The United Policyholders Group and others v The Attorney General of Trinidad and Tobago

The appellants, who are all residents of Trinidad and Tobago, are holders of life policies issued by the Colonial Life Insurance Company (CLICO). Their claim arises out of the banking crisis in early 2009 when CLICO was in serious financial difficulties. That claim is based on assurances of support for CLICO given by the then government, which they say created a “legitimate expectation” enforceable in law. They assert that, following the elections in May 2010, the new government failed to honour that expectation, and that they are entitled to relief accordingly. Their claim succeeded in the High Court but failed in the Court of Appeal.

CLICO is a company regulated by the Insurance Act and is a subsidiary of a holding company, CL Financial Ltd (CLF); one of CLICO’s subsidiaries is CLICO Investment Bank (CIB). As the Court of Appeal noted, CLF was the largest private conglomerate in Trinidad and Tobago. The reported value of its assets was equivalent to more than 70% of the country’s gross domestic product, and a substantial part of those assets are in the financial sector.

Under section 37(1) of the Insurance Act, CLICO was required to establish and maintain a statutory fund to cover its liabilities to policyholders under long-term insurance business. Section 37(4) of the same Act required CLICO to maintain a fund in Trinidad and Tobago with sufficient assets to cover its liability (including its contingency reserves) to policyholders’ resident in Trinidad and Tobago.The Insurance Act contained investment rules for such Funds, including, for example, in section 46 and Schedule 2, a prohibition on investment in more than 30% of the shares of a single company. There were also restrictive rules for the valuation of assets in such Funds. Under section 80(1) of the Insurance Act, policyholders protected by a Fund are to be treated in preference to other creditors in insolvency, and in such an event the assets and liabilities of the Fund are to be assessed separately and, with the exception of any surplus, used only to pay off liabilities of the Fund.

CLICO was under the general supervision of the Central Bank of Trinidad and Tobago. Under section 41 of the Insurance Act, it was required to report to the Central Bank with particulars of the assets and liabilities of the Fund. Section 44D of the Central Bank Act gave the Central Bank power to take control of a financial institution and to take all steps it considered necessary to protect the interests of depositors and creditors of the institution, where it was of the opinion that those interests are threatened, or that the institution was likely to be insolvent or that it was not maintaining high standards of probity or sound business practices. Section 44F(5) of the Central Bank Act required the Central Bank to comply with general or special directions of the Minister.

The appellants all held versions of a CLICO policy called the Executive Flexible Premium Annuity (EFPA). These policies were distinguished by the offer of relatively high interest rates paid out on the premium for a fixed initial period. They also included provision for an annuity on the life of the policyholder, and (as is now common ground) were thus within the definition of “long term insurance business” for the purposes of the Insurance Act.

In January 2009 the Central Bank became aware that CIB had serious liquidity problems, and that both CIB and CLICO were in substantial deficit. On 24 January 2009, the Central Bank advised the Minister of Finance and the Prime Minister that there would be a run on both institutions if the problems became public. On 30 January 2009, the Central Bank took control of CIB using its emergency powers under section 44D of the Central Bank Act.

On the same day the government and CLF came to an agreement which was set out in a Memorandum of Understanding (the MoU). The preamble of the MoU recorded that CLF had asked for the government’s intervention in the “rehabilitation” of CIB, CLICO and another subsidiary British American Insurance Company (BA). It stated:

CLF agreed to sell its shareholdings in a number of identified companies and such other assets as might be necessary, and to apply the proceeds (a) to correct CIB’s financial position, and (b) to ensure that CLICO’s and BA’s statutory fund requirements were satisfied.

The government agreed to provide collateralized loan financing to CLICO and BA to meet any residual statutory fund deficit which might still exist after this sale of CLF’s shareholdings and assets.

CLF agreed that CLICO and BA would restructure their operations to conform to traditional life insurance business in a manner to be approved by the Central Bank.

CLF was required to make full and fair disclosure of the liabilities and assets of companies within the group.

On 6 February 2009 the Central Bank Act was amended to extend the Central Bank’s powers of intervention under section 44D to cover insurance companies. On 13 February 2009 the Central Bank took control of CLICO. A statement issued by the Central Bank on that day said:

The appellants rely on a number of public statements, made at this time and over the following months, about the nature of the government’s support for CLICO. It is unnecessary to set them all out. Some were qualified specifically by reference to the conditions of the MoU, but others were not so qualified. For example, on 15 February 2009 a full-page advertisement was placed in local newspapers in the name of CLICO (signed by the new Chief Executive, Mr Musaib-Ali, appointed at or about the time of the takeover by the Central Bank).

In a Parliamentary Answer on 24 June 2009, the Minister of Finance referred to “the restructuring agreement … carried out in the Memorandum of Understanding signed with [CLF]” under which the government was “committed to restructuring [CLF] as a going concern” to ensure that other creditors and shareholders would “at the end of the day … get back and recoup all of their losses or potential losses”. But that was “not a guarantee to them”. She explained:

On 24 July 2009 the Cabinet approved funding of $5 billon to facilitate the restructuring of CLICO and BA. The initial injection through the Central Bank of $1.2 billon was later increased to $1.9 billon. Further funding of $3.1 billon was provided in the form of government bonds issued directly to CLICO and placed in the Fund to reduce the deficit. They were to be drawn upon to meet liabilities as they arose and which CLICO was not able to meet.

CLICO continued to trade under the control of the Central Bank, but problems continued into the following year. On 13 January 2010, in an interview, the Minister of Finance spoke of the wide implications of the CLICO intervention:

On 24 May 2010 there was a general election, and this led to a new administration. The new Minister of Finance, Mr Dookeran, stated in his evidence in these proceedings that CLICO was “the biggest and most difficult issue” that he had had to deal with in preparing the government’s first budget due in September. He noted also that the International Monetary Fund considered the stability of Trinidad and Tobago’s economic outlook to be heavily dependent on the resolution of CLICO’s restructuring.

In June 2010 the government appointed an expert select committee to provide recommendations on the way forward for CLICO and the Group. Its mandate required it to make recommendations on “a preferred solution, from a menu of options, for the repayment of CLICO’s traditional and non-traditional (EFPA) insurance liability products”, a financial reorganisation plan for CLF, and a “clear path and timetable” for the government to “exit its loan capital position and restore public confidence”.

At the end of July the committee reported on three options (in summary):

Provide no further funding and liquidate CLICO and BA;

Fully fund the entire asset shortfall of CLICO and BA and repay all creditors based on contractual terms (not just policyholders protected by the statutory fund);

Provide $75,000 to all EFPA and Mutual Fund policyholders (including non-residents not protected by the fund), and pay remaining liabilities by government bonds spread over a 20-year period.

On 8 September 2010, in his budget statement to the House of Assembly, Mr Dookeran reported on the position of CLICO. He made critical observations about the actions taken by the previous administration, which, he said, had treated it as a liquidity issue that could be covered in the short-to-medium term, but had done so without full information on the financial conditions of the companies. He described this as “a reckless assumption” which had cost the nation significant public funds involving more than 10% of the country’s gross domestic product, and affecting 250,000 of its citizens. He also made reference to the “numerous public statements” made by the then Minister of Finance, the Central Bank and CLICO, assuring depositors that their money was safe and would be protected by the government. Total funding provided up to May 2010 had amounted to approximately $7.3 billion.

Announcing the government’s proposals he said that the traditional insurance business would be separated from short term investment business, and that “the obligation to the 225,000 policyholders would be honoured, backed by the statutory fund. For short term investors (including EFPA holders), the government would make “an initial partial payment of a maximum of $75,000 … intended to bring relief to the small depositors”. This would result in full payment to “approximately 45% of the 25,000 investors in these products, including more than 140 credit unions and 15 trade unions”. Those whose principal balances exceeded $75,000 would be paid through “a government IOU amortized over 20 years at zero interest” structured in such a way that it could be traded on the secondary markets, thereby creating “a measure of immediate liquidity for the depositors”.

On 9 September 2010, pursuant to a direction of the Minister of Finance, CLICO placed a moratorium on all EFPA transactions and all payments to EFPA policyholders. It continued to collect premiums from traditional policyholders and to pay out claims.

In a media briefing on 28 September Mr Dookeran commented on suggestions that the government was not honouring the “guarantee” given by the previous administration. He accepted that there had been “utterances of guarantees”, but he contended that, in order to be effective, they would have required parliamentary appropriations which had not been obtained. Accordingly, these guarantees had “no backing, either in terms of the allocation of funds through the parliament or certainly in any other way”. He explained:

On 1 October 2010, in a speech to the House of Assembly, the Prime Minister also criticised the former administration’s handling of the CLICO matter. In particular she criticised it for adopting a “narrow view” directed solely to local investors covered by the statutory fund, and ignoring the 1,100 investors in EFPAs from outside the country, worth some £1.2 billon. The government was under “no legal obligation” to put in more public money to help those placed in this position because of “mismanagement on the part of the former administration” or of the companies themselves.

In the following year, on 29 April 2011 the appellants wrote to the Prime Minister challenging the legality of the government’s proposals, on the grounds (inter alia) that they would frustrate their legitimate expectations arising out of promises by the previous administration, and asking for further information about CLICO’s assets. Following further exchanges, in August 2011 they sent a draft application for judicial review.

Meanwhile the Cabinet was considering further options aimed at improving the return to policy-holders. On 14 September 2011 the Minister of Finance announced an “enhanced pay-out regime” for those EFPA policyholders with balances greater than $75,000. Under the new proposals, policyholders could exchange their bonds with maturities of 11 to 20 years for units in a new entity, thereby obtaining a value equivalent to the face value of those bonds. As a result, the average return would be in the order of 92% of the value of the principal balances of their holdings. The new entity was set up in 2012 and named the CLICO Investment Fund. On 20 September 2011 the Central Bank Act was amended to prohibit proceedings against an institution (such as CLICO) which had been taken under control under section 44D of that Act.

From the time the new bailout plan was devised (either in its original form or as revised in September 2011), it was a condition that investors should either give up their rights against the statutory fund and take up the government’s bailout plan or stand on their rights. The final deadline for acceptance was 30 November 2012. On the appellants’ case, up to two weeks before that deadline, and in spite of numerous requests, they were given no information as to the status of the fund and the rights that they were being asked to give up.

Since the High Court decision, on 27 and 28 March 2015 the Central Bank announced that policyholders would receive 85% of their contractual entitlements within three months (principal balances plus contractual interest), and those sums have been paid. As to the balance of 15%, this was said to depend on the sale of one of CLICO’s assets, Methanol Holdings International Ltd (“MHIL”). In 2009 CLF transferred shares in both Methanol Holdings (Trinidad) Ltd (“MHTL”) and MHIL to CLICO. These shares, together with the MHTL and MHIL shares already held by CLICO, constituted about a third of CLICO’s total assets. The transfer of the shares prompted arbitration proceedings, brought by the holders of preference rights in the shares. The proceedings were only resolved in 2013. The tribunal ordered that the MHTL shares be transferred for value to the holder of the preference rights. The value at which they were to be sold was determined in 2014.

The present proceedings were commenced by the appellants against the Attorney General (on behalf of the government) on 1 December 2011. On 18 April 2012 Charles J granted leave to apply for judicial review. She rejected the government’s argument that the appellants had been guilty of delay, on the ground that “significant changes” had been made to the government’s proposal between 8 September 2010 and 14 September 2011, so that the latter date should be taken as the point at which the decision subject to challenge was taken.

She heard the application over three days in November 2012, and she gave judgment on 12 March 2013 upholding the claim. She ordered the respondents to make arrangements to secure that the appellants recovered 100% of CLICO’s contractual liability to them. In July 2015 the Court of Appeal (Archie CJ, Narine and Smith JJA) heard the appeal, which they allowed on 23 June 2014 in a single judgment given by Narine JA.

Charles J’s approach to the claim started from the proposition, which had been conceded, that the appellants had a legitimate expectation to the effect that -

Mr Dookeran’s evidence gave no assessment of the public interest factors which at that time justified the breach of the legitimate expectation, having regard in particular to the “fiscal improvement” by then in both the country’s finances and CLICO’s balance sheet (para 74). She rejected the submission of Mr Newman QC for the Attorney General that the matter lay within “the macro-economic/political field” requiring deference to the government’s decision; it was for the court to examine all of the circumstances to determine whether there was “an overriding public interest” justifying the breach (para 76).

Further, Charles J said that the government had failed to take proper account of the promises made by the previous administration. Mr Dookeran’s affidavit spoke of “statements” made by the previous administration and the “expectation” of the policyholders; but failed to acknowledge that the statements amounted to promises which gave rise to a legitimate expectation or to address the fact that the new plan was in breach of this legitimate expectation” (para 78).

Narine JA took a different view from Charles J on each of the principal issues. He held that the statements relied on had not been “clear, unambiguous and devoid of relevant qualification”. He noted in particular that the statements failed to specify what was being guaranteed, whether principal alone, or principal plus interest; there was no indication when the payment was to be made; and the guarantee was premised upon the assets of CLF being sold and the proceeds deposited into the statutory fund, and on the restructuring of CLICO, and its solvency as a going concern. CLF had not carried out its obligations to sell assets and place the proceeds into the statutory fund, so that the government’s obligation to finance any deficit in the fund was not triggered (paras 55-56). Although he had earlier accepted that the CLICO full page advertisement of 15 February 2009 “appear[ed] to go further”, it had been placed in the local media by CLICO, not the government or the Central Bank (paras 45-46). Similarly, there was no evidence of breach; in the absence of a “specified time frame” there was no basis for saying that the guarantee had fallen due, or the preconditions had been fulfilled (paras 64-70).

In any event, the evidence filed on behalf of the Attorney General had established an “overriding public interest” for the solution that the government adopted, having regard to the importance of CLICO to the economy and to the “increasingly difficult” financial position in 2010 (paras 75-94). The time for considering that issue was September 2010 when the original plan was introduced, but the same conclusion would have applied in September 2011 (para 101). The evidence also showed that the statements made by the previous administration had been taken into account (para 114).

Finally, Narine JA considered that the decision was “firmly anchored in macro-economic and macro-political issues”, having regard (inter alia) to the scale of CLF’s assets (roughly equivalent to 70% of the Gross Domestic Product of the country, as mentioned above), the potential for systemic risk to the financial system, the concerns of the IMF, and the possible downgrading of the credit rating of the country by the international credit rating agencies, and the potential impact on moneys available for public services. This issue had been “glossed over” by the judge (paras 118-119). In this context it was for the claimants to show that the decision to adopt option (iii), as described in para 15 above, was “irrational”, which they had failed to do. Rather it was “methodical, reasonable and proportionate in the circumstances” (para 125).

On 3 November 2014 the Court of Appeal gave final leave to the appellants to appeal to the Privy Council. Although the grounds of appeal make a number of criticisms of the Court of Appeal’s reasoning, the issues can be summarised under two principal headings:

Did the representations made by or on behalf of the government during 2009 give rise to a “legitimate expectation” that resident EFPA holders in the position of the appellants would be fully protected?

If so, was the government entitled in law to resile from the expectation so created?

In his submissions before the Board, Mr Knox for the appellants generally supported the reasoning of Charles J. He conducted a meticulous review of the evidence relating to the representations made during 2009 and to the actions of the government in 2010 and 2011, with a view to showing (i) that the concession which had been made on behalf of the government before the judge that the appellants had a legitimate expectation founded on the assurances given by the previous administration set out in paras 7 to 11 above (“the Assurances”) was well-founded, and (ii) that the government’s subsequent decision not to comply with the Assurances and to pursue a different policy (“the 2010/2011 policy”) was, on analysis, unjustifiable in the light of that legitimate expectation and all other relevant factors - whether the date as at which to assess this point is September 2010, or (as he argued) September 2011.

Mr Stevens QC for the Attorney General supported the reasoning of the Court of Appeal. Like them he placed weight on the “macroeconomic” character of the issues facing the government in 2010, and the careful consideration given to their response. He also contended that the assurances on which the appellants relied were insufficiently unambiguous to found a claim based on legitimate expectation. Even if a claim based on legitimate expectation was in principle justified, he said that the government had taken proper account of the assurances given by its predecessor, but also of the serious and largely unpredicted financial problems affecting CLICO. These had to be balanced against the wider impact on the financial system and the economy. The correct date for assessment of its actions was, he said September 2010 when the government made clear that it would not regard itself as absolutely bound by the assurances of its predecessor. Within the bounds of rationality, it was entitled to take that position.

Before addressing the two questions identified in para 33 above, it is appropriate to summarise briefly the board’s understanding of the law relating to legitimate expectation.

In the broadest of terms, the principle of legitimate expectation is based on the proposition that, where a public body states that it will do (or not do) something, a person who has reasonably relied on the statement should, in the absence of good reasons, be entitled to rely on the statement and enforce it through the courts. Some points are plain. First, in order to found a claim based on the principle, it is clear that the statement in question must be “clear, unambiguous and devoid of relevant qualification”, according to Bingham LJ in R v Inland Revenue Comrs, Ex p MFK Underwriting Agents Ltd[1990] 1 WLR 1545, 1569, cited with approval by Lord Hoffmann in R (Bancoult) v Secretary of State for Foreign and Commonwealth Affairs (No 2)[2009] AC 453, para 60.

Secondly, the principle cannot be invoked if, or to the extent that, it would interfere with the public body’s statutory duty - see eg Attorney-General of Hong Kong v Ng Yuen Shiu [1983] 2 AC 629, 636, per Lord Fraser of Tullybelton. Thirdly, however much a person is entitled to say that a statement by a public body gave rise to a legitimate expectation on his part, circumstances may arise where it becomes inappropriate to permit that person to invoke the principle to enforce the public body to comply with the statement. This third point can often be elided with the second point, but it can go wider: for instance, if, taking into account the fact that the principle applies and all other relevant circumstances, a public body could, or a fortiori should, reasonably decide not to comply with the statement.

Quite apart from these points, like most widely expressed propositions, the broad statement set out at the beginning of para 37 above is subject to exceptions and qualifications. It is, for instance, clear that legitimate expectation can be invoked in relation to most, if not all, statements as to the procedure to be adopted in a particular context (see again Ng Yuen Shiu [1983] 2 AC 629, 636). However, it is unclear quite how far it can be applied in relation to statements as to substantive matters, for instance statements in relation to what Laws LJ called “the macro-political field” (in R v Secretary of State for Education and Employment, Ex p Begbie [2000] 1 WLR 1115, 1131), or indeed the macro-economic field. As the cases discussed by Lord Carnwath show, such issues have been considered by the Court of Appeal of England and Wales, perhaps most notably, in addition to Begbie, in R v North and East Devon Health Authority, Ex p Coughlan [2001] QB 213, R (Nadarajah) v Secretary of State for the Home Department [2005] EWCA Civ 1363, and R (Niazi) v Secretary of State for the Home Department[2008] EWCA Civ 755, and also by the Board in Paponette v Attorney General of Trinidad and Tobago [2012] 1 AC 1.

For present purposes, for reasons which should become clear from the ensuing part of this judgment, it is unnecessary for the Board to consider the law on this difficult and important topic more fully.

This question involves considering whether the Assurances gave rise in law to a legitimate expectation on the part of the appellants. Before the Board turns to the substantive issue, there is a procedural aspect which needs to be addressed.

In paras 34-35 of her judgment, Charles J recorded a concession made by Mr Newman QC for the Attorney General that the Assurances did in fact amount to “promises to the claimants which gave rise to a legitimate expectation”. The Attorney General was not held to this concession in the Court of Appeal. Indeed, Narine JA criticised the judge for accepting it, thereby “in effect surrender[ing] to the attorneys the very basis of her decision”, which he said was particularly inappropriate in the context of declaratory proceedings, in which it was for the court to “satisfy itself on the evidence and the law that there is a proper basis for granting such relief” (paras 62-63).

For the appellants Mr Knox argued that the Attorney General should not have been allowed to go back on his concession, made by senior counsel on instructions from the Attorney General, and presumably after careful consideration of the evidence and arguments.

The Board considers that it was harsh on the part of the Court of Appeal to criticise Charles J for accepting the concession. Indeed, in a case with experienced counsel making the concession, she might have been open to criticism for not accepting that concession.

However, while the Board accepts that the Court of Appeal’s decision to permit the government to withdraw their concession could be characterised as lenient, it is not the sort of decision with which an appellate tribunal should lightly interfere. While it may not have been a case management decision, it was a procedural, rather than a substantive, decision, which involved weighing up competing arguments, and it was therefore a decision which an appellate tribunal should respect, unless it was outside the bounds of reasonableness. The Board does not consider that it was outside those bounds. The appellants were not relevantly prejudiced by the withdrawal of the concession, as it had been made at a relatively late stage in the hearing before the judge, and the appellants had had the opportunity to call all the evidence, and make all the submissions, which they wished to call and make on this issue. The Board has also heard full argument on the facts and law relating to the issue.

In all the circumstances, the Board rejects the contention that the Attorney General should not have been allowed to challenge the appellants’ case that they had a legitimate expectation as described by the judge and quoted in para 27 above, and therefore turns to consider the issue.

As explained in para 35 above, Mr Stevens’ case for the Attorney General was that the Assurances were, as a matter of law, incapable of giving rise to a legitimate expectation because (i) they fell within the macro-economic area (and therefore also in the macro-political area), and (ii) in any event, they did not amount to a sufficiently clear and unambiguous representation or promise, particularly in the light of the terms of the MoU. To a significant extent, these two points are closely connected, as a major factor in the second point is that the MoU made it clear that substantial macro-economic issues were potentially engaged by any of the Assurances.

There is obvious force in these two points. The sums potentially involved in complying with the Assurances were enormous when viewed in context: as already mentioned, CLF was the largest private conglomerate in Trinidad and Tobago, and its assets had an apparent value of 70% of the GDP of the country. And the wider significance of the Assurances was made clear in the preamble to the MoU itself, in which the financial condition of CLICO and its group was said not only to threaten the interests of depositors and policy holders but to “pose danger of disruption or damage to the financial system of Trinidad and Tobago”. No interested investors can have failed to understand the policy background against which the decisions were being made and the assurances given.

However, even though the assurances had macro-economic (and therefore also macro-political) implications, the appellants can rely on the fact that they were made to the appellants as investors rather than as members of the public. Quite apart from this, there is obviously a case for saying that, as a matter of principle, the macro-economic implications of the Assurances ought not to go to the question whether they are capable of giving rise to a legitimate expectation: rather they should be relevant to the question whether it was permissible for the government not to adhere to any promises.

Mr Knox for the appellants relied on the Assurances made in apparently unqualified terms after the MoU was signed, most notably the full page advertisement of 15 February 2009 which asserted in terms that policyholders’ funds are being “guaranteed by the government”, and which made no reference to the MoU. However, it was not until the Minister’s statements in June 2009, that it was made clear that the “guarantee” was directed to a defined group, namely the resident policy-holders protected by the section 37(4) statutory fund, rather than investors in general, and those statements were linked in terms to the MoU.

In all the circumstances, the Board is prepared to assume, without deciding, that the appellants succeed in crossing the first of the two hurdles which their legitimate expectation argument faces. In other words, for the purposes of this appeal it is accepted that the appellants had a legitimate expectation, as a result of the Assurances, that the government would, as Charles J put it, “make good the deficit in CLICO’s Statutory Fund … and [the company] would be placed in a position to fulfil all of its obligations including that of the claimants.”

The first point to decide in this connection is as at what date the government’s decision not to adhere to the Assurances is to be assessed. The Board agrees with the Court of Appeal that the issue must be judged not later than September 2010, when the government decided and announced that the Assurances would not be regarded as binding, and that the 2010/2011 policy would be pursued. The government made it clear at that point that its primary duty was to the public at large, that the earlier Assurances to particular groups had to be viewed against that background, and that they need not - indeed they could not - regard themselves as precluded by those assurances from giving effect to what they regarded as the most appropriate policy in the national economic interest.

It is true that Charles J decided that the correct date was September 2011, but that was at the stage when she was deciding whether to refuse the appellants leave to seek judicial review on the ground of delay. More importantly, the essential question is when the decision not to adhere to the Assurances, or to put it another way, to abandon the previous administration’s policy, was made, and that is apparent from the documents. The decision was made, announced, explained and preliminarily implemented in the autumn of 2010 - see paras 17-20 above. The fact that this led to the 2010/2011 policy which was only finalised in September 2011 does not mean that the appellants can, as it were, elect for the later date.

If the 2010/2011 policy can be undermined by the appellants as at September 2010, then it would be unnecessary to decide if the appellants are right to contend that that is the wrong date. So it is only if that policy was valid as at September 2010 that the appellants’ contention that September 2011 is the correct date would come into play. But it seems to the Board to be unrealistic to contend that, having justifiably adopted a policy in September 2010, the government could be criticised for continuing with that policy in September 2011, because of assurances given in 2009, especially given that the implications of the policy were such that it took the best part of a year in working out how to implement it. Government business would be severely impeded if such an argument could succeed. As Mr Stevens put it, the appellants may, at least in theory, be entitled to challenge the decision to implement the 2010/2011 policy in September 2011 on normal judicial review grounds, but their challenge is to the decision not to adhere to the Assurances (and to embark on the policy) and that decision was made in September 2010. For the same reason, the fact that there were some changes in the policy between the time it was made in September 2010 and the time it was finalised in September 2011 is of no assistance to the appellants.

When it comes to the question whether the government was entitled to renege on the Assurances (which the Board is prepared to assume gave rise to a legitimate expectation on the part of the appellants), Mr Knox contended that the Court of Appeal ought not to have interfered with Charles J’s conclusion that the government was not so entitled, as it was based on findings of fact. The Board takes the view that there were a number of aspects on which the Court of Appeal was entitled to criticise the judge, and, as a result, to substitute its own view for that which she had reached. Although there may have been others, there were three principal aspects, the first two of which are connected.

First, Charles J did not give any real weight to the point made by the Prime Minister in her October 2010 speech (supported by Mr Dookeran’s evidence) that as “a responsible government”, the government could not “deprive 1.2m citizens … who did not invest in CLICO … of much needed expenditure, infrastructure for health, for education, for security”, and that “the whole budget” should not be “take[n] up” for that purpose”.

The judge also failed to take into account the fact that compliance with the assurances would not simply have meant topping up the Fund, but would have had serious cash-flow implications for the government. That is reflected in another passage in the Prime Minister’s speech (and Mr Dookeran’s evidence) to the effect that committing the government to paying $1 billon a year for ten years “would have crowded out the possibility of borrowing for the purposes of resuscitating the economy and financing the development of the nation”.

The judge further failed to take into account an impediment to an important component of the implementation of the arrangements which had been part of the basis upon which the Assurances had been given. The MHIL shares remained with CLICO. As per the announcement of the Central Bank in March 2015, it was envisaged that the MHIL shares would be sold to help fund the remaining 15% balance.

Particularly because the decision whether or not to adhere to the Assurances had potentially very serious implications for government finances, especially for government expenditure on public projects, many of which could be regarded as being of real general significance, it seems to the Board that these criticisms amply justified the Court of Appeal’s decision to form its own view as to the justifiability of the government’s decision not to comply with the Assurances. In fairness to the judge, it should be added that the way in which the Attorney General’s case was put before her appears to have been rather different from the way in which it was put before the Court of Appeal and the Board.

There is no real difference in principle between the parties so far as the law is concerned, and, while there are differences of emphasis, it does not appear to the Board that the differences in emphasis are of significance in the context of this appeal. On the assumption that the appellants had a legitimate expectation as claimed, Mr Stevens contends that, bearing in mind the factual position as at September 2010, the new administration was, to put it at its lowest, entitled not to adhere to the Assurances. Mr Knox accepts that the government was not absolutely bound by the assurances, but says that the factual basis for its justification for not being bound does not hold water. In effect, the appellants say Charles J was right on this issue, and the Attorney General contends that Narine JA was correct.

In summary terms, Mr Stevens contended that, as explained by the Prime Minister and Mr Dookeran to Parliament in 2010, and by Mr Dookeran in his affidavit in these proceedings, CLICO’s financial problems were found to be even more serious than originally anticipated. For example, the group assets were far more leveraged than at first realised such that they were not going to realise much cash in the short to medium term, the value of some of the assets had fallen, the deficit in the Fund was larger than envisaged, and the redemption of EFPAs had been much larger than expected (thereby increasing the cash flow demand on CLICO above that expected). Furthermore, there were particular problems in realising the value of the important MHTL holding which were not resolved until much later. It was not possible to divert all the assets to the Fund without threatening the viability of the company as a whole. Since the basis of the Memorandum was that CLICO would continue as a going concern, it was essential to consider the interests of other creditors.

Even assuming that there were promises which had been sufficiently clear and unqualified to satisfy the relevant test, the government was, said Mr Stevens, entitled to take account of the wider policy issues, when deciding how far to give effect to them. Unlike in Paponette where no explanation was given for the change of policy, in this case the government’s thinking was fully explained to Parliament at the time and later to the court. Mr Stevens accepted that it was somewhat disingenuous for the Minister to suggest that the promises might have been ineffective for lack of Parliamentary authorisation, given that they were made on behalf of a government with a Parliamentary majority; but it would be wrong to attach too much importance to political point-scoring of this kind. Mr Stevens contended that the important point is that the Minister took account of the assurances, but thought it wrong to use government funds to give undue priority to a particular group of investors, at the expense of the investors in general and the stability of the group as a whole. None the less they were given a large measure of protection, albeit less than 100% for the larger investors.

Accordingly, the case for the Attorney General, as accepted by the Court of Appeal, is that the government’s abandonment of the Assurances and adoption of the 2010/2011 policy cannot be challenged as irrational applying Wednesbury principles, especially when one bears in mind the heightened standard applicable to decisions of national economic policy. Mr Stevens contended that any legitimate expectation created by the Assurances given by the previous administration had to be seen in the wider policy context, and the government was entitled to depart from it for what it perceived to be good reasons in the national economic interest. He also commended the Court of Appeal’s assessment that the new administration’s approach was “methodical, reasonable and proportionate in the circumstances”.

Mr Knox’s assault on this submission was based on a number of points, which, he contended undermined the case made out on behalf of the government as to why it was entitled not to adhere to the assurances. The appropriate way of dealing with those points is to set each of them out in turn and then discuss them, albeit relatively summarily.

The first of the points is that there were, according to a report from Credit Suisse, sufficient assets in the Fund to cover its liabilities. The conclusion reached in the Credit Suisse report was, however, contradicted by a report produced by a select committee of experts appointed by the government. In the Board’s view, the government was entitled (to put it at its lowest) to rely on the latter report, especially as the former report was based on CLICO and BA management accounts, and did not take into account the problem with the MHTL shares referred to in paras 24, 58 and 61 above.

More broadly, the appellants challenged the government’s contention that it had become apparent by September 2010 that CLICO’s financial position was significantly worse than it appeared when the Assurances were given in 2009. However, Mr Dookeran, the Finance Minister and Mr Williams, the Governor of the Central Bank, each stated in evidence that CLICO’s position was worse than it had previously appeared, and there is contemporaneous support for this in what was said during 2010 - even statements from members of the previous administration: see para 12 above. Further, in their evidence, Mr Dookeran and Mr Williams gave a number of reasons for this perceived deterioration, including the facts that CLICO was far more leveraged than had been appreciated, the value of some of its assets had fallen, the deficit in the Fund appeared larger than had been thought, redemption of EFPAs had been more widespread than anticipated, and the government’s injection of capital, which had been expected to be sufficient, did not avoid a shortfall - to which can be added the problem already mentioned in connection with the MHTL shares. The Board can see no reason to suspect that the Court of Appeal went wrong in accepting this evidence.

The appellants also contended that the government did not explain why the Fund could not be topped up, effectively on an ad hoc basis, in September 2010 (and indeed the following year). In so far as this contention would have involved the government amending the statutory provisions relating to the statutory funds, it seems to the Board that such a proposal could, to put it at its lowest, have been reasonably rejected by the government. It would have involved the rules relating to such funds being less prudent than they were - a recipe, many may think, for a revival of the very sort of problem that gave rise to the Assurances in the first place. In any event, there is no suggestion that the Central Bank would have approved any such amendments.

In so far as it is suggested that the government should have topped up the Fund, as and when it needed topping up, from its own moneys, the evidence was that this would have required a commitment over the years to inject around $7 billon, which gives rise to the sort of macro-economic/political considerations which are to be found in the remarks of the Prime Minister quoted in paras 20, 56 and 57 above. The government’s evidence before the Court of Appeal set out in some detail how difficult the future financial prospects of Trinidad and Tobago were in September 2010. Mr Dookeran explained how important it was that the IMF approved, as it duly did, his budget for 2010/2011 as “appropriately geared towards supporting the recovery”, and that the credit ratings agencies gave, as they did, a satisfactory assessment of the debt of Trinidad and Tobago. He said that there would have been “catastrophic consequences for the economy of Trinidad and Tobago” if all policyholders were “compensate[d] … fully”. The effect of his evidence was that considerations of this sort played a very prominent part in the government’s decision in September 2010 to abandon the Assurances and embark on the 2010/2011 policy. It is hard to see how the Court of Appeal could be criticised for accepting this evidence and giving it great weight in deciding whether it was reasonable for the government to abandon the Assurances.

Mr Knox suggested that the financial situation had not changed much between the time that the assurances were given in 2009 and September 2010. Even if that is true, the change of perception as to the national economic prospects, risks and priorities, owing in part to the passage of time and in part on a change of government is unsurprising. Such issues are ones in which any court should be very diffident indeed about interfering. In any event, as already mentioned in para 66 above, there was both significant contemporary evidence and significant authoritative sworn evidence in these proceedings to justify the Court of Appeal’s acceptance of the government’s case on this point.

The Board is also unimpressed with the appellants’ contention that it would have been cheaper for the government to abide by the Assurances to pay out the policyholders protected by the Fund as opposed to paying all policyholders, which is what it proposed. Ignoring the fact that this is an oversimplification of the government’s 2010/2011 policy when put forward in 2010, and when crystallised in 2011, it overlooks the fact that those proposals involve payments being made over a much longer period than would have been involved if the Assurances had been complied with. In any event, as explained in para 68 above, a very significant reason for abandoning the Assurances and adopting the 2010/2011 policy was to ensure IMF approval and the confidence of the international financial community, and that is by no means necessarily consistent with opting for the cheapest scheme.

It was also contended by the appellants that, when deciding to abandon the Assurances and to adopt the 2010/2011 policy, the government did not take into account the fact that the Assurances had created a legitimate expectation with all that that entailed. It is true that in the course of 2010 after the election, both the Prime Minister and Mr Dookeran denied that there had been “guarantees” to any policyholders, and on one occasion Mr Dookeran referred to the Assurances as “empty, hollow statement[s]”. A judge has to be rather careful of treating a statement made in a political context as if it was intended to have the meaning which it would have if it had been given under oath in court or in a legal document. Furthermore, it is technically right to say that, even if they gave rise to a legitimate expectation, the Assurances would have fallen short of being what a lawyer, or even someone engaged in commerce, would regard as a “guarantee”.

The notion that insufficient attention was given to the Assurances as a factor against adopting the 2010/2011 policy has a little more attraction. If, as the Board is assuming, there was a legitimate expectation in the light of the assurances, there is obvious force in the point that the government should have considered very seriously the fact that it would be reneging on the Assurances before embarking on the 2010/2011 policy. The history as summarised in paras 13 to 21 above does give rise to a concern whether the implications of departing from the Assurances were given proper weight. In the end, however, the Board is satisfied that they were.

First, the fact that the Assurances were swept aside in public does not mean that they were not given proper consideration by the government when making decisions. The evidence from the government, and in particular from Mr Dookeran, supports the contention that they were. He stated that the government “gave careful consideration to the expectations engendered by those statements [sic the Assurances] and the negative effect that might be suffered by those who had relied on those statements”. Of course, courts should not always take what is said in an affidavit or witness statement as accurate simply because it is sworn to be true. However, where the deponent is the Finance Minister, and there is no convincing evidence to contradict what he has said, then it seems to the Board that a tribunal, in this case, the Court of Appeal, is entitled to accept what he says as being accurate. It is true, as Mr Knox says, that Charles J took a different view, but, as explained above, the Court of Appeal was entitled to set aside her conclusions, and embark on its own assessment of the facts.

Secondly, and in any event, the proper weight to be given to a promise giving rise to a legitimate expectation before departing from it must depend on all the circumstances of the particular case. In this case, the Assurances had macro-economic (and therefore macro-political) implications, and it is clear from the evidence that the committee which came up with the three options and recommendation identified in para 15 above gave careful consideration to the implications of all possible courses, and it is also clear that the time pressures were considerable. In those circumstances, on the totality of the evidence, the Court of Appeal was certainly entitled to reach the conclusion that appropriate attention was given by the government to the existence and effect of the Assurances before they were abandoned.

The appellants also suggested that the 2010/2011 policy was unlawful in that (i) it required the Fund to pay out moneys at a time when it was in deficit contrary to section 37 of the Act, and (ii) it required the trustees of the Fund to pay out moneys without exercising their own discretion as to whether it should be paid out. In the Board’s view, point (i) overstates the effect of section 37, although it is a fair point that failure to make up a deficit in the Fund represents a breach of duty. But such a breach has long existed and would exist on any view. Further, especially in the light of that fact, any breach of section 37 could not invalidate the 2010/2011 policy. As to point (ii), a problem could possibly arise if the trustees refused to pay out in accordance with the 2010/2011 policy (assuming they would be entitled to do so), but it is hard to see how that possibility could render the policy unlawful. At worst, it represents a risk for the implementation of the policy, but, particularly as there has been no suggestion of the trustees taking such a course, it would appear to be a risk which the government could reasonably consider was one which was worth taking.

Another problem with the policy according to the appellants is that it was unlawful in discriminating between resident EFPA policyholders (who were to be paid 75% at least according to the appellants’ evidence, although the government say the figure is 92%) and traditional policyholders (who were to be paid in full). There is a difference between these two classes of policyholders, and therefore the Board has some doubt as to whether the discrimination argument is well based in any event. But even if it is, as Mr Stevens points out, the Central Bank, acting under the direction of the government, is given very wide powers to deal with companies under sections 44D and 44E of the Central Bank Act, and the Board considers that those provisions cover the implementation of the 2010/2011 policy.

Finally, the appellants relied on the fact that, as they saw it, the government had wholly unreasonably refused to supply them with information since September 2011. It is unnecessary at any rate at this stage to decide whether the factual basis for this contention is correct (as Charles J found that it was). Even if that were true, the Board cannot see how it assists the appellants in contending that the decision to abandon the Assurances was unjustified in law. At best, it might go to costs. In any event, once these proceedings had been started, the appellants could have asked the court to order disclosure of any documents which the court considered that they should see.

For these reasons the Board upholds the judgment of the Court of Appeal, and dismisses the appeal.

I agree that the appeal should be dismissed for the reasons given by Lord Neuberger. This is the second case before the Board in recent years concerning the law of legitimate expectation. In this, as in the previous case (Paponette v Attorney General of Trinidad and Tobago [2012] 1 AC 1), the parties have been content generally to adopt the law as stated in the judgment of Lord Woolf MR in R v North and East Devon Health Authority, Ex p Coughlan [2001] QB 213, without detailed argument. As Lord Neuberger explains, that has been sufficient for our decision in the appeal.

In the previous case, Lord Brown (Paponette at para 61) referred to the controversy, but felt unable in a dissenting opinion to conduct a detailed re-examination of the law. He noted, with evident approval, an “illuminating” article by Jack Watson (December 2010 SLS Legal Studies Clarity and ambiguity: a new approach to the test of legitimacy in the law of legitimate expectations, vol 30, No 4, p 633), in which the author had described the law as “a patchwork of possible elements to consider” rather than an organised system of rules, and “little more than a mechanism to dispense palm-tree justice”, and had called for reconsideration by the Supreme Court (para 61). This is not the occasion for detailed reconsideration, but it may be helpful in a concurring judgment to offer some thoughts as to the present state of the law.

The issues are discussed in detail in recent editions of several leading text-books, including Craig Administrative Law 7th ed (2012), chapter 22; De Smith’s Judicial Review 7th ed (2015), para 12-001ff; Wade and Forsyth Administrative Law 11th ed (2014), p 450ff. They have also been the subject of numerous academic articles (see the list of some 20 articles published between 1986 and 2011, in De Smith para 12-018 n 47). There is valuable up-to-date review of the cases and the academic commentary by Professor Mark Elliott From Heresy to Orthodoxy: Substantive Legitimate Expectations in English Public Law (Cambridge Legal Research Paper No 5/2016).

It seems that the expression “legitimate expectation” first appeared in the domestic case law in the context of procedural fairness (see eg Schmidt v Secretary of State for Home Affairs [1969] 2 Ch 149, 170F per Lord Denning MR). It was in that context also that it was recognised by the Privy Council in Attorney-General of Hong Kong v Ng Yuen Shiu [1983] 2 AC 629 (“the Hong Kong case”), where Lord Fraser said of the principle:

The possible extension of the same principle to substantive rather than merely procedural benefits emerged later in the law of England and Wales (De Smith at para 12-012). Two decisions are commonly cited in this connection: R v Secretary of State for the Home Department, Ex p Asif Mahmood Khan [1984] 1 WLR 1337 CA, and R v Secretary of State for the Home Department, Ex p Ruddock [1987] 1 WLR 1482 (Taylor J). Both concerned the application of published government policy criteria (relating respectively to entry clearance for children, and telephone tapping). In the former the criteria were equated to procedural benefits, applying the principles stated by Lord Fraser in the Hong Kong case (pp 1346-1347 per Parker LJ). In the latter, Taylor J (at p 1497) treated the principle as one of “fairness” not limited to procedural benefits, but he dismissed the claim on the facts.

The extension to substantive benefits remained controversial for some years: see eg R v Secretary of State for the Home Department, Ex p Hargreaves [1997] 1 WLR 906, 921, where the suggested departure from Wednesbury principles (as advocated by Sedley J in R v Ministry of Agriculture, Fisheries and Food, Ex p Hamble (Offshore) Fisheries Ltd [1995] 2 All ER 714)was described by the Court of Appeal as “heretical”. It was evidently for this reason that in Coughlan in 1999 the Court of Appeal (Lord Woolf MR, Mummery and Sedley LJJ) took the opportunity to undertake a comprehensive analysis of the authorities and the principles underlying them.

The principle has since been recognised as “well-established” in the House of Lords, but without detailed argument: see R v Ministry of Defence, Ex p Walker[2000] 1 WLR 806; R v Secretary of State for the Home Department, Ex p Zeqiri[2002] UKHL 3; [2002] Imm AR 296 para 44; R (Reprotech (Pebsham) Ltd) v East Sussex County Council [2003] 1 WLR 348 para 34. In R (Bancoult) v Secretary of State for Foreign and Commonwealth Affairs (No 2)[2008] UKHL 61; [2009] AC 453, para 60, Lord Hoffmann summarised the requirements of legitimate expectation, but again without detailed discussion:

That case concerned an alleged promise by the Secretary of State to the Chagos Islanders not to restrict their return to the islands from which they had been evicted for military purposes. The majority held on the facts there had been no such “clear and unambiguous” promise, and that in any event there was “a sufficient public interest justification for the adoption of a new policy in 2004”, having regard (inter alia) to the fact that “the decision was very much concerned with the ‘macro-political field’” (paras 61, 63, per Lord Hoffmann). Lord Carswell (also in the majority) noted that the principles governing “substantive legitimate expectation” as outlined in Coughlan had not yet been considered in depth by the House (para 133). Lord Mance (dissenting on this issue) recorded that counsel for the Foreign Secretary had reserved his right to argue in another case that Coughlan was wrongly decided (para 177).

The issues arising from Coughlan including its basis in legal principle,have been subject to significant discussion by Laws LJ in three subsequent judgments in the Court of Appeal: R v Secretary of State for Education and Employment, Ex p Begbie [2000] 1 WLR 1115; R (Nadarajah) v Secretary of State for the Home Department[2005] EWCA Civ 1363; R (Bhatt Murphy) v Independent Assessor (also cited as R (Niazi) v Secretary of State for the Home Department)[2008] EWCA Civ 755). As will be seen, Laws LJ made some critical observations on the reasoning in Coughlan and took the opportunity to develop in some detail his views as to the scope and content of the principle.

To set his observations in context, it is necessary to start by considering the factual background of the Coughlan case, the authorities on which the court relied, and how the principles so extracted were applied to the case.

The applicant was severely disabled. In 1993 she and seven other patients were moved from a hospital which the health authority wished to close to a house known as Mardon House. They agreed to the move on the basis of assurances by the authority that Mardon House would be their home for life. In 1998, in the light of new government guidance and following public consultation, the health authority decided to close Mardon House. It proposed to transfer responsibility for long-term nursing care of the applicant and her colleagues to the local authority, but without at that stage identifying any specific alternative accommodation. On her application for judicial review the judge quashed the decision on the grounds that she had been given a clear promise, and that no overriding public interest had been identified to justify breaking it. That decision was upheld by the Court of Appeal.

Giving the judgment of the court, Lord Woolf MR referred to the continuing controversy over the court’s role when a member of the public “as a result of a promise or other conduct, has a legitimate expectation that he will be treated in one way and the public body wishes to treat him or her in a different way” (para 56). It was for the court in such a case to determine the nature of the legitimate expectation, having regard to the precise terms of the representation, the circumstances in which it was made, and the nature of the statutory or other discretion; leading to one of “three possible outcomes”:

In respect of the second and third categories, he noted the difficulty of “segregating the procedural from the substantive” but added that most cases in the third category were likely to be cases where the expectation is “confined to one person or a few people, giving the promise or representation the character of a contract”. He recognised that the court’s role in relation to the third category was “still controversial”; but as he hoped to show “now clarified by authority”. (para 59)

The authorities on which he relied as supporting, or “clarifying”, the third category are apparent from the following discussion (paras 61-82). He relied in particular on the decision of the House of Lords in R v Inland Revenue Comrs, Ex p Preston[1985] AC 835, as applied in R v Inland Revenue Comrs, Ex p MFK Underwriting Agents Ltd [1990] 1 WLR 1545 and R v Inland Revenue Comrs, Ex p Unilever Plc [1996] STC 681. All three were tax cases, but Lord Woolf dismissed any suggestion that “special principles of public law” might apply to the Inland Revenue or to taxpayers. It is from the first of these cases that Lord Woolf evidently derived his reference to a promise “having the character of a contract”.

He referred to Lord Scarman’s concurring speech in Preston emphasising the place of “the principle of fairness” in the law of judicial review, so that “unfairness in the purported exercise of a power can be such that it is an abuse or excess of power” ([1985] AC 835, 851), and to similar passages in the leading speech of Lord Templeman. (Lord Templeman had also stated that it was only in “exceptional circumstances” that the court could decide to be unfair “that which the commissioners by taking action against the taxpayer have determined to be fair”: p 864F).

Lord Woolf quoted in full the final paragraph of this discussion (pp 866G-867C), leading to Lord Templeman’s conclusion that Mr Preston would have been entitled to relief by way of judicial review for -

In Preston, the taxpayer’s claim failed on the facts. No connection was made in the argument or speeches with the principles of legitimate expectation, as recently expounded in the Hong Kong case. That connection was made four years later in the second of Lord Woolf’s tax cases, Ex p MFK (1989). It was another unsuccessful taxpayer claim, this time based on a change of revenue practice. Giving the leading judgment of the Divisional Court, Bingham LJ spoke of his wish not to diminish the “valuable, developing doctrine of legitimate expectation”, but emphasised the need for the promise to be “clear, unambiguous and devoid of relevant qualification”, and said:

Unlike the other two, Lord Woolf’s third tax case, Ex p Unilever (1996), was a successful claim. It was based on the Revenue’s departure, without notice, from its long standing practice (over some 20 years) in dealing with the taxpayer’s tax computations. Applying the principles stated in Preston and MFK,the Court of Appeal accepted on the “unique” facts of the case that the Revenue’s action was “so unfair as to amount to an abuse of power” (p 4691G-H, per Sir Thomas Bingham MR).

Lord Woolf concluded his review of the authorities by emphasising the importance of the statutory context, and the need to “avoid jeopardising the important principle that the executive’s policy-making powers should not be trammelled by the courts” (citing Hughes v Department of Health and Social Security[1985] AC 766, 788, per Lord Diplock). He added that, policy being “for the public authority alone” would be accepted by the courts as “part of the factual data - in other words, as not ordinarily open to judicial review”. The court’s task was then limited to asking whether the application of the policy to an individual who has been led to expect something different is “a just exercise of power” (para 82).

In applying these principles to the facts of the case, the court noted the features of the “home for life” promise: “an express promise or representation made on a number of occasions in precise terms”, made to “a small group of severely disabled individuals …”, “specifically related to identified premises …”, and “in unqualified terms”. It had been made by the previous authority “for its own purposes” and “relied on by Miss Coughlan”. The court concluded:

Although the health authority had treated the promise as a starting point, and a factor to which they should give “considerable weight”, they had not considered it as “a legal responsibility or commitment to provide a home,as distinct from care or funding of care”. They had thus missed “the essential point of the promise”; their undertaking to fund her care for the remainder of her life was “substantially different in nature and effect from the earlier promise that care for her would be provided at Mardon House” (para 88, Lord Woolf’s emphasis). Further the quality of the alternative accommodation was as yet unknown. The court would not prejudge the effect of an offer of accommodation reasonably equivalent to Mardon House. “Absent such an offer, here there was unfairness amounting to an abuse of power by the health authority” (para 89).

As already noted, Laws LJ commented in later cases on the limitations of the Coughlan doctrine, and in particular on the need for caution in applying it in cases involving wider policy issues. The first was Ex p Begbie (2000), which concerned assurances allegedly given before a change of government and a consequent change of education policy, as to the continuation of the claimant’s assisted place at a particular school. The claim failed on the facts (see pp 1125-1127, per Peter Gibson LJ), but in a concurring judgment (pp 1129-1130), Laws LJ took the opportunity to comment more generally on the principles derived from Coughlan. He acknowledged the role of abuse of power as the “root concept” governing general principles of public law, but he saw difficulty in translating the root concept into “hard clear law”. In his view, the first and third categories explained in Coughlan were not “hermetically sealed”; the facts of the case, viewed in their statutory context, would steer the court to “a more or less intrusive quality of review”. He continued (in a passage relied on by the Court of Appeal in the present case):

He returned to the same theme in two later judgments, Nadarajah (2005) and Bhatt Murphy (2008). It is unnecessary to consider the facts or decisions in either case in any detail. Laws LJ’s purpose was, as he put it -to “move the law’s development a little further down the road …” (para 66ff), and -

In Bhatt Murphy (by contrast with Begbie and Nadarajah)he placed more emphasis on the differences between procedural and substantive expectation. The “core” distinction was that in the latter the court was concerned with a promise or practice “of present and future substantive policy” (para 33). This type of legitimate expectation was therefore concerned with an “exceptional situation” (citing Lord Templeman in Preston, p 864 - see para 93 above), because a public authority would not often be required by the law “to maintain in being a policy which on reasonable grounds it has chosen to alter or abandon”. He added:

Accordingly, authority showed that:

Before commenting on the present state of the law, I return to the decision of the Board in Paponette (Paponette v Attorney General of Trinidad and Tobago [2012] 1 AC 1) in which the judgment of the majority was given by Lord Dyson.

The claimants were members of the Maxi-Taxi Association operating in Port-of-Spain, Trinidad. Until 1995 they operated from a taxi stand which they controlled themselves and for which they paid no fee. In 1995 they reluctantly agreed government proposals to move them to the “City Gate” site, owned by the Public Transport Service Corporation (PTSC), the body which operated the country’s bus service. They did so in reliance on assurances by the Minister that they would not be under the control of the PTSC (regarded as a competitor), but would in due course be enabled to manage it themselves. Contrary to those assurances in 1997 the government decided to give management to the PTSC, and for this purpose promoted new statutory regulations, which also conferred on PTSC the power to grant permits and to charge for its use. The members of the association filed a constitutional motion in the High Court, claiming (inter alia) that the actions of the state had frustrated their legitimate expectations of a substantive benefit, in a way which affected their property rights protected under section 4(a) of the Constitution. They succeeded in the High Court but failed in the Court of Appeal.

On that issue Lord Dyson referred to the “convenient summary” given by Lord Hoffmann in Bancoult (cited at para 85 above). As to whether the representation was “clear, unambiguous and devoid of relevant qualification” he said the question was “how on a fair reading of the promise it would have been reasonably understood by those to whom it was made”. The members would reasonably have understood the representations as reassuring them that they would be able to continue to control and manage their own affairs if they moved, and in particular that they would not have to seek permits or pay a fee. It did not matter that there might have been “some uncertainty as to precisely what management entailed” (para 30).

On the “more difficult question” whether the government was entitled to frustrate the legitimate expectation so created, Lord Dyson referred to Coughlan, quoting Lord Woolf’s formulation of the test appropriate to legitimate expectation of a benefit “which is substantive, not simply procedural” (his category (c) see para 90 above): in short whether to frustrate the expectation is “so unfair (as to) … amount to an abuse of power”, weighing the requirements of fairness against “any overriding interest relied upon for the change of policy”. Lord Dyson noted that it was not in dispute that this was the test applicable to the case before the Board. The critical question was whether there was a sufficient public interest to override the legitimate expectation. This in turn raised the further question as to the burden of proof in such a case (paras 34-36).

The initial burden lay on an applicant to prove the legitimacy of his expectation, and so far as necessary his reliance on the promise. But once these elements had been proved, the onus shifted to the authority to justify the frustration, and to identify any overriding interest on which it relied (following Laws LJ in Nadarajah, para 68). It was then for the court “to weigh the requirements of fairness against that interest”.

In the present case the respondent had placed no evidence before the judge or the Court of Appeal to explain why the 1997 Regulations were made. Although some evidence had been admitted by the Court of Appeal, it failed to say why the 1997 Regulations were made despite the representations, nor did it indicate that in making them the minister took the representations into account (paras 37-40).

This review of Coughlan and the later cases reveals a striking contrast between, on the one hand, the relatively narrow scope of the actual decision in that case, and, on the other, the wide-ranging and open-ended nature of the legal discussion. Similarly wide-ranging has been the intense judicial and academic controversy, which in spite of the efforts of Laws LJ and others, remains unresolved more than 15 years later.

As the court emphasised in its application of legal principle to the facts, Coughlan concerned an express promise by the authority for its own purposes, made in unqualified terms to a small group of people with whom it had an established relationship, and relied on by them, and given for the specific purpose of persuading them to move out of premises which the authority wished to have available for other purposes. The promise was thus designed to promote the authority’s statutory functions, and keeping the promise would involve no inconsistency with them. As Laws LJ said in Begbie at p 1131,the facts in Coughlan were seen by the court as “discrete and limited” raising “no wide-ranging issues of general policy”. (In considering the court’s reasoning in Coughlan, it is irrelevant that this simple characterisation of the facts may be open to question: see Elliott op cit p 7.) There was also mutuality of commitment: a specific promise by the authority (a home for life) in return for specific action by the promisees (moving out). As Lord Woolf said in Coughlan at para 54, failing to honour it would indeed be “equivalent to breach of contract”.

With hindsight, it appears that the court in Coughlan may have been unnecessarily ambitious in seeking a grand unifying theory for all the authorities loosely grouped under the general heading of legitimate expectation. The first problem arises from the court’s apparent anxiety to assimilate the by then well-established principles relating to expectations of procedural benefits, with those relating to substance. This approach tended to underplay the important difference that, as Laws LJ emphasised in Bhatt Murphy,the latter involves potential conflict with the discretion of public authorities to “formulate and reformulate” policy (para 41). The court touched on this point at the end of its discussion, accepting that the formulation of policy was a matter for the executive, but it appears to have regarded its application to particular cases as no more than one aspect of the court’s assessment by reference to a broad test of justice or fairness (para 82).

Laws LJ by contrast saw it as “an exceptional situation”. In this he is supported by Wade & Forsyth (op cit p 460) who observe that it “involves the court in aspects of the decision-making process (eg the allocation of resources) generally considered beyond the bounds of judicial competence and authority”. The result, they say, is that the doctrine is to be “narrowly construed”, and in practice “only exceptionally” leads to a successful claim (as demonstrated by a long list of almost exclusively failed claims, in their footnote 432).

A related issue is the court’s reliance on Preston andthe other tax cases. It is of course true that the Revenue is not governed by special principles of public law. But those principles take effect in a special context. The Revenue has no free-standing policy-making function. So the potential for conflict with executive policy does not arise in the same way as in relation to most other government departments or local authorities. Tax policy is set by the legislature. The Revenue’s function is not to make the policy, but to collect the tax. It has a wide managerial discretion (see per Lord Templeman in Preston at p 862), and for that purpose may give general guidance or specific assurances to individual taxpayers. Even in that context, it is only in “exceptional circumstances” that the court will overrule the exercise of discretion by the commission (see para 93 above). As has been seen, it was the “unique” facts of Unilever which enabled the court to treat it as such an exceptional case.

That approach has been refined in more recent cases, where the giving, and honouring, of guidance or more specific undertakings by the Revenue has been seen as -

Outside the tax field, of course, published policy or guidance may have an equally important role, for example in relation to immigration. But there again the law has moved on since Coughlan. It is no longer necessary to find all the answers in the law of legitimate expectation. This accords with the view of Wade & Forsyth (op cit p 456) that -

Laws LJ was right in Begbie to contrast the narrow basis of the decision in Coughlan with cases at the opposite end of the spectrum involving much wider issues such as matters of national economic policy, exemplified by R v Secretary of State for the Environment, Ex p Hammersmith and Fulham London Borough Council [1991] 1 AC 521. In such cases, the court will not intervene outside of “the extremes of bad faith, improper motive or manifest absurdity” (R v Secretary of State for the Environment, Ex p Hammersmith and Fulham London Borough Council [1991] 1 AC 521, 596-597 per Lord Bridge of Harwich; see also Kennedy v Information Comr (Secretary of State for Justice intervening) [2014] UKSC 20; [2015] AC 455 para 53 per Lord Mance).

Laws LJ’s search for a constitutional foundation for the principle of legitimate expectation finds an echo in Professor Elliott’s recent article, where he suggests that “one of the key constitutional values at stake in such cases is the rule of law principle of legal certainty” (Elliott op cit p 16). He cites the recent judgment of the Supreme Court in Pham v Secretary of State for the Home Department (Open Society Justice Initiative intervening) [2015] UKSC 19; [2015] 1 WLR 1591, as indicative of “a more subtle understanding” (compared to the “traditional view” exemplified by R v Secretary of State for the Home Department, Ex p Brind [1991] 1 AC 696) of the way in which “constitutional and institutional concerns ought to shape the doctrinal architecture of substantive judicial review”. There the court accepted that given the “fundamental” status of the right to citizenship which was in issue “a strict standard of judicial review” was warranted as a matter of domestic law, which, for Lord Mance at least, was supplied by explicit adoption of the proportionality test (see Pham at para 98). On this view, says Elliott -

Comparison with the approach adopted in other common law jurisdictions is also revealing. In De Smith (para 12-081) there is a comparative review of the approach of the law to legitimate expectation in a number of countries (Australia, Canada, India, Ireland, New Zealand, South Africa) (para 12-081). None show support for the application of Coughlan or analogous principles in the context of substantive as opposed to procedural expectations. In a judgment in the Canadian Supreme Court, given shortly after Coughlan (Mount Sinai Hospital Center v Quebec (Minister of Health and Social Services)[2001] 2 SCR 281), Binnie J (in a concurring opinion agreed by McLachlin CJ) drew attention to the problems of a test based on “fairness”:

A narrower approach is also consistent with the Board’s decision in Paponette. Although the group involved was much larger than in Coughlan, there was similar mutuality of specific commitments. It involved a clear promise by the authority, made to a defined group in return for specific action by them within a defined time scale, and designed to further the authority’s own purposes. There was no argument that it raised wider political or economic considerations. The legitimate expectation having been established, it was for the authority to justify its departure from it, applying an approach which (although not so described by Lord Dyson) can in its practical effect be equated with a proportionality test (as proposed by Laws LJ and Professor Elliott).

In summary, the trend of modern authority, judicial and academic, favours a narrow interpretation of the Coughlan principle, which can be simply stated. Where a promise or representation, which is “clear, unambiguous and devoid of relevant qualification”, has been given to an identifiable defined person or group by a public authority for its own purposes, either in return for action by the person or group, or on the basis of which the person or group has acted to its detriment, the court will require it to be honoured, unless the authority is able to show good reasons, judged by the court to be proportionate, to resile from it. In judging proportionality the court will take into account any conflict with wider policy issues, particularly those of a “macro-economic” or “macro-political” kind. By that test, for the reasons given by Lord Neuberger, the present appeal must fail.