit. There was provision for consultation. Tenderers were obliged to produce outline proposals for the work: clause 3.6.8. The first respondents were obliged to conduct consultations with all tenderers to determine the acceptability or otherwise of the tenderers' outline proposals: clause 3.6.2. There was specification of the matters to be discussed in the course of the consultation process: clause 3.6.7. There were detailed provisions regulating the submission of tenders: clause 3.7. The last date for submission of tenders, as finally revised, was noon on 30 October 2000: clause 3.7.3.

[9] Section 4 of the instructions, so far as material, was in the following terms:

"4. ASSESSMENT OF TENDERS

The Assessment Procedure

This section 4 explains for the guidance of Tenderers the basis on which the Scottish Ministers intend to assess the Tenders.
The Contract shall be awarded on the basis of the conforming Tender, which is the most economically advantageous Tender to the Scottish Ministers. The Scottish Ministers shall assess the Tenders on the basis of the assessment criteria specified in section 4.2 of these Instructions for Tendering.
For the avoidance of doubt and in order to demonstrate to Tenderers the transparency of the Assessment Process, The Scottish Ministers shall notify all Tenderers in writing by post of the base quantities (which for the avoidance of doubt shall include numerical and monetary quantities) that they intend to use as part of the assessment process. ..... The details of the base quantities shall be posted to each Tenderer on the 30 October 2000. To ensure the probity of the assessment procedure no Tenders shall be opened earlier than the 30 October 2000. Where there shall be any delay in the posting to each Tenderer the base quantities referred to earlier in this paragraph 4.1.3. the Scottish Ministers shall not open any Tenders until the day after the day on which the said quantities were posted to Tenderers.

Assessment Criteria

4.2.1 The criteria that shall be used to determine which Tender shall be the most economically advantageous in descending order of importance shall be as follows:

(i) (a) Appendices A to I inclusive to the Form of Tender

(b) Appendix G to the Form of Tender ......

(c) Appendix A to Instructions for Tendering - Tender Submission incorporating Annexes (I) to (xviii) inclusive....

The lowest "overall comparative cost of Tender" resulting from the assessment procedure carried out by the Scottish Ministers and specified in this section 4 of these Instructions for Tendering utilising the assessment criteria contained herein applied to the base quantities referred to in paragraph 4.1.3. of these Instructions for Tendering shall be deemed by the Scottish Ministers to be the most economically advantageous Tender.

4.2.2 The detail of the assessment criteria that shall be used to determine whether or not the lowest "comparative cost of Tender" would remain the most economically advantageous Tender to the Scottish Ministers shall be as follows:

Variable items

(a) the base quantities referred to in paragraph 4.1.3. of these Instructions for Tendering shall be varied within the range 70 percent to 130 percent inclusive for the following Series ...

(There followed a list of base quantities within the variable classes)

(b) the base quantities in paragraph 4.1.3. of these Instructions for Tendering shall not be varies for the following Series ...

(There followed a list of base quantities within the variable class)

(c) the base quantities shall be varied within the range 0 percent to 130 percent inclusive for the following Series...

(There followed a statement of the items within this class)

...........

Once these "adjusted comparative cost of Tenders" have been established through the application of the sensitivities within the ranges specified earlier in this paragraph 4.2.2. the Scottish Ministers shall apply weightings to each of the "comparative cost of Tender" and to each of the "adjusted comparative cost of Tenders" after which the sum of each such weighted "comparative cost of tender" and the "adjusted cost of Tender" shall for the purposes of these Instructions for Tendering become the "overall comparative cost of Tender"

If, the Scottish Ministers having carried out the sensitivity assessments equally in respect of all Tenders to establish each "overall comparative cost of Tender, consider that a Tender has been submitted which because of its pricing structure could prove to be more costly when compared with other Tenders assessed on the same basis under any combination of the sensitivity ranges or weightings referred to within this paragraph 4.2.2, then the Scottish Ministers may reject such Tender and give it no further consideration even though it may have the lowest "overall comparative cost of Tender."

It is inappropriate for present purposes to make any general comment on these provisions or their draftsmanship. I shall comment on them only in so far as is necessary for the purposes of this opinion.

[10] Queries from interested contractors were entertained by the first respondents, and General Bulletins were issued to interested parties to answer appropriate queries. In General Bulletin No 9, the first respondents recorded the following question and answer:

"Question 31: What is included in the approximate works values per annum as published in the OJEC Notice?

Response 31: The approximate works value per annum that an Operating Company could undertake is noted below. The values given are dependent upon certain criteria including but not limited to:

budget availability;

Operating Company performance.

On that basis the approximate range of works values could be as follows:

...

South West £16,000,000 - £26,000,000..."

All of the selected tenderers received the same information on the prospective values of the works. It is not entirely clear that response 31 provided information of the kind a contractor might reasonably have anticipated in answer to the question. But on any view of the material in the notice and the bulletin it is clear that contractors were informed that budgetary limits could affect the works instructed, and that the range of values per annum projected for the relevant unit was from £16M to £26M.

[11] Regulation 20 of the 1991 Regulations provides, so far as material:

"(1) ... [A] contracting authority shall award a public works contract on the basis of the offer which-

(a) offers the lowest price, or

(b) is the most economically advantageous to the contracting authority.

(2) The criteria which a contracting authority may use to determine that an offer is the most economically advantageous include price, period for completion, running costs, profitability and technical merit.

(3) Where a contracting authority intends to award a public works contract on the basis of the offer which is the most economically advantageous it shall state the criteria on which it intends to base its decision, where possible in descending order of importance, in the contract notice or in the contract documents."

It was common ground that the first respondents had indicated their intention to award the contract on the basis of the most economically advantageous test.

[12] It was agreed that the Directive has direct effect: Gebroeders Beentjes BV v The Netherlands (1988) E.C.R.4635. But parties presented much of the argument with reference to the domestic regulation. For the petitioners, Mr Glennie argued that the principal purpose of the regulatory scheme, as reflected in the preamble to the 1993 Directive, so far as material for present purposes, was to ensure that the procedures devised for the conduct of tendering exercises for relevant contracts and for the assessment of competing tenders provided for the equal treatment of competing tenderers, objectivity, the elimination of arbitrary choice and subjective assessment factors, and transparency. From the Beentjes case it appeared clearly that while the factors mentioned in Regulation 20 paragraph (1) (b) were not exhaustive, the various criteria which an awarding authority might apply, and state that they intended to apply, were limited to the identification of the economic advantage offered by competing tenders: page 4657 paragraph 19. The court's opinion reflected the views of Advocate General Darmon at page 4649 that the criteria must share a common factor: they must concern the nature of the work to be carried out or the manner of execution of that work to the exclusion of considerations personal to the contractor. The requirement to ensure the equal treatment of tenderers was vouched by The Commission v Kingdom of Denmark (1993) E.C.R. 3353, page 3393 paragraph 33 and page 3394 paragraph 37. There was an obligation to ensure transparency: The Commission v Kingdom of Belgium (1995) E.C.R. 2943. Though that case related to a different Directive it was instructive of a general principle. It was held that a contracting entity had to comply at every stage of comparing tenders with the principle of equal treatment of tenderers and the principle of transparency. At page 2093, paragraph 88 it is stated, in relation to the regulation in question:

"The requirement .. for the contracting entities to state 'in the contract documents or in the tender notice all the criteria they intend to apply to the award, where possible in descending order of importance' is intended precisely to inform potential tenderers of the features to be taken into account in identifying the economically most advantageous offer. All the tenderers are thus aware of the award criteria. Moreover, that requirement ensures the observance of the principles of equal treatment and of transparency."

[13] Against the background of these general observations on the interpretation of the provisions, Mr Glennie argued that while, in some cases, there might be room for variation in contract provisions or the scope of work at a later stage, where a contracting authority intended to apply Regulation 20 (1) (b) in assessing tenders the relevant assessment criteria had to be made known to tenderers in the official notice or in the tender documents. Tenders were time consuming and expensive documents to prepare, and tenderers had to be informed what was required of them. Regulation 20 identified certain criteria which might be applied, but the list was not exhaustive. Other criteria might be selected provided that they shared the common feature that they referred to the offer and not to the personal characteristics of the offeror. The factors must be objective. A wide discretion in selection was not valid. The criteria had to be specified in the relevant documents. If they were not they could not be relied on. There had to be transparency. The objective character of the criteria had to be clear from their terms. Where the expressed criteria were capable of being construed as containing subjective elements it was always possible for the authority to assert that nevertheless they had been applied in an objective way. That assertion was very difficult to disprove. The purpose of the regulatory scheme was to avoid that difficulty and to ensure that the court tested the procedures by reference to the terms of the relevant documents. If on a proper construction those documents allowed arbitrary choice, or if a decision could be taken on a basis that was not objective, the procedures were defective and they must be set aside even if it might established on full proof that there had been no discrimination. If that were not so the policy of the regulations would be frustrated. He referred to The Commission v Kingdom of Belgium at page 2093 paragraphs 88 and 89. In this case there was failure to specify the base data at all. The specification of the sensitivity percentages in clause 4.2.2. was insufficient to provide an objective standard. In any event the specification of sensitivity adjustment values made no sense without specification of the basic data to which they were to be applied. If there were inadequacies in the specification of the criteria, the lowest price tendered ruled: R v Portsmouth City Council (1997) 9 Admin. Law Reports 535. The material provided in the information pack showed levels of expenditure. Men of skill could use such information to form views of the likely quantities of work involved. The petitioners had been involved in the work and could make such assessments as could other tenderers. The first respondents should be in the same position. But the consultation and assessment processes were obscure. If quality aspects were taken into account in the assessment one could not know how that aspect of the assessment was done. Clause 3.6.7 mentioned without specifying any order of importance a number of qualitative factors. It appeared to be an irrational attempt to meet minimum requirements in an area in which there would inevitably be differences between competing tenderers. At the end of the day the awarding authority would know the approach it had adopted, but the tenderers would not. If the exercise were truly based on an objective analysis and comparison of cost data it was difficult to understand how these factors could be accommodated. The estimate of the quantities in this case was material. The preliminaries section of the schedule was restricted to four heads. Many of the items usually found in preliminaries in similar contracts had to be recovered by contractors through the tender rates. Knowledge of quantities was therefore essential. It was accepted that any tenderer would have appreciated that quantitative data was not going to be provided until after tenders were lodged, and that at the assessment stage disclosure of base data satisfied, to that extent, the requirement of transparency. But it was all the more important that tenderers should be able to gauge in advance, using their best endeavours, the quantities likely to be used by the first respondents. The tenderer was entitled to expect that the values introduced by the first respondents into the assessment exercise would reflect the sort of values that could reasonably be deduced from the information provided by them to contractors. If it were possible for tenderers who gauged quantities in a reasonable and skilled fashion from the information provided to arrive at values which were wholly out of line with those used by the first respondents that demonstrated that the process was skewed, and that the tendering process could not be wholly objective, and was useless. If the quantities used in the assessment were out of line with the reasonable contemplation of all tenderers that would inevitably be a distorting factor. It could not be said by the petitioners that the base values for assessment were so out of line with reality that they could never be justified. But there was a narrower basis of attack. The petitioners were entitled to have had a reasonable expectation that the quantities would not be significantly out of line with the basis of the tender. A reasonable relationship was necessary if the public policy objective of furthering competitiveness were to be served. If the awarding authority judged tenders by different yardsticks from those advertised, the result would be that the tenders were not compared on a fair basis, and the process would be arbitrary. The primary exercise in clause 4.2.1 arrived at a price for the base quantities. Most of the criteria specified had no place in that exercise, and could have no bearing on the assessment of value. The economic advantage of the tenders was to be related only to the application to the base quantities of the quoted prices. Although those base quantities were not identified as a criterion, they were the most significant factor in the exercise. The tenderers could not assess the role of the base quantities without knowing the values the first respondents would apply. Moving on to the second stage, the document did not indicate how the sensitivity adjustment was to be carried out. There was no indication that there would be a uniform application of any one percentage variation. Within the permitted range there could be different variations for different items. There was a potential for arbitrariness inherent in the procedure. To be acceptable the criteria had to be pre-determined and specified in the tender documents. There was no indication whatsoever how the weighting procedure was to be carried out. There was no objective criterion by reference to which the adjustment could be made. The procedure permitted a wholly subjective assessment to be made: the choice of values could be wholly arbitrary. Uncontrolled choice was necessarily arbitrary. The figures could be manipulated to achieve a subjectively desirable result. The final over-riding discretionary power to reject a tender appeared to bite at the whole assessment procedure. One was entitled to wonder why there was such a number of control mechanisms. In combination they clearly gave rise to the risk of arbitrariness. Every stated criterion had to be taken into account. In this case it appeared that having stated certain qualitative criteria, the first respondents finally made an assessment exclusively on price alone. On the application of the statutory material, the petitioners' arguments were: (1) the failure to specify the base quantities which were in fact used and intended to be used as assessment criteria was a fatal omission: sensitivity was clearly a criterion, and it could not fulfill that role unless the data to which it applied were similarly a criterion; (2) there was the potential for arbitrary choice inherent in the selection and application of sensitivity factors; and (3) the weighting exercise was wholly unconstrained by any advertised values or range of values and was therefore wholly arbitrary. The procedures infringed the regulations accordingly.

[14] In addition to his primary argument, Mr Glennie submitted that a decision to award a contract on an assessment of tenders in pursuance of the tender procedures on a basis wholly inconsistent with the reasonable understanding of the information provided to tenderers would be irrational in the Wednesbury sense, and would be contrary to the legitimate expectations of the petitioners. He referred to the formulations of principle in Council of Civil Service Unions v Minister for the Civil Service [1985] AC 374; the opinion of Simon Brown L.J. in R v Devon County Council ex parte Baker [1995] 1 All ER 73 and Stannifer v Glasgow Development Agency 1999 S.C. 156. The factual basis of the argument was substantially the same as that relied on in relation to the argument based on the regulations. The point was self-explanatory. The procedures adopted applied to all tenderers equally. But that did not demonstrate that they were fair. To apply a quantitative analysis that had not been communicated to tenderers was unfair and irrational. Further it was contrary to the petitioners' legitimate expectations. The discrete problem was not apparent until the base quantities selected by the first respondents had been intimated and analysed by the petitioners. The consultation process aggravated the problem. It was obvious that the process involved the application of different qualitative factors as among the contending tenderers. But to what effect these differences tended it was impossible to say. Cross-fertilisation could not be ruled out. One could not argue that there should not be discussion, nor that the commercial confidentiality of the individual tenderers should not be respected. But the end product of the assessment exercise was of necessity a money-based comparison. One could not ascertain from that the role played by qualitative factors emerging in the course of the consultation process. Properly construed, what one had in clause 4 of the tender instructions paid lip service to the requirements of the regulatory scheme. But the stipulations had no rational relationship to the assessment exercise. The petitioners were entitled to a remedy on this basis also.

[16] For the third respondents, Mr Cullen adopted the arguments advanced by Mr Anderson, and added independent submissions. He argued that the petitioners' averments were irrelevant. They reflected an error of approach to Community law. The domestic regulation had faithfully implemented the Directive. The court was bound to follow and apply Court of Justice decisions on the meaning and scope of the Directive. On a proper view of the requirements of the European jurisprudence there was no relevant basis for the petitioners' complaint. The recitals in the 1993 Directive demonstrated that the Directive had no impact on the choice of form of contract. That was left to national authorities. In essence the petitioners' complaints related to the form of the contract documents and were unrelated to the objectives the Directive was designed to secure. Beentjes explained clearly the policy of the Directive. The competitive assessment of contracts was permitted provided that an objective system was adopted: paragraphs 35 - 38 of the Advocate General's opinion; paragraphs 15 and 27 of the Court's opinion. The emphasis was on the objective assessment of the provisions of the contract and how it was intended to be carried out rather than on the particular contractors. The test for a national court was whether the policy objectives of the Directive had been satisfied. In this case there had been equal treatment of all tenderers. The objective of the promotion of competition was satisfied. The petitioners' criticisms depended on reading far more into the Directive than the language and the authorities permitted. The Directive was not concerned with the form and terms of contract, which were left to national states. Whatever detailed criticisms might be made about such matters, whatever inconsistencies of language might be highlighted, such matters were exclusively for domestic law: the Directive was not relevant. The Directive was concerned only with the treaty objectives of freedom of establishment and effectiveness of competition. In regard to those matters the first respondents' procedures complied. A complaint that the application of base quantities was unfair could not be brought relevantly within the scope of the objectives. The objections to the application of sensitivity factors and weighting were even further removed from the scope of the Directive. Mr Cullen contended that the petition set out a multiplicity of complaints which lacked logical consistency. It appeared that the pleader had cast around diffusely and picked up any conceivable basis for criticism. Properly understood the petition complained that the first respondents were not entitled to enter into a contract using the procedures they had adopted. It appeared that the petitioners contended that unless the quantities were disclosed at the outset the procedures could not satisfy the regulations. There was no relevant basis for such a contention. The very diffuse nature of the approach was a sign of weakness. As one saw from Stannifer the court should be very wary of this type of complaint from a disappointed or disaffected tenderer. Mr Cullen made a detailed analysis of the averments to highlight his criticism that there was simply an aggregation of diffuse and irrelevant material lacking rigorous analysis, and characteristic of a weak and insubstantial case. There was no obligation in the Directive to state the base quantities in the contract documents. Whether such a mechanism was to be used or not, in determining whether there had been a breach of the regulatory code what one had to do was to consider the terms of the rules. The only obligation incumbent on a contracting authority was to state the criteria by which the assessment was to be made. The authority was not obliged to set out the mechanisms. So far as arbitrariness was concerned, there was, in fact, no allegation of any arbitrary exercise of power. What was said was that on a construction of the documents there was a potential for arbitrary decision. That was no more than a statement of a theoretical possibility. It was irrelevant that the expression of terms might allow an arbitrary exercise of power. What would be required would be a positive averment of arbitrary action. The petitioners clearly could not make such an averment or advance such an argument. The complaints about the application of quantity sensitivities and weighting merely demonstrated that these were pre-determined, and forecast, steps which were carried out in a uniform and equal way. This was the antithesis of an arbitrary exercise. The petitioners had failed to identify any breach of any provision of the regulatory scheme. Further the petitioners' reliance on a discrete principle or requirement for transparency was misconceived. There was no such principle or requirement in the 1993 Directive. The Kingdom of Belgium case related to a different Directive which made express provision for an obligation to demonstrate a minimum level of transparency. The observations of the Court of Justice had to be understood in context. There was nothing to support the generality of the obligation founded on in this case. So far as the common law case was concerned, the petitioners had failed to make out any case based on legitimate expectation. Further in argument the petitioners had given no content to the term "transparency". It was for the petitioners to explain what criteria of transparency they relied on as a measure of the first respondents' obligation, and to spell out the respects in which the conduct of the first respondents failed to meet their international obligations. A comparison of the two relevant directives showed that there was no legislative basis for an obligation of transparency in this case. If there were a principle of transparency, the petitioners had to explain its content. It appeared that the petitioners considered that transparency was bound up with the tender appraisal procedure. But that was outwith the scope of the Directive objectives. There was no measure of any obligation of the first respondents to disclose the detail of their procedures. There were no clear limits to the idea relied on. In the Kingdom of Belgium case the transparency required was of very limited extent: page 2093 paragraphs 88-89. It extended to the disclosure of criteria, not the mechanics of assessment by reference to such criteria. This court should be cautious in relying on observations in a case which could so readily be distinguished. None of the other cases referred to supported a general and wide ranging obligation of transparency such as the petitioners required to support their case. The petitioners' common law case was fundamentally irrelevant. There was no rule requiring authorities to act fairly in their commercial dealings. Mr Cullen discussed the case of Stannifer and submitted that in the absence of a positive rule requiring authorities to act fairly in commercial relations with third parties a case based on legitimate expectation required specific averments of the basis on which the independent obligation arose. There was no general duty to act fairly. The petitioners had failed in averment and argument to make out any basis for an independent obligation in this case. The petition should be dismissed.

[17] A proper understanding of the factual background and of the framework of the tender documents is essential in disposing of the issues before me. The debate of necessity took place on incomplete pleadings, and perhaps reflected at some stages an incomplete understanding of the facts. It is necessary to set out at the outset of the discussion of the arguments what I understand the position to be so that parties have a proper basis for consideration of the views I have reached on the law, and, if necessary, preparing a more comprehensive or corrected factual picture before any further proceedings in the case. It was clear that the contracting authority might adopt any one of a range of solutions in constructing a tender package for a contract for future trunk roads maintenance work. In the nature of things the scope of the work cannot be forecast precisely. There are certain items which can be predicted, because the content of the work can be prescribed by reference to factors other than emerging need. One can require verges to be cut at specified intervals during growing seasons. One can require drains to be cleaned at specified intervals according to season throughout the year. But the need for carriageway repair is likely to depend on a range of factors, including the effectiveness of the original construction, the suitability of the materials used, the effects of weather and the impact of traffic, and so on, which in combination make the precise scope of the work casual rather than certain or predictable.

[18] In such circumstances a contracting authority cannot reasonably be criticised for avoiding the prescription of quantities as a factor in inviting a tender for rates. Mr Anderson was, in my view, correct to emphasise that this was not a tendering process aimed at producing a contract price for specified work. It requires little imagination to anticipate the scope for disputes about the continued validity of rates in changed circumstances if the authority were to attempt such an exercise in an area which was so wholly unpredictable. On the other hand, there is little difficulty in understanding the wish of a contractor for some measure of anticipated quantities as a basis for estimating rates. The recovery of overheads and the generation of profit are major considerations in deciding the balance of rates within a tender, and in pricing individual items. A contractor who must use capital plant to perform particular operations but who cannot be assured of sufficient work of the relevant class to ensure recovery of depreciation from that class of work must include some element of that charge in his general rates. Forecast of likely quantities across the tender creates risk for the contractor. A contractor faced with a description of works without quantities and an invitation to tender rates may reject the invitation or load all rates to compensate for the risk. Inviting quotations and quoting rates for unquantified work exposes the employer and the contractor to risk. But that is of the essence of the commercial process. Where there are several tenderers all parties are necessarily engaged in a competitive process in which professional and commercial judgment contribute to the solution proposed by each tenderer and evaluated by the prospective employer. The final paragraph of clause 4.2.2.makes it quite clear that the first respondents appreciated that whatever the mechanics employed in arithmetical assessment of tenders situations could arise in which a particular tenderer's solution involved disproportionate risk for the employer even where on all of the arithmetical tests the relevant tender appeared to offer best economic advantage.

[19] Against this background, and before considering the European authorities and the regulatory framework, there is a preliminary question, whether there is anything obviously or inherently wrong in a prospective employer inviting tenders which require the competing tenderers to make their own assessments, on a common pool of information, of the likely work content of variable items within the description of works in deciding on the balance and amount of their rates. I understand at least some of the risks this creates for the tenderers. But I cannot identify any inherent unfairness or illegality in the procedure as a matter of domestic contract law. Next, it appears to me to make no difference that the prospective employer reserves to himself the right to make an independent assessment of the quantities likely to be involved, and to use that assessment in weighing the attractiveness of competing tenders. It is almost inevitable that the employer must make such an assessment. The alternative would be that the competing tenderers' assessments, individually or collectively or by some arithmetical process, would come to dictate an important factor in the competition. It is not usually wise to allow the players to dictate the rules of the game. Premature disclosure of the employer's assessment could easily distort the competition. Looking at the background, therefore, in a purely domestic context, there is no obvious ground for criticism of the general structure employed in this case. I shall return to the petitioners' common law case later. But I have formed these general views with that in mind as well as the arguments under the regulations.

[20] These views are wholly consistent with the attitudes adopted by the tenderers in this case. The petitioners and the third respondents lodged tenders. There was no objection to the structure of the tender documents nor to the procedures set out. It was not stated at any time prior to the written notice of these proceedings that there was any difficulty in completing tenders or in making the necessary estimates of the likely pattern of work or of the quantities necessarily assumed to enable a rational decision to be reached on the individual rates or on the balance within the tender.

[21] Turning to the tender documents, it is plain that there are a number of criticisms that might reasonably be made. The interpretation of section 4 is central to the dispute, and there are a number of observations to be made in respect of the clause. It is, in my opinion, plain from the terms of clauses 3.7.3 and 4.1.3 that it was at all times the intention of the first respondents, clearly focused in the tender documents, that there should be no disclosure to competing tenderers of the quantities of work which might be comprised in individual work units or series of units, in the view of the first respondents, before tenders were completed and submitted. Until that stage, tenderers were to be dependent on their own skill and experience in forming views about the balance within the several classes of work involved. Secondly, it is plain that the first respondents intended to use base quantities, in the terms to be intimated to the tenderers on the date or dates identified in terms of clause 4.1.3., as part of the process of assessment of competing tenders. Any reasonably experienced and skilled contractor would have understood from these provisions that the assessment procedure would involve the application of the rates submitted in his tender to base quantities identified for each relevant head of work by the first respondents. It would have been obvious that one result of that exercise would have been a series of products which by addition would have produced a sum which could be represented as a price for those quantities of work at the contractor's rates, but which could be different, at least within certain parameters, from the price which would have emerged from a calculation which used only the rates and the contractor's own estimate of the likely quantities.

[22] The express provisions of clause 4.2.1. provide a vocabulary to express the initial assessment process. The arithmetical process is unduly complicated by the aggregation of financial and non-financial factors in the listing of the assessment criteria. It is not at all clear how one would apply non-financial criteria "to" the base quantities to arrive at an overall cost of tender sum. Apart from appendix E to the Form of Tender, the scope for arithmetical correlation of the assessment criteria and any money sum reflecting the tender cost is difficult to understand or express. But the obscurity does not appear to me to be of central importance. On Mr Anderson's account of events it may not been a real issue in any event. But apart from drawing attention to the conceptual difficulty, the petitioners have not specified any complaint about the actual process, and have not identified any step which might have been taken in fact in making the assessment which could be open to criticism.

[23] Having made provisions for ascertaining the "overall comparative cost" of the tenders, the clause proceeds to identify the lowest sum as the most economically advantageous tender. Having regard to the terms of clause 4.1.2 one might have expected that to be the end of the exercise. But, somewhat illogically, the document proceeds to consider whether the lowest "comparative cost" (not apparently the "overall comparative cost") of tender would remain the same after adjustment. There is an unfortunate change of terminology. Clause 4.2.1. does not use the expression "comparative cost of Tender" which is assumed in clause 4.2.2. It refers to a value described as the "overall comparative cost of Tender" which is only picked up again at the end of clause 4.2.2 where substituted values are provided for that expression. But despite the inelegance of the result, I consider that the scheme of the provisions is reasonably clear. The values first brought out are subjected to a process of adjustment to ascertain the effect of varying the quantitative data for variable items within specified parameters. The basis on which the parameters had been selected is not specified. There is an obscure provision (which I have not quoted) relating to part of this exercise which was not explained, but no party relied on it as a ground of criticism.

[24] The result of the adjustment exercise is then to be assessed by weighting the values brought out, and the sum of the several costs is expressed as the new "overall comparative cost of Tender". This is, on any view, a peculiar use of language, since the sum of a number of values each of which results from costing the works can hardly in common parlance be described in any sense as a cost of tender. However, I have come to the conclusion that the obvious difficulties arise from the application of what, in other circumstances, would be readily understood language to wholly artificial figures, which in themselves are readily understood, and that the obscurities disappear when one analyses out the steps required to enable the arithmetical comparison to be completed.

[25] In summary, the adjustment exercise involves two distinct steps. Within the overall range of parameters specified, the first respondents intimated an intention to compute values for the works at the contractors' specified rates for adjusted volumes of work within the specified groups or series. There was then to be a weighting of the results, which might involve two, three or more calculations, on an unspecified basis selected by the first respondents. The resulting weighted values were to be aggregated and compared. The aggregates were artificially defined as a tender sum, which they could never be, but their sole function was to allow a comparison which took account of the effect of varying the quantitative data. That allowed the balance within competing tenders to be tested. So far there is no irrationality in the process. It is understandable that different contractors might differ in the balance struck by them in pricing a tender between different groups of items according to their understanding of the work involved, their individual capacities to perform the work, and other factors, some of which might reflect their understanding of where profit might be earned.

[26] As already mentioned the final provision of clause 4.2.2. reflected the risk to the employer of the last point. A contractor might so distort the balance of pricing of individual items that, even though his tender resulted in the lowest price on any particular valuation, or in the aggregate, a careful employer would notice that the outcome of applying the distorted rates in foreseeable events might be a higher cost overall. The problem is more easily understood in cases of full bills of quantities where rates for small items, or allowances, might in the out-turn wholly distort the contract price. A contractor who has knowledge of a site may know that the quantities used in the bill are unlikely to be reflected in the execution of the contract. A rate selected to meet the quantity surveyor's estimate, and apparently reasonable, may be thought grossly excessive when it emerges that the actual volume is many times greater. The provision is perfectly intelligible.

[27] It would be impossible for any contractor reading the provisions, however, to forecast what values the first respondents would apply at any stage of the adjustment exercise, either in adjusting for the "sensitivities", or in selecting appropriate weightings. Further, as a matter of language, the first respondents would be under no obligation to apply the same percentage adjustments throughout the exercise. On the language of clause 4.2.2. the first respondents might have applied positive percentage adjustments to some series or items within the same series, and negative percentage adjustment to others. The final power of rejection of distorted pricing structures is not qualified by any arithmetical factor, nor made a function of any test other than the first respondents' judgment. The successful tenderer and the first respondents may find some of these provisions to be the source of difficulty or conflict in the course of carrying out the works. It is not of any moment for present purposes. It is on no view a requirement of the regulations, nor of procedural fairness in the sense of the domestic authorities, that contract documents be elegant or even readily intelligible.

[28] So far as the regulatory scheme is concerned, the first question is whether there has been any failure in implementing the requirements of regulation 20. That there has been less than comprehensive disclosure of the precise arithmetic of the assessment procedure is clear. Without knowledge of the values to be adopted for the base quantities, or the approach to be adopted to the sensitivity adjustments, or the values to be applied in weighting, there could not be such a disclosure. But, in my view, Mr Cullen was correct in his contentions that that was not what the regulatory scheme sought to ensure. The primary requirements of a valid contractual scheme, as identified in paragraphs 18 and 19 of the Beentjes judgment are that the requirements to be applied are restricted to requirements aimed at identifying the offer which is economically the most advantageous. Provided that contracting authorities restrict their requirements to criteria having that characteristic, the scope and definition of the requirements are matters for the national authority within the framework of contract law available to it. As it stated in paragraph 20 of the judgment:

"Furthermore the directive does not lay down a uniform and exhaustive body of Community rules; within the framework of the common rules which it contains, the Member States remain free to maintain and adopt substantive and procedural rules in regard to public works contracts on condition that they comply with all the relevant provisions of Community Law..."

[29] There has clearly been no failure to ensure freedom to enter the competition. There has been no criticism of the official notice in the Journal. In my opinion, there has been no relevant attack on the steps taken to ensure effective competition, in the context of the rules. There has clearly been equality in treatment. The most substantial issue raised by Mr Glennie, in my view, was whether, having regard to the obscurity of the provisions, and the lack of information about base quantities and the limiting values in the adjustment process (in the case of weighting) and the uncertainty of application of the sensitivity adjustments, there could not be effective competition.

As a matter of language at least, there was force in his submission that there was not effective competition where determinative values were undisclosed and unknown and could be selected without prior disclosure. As he said, that process could be said to be by definition arbitrary. In terms of analysis of the tender documents, there is substance in that criticism. But in the circumstance of this case its weight is undermined by two significant factors. In the first place the petitioners acknowledge that they cannot criticise the values adopted by the first respondents as being outwith the range of values that might be selected on a reasonably competent assessment of the probabilities by skilled and experienced people knowledgeable of the business. Secondly, the petitioners have identified as open to criticism on their own approach a very small number of items, only two of which are within the variable class. Quite apart from the fallacy inherent in this approach as a matter of traditional formal logic, it is not contended that the examples selected for attack can properly be said to be typical of the assessment as a whole. Nor is it said, in averment or argument, that the examples illustrate a general error of approach or of principle which undermines the whole exercise. In my view the respondents have correctly argued that the material averments, given specification in the petitioners' note of adjustments, are irrelevant.

[30] Further, although as a matter of construction it is possible to identify scope for purely arbitrary factors to enter into the adjustment process, there is no basis in averment or argument for suggesting that that has at any time been more than a theoretical possibility either in absolute terms, or in terms of altering the balance among tenderers within the process. The adjustment for sensitivities, on the narrative for the first respondents, amounted to no more than a uniform adjustment of variable items to reflect the likely range of sums of money available within the first respondents' budget. The information was disappointing, since the provision would on any view have allowed for a much more imaginative exercise, on the one hand, and, on the other, if that had been the extent of the intention it could have been expressed with rather less mystery and rather more clarity. But, once explained, and in the absence of challenge, the exercise simply reflected the common sense fact, anticipated in the official notice and intimated in the bulletin, that greater or lesser budget allocations would influence the amount of variable work that could be done during the contract period. There remained more mystery about the weighting. However, it is understandable that the first respondents would be unwilling to forecast publicly their estimate of the chances of the roads budget being increased or decreased over the next five years. Acknowledging such a possibility is a fact of political life. Putting a figure on the prospects might risk offering a hostage to fortune in the political arena and be a step which the first respondents could be required to take, if at all, only in a political forum. Nothing was said to give substance to any suggestion that the first respondents were in any way affected by subjective considerations relating to the tenderers or any of them.

[31] The independent argument that the procedures lacked necessary transparency focuses on a different aspect of the issue. Mr Glennie's contention that, even if everything was in the event done properly, the procedures lacked transparency raises, in the first place, a question whether there is an obligation of transparency. The question is difficult not least because the first respondents clearly consider that there is such an obligation, and maintain that they have fulfilled it. The first respondents acknowledged in the contract documents an obligation to demonstrate transparency: Tender Instructions clause 4.1.3. It was the third respondents who argued that the obligation does not exist. The third respondents would be the party most interested in usual course in asserting and enforcing such an obligation if they were in the event the successful tenderer. For present purposes, it is sufficient that the basis of the petitioners' contention, the Kingdom of Belgium case, was conclusively distinguished by Mr Cullen. Whether there is an independent obligation to ensure transparency, and what the content of that obligation might be, as issues logically prior to the question whether there was any infringement in this case, are matters that were not fully debated before me. I cannot express any concluded view on the issue. However, to the extent that there might be such an obligation, for present purposes its scope could not be wider than disclosure of factors bearing on compliance with primary obligations under the regulatory scheme. Since in my view no breach of any such obligation has been made out, either in fact or as a matter of probability or likelihood, there is no basis in which one could hold rationally that there was any material failure in ensuring transparency.

[32] Turning to the common law arguments, I refer first to that based on legitimate expectation. It is clear from the adoption of Simon Brown LJ's observations by the second division in Stannifer that an independent or free-standing remedy based on breach of a legitimate expectation of some procedural practice must be founded on a specific promise or practice or implication from some established practice. No such promise or practice was relied on. The factual circumstances on the contrary demonstrated that this was an unprecedented exercise arising from a new approach to the arrangements for performance of the first respondents' duties as trunk roads authority, involving a fresh definition of unit areas, and the first assessment of quantities overall, as distinct from the previous mixed system of financial control. Nothing was brought to my attention that would instruct a relevant duty on the first respondents which they could be said to have infringed.

[33] Stannifer also undermines the case based on Wednesbury unreasonableness to a considerable extent. It shows that there is in domestic law no general obligation on a contracting authority to act fairly in its commercial affairs. The argument in other respects depended on the analysis of the procedures adopted, and on the contract documents discussed in relation to the regulatory scheme. As already indicated, I consider that the documents lack clarity in certain respects. But there is no basis in law for considering such failures of analysis and draftsmanship to instruct a failure of the kind which gives rise to a remedy at common law.

[34] In the whole circumstances I consider that the respondents have demonstrated that the petition is fundamentally irrelevant and that it should be dismissed. It is therefore unnecessary to consider the first respondents' argument that the proceedings are out of time. I am of opinion that that issue is in any event dependent on the view one forms about the significance of the timing of publication of the base quantities data. But it would be inappropriate to deal with the issue more fully. It is also unnecessary to hear the parties further on the reserved matter of remedy. Since I have not heard submissions, I shall make no comment on that topic.