COUNCIL OF CIVIL SERVICE UNIONS AND OTHERS APPELLANTS AND MINISTER FOR THE CIVIL SERVICE RESPONDENT
22 November. LORD FRASER OF TULLYBELTON. My Lords, Government Communications Headquarters (“GCHQ”) is a branch of the public service under the Foreign and Commonwealth Office, the main functions of which are to ensure the security of the United Kingdom military and official communications and to provide signals intelligence for the Government. These functions are of great importance and they involve handling secret information which is vital to the national security. The main establishment of GCHQ is at Cheltenham where over 4,000 people are employed. There are also a number of smaller out-stations one of which is at Bude in Cornwall.
Since 1947, when GCHQ was established in its present form, all the staff employed there have been permitted, and indeed encouraged, to belong to national trade unions, and most of them did so. Six unions were represented at GCHQ. They were all members, though not the only members, of the Council of Civil Service Unions (“CCSU
”), the first appellant. The second appellant is the Secretary of CCSU
. The other appellants are individuals who are employed at GCHQ and who were members of one or other of the unions represented there. A departmental Whitley Council was set up in 1947 and, until the events with which this appeal is concerned, there was a well-established practice of consultation between the official side and the trade union side about all important alterations in the terms and conditions of employment of the staff.
On 25 January 1984 all that was abruptly changed. The Secretary of State for Foreign and Commonwealth Affairs announced in the House of Commons that the Government had decided to introduce with immediate effect new conditions of service for staff at GCHQ, the effect of which was that they would no longer be permitted to belong to national trade unions but would be permitted to belong only to a departmental staff association approved by the director of GCHQ. The announcement came as a complete surprise to the trade unions and to the employees at GCHQ, as there had been no prior consultation with them. The principal question raised in this appeal is whether the instruction by which the decision received effect, and which was issued orally on 22 December 1983 by the respondent (who is also the Prime Minister), is valid and effective in accordance with article 4 of the Civil Service Order in Council 1982. The respondent maintains that it is. The appellants maintain that it is invalid because there was a procedural obligation on the respondent to act fairly by consulting the persons concerned before exercising her power under article 4 of the Order in Council, and she has failed to do so. Underlying that question, and logically preceding it, is the question whether the courts, and your Lordships' House in its judicial capacity, have power to review the instruction on the ground of a procedural irregularity, having regard particularly to the facts (a) that it was made in the exercise of a power conferred under the royal prerogative and not by statute, and (b) that it concerned national security.
It is necessary to refer briefly to the events which led up to the decision on 22 December 1983. Between February 1979 and April 1981 industrial action was taken at GCHQ on seven occasions. The action took various forms — one-day strikes, work to rule, and overtime bans. The most serious disruption occurred on 9 March 1981 when about 25 per cent. of the staff went on one-day strike and, according to Sir Robert Armstrong, the Secretary to the Cabinet, who made an affidavit in these proceedings, parts of the operations at GCHQ were virtually shut down. The appellants do not accept the respondent's view on the seriousness of the effects of industrial action upon the work at GCHQ. But clearly it must have had some adverse effect, especially by causing some interruption of the constant day and night monitoring of foreign signals communications. The industrial action was taken mainly in support of national trade unions, when they were in dispute with the Government about conditions of service of civil servants generally, and not about local problems at GCHQ. In 1981 especially it was part of a campaign by the national trade unions, designed to do as much damage as possible to government agencies including GCHQ. Sir Robert Armstrong in his affidavit refers to several circular letters and “campaign reports” issued by CCSU
and some of its constituent unions, which show the objects of the campaign. One of these is a circular letter dated 10 March 1981 from the Society of Civil and Public Servants. In a paragraph headed “Selective Strikes” the letter states as follows:
“Union members at certain key Government sites are now on permanent strike. This is the first phase of the selective action: it includes naval supplies and dockyards, locations where the Government finance machine can be disrupted, a Government surveillance centre and the DHSS contributions records computer.” (Emphasis added.)
Among the selective strike areas referred to in the list appended to the letter is “GCHQ Bude, Cornwall.” The seriousness of the intended challenge to the security system of this country can be gauged from the literature issued at the time by the CCSU
, of which the following are examples:
“Our ultimate success depends upon the extent to which revenue collection is upset, defence readiness hampered, and trading relations disrupted by this and future action.”
“Walk-outs in key installations have affected Britain's defence capability in general, and crippled the UK contribution to the NATO exercise ‘Wintex.’”
“another vital part of the Government's Composite Signals Organisation … is to be hit by a strike from Friday, 3 April.”
“48-hour walk-outs have severely hit secret monitoring stations belonging to the Composite Signals Organisation. The Government is clearly worried and will be subject to huge pressure from NATO allies.”
“Defence plans have been upset by the continuing action at naval supplies depots, dock-yards, and other crucial establishments.”
Approaches were made on behalf of the Government to local union officials, and later to national CCSU
officials, to dissuade them from action which would directly adversely affect operations at GCHQ. Some co-operation was given by the local officials, but none at all by national officers. Sir Brian Tovey (former director of GCHQ) gave evidence to the Employment Committee of the House of Commons on 8 February 1984 and told them that, after one of his subordinates had sought to explain to the general Secretary of one of the trade unions the serious consequences that might follow from disruption of certain parts of GCHQ work, the answer was “Thank you. You are telling me where I am hurting Mrs. Thatcher the most.”
In 1982 the Government considered whether measures should be taken to prevent the recurrence of such disruptive action. But at that time the intelligence functions of GCHQ had not been publicly acknowledged by the Government, although they had already been referred to in the newspapers, and it was decided that no action which would involve public acknowledgement of the activities should be taken. In May 1983 following the report of the Security Commission in the case of Geoffrey Prime who had been convicted of espionage at GCHQ, the intelligence role of GCHQ was for the first time publicly acknowledged, and the reason for avoiding public action to deal with disruption was thus removed. The report of the Security Commission on the Prime case is also relevant to this appeal in another way, because it recommended that a pilot scheme should be undertaken to test the feasibility of polygraph security screening at intelligence agencies including GCHQ. The CCSU
were opposed to this recommendation and several meetings were held between their representatives and the Cabinet Office officials to discuss the matter. CCSU
were concerned that the polygraph might be introduced without adequate consultation and on 9 January 1984 Sir Robert Armstrong wrote to the chairman of their general policy committee explaining that before a decision was taken for the definitive introduction of polygraph, as distinct from the experimental pilot scheme, there would certainly need to be consultations. That was the last word on the polygraph question before the announcement on 25 January 1984 that national trade unions were to be excluded from GCHQ. Their exclusion would necessarily prevent their playing any part in further consultations on the polygraph and that was one of their reasons for resenting the decision of 22 December 1983.
Course of the proceedings
The trade unions, and some at least of the employees at GCHQ, objected strongly to the decision made on 22 December 1983 and announced on 25 January 1984. Representatives of the trade unions met the Minister for the Civil Service on two occasions in February 1984 to express their objections. They also met Sir Robert Armstrong several times. They presented a draft agreement to prevent disruption at certain parts of GCHQ but the draft was rejected by the Government and no agreement was reached about changing the Government's decision. Eventually the first and second appellants obtained leave from Glidewell J. on 8 March 1984 to bring proceedings for judicial review against the Minister for the Civil Service in respect of the instruction of 22 December 1983 and against the Foreign Secretary in respect of certificates which he had issued under the Employment Protection Act 1975, section 121(4), and the Employment Protection (Consolidation) Act 1978, section 138(4), to give effect to the instruction by discontinuing, on national security grounds, the right of staff to appeal to industrial tribunals. The attack on these certificates has been abandoned, and the attack on the instruction is now limited to seeking a declaration that it is invalid; the remedy of certiorari is no longer sought.
Glidewell J. granted a declaration that
“the instruction purportedly issued by the Minister for the Civil Service on 22 December 1983 that the terms and conditions of service of civil servants serving at GCHQ should be revised so as to exclude membership of any trade union other than a departmental staff association approved by the Director of GCHQ was invalid and of no effect.”
His reason for granting the declaration was that there had been a procedural irregularity in failing to consult before issuing the instruction. I take this opportunity of expressing my respectful admiration for the carefully reasoned opinion of the learned judge which has substantially assisted me and, I believe, my noble and learned friends.
Against that declaration the respondent appealed. The Court of Appeal (Lord Lane C.J., Watkins and May L.JJ.) reversed the judge's decision and dismissed the appellants' application for judicial review. They also dismissed a cross-appeal by the appellants.
The appeal raises a number of questions. I shall consider first the question which I regard as the most important and also the most difficult. It concerns the royal prerogative.
The Royal Prerogative
The mechanism on which the Minister for the Civil Service relied to alter the terms and conditions of service at GCHQ was an “instruction” issued by her under the Order in Council of 1982, article 4. That article so far as relevant provides as follows:
“As regards Her Majesty's Home Civil Service — ( a) the Minister for the Civil Service may from time to time make regulations or give instructions — … (ii) for controlling the conduct of the service, and providing for the classification of all persons employed therein and … the conditions of service of all such persons; …”
The Order in Council was not issued under powers conferred by any Act of Parliament. Like the previous Orders in Council on the same subject it was issued by the sovereign by virtue of her prerogative, but of course on the advice of the government of the day. In these circumstances Mr. Alexander submitted that the instruction was not open to review by the courts because it was an emanation of the prerogative. This submission involves two propositions: (1) that prerogative powers are discretionary, that is to say they may be exercised at the discretion of the sovereign (acting on advice in accordance with modern constitutional practice) and the way in which they are exercised is not open to review by the courts; (2) that an instruction given in the exercise of a delegated power conferred by the sovereign under the prerogative enjoys the same immunity from review as if it were itself a direct exercise of prerogative power. Mr. Blom-Cooper contested both of these propositions, but the main weight of his argument was directed against the second.
The first of these propositions is vouched by an impressive array of authority, which I do not propose to cite at all fully. Starting with Blackstone's Commentaries
, 15th ed. (1809), p. 251 and Chitty's Prerogatives of the Crown
(1820), pp. 6–7 they are at one in stating that, within the sphere of its prerogative powers, the Crown has an absolute discretion. In more recent times the best known definition of the prerogative is that given in Dicey, Law of the Constitution
, 8th ed. (1915), p. 421 which is as follows:
“The prerogative is the name for the remaining portion of the Crown's original authority, and is therefore, as already pointed out, the name for the residue of discretionary power left at any moment in the hands of the Crown, whether such power be in fact exercised by the King himself or by his ministers.”
Dicey's definition was quoted with approval in this House in Attorney-general v. De Keyser's Royal Hotel Ltd.  AC 508
, 526 by Lord Dunedin and was impliedly accepted by the other Law Lords in that case. In Burmah Oil Co. Ltd. v. Lord Advocate, 1964 SC (HL) 117
Lord Reid, at p. 120, referred to Dicey's definition as being “always quoted with approval” although he said it did not take him very far in that case. It was also referred to with apparent approval by Roskill L.J. (as my noble and learned friend then was) in Laker Airways Ltd. v. Department of Trade  QB 643
, 719. As De Keyser's case shows, the courts will inquire into whether a particular prerogative power exists or not, and, if it does exist, into its extent. But once the existence and the extent of a power are established to the satisfaction of the court, the court cannot inquire into the propriety of its exercise. That is undoubtedly the position as laid down in the authorities to which I have briefly referred and it is plainly reasonable in relation to many of the most important prerogative powers which are concerned with control of the armed forces and with foreign policy and with other matters which are unsuitable for discussion or review in the law courts. In the present case the prerogative power involved is power to regulate the Home Civil Service, and I recognise there is no obvious reason why the mode of exercise of that power should be immune from review by the courts. Nevertheless to permit such review would run counter to the great weight of authority to which I have briefly referred. Having regard to the opinion I have reached on Mr. Alexander's second proposition, it is unnecessary to decide whether his first proposition is sound or not and I prefer to leave that question open until it arises in a case where a decision upon it is necessary. I therefore assume, without deciding, that his first proposition is correct and that all powers exercised directly under the prerogative are immune from challenge in the courts. I pass to consider his second proposition.
The second proposition depends for its soundness upon whether the power conferred by article 4 of the Order in Council of 1982 on the Minister for the Civil Service of “providing for … the conditions of service” of the Civil Service is subject to an implied obligation to act fairly. (Such an obligation is sometimes referred to as an obligation to obey the rules of natural justice, but that is a less appropriate description, at least when applied, as in the present case, to a power which is executive and not judicial.) There is no doubt that, if the Order in Council of 1982 had been made under the authority of a statute, the power delegated to the Minister by article 4 would have been construed as being subject to an obligation to act fairly. I am unable to see why the words conferring the same powers should be construed differently merely because their source was an Order in Council made under the prerogative. It is all the more difficult in the face of article 6(4) of the Order in Council of 1982 which provides that the Interpretation Act 1978 shall apply to the Order; it would of course apply to a statutory order. There seems no sensible reason why the words should not bear the same meaning whatever the source of authority for the legislation in which they are contained. The Order in Council of 1982 was described by Sir Robert Armstrong in his first affidavit as primary legislation; that is, in my opinion, a correct description, subject to the qualification that the Order in Council, being made under the prerogative, derives its authority from the sovereign alone and not, as is more commonly the case with legislation, from the sovereign in Parliament. Legislation frequently delegates power from the legislating authority — the sovereign alone in one case, the sovereign in Parliament in the other — to some other person or body and, when that is done, the delegated powers are defined more or less closely by the legislation, in this case by article 4. But whatever their source, powers which are defined, either by reference to their object or by reference to procedure for their exercise, or in some other way, and whether the definition is expressed or implied, are in my opinion normally subject to judicial control to ensure that they are not exceeded. By “normally” I mean provided that considerations of national security do not require otherwise.
The courts have already shown themselves ready to control by way of judicial review the actions of a tribunal set up under the prerogative. Reg. v. Criminal Injuries Compensation Board, Ex parte Lain  2 Q.B. 864 was such a case. In that case Lord Parker C.J. said, at p. 881:
“I can see no reason either in principle or in authority why a board set up as this board was set up is not a body of persons amenable to the jurisdiction of this court. True it is not set up by statute but the fact that it is set up by executive government, i.e., under the prerogative, does not render its acts any the less lawful. Indeed, the writ of certiorari has issued not only to courts set up by statute but to courts whose authority is derived, inter alia, from the prerogative. Once the jurisdiction is extended, as it clearly has been, to tribunals as opposed to courts, there is no reason why the remedy by way of certiorari cannot be invoked to a body of persons set up under the prerogative.”
That case was concerned with the actions of a board or tribunal exercising functions of a judicial character, but it is now established that certiorari is not limited to bodies performing judicial functions. In Reg. v. Secretary of State for Home Affairs, Ex parte Hosenball  1 W.L.R. 766 which was concerned with the actions of the Secretary of State himself in refusing to give information about the reasons for making a deportation order against an alien, the Divisional Court and the Court of Appeal refused to make an order of certiorari because the refusal had been based on grounds of national security but, if it had been made in what Lord Denning M.R., at p. 778, called an “ordinary case” — that is, one in which national security was not involved — the position would have been different. Lord Denning M.R. said, at p. 781:
“if the body concerned, whether it be a minister or advisers, has acted unfairly, then the courts can review their proceedings so as to ensure, as far as may be, that justice is done.”
Accordingly I agree with the conclusion of Glidewell J. that there is no reason for treating the exercise of a power under article 4 any differently from the exercise of a statutory power merely because article 4 itself is found in an order issued under the prerogative.
It follows, in my opinion, that some of the reasoning in Reg. v. Secretary of State for War  2 Q.B. 326 and Griffin v. Lord Advocate, 1950 S.C. 448 is unsound, although the decisions themselves might perhaps be supported on the ground that they related to actions by the Crown connected with the armed forces. The former case was of course decided long before the modern development of judicial review and the latter, which was a decision of Lord Sorn in the Outer House, mainly followed it.
The duty to consult
Mr. Blom-Cooper submitted that the Minister had a duty to consult the CCSU
, on behalf of employees at GCHQ, before giving the instruction on 22 December 1983 for making an important change in their conditions of service. His main reason for so submitting was that the employees had a legitimate, or reasonable, expectation that there would be such prior consultation before any important change was made in their conditions.
It is clear that the employees did not have a legal right to prior consultation. The Order in Council confers no such right, and article 4 makes no reference at all to consultation. The Civil Service handbook ( Handbook for the new civil servant
, 1973 ed. as amended 1983) which explains the normal method of consultation through the departmental Whitley Council, does not suggest that there is any legal right to consultation; indeed it is careful to recognise that, in the operational field, considerations of urgency may make prior consultation impracticable. The Civil Service Pay and Conditions of Service Code expressly states:
“The following terms and conditions also apply to your appointment in the Civil Service. It should be understood, however, that in consequence of the constitutional position of the Crown, the Crown has the right to change its employees' conditions of service at any time, and that they hold their appointments at the pleasure of the Crown.”
But even where a person claiming some benefit or privilege has no legal right to it, as a matter of private law, he may have a legitimate expectation of receiving the benefit or privilege, and, if so, the courts will protect his expectation by judicial review as a matter of public law. This subject has been fully explained by my noble and learned friend, Lord Diplock, in O'Reilly v. Mackman  2 AC 237
and I need not repeat what he has so recently said. Legitimate, or reasonable, expectation may arise either from an express promise given on behalf of a public authority or from the existence of a regular practice which the claimant can reasonably expect to continue. Examples of the former type of expectation are Reg. v. Liverpool Corporation, Ex parte Liverpool Taxi Fleet Operators' Association  2 Q.B. 299 and Attorney-General of Hong Kong v. Ng Yuen Shiu  2 AC 629
. (I agree with Lord Diplock's view, expressed in the speech in this appeal, that “legitimate” is to be preferred to “reasonable” in this context. I was responsible for using the word “reasonable” for the reason explained in Ng Yuen Shiu, but it was intended only to be exegetical of “legitimate”.) An example of the latter is Reg. v. Board of Visitors of Hull Prison, Ex parte St. Germain  Q.B. 425 approved by this House in O'Reilly, at p. 274D. The submission on behalf of the appellants is that the present case is of the latter type. The test of that is whether the practice of prior consultation of the staff on significant changes in their conditions of service was so well established by 1983 that it would be unfair or inconsistent with good administration for the Government to depart from the practice in this case. Legitimate expectations such as are now under consideration will always relate to a benefit or privilege to which the claimant has no right in private law, and it may even be to one which conflicts with his private law rights. In the present case the evidence shows that, ever since GCHQ began in 1947, prior consultation has been the invariable rule when conditions of service were to be significantly altered. Accordingly in my opinion if there had been no question of national security involved, the appellants would have had a legitimate expectation that the minister would consult them before issuing the instruction of 22 December 1983. The next question, therefore, is whether it has been shown that consideration of national security supersedes the expectation.
The issue here is not whether the minister's instruction was proper or fair or justifiable on its merits. These matters are not for the courts to determine. The sole issue is whether the decision on which the instruction was based was reached by a process that was fair to the staff at GCHQ. As my noble and learned friend Lord Brightman said in Chief Constable of the North Wales Police v. Evans  1 WLR 1155
, 1173: “Judicial review is concerned, not with the decision, but with the decision-making process.”
I have already explained my reasons for holding that, if no question of national security arose, the decision-making process in this case would have been unfair. The respondent's case is that she deliberately made the decision without prior consultation because prior consultation “would involve a real risk that it would occasion the very kind of disruption [at GCHQ] which was a threat to national security and which it was intended to avoid.” I have quoted from paragraph 27(i) of the respondent's printed case. Mr. Blom-Cooper conceded that a reasonable minister could reasonably have taken that view, but he argued strongly that the respondent had failed to show that that was in fact the reason for her decision. He supported his argument by saying, as I think was conceded by Mr. Alexander, that the reason given in paragraph 27(i) had not been mentioned to Glidewell J. and that it had only emerged before the Court of Appeal. He described it as an “afterthought” and invited the House to hold that it had not been shown to have been the true reason.
The question is one of evidence. The decision on whether the requirements of national security outweigh the duty of fairness in any particular case is for the Government and not for the courts; the Government alone has access to the necessary information, and in any event the judicial process is unsuitable for reaching decisions on national security. But if the decision is successfully challenged, on the ground that it has been reached by a process which is unfair, then the Government is under an obligation to produce evidence that the decision was in fact based on grounds of national security. Authority for both these points is found in The Zamora  2 AC 77
. The former point is dealt with in the well known passage from the advice of the Judicial Committee delivered by Lord Parker of Waddington, at p. 107:
“Those who are responsible for the national security must be the sole judges of what the national security requires. It would be obviously undesirable that such matters should be made the subject of evidence in a court of law or otherwise discussed in public.”
The second point, less often referred to, appears at p. 106 and more particularly at p. 108 where this passage occurs:
“In their Lordships' opinion the order appealed from was wrong not because, as contended by the appellants, there is by international law no right at all to requisition ships or goods in the custody of the court, but because the judge had before him no satisfactory evidence that such a right was exercisable.” (Emphasis added.)
What was required was evidence that a cargo of copper in the custody of the Prize Court was urgently required for national purposes, but no evidence had been directed to that point. The claim on behalf of the Crown that it was entitled to requisition the copper therefore failed; considering that the decision was made in 1916 at a critical stage of the 1914–1918 war, it was a strong one. In Chandler v. Director of Public Prosecutions  AC 763
, which was an appeal by persons who had been convicted of a breach of the peace under section 1 of the Official Secrets Act 1911 by arranging a demonstration by the Campaign for Nuclear Disarmament on an operational airfield at Wethersfield, Lord Reid said, at p. 790:
“The question more frequently arises as to what is or is not in the public interest. I do not subscribe to the view that the Government or a minister must always or even as a general rule have the last word about that. But here we are dealing with a very special matter — interfering with a prohibited place which Wethersfield was.”
But the court had had before it evidence from an Air Commodore that the airfield was of importance for national security. Both Lord Reid and Viscount Radcliffe, at p. 796, referred to the evidence as being relevant to their refusal of the appeal.
The evidence in support of this part of the respondent's case came from Sir Robert Armstrong in his first affidavit, especially at paragraph 16. Mr. Blom-Cooper rightly pointed out that the affidavit does not in terms directly support paragraph 27(i), ante pp. 401H–402A. But it does set out the respondent's view that to have entered into prior consultation would have served to bring out the vulnerability of areas of operation to those who had shown themselves ready to organise disruption. That must be read along with the earlier parts of the affidavit in which Sir Robert had dealt in some detail with the attitude of the trade unions which I have referred to earlier in this speech. The affidavit, read as a whole, does in my opinion undoubtedly constitute evidence that the Minister did indeed consider that prior consultation would have involved a risk of precipitating disruption at GCHQ. I am accordingly of opinion that the respondent has shown that her decision was one which not only could reasonably have been based, but was in fact based, on considerations of national security, which outweighed what would otherwise have been the reasonable expectation on the part of the appellants for prior consultation. In deciding that matter I must with respect differ from the decision of Glidewell J. but, as I have mentioned, I do so on a point that was not argued to him.
The judge held that had the prior consultations taken place they would not have been so limited that he could confidently say that they would have been futile. It is not necessary for me to reach a concluded view on this matter, but as at present advised I am inclined to differ from the learned judge, especially because of the attitude of two of the trade union members of CCSU
which declared that they were firmly against any no-strike agreement.
The Court of Appeal considered the proper construction of certain international labour conventions which they cite. I respectfully agree with Lord Lane C.J. who said that “the correct meaning of the material articles of the Conventions is by no means clear,” but I do not propose to consider the matter as the Conventions are not part of the law in this country.
Mr. Blom-Cooper submitted that the oral direction did not qualify as an “instruction” within the meaning of article 4, and that for two reasons. First he said that there was no sufficient evidence of any instruction. In my opinion there is no substance in this ground. There is ample evidence in a letter dated 7 February 1984 from Sir Robert Armstrong to the Director of GCHQ and also in the General Notice 100/84 and a covering letter issued by the Director to all employees at GCHQ. Secondly counsel said that the instruction did not sufficiently specify conditions that were being altered, but I agree with Glidewell J., and with the Court of Appeal, that the Minister's direction on 22 December 1983 did give “instructions … providing for … the conditions of service” of employees at GCHQ in the sense of article 4 of the Order in Council of 1982. There was no obligation to put the instructions in writing, although that might perhaps have been expected in a matter so important as this. Nor was there any obligation to couch the instructions in any particular form. Accordingly I reject this submission.
For these reasons I would dismiss the appeal.
LORD SCARMAN. My Lords, I would dismiss this appeal for one reason only. I am satisfied that the respondent has made out a case on the ground of national security. Notwithstanding the criticisms which can be made of the evidence and despite the fact that the point was not raised, or, if it was, was not clearly made before the case reached the Court of Appeal, I have no doubt that the respondent refused to consult the unions before issuing her instruction of the 22 December 1983 because she feared that, if she did, union-organised disruption of the monitoring services of GCHQ could well result. I am further satisfied that the fear was one which a reasonable minister in the circumstances in which she found herself could reasonably entertain. I am also satisfied that a reasonable minister could reasonably consider such disruption to constitute a threat to national security. I would, therefore, deny relief to the appellants upon their application for judicial review of the instruction, the effect of which was that staff at GCHQ would no longer be permitted to belong to a national trade union.
The point of principle in the appeal is as to the duty of the court when in proceedings properly brought before it a question arises as to what is required in the interest of national security. The question may arise in ordinary litigation between private persons as to their private rights and obligations: and it can arise, as in this case, in proceedings for judicial review of a decision by a public authority. The question can take one of several forms. It may be a question of fact which Parliament has left to the court to determine: see for an example section 10 of the Contempt of Court Act 1981. It may arise for consideration as a factor in the exercise of an executive discretionary power. But, however it arises, it is a matter to be considered by the court in the circumstances and context of the case. Though there are limits dictated by law and common sense which the court must observe in dealing with the question, the court does not abdicate its judicial function. If the question arises as a matter of fact, the court requires evidence to be given. If it arises as a factor to be considered in reviewing the exercise of a discretionary power, evidence is also needed so that the court may determine whether it should intervene to correct excess or abuse of the power.
Let me give three illustrations taken from the case law of the 20th century. First, The Zamora  2 AC 77
— surely one of the more courageous of judicial decisions even in our long history. In April 1916 a question of national security came before the Judicial Committee of the Privy Council sitting in Prize. The Crown's role in the Prize Court was that of a belligerent power having by international law the right to requisition vessels or goods in the custody of its Prize Court. A neutral vessel carrying a cargo of copper (contraband) had been stopped at sea by the Royal Navy and taken to a British port. No decree of condemnation of the cargo had yet been made by the Prize Court, when the Crown intervened by summons to requisition the cargo then in the custody of the court. Lord Parker of Waddington, who delivered the judgment of the Judicial Committee, concluded, at p. 106:
“A belligerent power has by international law the right to requisition vessels or goods in the custody of its Prize Court pending a decision of the question whether they should be condemned or released, but such right is subject to certain limitations. First, the vessel or goods in question must be urgently required for use in connection with the defence of the realm, the prosecution of the war, or other matters involving national security. Secondly, there must be a real question to be tried, so that it would be improper to order an immediate release. And, thirdly, the right must be enforced by application to the Prize Court, which must determine judicially whether, under the particular circumstances of the case, the right is exercisable.”
Discussing the first limitation, Lord Parker of Waddington observed that the judge ought, “ as a rule,” to treat the statement of the proper officer of the Crown that the vessel or goods were urgently required for national security reasons as conclusive of the fact. And it was in this context that he delivered his famous dictum, at p. 107: “Those who are responsible for the national security must be the sole judges of what the national security requires.” These words were no abdication of the judicial function, but were an indication of the evidence required by the court. In fact the evidence adduced by the Crown was not sufficient, and the court ruled that the Crown had no right to requisition. The Crown's claim was rejected “because the judge had before him no satisfactory evidence that such a right was exercisable” (p. 108). The Prize Court, therefore, treated the question as one of fact for its determination and indicated the evidence needed to establish the fact. The true significance of Lord Parker's dictum is simply that the court is in no position to substitute its opinion for the opinion of those responsible for national security. But the case is a fine illustration of the court's duty to ensure that the essential facts to which the opinion or judgment of those responsible relates are proved to the satisfaction of the court.
My second illustration is Chandler v. Director of Public Prosecutions  AC 763
. In this case the interest of national security came into court as a matter of fact to be established by evidence to the satisfaction of a jury in a criminal case. The appellants were convicted of conspiring to commit a breach of section 1 of the Official Secrets Act 1911, “namely, for a purpose prejudicial to the safety or interests of the state to enter a Royal Air Force station … at Wethersfield.” There was evidence from an officer of air rank that the airfield was of importance for national security: and, as my noble and learned friend Lord Fraser of Tullybelton has pointed out, Lord Reid and Viscount Radcliffe treated his evidence as relevant to the dismissal of the appeal. Lord Devlin developed the point taken in the case on national security in a passage beginning at p. 809 which, with all respect to those who take a different view, I believe to be sound law. Having referred to the undoubted principle that all matters relating to the disposition and armament of the armed forces are left to the unfettered control of the Crown, he made three comments. First, he put the Zamora dictum into its true context. Secondly, he observed that, when a court is faced with the exercise of a discretionary power, inquiry is not altogether excluded: the court will intervene to correct excess or abuse. His third and, as he said, his “most significant” comment was as to the nature and effect of the principle. “Where it operates, it limits the issue which the court has to determine; it does not exclude any evidence or argument relevant to the issue” (p. 810).
As I read the speeches in Chandler's case, the House accepted that the statute required the prosecution to establish by evidence that the conspiracy was to enter a prohibited place for a purpose prejudicial to the safety or interests of the state. As Parliament had left the existence of a prejudicial purpose to the decision of a jury, it was not the Crown's opinion as to the existence of prejudice to the safety or interests of the state but the jury's which mattered: hence, as Lord Devlin, at p. 811, remarked, the Crown's opinion on that was inadmissible but the Crown's evidence as to its interests was an “entirely different matter.” Here, like Lord Parker in the Zamora, Lord Devlin was accepting that the Crown, or its responsible servants, are the best judges of what national security requires without excluding the judicial function of determining whether the interest of national security has been shown to be involved in the case.
Finally, I would refer to Secretary of State for Defence v. Guardian Newspapers Ltd.  A.C. 339, a case arising under section 10 of the Act of 1981. As in Chandler's case, the interest of national security had to be considered in proceedings where it arose as a question of fact to be established to the satisfaction of a court. Though the House was divided as to the effect of the evidence, all their Lordships held that evidence was necessary so that the court could be judicially satisfied that the interest of national security required disclosure of the newspaper's source of information.
My Lords, I conclude, therefore, that where a question as to the interest of national security arises in judicial proceedings the court has to act on evidence. In some cases a judge or jury is required by law to be satisfied that the interest is proved to exist: in others, the interest is a factor to be considered in the review of the exercise of an executive discretionary power. Once the factual basis is established by evidence so that the court is satisfied that the interest of national security is a relevant factor to be considered in the determination of the case, the court will accept the opinion of the Crown or its responsible officer as to what is required to meet it, unless it is possible to show that the opinion was one which no reasonable minister advising the Crown could in the circumstances reasonably have held. There is no abdication of the judicial function, but there is a common sense limitation recognised by the judges as to what is justiciable: and the limitation is entirely consistent with the general development of the modern case law of judicial review.
My Lords, I would wish to add a few, very few, words on the reviewability of the exercise of the royal prerogative. Like my noble and learned friend Lord Diplock, I believe that the law relating to judicial review has now reached the stage where it can be said with confidence that, if the subject matter in respect of which prerogative power is exercised is justiciable, that is to say if it is a matter upon which the court can adjudicate, the exercise of the power is subject to review in accordance with the principles developed in respect of the review of the exercise of statutory power. Without usurping the role of legal historian, for which I claim no special qualification, I would observe that the royal prerogative has always been regarded as part of the common law, and that Sir Edward Coke had no doubt that it was subject to the common law: Prohibitions del Roy
(1608) 12 Co. Rep. 63 and the Proclamations Case(1611) 12 Co Rep 74
. In the latter case he declared, at p. 76, that “the King hath no prerogative, but that which the law of the land allows him.” It is, of course, beyond doubt that in Coke's time and thereafter judicial review of the exercise of prerogative power was limited to inquiring into whether a particular power existed and, if it did, into its extent: Attorney-General v. De Keyser's Royal Hotel Ltd.  AC 508
. But this limitation has now gone, overwhelmed by the developing modern law of judicial review: Reg. v. Criminal Injuries Compensation Board, Ex parte Lain  2 Q.B. 864 (a landmark case comparable in its generation with the Proclamations Case
, 12 Co Rep. 74) and Reg. v. Secretary of State for Home Affairs, Ex parte Hosenball  1 W.L.R. 766. Just as ancient restrictions in the law relating to the prerogative writs and orders have not prevented the courts from extending the requirement of natural justice, namely the duty to act fairly, so that it is required of a purely administrative act, so also has the modern law, a vivid sketch of which my noble and learned friend Lord Diplock has included in his speech, extended the range of judicial review in respect of the exercise of prerogative power. Today, therefore, the controlling factor in determining whether the exercise of prerogative power is subject to judicial review is not its source but its subject matter.
Subject to these few comments, I agree with the speeches delivered by my noble and learned friends Lord Diplock and Lord Roskill. I am in favour of dismissing the appeal only because the respondent has established by evidence that the interest of national security required in her judgment that she should refuse to consult the unions before issuing her instruction. But for this I would have allowed the appeal on the procedural ground that the respondent had acted unfairly in failing to consult unions or staff before making her decision.
LORD DIPLOCK. My Lords, the English law relating to judicial control of administrative action has been developed upon a case to case basis which has virtually transformed it over the last three decades. The principles of public law that are applicable to the instant case are in my view well established by authorities that are sufficiently cited in the speech that will be delivered by my noble and learned friend, Lord Roskill. This obviates the necessity of my duplicating his citations: though I should put on record that after reading and rereading Lord Devlin's speech in Chandler v. Director of Public Prosecutions  AC 763
, I have gained no help from it, for I find some of his observations that are peripheral to what I understand to be the ratio decidendi difficult to reconcile with the actual decision that he felt able to reach and also with one another.
The only difficulty which the instant case has presented upon the facts as they have been summarised by my noble and learned friend, Lord Fraser of Tullybelton, and expanded in the judgment of Glidewell J. has been to identify what is, in my view, the one crucial point of law on which this appeal turns. It never was identified or even adumbrated in the respondent's argument at the hearing before Glidewell J. and so excusably, finds no place in what otherwise I regard as an impeccable judgment. The consequence of this omission was that he found in favour of the applicants. Before the Court of Appeal the crucial point was advanced in argument by the Crown in terms that were unnecessarily and, in my view, unjustifiably wide. This stance was maintained in the appeal to this House, although, under your Lordships' encouragement, the narrower point of law that was really crucial was developed and relied on by the respondent in the alternative. Once that point has been accurately identified the evidence in the case in my view makes it inevitable that this appeal must be dismissed. I will attempt to state in summary form those principles of public law which lead me to this conclusion.
Judicial review, now regulated by R.S.C., Ord. 53, provides the means by which judicial control of administrative action is exercised. The subject matter of every judicial review is a decision made by some person (or body of persons) whom I will call the “decision-maker” or else a refusal by him to make a decision.
To qualify as a subject for judicial review the decision must have consequences which affect some person (or body of persons) other than the decision-maker, although it may affect him too. It must affect such other person either:
(a) by altering rights or obligations of that person which are enforceable by or against him in private law; or
(b) by depriving him of some benefit or advantage which either (i) he had in the past been permitted by the decision-maker to enjoy and which he can legitimately expect to be permitted to continue to do until there has been communicated to him some rational grounds for withdrawing it on which he has been given an opportunity to comment; or (ii) he has received assurance from the decision-maker will not be withdrawn without giving him first an opportunity of advancing reasons for contending that they should not be withdrawn. (I prefer to continue to call the kind of expectation that qualifies a decision for inclusion in class (b) a “legitimate expectation” rather than a “reasonable expectation,” in order thereby to indicate that it has consequences to which effect will be given in public law, whereas an expectation or hope that some benefit or advantage would continue to be enjoyed, although it might well be entertained by a “reasonable” man, would not necessarily have such consequences. The recent decision of this House in In re Findlay  A.C. 318 presents an example of the latter kind of expectation. “Reasonable” furthermore bears different meanings according to whether the context in which it is being used is that of private law or of public law. To eliminate confusion it is best avoided in the latter.)
For a decision to be susceptible to judicial review the decision-maker must be empowered by public law (and not merely, as in arbitration, by agreement between private parties) to make decisions that, if validly made, will lead to administrative action or abstention from action by an authority endowed by law with executive powers, which have one or other of the consequences mentioned in the preceding paragraph. The ultimate source of the decision-making power is nearly always nowadays a statute or subordinate legislation made under the statute; but in the absence of any statute regulating the subject matter of the decision the source of the decision-making power may still be the common law itself, i.e., that part of the common law that is given by lawyers the label of “the prerogative.” Where this is the source of decision-making power, the power is confined to executive officers of central as distinct from local government and in constitutional practice is generally exercised by those holding ministerial rank.
It was the prerogative that was relied on as the source of the power of the Minister for the Civil Service in reaching her decision of 22 December 1983 that membership of national trade unions should in future be barred to all members of the home civil service employed at GCHQ.
My Lords, I intend no discourtesy to counsel when I say that, intellectual interest apart, in answering the question of law raised in this appeal, I have derived little practical assistance from learned and esoteric analyses of the precise legal nature, boundaries and historical origin of “the prerogative,” or of what powers exercisable by executive officers acting on behalf of central government that are not shared by private citizens qualify for inclusion under this particular label. It does not, for instance, seem to me to matter whether today the right of the executive government that happens to be in power to dismiss without notice any member of the home civil service upon which perforce it must rely for the administration of its policies, and the correlative disability of the executive government that is in power to agree with a civil servant that his service should be on terms that did not make him subject to instant dismissal, should be ascribed to “the prerogative” or merely to a consequence of the survival, for entirely different reasons, of a rule of constitutional law whose origin is to be found in the theory that those by whom the administration of the realm is carried on do so as personal servants of the monarch who can dismiss them at will, because the King can do no wrong.
Nevertheless, whatever label may be attached to them there have unquestionably survived into the present day a residue of miscellaneous fields of law in which the executive government retains decision-making powers that are not dependent upon any statutory authority but nevertheless have consequences on the private rights or legitimate expectations of other persons which would render the decision subject to judicial review if the power of the decision-maker to make them were statutory in origin. From matters so relatively minor as the grant of pardons to condemned criminals, of honours to the good and great, of corporate personality to deserving bodies of persons, and of bounty from moneys made available to the executive government by Parliament, they extend to matters so vital to the survival and welfare of the nation as the conduct of relations with foreign states and — what lies at the heart of the present case — the defence of the realm against potential enemies. Adopting the phraseology used in the European Convention on Human Rights 1953 (Convention for the Protection of Human Rights and Fundamental Freedoms (1953) (Cmd. 8969)) to which the United Kingdom is a party it has now become usual in statutes to refer to the latter as “national security.”
My Lords, I see no reason why simply because a decision-making power is derived from a common law and not a statutory source, it should for that reason only be immune from judicial review. Judicial review has I think developed to a stage today when without reiterating any analysis of the steps by which the development has come about, one can conveniently classify under three heads the grounds upon which administrative action is subject to control by judicial review. The first ground I would call “illegality,” the second “irrationality” and the third “procedural impropriety.” That is not to say that further development on a case by case basis may not in course of time add further grounds. I have in mind particularly the possible adoption in the future of the principle of “proportionality” which is recognised in the administrative law of several of our fellow members of the European Economic Community; but to dispose of the instant case the three already well-established heads that I have mentioned will suffice.
By “illegality” as a ground for judicial review I mean that the decision-maker must understand correctly the law that regulates his decision-making power and must give effect to it. Whether he has or not is par excellence a justiciable question to be decided, in the event of dispute, by those persons, the judges, by whom the judicial power of the state is exercisable.
By “irrationality” I mean what can by now be succinctly referred to as “Wednesbury unreasonableness” (Associated Provincial Picture Houses Ltd. v. Wednesbury Corporation  1 KB 223
). It applies to a decision which is so outrageous in its defiance of logic or of accepted moral standards that no sensible person who had applied his mind to the question to be decided could have arrived at it. Whether a decision falls within this category is a question that judges by their training and experience should be well equipped to answer, or else there would be something badly wrong with our judicial system. To justify the court's exercise of this role, resort I think is today no longer needed to Viscount Radcliffe's ingenious explanation in Edwards v. Bairstow  AC 14
of irrationality as a ground for a court's reversal of a decision by ascribing it to an inferred though unidentifiable mistake of law by the decision-maker. “Irrationality” by now can stand upon its own feet as an accepted ground on which a decision may be attacked by judicial review.
I have described the third head as “procedural impropriety” rather than failure to observe basic rules of natural justice or failure to act with procedural fairness towards the person who will be affected by the decision. This is because susceptibility to judicial review under this head covers also failure by an administrative tribunal to observe procedural rules that are expressly laid down in the legislative instrument by which its jurisdiction is conferred, even where such failure does not involve any denial of natural justice. But the instant case is not concerned with the proceedings of an administrative tribunal at all.
My Lords, that a decision of which the ultimate source of power to make it is not a statute but the common law (whether or not the common law is for this purpose given the label of “the prerogative”) may be the subject of judicial review on the ground of illegality is, I think, established by the cases cited by my noble and learned friend, Lord Roskill, and this extends to cases where the field of law to which the decision relates is national security, as the decision of this House itself in Burmah Oil Co. Ltd. v. Lord Advocate, 1964 SC (HL) 117
shows. While I see no a priori reason to rule out “irrationality” as a ground for judicial review of a ministerial decision taken in the exercise of “prerogative” powers, I find it difficult to envisage in any of the various fields in which the prerogative remains the only source of the relevant decision-making power a decision of a kind that would be open to attack through the judicial process upon this ground. Such decisions will generally involve the application of government policy. The reasons for the decision-maker taking one course rather than another do not normally involve questions to which, if disputed, the judicial process is adapted to provide the right answer, by which I mean that the kind of evidence that is admissible under judicial procedures and the way in which it has to be adduced tend to exclude from the attention of the court competing policy considerations which, if the executive discretion is to be wisely exercised, need to be weighed against one another — a balancing exercise which judges by their upbringing and experience are ill-qualified to perform. So I leave this as an open question to be dealt with on a case to case basis if, indeed, the case should ever arise.
As respects “procedural propriety” I see no reason why it should not be a ground for judicial review of a decision made under powers of which the ultimate source is the prerogative. Such indeed was one of the grounds that formed the subject matter of judicial review in Reg. v. Criminal Injuries Compensation Board, Ex parte Lain  2 Q.B. 864. Indeed, where the decision is one which does not alter rights or obligations enforceable in private law but only deprives a person of legitimate expectations, “procedural impropriety” will normally provide the only ground on which the decision is open to judicial review. But in any event what procedure will satisfy the public law requirement of procedural propriety depends upon the subject matter of the decision, the executive functions of the decision-maker (if the decision is not that of an administrative tribunal) and the particular circumstances in which the decision came to be made.
My Lords, in the instant case the immediate subject matter of the decision was a change in one of the terms of employment of civil servants employed at GCHQ. That the executive functions of the Minister for the Civil Service, in her capacity as such, included making a decision to change any of those terms, except in so far as they related to remuneration, expenses and allowances, is not disputed. It does not seem to me to be of any practical significance whether or not as a matter of strict legal analysis this power is based upon the rule of constitutional law to which I have already alluded that the employment of any civil servant may be terminated at any time without notice and that upon such termination the same civil servant may be re-engaged on different terms. The rule of terminability of employment in the civil service without notice, of which the existence is beyond doubt, must in any event have the consequence that the continued enjoyment by a civil servant in the future of a right under a particular term of his employment cannot be the subject of any right enforceable by him in private law; at most it can only be a legitimate expectation.
Prima facie, therefore, civil servants employed at GCHQ who were members of national trade unions had, at best, in December 1983, a legitimate expectation that they would continue to enjoy the benefits of such membership and of representation by those trade unions in any consultations and negotiations with representatives of the management of that government department as to changes in any term of their employment. So, but again prima facie only, they were entitled, as a matter of public law under the head of “procedural propriety,” before administrative action was taken on a decision to withdraw that benefit, to have communicated to the national trade unions by which they had theretofore been represented the reason for such withdrawal, and for such unions to be given an opportunity to comment on it.
The reason why the Minister for the Civil Service decided on 22 December 1983 to withdraw this benefit was in the interests of national security. National security is the responsibility of the executive government; what action is needed to protect its interests is, as the cases cited by my learned friend, Lord Roskill, establish and common sense itself dictates, a matter upon which those upon whom the responsibility rests, and not the courts of justice, must have the last word. It is par excellence a non-justiciable question. The judicial process is totally inept to deal with the sort of problems which it involves.
The executive government likewise decided, and this would appear to be a collective decision of cabinet ministers involved, that the interests of national security required that no notice should be given of the decision before administrative action had been taken to give effect to it. The reason for this was the risk that advance notice to the national unions of the executive government's intention would attract the very disruptive action prejudicial to the national security the recurrence of which the decision barring membership of national trade unions to civil servants employed at GCHQ was designed to prevent.
There was ample evidence to which reference is made by others of your Lordships that this was indeed a real risk; so the crucial point of law in this case is whether procedural propriety must give way to national security when there is conflict between (1) on the one hand, the prima facie rule of “procedural propriety” in public law, applicable to a case of legitimate expectations that a benefit ought not to be withdrawn until the reason for its proposed withdrawal has been communicated to the person who has theretofore enjoyed that benefit and that person has been given an opportunity to comment on the reason, and (2) on the other hand, action that is needed to be taken in the interests of national security, for which the executive government bears the responsibility and alone has access to sources of information that qualify it to judge what the necessary action is. To that there can, in my opinion, be only one sensible answer. That answer is “Yes.”
I agree with your Lordships that this appeal must be dismissed.
LORD ROSKILL. My Lords, this appeal arises out of the exercise by the respondent, the Minister for the Civil Service, of a specific power vested in her by article 4 of the Civil Service Order in Council 1982. That specific power purported to be exercised orally on 22 December 1983. The terms in which it is claimed to have been exercised are contained in a letter dated 7 February 1984 from Sir Robert Armstrong writing as Head of the Civil Service to the Director of the Government Communications Headquarters at Cheltenham (“GCHQ”). The exercise of the power took the form of:
“instructions that the conditions of service under which civil servants are employed as members of the staff of the Government Communications Headquarters shall be varied so as to provide that such civil servants shall not be members of any trade union other than a departmental staff association approved by yourself.”
The making of this change in the conditions of service of civil servants employed at GCHQ was announced in the House of Commons by the Secretary of State for Foreign and Commonwealth Affairs on 25 January 1984 and on the same day he issued certificates under section 121(4) of the Employment Protection Act 1975 and under section 138(4) of the Employment Protection (Consolidation) Act 1978 certifying that employment at GCHQ was to be excepted from those sections “for the purpose of safeguarding national security.” On the same day the Director of GCHQ informed his staff in writing of the decision, of the issue of the certificates and of the various options which were thereafter to remain open to them.
My Lords, the background to these actions in December 1983 and January 1984 is fully set out in the speech of my noble and learned friend, Lord Fraser of Tullybelton, which I gratefully adopt. It requires no repetition. Nor does the history of the antecedent rights of those concerned to join trade unions. That the instructions thus given and the certificates thus issued drastically altered the trade union rights of those civil servants concerned cannot be doubted. Nor can it be doubted that the issue of the instructions and of the certificates without prior warning or consultation of any kind with the various trade unions concerned either at a national or at a local level involved a complete departure from the normal manner in which relations between management and staff had hitherto been conducted and was bitterly resented by some of those immediately involved on the staff side.
My Lords, with matters of that kind your Lordships are in no way concerned. This appeal is concerned with and only with judicial review. Judicial review, as my noble and learned friend Lord Brightman stated in Chief Constable of the North Wales Police v. Evans  1 WLR 1155
, 1174, “is not an appeal from a decision, but a review of the manner in which the decision was made.” It is the appellants' case, stated in a sentence, that the oral instruction of 22 December 1983 should be judicially reviewed and declared invalid because of the manner in which the decision which led to those instructions being given was taken, that is to say without prior consultation of any kind with the appellants or indeed others. Initially the respondents also sought judicial review of the two certificates to which I have referred but that claim has been abandoned.
Before considering the rival submissions in more detail, it will be convenient to make some general observations about the process now known as judicial review. Today it is perhaps commonplace to observe that as a result of a series of judicial decisions since about 1950 both in this House and in the Court of Appeal there has been a dramatic and indeed a radical change in the scope of judicial review. That change has been described — by no means critically — as an upsurge of judicial activism. Historically the use of the old prerogative writs of certiorari, prohibition and mandamus was designed to establish control by the Court of King's Bench over inferior courts or tribunals. But the use of those writs, and of their successors the corresponding prerogative orders, has become far more extensive. They have come to be used for the purpose of controlling what would otherwise be unfettered executive action whether of central or local government. Your Lordships are not concerned in this case with that branch of judicial review which is concerned with the control of inferior courts or tribunals. But your Lordships are vitally concerned with that branch of judicial review which is concerned with the control of executive action. This branch of public or administrative law has evolved, as with much of our law, on a case by case basis and no doubt hereafter that process will continue. Thus far this evolution has established that executive action will be the subject of judicial review on three separate grounds. The first is where the authority concerned has been guilty of an error of law in its action as for example purporting to exercise a power which in law it does not possess. The second is where it exercises a power in so unreasonable a manner that the exercise becomes open to review upon what are called, in lawyers' shorthand, Wednesbury principles (Associated Provincial Picture Houses Ltd. v. Wednesbury Corporation  1 KB 223
). The third is where it has acted contrary to what are often called “principles of natural justice.” As to this last, the use of this phrase is no doubt hallowed by time and much judicial repetition, but it is a phrase often widely misunderstood and therefore as often misused. That phrase perhaps might now be allowed to find a permanent resting-place and be better replaced by speaking of a duty to act fairly. But that latter phrase must not in its turn be misunderstood or misused. It is not for the courts to determine whether a particular policy or particular decisions taken in fulfilment of that policy are fair. They are only concerned with the manner in which those decisions have been taken and the extent of the duty to act fairly will vary greatly from case to case as indeed the decided cases since 1950 consistently show. Many features will come into play including the nature of the decision and the relationship of those involved on either side before the decision was taken.
My noble and learned friend, Lord Diplock, in his speech has devised a new nomenclature for each of these three grounds, calling them respectively “illegality,” “irrationality” and “procedural impropriety” — words which, if I may respectfully say so, have the great advantage of making clear the differences between each ground.
In the present appeal your Lordships are not concerned with the first two matters already mentioned, with the exercise of a power which does not exist or with Wednesbury principles. But this appeal is vitally concerned with the third, the duty to act fairly.
The particular manifestation of the duty to act fairly which is presently involved is that part of the recent evolution of our administrative law which may enable an aggrieved party to evoke judicial review if he can show that he had “a reasonable expectation” of some occurrence or action preceding the decision complained of and that that “reasonable expectation” was not in the event fulfilled.
The introduction of the phrase “reasonable expectation” into this branch of our administrative law appears to owe its origin to Lord Denning M.R. in Schmidt v. Secretary of State for Home Affairs  2 Ch 149
, 170 (when he used the phrase “legitimate expectation”). Its judicial evolution is traced in the opinion of the Judicial Committee delivered by my noble and learned friend, Lord Fraser of Tullybelton, in Attorney-General of Hong Kong v. Ng Yuen Shiu  2 AC 629
, 636–638. Though the two phrases can, I think, now safely be treated as synonymous for the reasons there given by my noble and learned friend, I prefer the use of the adjective “legitimate” in this context and use it in this speech even though in argument it was the adjective “reasonable” which was generally used. The principle may now be said to be firmly entrenched in this branch of the law. As the cases show, the principle is closely connected with “a right to be heard.” Such an expectation may take many forms. One may be an expectation of prior consultation. Another may be an expectation of being allowed time to make representations especially where the aggrieved party is seeking to persuade an authority to depart from a lawfully established policy adopted in connection with the exercise of a particular power because of some suggested exceptional reasons justifying such a departure.
The appellants say that the relationship between management and staff over many years gave rise to a legitimate expectation of consultation before action involving so drastic a curtailment of trade union rights as that taken on 22 December 1983 was decreed. It is of the deprivation of that legitimate expectation that they now principally complain. They say, that deprivation entitles them to judicial review.
In a judgment which, if I may respectfully say so, I have read and reread with increasing admiration for its thoroughness and clarity, Glidewell J., while in my view correctly rejecting all the other arguments of the appellants, accepted this submission. The Court of Appeal (Lord Lane C.J., Watkins and May L.JJ.) in a single judgment delivered by the Lord Chief Justice was of a different opinion. But it is right to say that the submission on which Mr. Alexander Q.C. for the respondent finally and principally rested was never advanced at all before Glidewell J. and though advanced for the first time in the Court of Appeal does not seem to have been advanced even there in entirely the same way as in argument before this House for it was advanced there on a considerably wider basis than that upon which Mr. Alexander ultimately came to rest. Mr. Blom-Cooper Q.C. for the appellants understandably made skilful forensic play with this failure to advance this crucial submission before the learned judge. Thus the House has not got the benefit of the views of Glidewell J. upon what I regard as the crucial issue for the determination of this appeal.
My Lords, before considering this issue it is necessary to consider a further important question which arises by reason of the fact that the instructions given under article 4 of the Order in Council of 1982 were by means of the exercise of a prerogative power. The appellants in their printed case invited the House to consider and if necessary to reconsider the reviewability of executive acts done under the prerogative. Mr. Alexander for the respondent understandably did not press the argument that no action taken under the prerogative could ever be the subject of judicial review. But, helpfully, he thought it right to make available to your Lordships a selection from the classic pronouncements of many famous writers in this field from Locke through Blackstone and Chitty to Dicey and from the writings of distinguished modern authorities including de Smith, Wade, Hood Phillips and Heuston designed to show first the historic view that acts done under the prerogative were never reviewable and secondly the extent to which that classic doctrine may at least in this century be said to have been diluted.
Dicey's classic statement in Law of the Constitution
, 10th ed. (1959), p. 424, that the prerogative is “the residue of discretionary or arbitrary authority, which at any given time is legally left in the hands of the Crown” has the weight behind it not only of the author's own authority but also of the majority of this House in Burmah Oil Co. Ltd. v. Lord Advocate  AC 75
: see per
Lord Reid, at p. 99. But as Lord Reid himself pointed out this definition “does not take us very far.” On the other hand the attempt by Lord Denning M.R. in Laker Airways Ltd. v. Department of Trade  QB 643
, 705 (obiter since the other members of the Court of Appeal did not take so broad a view) to assert that the prerogative “if … exercised improperly or mistakenly” was reviewable is, with great respect, far too wide. The Master of the Rolls sought to support his view by a quotation from Blackstone's Commentaries
, 15th ed., vol. 1, p. 252. But unfortunately and no doubt inadvertently he omitted the opening words of the paragraph:
“In the exercise therefore of those prerogatives, which the law has given him, the King is irresistible and absolute, according to the forms of the constitution. And yet, if the consequence of that exertion be manifestly to the grievance or dishonour of the kingdom, the parliament will call his advisers to a just and severe account.”
In short the orthodox view was at that time that the remedy for abuse of the prerogative lay in the political and not in the judicial field.
But fascinating as it is to explore this mainstream of our legal history, to do so in connection with the present appeal has an air of unreality. To speak today of the acts of the sovereign as “irresistible and absolute” when modern constitutional convention requires that all such acts are done by the sovereign on the advice of and will be carried out by the sovereign's ministers currently in power is surely to hamper the continual development of our administrative law by harking back to what Lord Atkin once called, albeit in a different context, the clanking of mediaeval chains of the ghosts of the past: see United Australia Ltd. v. Barclays Bank Ltd.  A.C. 1, 29. It is, I hope, not out of place in this connection to quote a letter written in 1896 by the great legal historian F. W. Maitland to Dicey himself: “The only direct utility of legal history (I say nothing of its thrilling interest) lies in the lesson that each generation has an enormous power of shaping its own law”: see Richard A. Cosgrove, The Rule of Law: Albert Venn Dicey, Victorian Jurist (1980), p. 177. Maitland was in so stating a greater prophet than even he could have foreseen for it is our legal history which has enabled the present generation to shape the development of our administrative law by building upon but unhampered by our legal history.
My Lords, the right of the executive to do a lawful act affecting the rights of the citizen, whether adversely or beneficially, is founded upon the giving to the executive of a power enabling it to do that act. The giving of such a power usually carries with it legal sanctions to enable that power if necessary to be enforced by the courts. In most cases that power is derived from statute though in some cases, as indeed in the present case, it may still be derived from the prerogative. In yet other cases, as the decisions show, the two powers may coexist or the statutory power may by necessary implication have replaced the former prerogative power. If the executive in pursuance of the statutory power does an act affecting the rights of the citizen, it is beyond question that in principle the manner of the exercise of that power may today be challenged on one or more of the three grounds which I have mentioned earlier in this speech. If the executive instead of acting under a statutory power acts under a prerogative power and in particular a prerogative power delegated to the respondent under article 4 of the Order in Council of 1982,so as to affect the rights of the citizen, I am unable to see, subject to what I shall say later, that there is any logical reason why the fact that the source of the power is the prerogative and not statute should today deprive the citizen of that right of challenge to the manner of its exercise which he would possess were the source of the power statutory. In either case the act in question is the act of the executive. To talk of that act as the act of the sovereign savours of the archaism of past centuries. In reaching this conclusion I find myself in agreement with my noble and learned friends Lord Scarman and Lord Diplock whose speeches I have had the advantage of reading in draft since completing the preparation of this speech.
But I do not think that that right of challenge can be unqualified. It must, I think, depend upon the subject matter of the prerogative power which is exercised. Many examples were given during the argument of prerogative powers which as at present advised I do not think could properly be made the subject of judicial review. Prerogative powers such as those relating to the making of treaties, the defence of the realm, the prerogative of mercy, the grant of honours, the dissolution of Parliament and the appointment of ministers as well as others are not, I think susceptible to judicial review because their nature and subject matter are such as not to be amenable to the judicial process. The courts are not the place wherein to determine whether a treaty should be concluded or the armed forces disposed in a particular manner or Parliament dissolved on one date rather than another.
In my view the exercise of the prerogative which enabled the oral instructions of 22 December 1983 to be given does not by reason of its subject matter fall within what for want of a better phrase I would call the “excluded categories” some of which I have just mentioned. It follows that in principle I can see no reason why those instructions should not be the subject of judicial review.
My Lords, I am not conscious of any previous decision of this House which is inconsistent with the principles I have just endeavoured to state. It may well be that there are decisions or dicta of other courts which are inconsistent. Reg. v. Secretary of State for War  2 Q.B. 326 arose in connection with the armed forces with which this appeal is not concerned, but even so, some of the reasoning cannot I think now be supported. There are also passages in the judgments of the Court of Appeal in Commissioners of Crown Lands v. Page  2 Q.B. 274 and in the opinion of Lord Sorn in Griffin v. Lord Advocate, 1950 S.C. 448 (to mention but two decisions) which require reconsideration in the light of the decision of this House in this appeal: in the latter case, Lord Sorn mainly followed the first of these three cases.
I find considerable support for the conclusion I have reached in the decision of the Divisional Court (Lord Parker C.J., Diplock L.J. (as my noble and learned friend then was) and Ashworth J. in Reg. v. Criminal Injuries Compensation Board, Ex parte Lain  2 Q.B. 864, the judgments in which may without exaggeration be described as a landmark in the development of this branch of the law. The board had been set up not by statute but by executive action under, as I think and as Lord Parker C.J. stated, the prerogative. It was strenuously argued that the board was not subject to the jurisdiction of the courts since it did not have what was described as legal authority in the sense of statutory authority. This argument by Mr. Nigel Bridge, as he then was, was emphatically and unanimously rejected. I will quote one passage from the judgment of Lord Parker C.J., at p. 881:
“I can see no reason either in principle or in authority why a board set up as this board was set up is not a body of persons amenable to the jurisdiction of this court. True it is not set up by statute but the fact that it is set up by executive government, i.e., under the prerogative, does not render its acts any the less lawful. Indeed, the writ of certiorari has issued not only to courts set up by statute but to courts whose authority is derived, inter alia, from the prerogative. Once the jurisdiction is extended, as it clearly has been, to tribunals as opposed to courts, there is no reason why the remedy by way of certiorari cannot be invoked to a body of persons set up under the prerogative. Moreover the board though set up under the prerogative and not by statute had in fact the recognition of Parliament in debate and Parliament provided the money to satisfy its awards.”
I would also refer, albeit without citation, to the entirety of the judgment delivered by my noble and learned friend, Lord Diplock.
It follows from what I have said thus far that in principle I am of the clear opinion that the respondent's oral instructions of 22 December 1983 are amenable to judicial review and are not immune from such review because the instructions were given pursuant to prerogative powers.
The next question is whether they are susceptible of successful challenge on the third of the grounds mentioned earlier, namely that the appellants had “a legitimate expectation” of consultation prior to any such instructions being given which radically affected the long-established rights of the staff at GCHQ to be members of trade unions.
It was common ground before your Lordships, though it was not common ground below, that there was no contractual relationship between the Crown and the staff at GCHQ. Mr. Alexander accepted that the absence of a contractual relationship and thus of a remedy in private law did not preclude the possibility of a remedy in public law if a legitimate expectation of consultation were established. But he suggested that the absence of such a relationship in private law made it difficult to establish a legitimate expectation justiciable in the field of public law without eroding the basic principle that, at least in theory, civil servants are dismissible by the Crown at will and thus have no remedy in private law. He further argued that even if in principle there were a legitimate expectation of the nature for which the appellants contended, that legitimate expectation could not exist when the government of the day considered that their duty in the field of national security required them not to give effect to any such legitimate expectation as might otherwise exist. Once, he contended, the respondent on the material before her could conclude that consultations of the kind contended for by the appellants could and indeed would damage national security, any obligation to consult the appellants prior to taking the decision disappeared. Indeed Mr. Alexander went so far as to contend that in such circumstances the respondent was under a duty not to consult the appellants lest otherwise the very mischief which he feared might arise would arise.
My Lords, if no question of national security were involved I cannot doubt that the evidence and the whole history of the relationship between management and staff since 1919 shows that there was a legitimate expectation of consultation before important alterations in the conditions of service of civil servants were made. No doubt in strict theory civil servants are dismissible at will and the various documents shown to your Lordships seek to preserve the strict constitutional position. But in reality the management-staff relationship is governed by an elaborate code to which it is unnecessary to refer in detail. I have little doubt that were management to seek to alter without prior consultation the terms and conditions of civil servants in a field which had no connection whatever with national security or perhaps, though the matter does not arise in this appeal, with urgent fiscal emergency, such action would in principle be amenable to judicial review.
But that is not the present issue. It is asserted on behalf of the respondent that the reason for the instructions being given without prior consultation was that it was feared that so to consult would have given rise to grave risk of industrial action through the reaction of the appellants and others and thus have brought about the very situation which the oral instructions were themselves designed to avoid, namely the risk of industrial action by the staff at GCHQ caused or at least facilitated by a membership of trade unions, and damaging to national security. GCHQ was, it was said, and is, highly vulnerable to industrial action and prior consultation would have revealed to those who had previously organised disruption that high degree of vulnerability.
My Lords, the conflict between private rights and the rights of the state is not novel either in our political history or in our courts. Historically, at least since 1688, the courts have sought to present a barrier to inordinate claims by the executive. But they have also been obliged to recognise that in some fields that barrier must be lowered and that on occasions, albeit with reluctance, the courts must accept that the claims of executive power must take precedence over those of the individual. One such field is that of national security. The courts have long shown themselves sensitive to the assertion by the executive that considerations of national security must preclude judicial investigation of a particular individual grievance. But even in that field the courts will not act on a mere assertion that questions of national security were involved. Evidence is required that the decision under challenge was in fact founded on those grounds. That that principle exists is I think beyond doubt. In a famous passage in The Zamora  2 AC 77
, 107 Lord Parker of Waddington, delivering the opinion of the Judicial Committee. said:
“Those who are responsible for the national security must be the sole judges of what the national security requires. It would be obviously undesirable that such matters should be made the subject of evidence in a court of law or otherwise discussed in public.”
The Judicial Committee were there asserting what I have already sought to say, namely that some matters, of which national security is one, are not amenable to the judicial process. The force of the passage I have quoted is in no way diminished by the fact, much relied on by Mr. Blom-Cooper, that in that case the Crown failed because they had failed to adduce before the Prize Court the requisite evidence of urgent necessity, proof of which was essential if the right of angary were to be successfully invoked in relation to a cargo in the custody of the Prize Court. This last mentioned fact merely reinforces what I have just said — that evidence and not mere assertion must be forthcoming.
A similar problem arose in Chandler v. Director of Public Prosecutions  AC 763
, a case under section 1 of the Official Secrets Act 1911. Lord Reid, at p. 790, expressly stated that he did not “subscribe to the view that the Government or a minister must always or even as a general rule have the last word” about the safety or interests of the state. But he agreed, in common with all the other members of the House, that cross-examination was not permissible to challenge the evidence of a senior Air Force officer that a proposed obstruction of an airfield was contrary to the “safety or interests of the state” which were the relevant words of the statute.
“The defence of the State from external enemies is a matter of real concern, in time of peace as in days of war. The disposition, armament and direction of the defence forces of the State are matters decided upon by the Crown and are within its jurisdiction as the executive power of the State. So are treaties and alliances with other states for mutual defence …” (p. 796).
The other noble and learned Lords then sitting shared Lord Reid's view, though I venture most respectfully to question one observation of Lord Devlin's, where after referring to The Zamora the learned Lord said, at p. 810:
“It is said that in such cases the minister's statement is conclusive. Certainly: but conclusive of what? Conclusive, in the absence of any allegation of bad faith or abuse, that he does think what he says he thinks. The court refrains from any inquiry into the question whether the goods are, in fact, necessary, not because it is bound to accept the statement of the Crown that they are, and to find accordingly, but because that is not the question which it has to decide.”
I respectfully suggest that that passage is out of line with the views expressed by the other noble and learned Lords then sitting.
The same problem arose in Reg. v. Secretary of State for Home Affairs, Ex parte Hosenball  1 W.L.R. 766 where the Court of Appeal and in particular Lord Denning M.R., at p. 778, accepted that if the case had been one “in which the ordinary rules of natural justice were to be observed, some criticism could be directed upon it” but held that the interests of national security must override the appellants' private rights and that where compliance with the requirements of natural justice would itself have revealed that which it was in the interests of national security not to reveal, private rights must yield to the public interest: see especially pp. 782–783.
My Lords, I venture to think that today this principle cannot be disputed. The question is whether, on the evidence before your Lordships, the respondent is entitled to assert that it was for fear of revealing the vulnerability of GCHQ to industrial action that it was decided that advance consultation could not take place. Mr. Blom-Cooper did not contest that there was evidence upon which a reasonable minister might have taken that view or, indeed, that the respondent as a reasonable minister might have taken that view. His main contention was that the submission on behalf of the respondent to be found encapsulated in paragraph 27(i) of the respondent's case thus:
“It was considered that consultation would involve a real risk that it would occasion the very kind of disruption which was a threat to national security and which it was intended to avoid. Having regard to these factors a reasonable minister could properly take the decision without consultation.”
was an afterthought and unjustified by the evidence adduced on the respondent's behalf.
In their judgment, the Court of Appeal set out three of the assertions by or on behalf of the trade unions concerned regarding the possibility of and the effect of disruption at GCHQ by industrial action. There are many other similar statements in the evidence. I refer only to two of these other statements. The first is:
“Walk-outs in key installations have affected Britain's defence capability in general, and crippled the UK contribution to the NATO exercise ‘Wintex.’”
The other, under the heading “Government Communications,” is:
“48-hour walk-outs have severely hit secret monitoring stations belonging to the Composite Signals Organisation. The Government is clearly worried and will be subject to huge pressure from NATO allies ….”
Nevertheless, Mr. Blom-Cooper claimed that careful reading of Sir Robert Armstrong's first affidavit, and in particular paragraph 16 of that affidavit, did not support the view that this was a consideration which the respondent had ever had in mind. My Lords, with all respect, paragraph 16 must not be divorced from its contents or read in isolation from the paragraphs which both precede it and follow it. Paragraphs 13 to 18 inclusive must all be read together. In those paragraphs I read Sir Robert as explaining why the possibility of negotiating a non-disruption agreement was considered and rejected. I draw particular attention to the final sentence in paragraph 16 which reads:
“To have entered such consultations would have served to bring out the vulnerability of areas of operations to those who had shown themselves ready to organise disruption and consultation with individual members of staff at GCHQ would have been impossible without involving the national unions.”
Ministers also were of the view that the importance of the decision was such as to warrant its first being announced in Parliament. This passage read in the context of the documentary evidence exhibited to Sir Robert's affidavit to which I have already referred seems to me to make abundantly clear why the respondent and other ministers declined to engage in consultations in advance of issuing the instructions. It was argued that such consultation might have led to a non-disruption agreement such as was later suggested on behalf of the appellants. But the draft of that agreement clearly does not achieve that which the respondent sought to achieve by the instructions and the evidence clearly shows that the national unions, without whose co-operation a non-disruption agreement would have been valueless, were not prepared to countenance such an agreement. It was also suggested that if consultation had taken place regarding the polygraph there was no reason why consultation should not take place regarding the intended instructions. My Lords, the short answer to that is that the two are not comparable.
My Lords, I have therefore reached the clear conclusions, first, that the respondent has established that the work at GCHQ was a matter of grave national security, second, that that security would have been seriously compromised had industrial action akin to that previously encountered between 1979 and 1981 taken place, third, that consultation with the appellants prior to the oral instructions would have served only further to reveal the vulnerability of GCHQ to such industrial action, fourth, that it was in the interests of national security that that should not be allowed to take place, and fifth, that accordingly the respondent was justified in the interests of national security in issuing the instructions without prior consultation with the appellants.
That conclusion accords with the conclusion reached by the Court of Appeal and must lead to the result that the appeal should be dismissed. I would only add, again in agreement with the Court of Appeal, that had the matter been argued before the learned judge, as it was in the Court of Appeal and before this House, he might well have reached a different conclusion from that which he reached.
For the sake of completeness I would add that I reject Mr. Blom-Cooper's separate argument that the oral instructions were in any event bad as insufficiently specific or precise. I am in complete agreement with the views of both courts below on that submission.
I do not find it necessary to say anything about what became known as the “futility argument,” that is to say that even if consultation were required it would have been futile because it would have been of no effect. On the view I take, that matter does not arise for decision.
LORD BRIGHTMAN. My Lords, I also would dismiss this appeal for one reason only, namely, on the ground of national security. The evidence is compelling that the Minister for the Civil Service acted without prior consultation with the unions concerned because she believed, and reasonably believed, that such process of prior consultation might result in disruption that would pose a threat to the security of the nation. This factor overrode the right in public law which the unions would otherwise have had, on the facts of this particular case, to be consulted before the instruction of 22 December 1983 was given.
There is nothing which I can usefully add to the comprehensive survey which your Lordships have already made of the authorities on the reviewability of decisions taken under the royal prerogative. There is no difference between the conclusions reached by your Lordships except on one isolated point: whether the reviewability of an exercise of a prerogative power is limited to the case where the power has been delegated to the decision-maker by Order in Council, so that the decision-making process which is sought to be reviewed arises under and must be exercised in accordance with the terms of that order; or whether reviewability may also extend, in an appropriate case, to a direct exercise of a prerogative power. Like my noble and learned friend, Lord Fraser of Tullybelton, I would prefer to leave the resolution of that question to a case where it must necessarily be determined.
For the reason indicated, I would dismiss this appeal.
No order as to costs in House of Lords or below.