McPhail v Doulton revisited: Shades of Uncertainty*
Moshiuzzaman**
This paper seeks to revisit Lord Wilberforce’s infamous statement from the McPhail v Doulton case vis-à-vis the test for ascertaining the beneficiaries of a trust. The paper is broken down into three segments which follows discussion and distinction between pairs of fixed and discretionary trusts and, then pairs of discretionary trusts and powers of appointment. In addition, the article further provides a comprehensive and well-resourced argument for the application of the “Is or is not” test for discretionary trusts considering the prior discussion.
The paper concludes that discretionary trusts are wholly different from fixed trusts because they profess different requirements provided by the settlor. As such, the ‘complete list’ is better suited for fixed trusts where the people are limited and ascertainable, particularly in the case of familial situations. Conversely, despite sharing a subtle difference with powers, discretionary trusts by themselves, do not segregate themselves completely from powers. A discretionary trust and a mere power share a lot more in common than they do apart, as such the test that was originally taken for powers from Re Gulbenkian could be adapted for a discretionary trust, as expressed by Lord Wilberforce himself.
For a valid private[1] express trust to exist, three certainties must be met,[2] namely, the certainty of intention, the certainty of subject-matter and the certainty of objects;[3] absence of any of these three certainties would render the trust void.[4] As a trust imposes a legal obligation on the trustees to enforce the trust, the parties who are to receive the benefit, must be ascertained. For this reason, we must turn to the last limb of the three certainties; certainty of objects – the beneficiaries. To ascertain the beneficiaries however, depending on the nature of the trust or power in question, different tests are applied.
This paper, therefore, considering Lord Wilberforce’s statement[5] in McPhail v Doulton[6] seeks to determine whether the tests for ascertaining objects should be the same between pairs of fixed and discretionary trusts and pairs of discretionary trusts and powers of appointments. And finally, this answer will provide an explanation for the application of the ‘is or is not test’ for discretionary trusts in light of the two discussions prior to it. This discussion thus will be done in three folds.
- Should the test of certainty of objects be the same for fixed and discretionary trusts?
In order to address the first part of the question, it is of utmost significance to distinguish between the fixed trust and the discretionary trust. In a fixed trust, trustees are provided with instructions about the proportions in which the trust property would be divided amongst the beneficiaries;[7] the trustees have no discretion on who gets what and at what proportions they get it in. Conversely, in a discretionary trust, also known as ‘trust powers’,[8] trustees are required to use their discretion to execute their trust obligations;[9] they reserve the right to divide and quantify the trust property and distribute the property as they please.
As the obligations of the trustees differ in manner, the test that determines the way the distribution takes place is subsequentially different. As such the difference is crucial. In a fixed trust, the trustees have no discretion to distribute the trust property, therefore everyone that holds a beneficial interest must be identified.[10] This method of identification is referred to as the ‘complete list’ test.[11]
On the other hand, in the case of discretionary trusts, the trustees have the right to distribute the trust property in whichever manner they choose. As such, all the beneficiaries need not be identified altogether at once,[12] although those seeking to exercise a beneficial interest of the trust property must be ascertainable ‘if and when…’ they choose to exercise their beneficial interest.[13] For this purpose, the ‘is or is not test’[14] has been adopted for discretionary trusts.[15] Of course, it should be noted that prior to IRC v Broadway Cottages,[16] the test for discretionary trust was the same as for fixed trusts – the ‘complete list’ test. It was not until Lord Wilberforce’s judgement in McPhail[17] that the position had changed.
This leads to Lord Wilberforce’s judgement in McPhail v Doulton. In that case, clause 9(a) of the trust instrument had provided that the trustees of the trust will exercise their absolute discretion to benefit the officers and employees, and ex-officers and ex-employees of the company and any of the relatives and dependents[18] with the net income of the fund in question. The problem arose with the terms ‘relatives’ and ‘dependents’, because it was uncertain as to who the relatives and/or dependents were.[19] By a majority[20] it was held that clause 9(a) of the trust instrument was a trust and not a power and the test for determining certainty of beneficiaries was as provided in Re Gulbenkian;[21]the trust would be valid if it could be said with certainty that any individual was or was not a member of the class.[22]
Lord Wilberforce assessed whether the ‘complete list’ test from Broadway Cottages was the right test to be applied or the ‘is or is not’ test for powers as provided in Re Gulbenkian.[23]His assessment led him to conclude that the ‘complete list’ test was not the appropriate test to be applied to McPhail.[24]
Lord Wilberforce argued on various points regarding the non-applicability of the ‘complete list’ test to discretionary trusts in, but most notably he reiterated Jenkins LJ’s words from Morice.[25]He echoed that should the trustees fail to comply with their duties, the courts must be able to execute the trust themselves. But he argues on this and says, it does not imply that execution would be impossible without equal division.[26]Furthermore, he points to fact that the ‘complete list’ test is better suited and thus best applicable to fixed trust where the trustees are identifiable through the trust instrument and are usually limited to family[27] cases where the beneficiaries are smaller as opposed to discretionary trusts which have a wider class of beneficiaries.[28]
He proceeded by making an important distinction. In Broadway Cottages, he stated that the ‘complete list’ test applied not only to a fixed trust of capital, which involved equal division but also, a discretionary trust of income.[29]The Court of Appeal’s contention that a trust income could only be executed based on equal division was rejected by Lord Wilberforce.[30]He rejected all arguments that postulated the application of the ‘complete list’ test based on Broadway Cottages for discretionary trusts.[31]
As per the aforementioned discussion, it is quite evident that a fixed trust and a discretionary trust are fundamentally different. Although, they create an obligation to distribute the trust property for the benefit of the beneficiaries, the certainty of the who the beneficiaries are depend primarily on the wording of the settlor or testator and the nature of the trust itself. It has been seen that a fixed trust is better applicable for a family context where the beneficiaries are limited and hence identifiable from the face of the trust instrument, as opposed to a commercial context with a wide a range of beneficiary. Thus, it can be concluded that the test for the two trusts are different and should not be made the same.
- Should the test of certainty of objects be the same for discretionary trusts and powers of appointment?
As we have already discussed a trustee bears the obligation[32] to distribute the trust property as stated in the trust instrument. This is the same for both a fixed and a discretionary trust. A similar obligation might arise under a power of appointment. Essentially, a power is an authority that is vested to a person to deal with or dispose of property which is not his own.[33]However, as Pettit[34] has correctly noted, points to a fundamental difference between the two. He states that a trust is imperative, while a power is permissive[35] or discretionary.[36] This infers that a trust provides an obligation that must be carried out whereas a power is an obligation that is not necessarily mandatory to be carried out,[37] as they are ancillary to the implementation of the grantor’s wishes.[38]
The next questions that comes is how the trustees’ duties are ensured to be properly administered. This is done by the courts.[39]Trustees with identifiable beneficiaries in a trust can be supervised by the courts and where the trustees to fail to act in accordance to the trust instrument,[40] the courts can always enforce the trust themselves.[41]This rule is equally applicable for powers too.[42] Although, typically there is no redress against an appointee under a power of appointment who disregards to carry out his permissive obligations, unless there has been excessive or fraudulent appointment.[43]
This leads to some important distinctions that need to be drawn from the two instruments. Spearing,[44]argues that the beneficiary of a discretionary trust is in a better position than the object(s) of a power of appointment.[45] He reasons, that the purpose of a discretionary trust is flexibility,[46]which can only be achieved if the settlor endows his trustees with a power of disposition over the trust property.[47]He continues further, and states that a well drafted trust instrument will bear a combination[48] of a discretionary trust with powers of appointment and powers to include beneficiaries or objects to facilitate maximum benefits from the available fund and to distribute in accordance with the intentions of the settlor.[49]This would naturally imply that unless the trustees exercise their discretion, no beneficiary can be sure of their beneficial entitlement.[50]
This part of the essay will now return to Lord Wilberforce’s judgment in McPhail v Doulton where his Lordship extended the rule of ascertaining beneficiaries of mere powers adopted from Re Gulbenkian’s Settlements[51] to trust powers i.e. discretionary trusts. Currently, the test for ascertaining beneficiaries for discretionary trusts is the ‘Is or is not’ test. It is no longer necessary to determine all the objects of the class in order to execute the trust.[52]Donnell, argues that the Lord Wilberforce’s reasoning and the decision reached in McPhail was ‘pragmatic but not clear’.[53]
John Hopkins states that most leading judges do not consider there to be a ‘fundamental’ difference between a power and trust.[54]In Re Gulbenkian, Lord Denning M.R., stated that trust claims should be ‘brought into line’[55] with powers.[56]When this case proceeded to the House of Lords, Lord Reid added to this point[57] and stated that ‘donees’[58] often owe a fiduciary duty in application of powers.[59]Donnell contends that these opinions point to the distinction between a trust power and mere power as an artificial construction of the law.[60] In fact Lord Wilberforce, himself had called it ‘artificial’.[61]These commentaries from the McPhail[62]seem to suggest that distinction is artificial when the case concerns a discretionary trust.[63]With the superadded fact that the distinction was never ‘fundamental’, it demonstrates how the Lord Wilberforce’s remarks in McPhail were pragmatic[64] and yet indecisive.[65]
Donnell proceeds to argue the case based on equal distribution.[66]Referring to John Hopkins again, he explains that argument of ascertainability will only lead to enforceability by the court which is robust.[67] And it is sensible argument indeed. However, Palmer reasons that the discretion bestowed upon the trustees are for a specific purpose;[68] allowing the courts to execute the trust by equal division would act contrary to the intent of the settlor.[69]These arguments however, opposes Lord Wilberforce’s directions in McPhail. Indeed, his Lordship had stated that distribution does not have to be in equal shares. Given the right circumstances arise, he argues, the courts can always distribute the shares that is not equal[70] as per the perceived intention of the settlor.[71]
Lord Wilberforce towards his conclusion cites, Warbuton v Warbuton.[72]Using this he argues that the assumption of equal distribution was untrue,[73] and unequal distribution is nothing unprecedented or ‘revolutionary’.[74] In light of the discussion above, it seems evident that although there is a subtle difference between trust powers and mere powers and their tests’, there is no strong nor conclusive evidence that can suggest that the two instruments should have distinct tests. Although, if opted for, it can be made the same.
- In light of your answers to (i) and (ii), do you think “is or is not test” as discussed in McPhail v Doulton is the appropriate rule for discretionary trusts?
This part of the essay will seek to verify whether the ‘is or is not test’ taken from McPhail v Doulton is the correct test for determining discretionary trusts. Previously, we have discussed that a fixed and discretionary trust are fundamentally different and therefore their tests are inherently required to be different. On the contrary, a mere power and trust power is not ‘fundamentally’ distinct therefore they do not require their tests to be different, although if they are contested for the same status, there is nothing that bars them from being the same. Consequentially, now we consider the “is or is not test” and it application on discretionary trusts.
Having decided that the ‘is or is not test’ is the correct test for discretionary trusts, the House of Lords in McPhail remitted the case to the High Court’s Chancery Division, which on appeal became Re Baden’s Deed Trusts (No 2).[75]However, the application of it turned to be more difficult than expected.[76]The ‘class test’, the ‘any given postulant test’, or the ‘is or is not test’ dictates whether a given individual can be said with certainty to belong to a class.[77] But this lead to different interpretations by the three judges of the Court of Appeal.
Although, Penner states that there was no difficulty in ascertaining the meaning of ‘dependents’[78] there was considerable debate with the meaning of ‘relatives’.[79] He stated that ‘relatives’ meant having a descendent from a common ancestor. However, as he would put it, everyone’s lineage, at least at a scientific and practical level can be traced backed the ‘original Australopithecines’.[80] Of course, there was nothing wrong with this because it was conceptually certain. However, problem arose when proving the relation – evidential uncertainty.[81]
Stamps LJ refused to allow evidential certainty to rule out the problems posed by conceptual certainty.[82]Penner, argues that it cannot water down the test to a dependent upon a burden of proof.[83]Indeed, as his Lordship stated that it was not enough for trustees to distribute funds among beneficiaries who happen to be present or present themselves.[84]However, he found an authority[85] to support ‘relatives’ to mean ‘next of kin’ and thus held the trust to be valid.[86]
On a different prospective, Megaw LJ stated that a ‘substantial’[87] number of people need to be surveyed to determine whether are ‘in’ or ‘out’ of the class of beneficiaries.[88] So, blatantly his Lordship rejects the need for holding a difference between conceptual and evidential certainty. This however, is a problem because the meaning of ‘substantial’ is a debate itself.[89] Consequently, Megaw LJ sided with Sach LJ’s definition.
This then leads to Sach LJ’s interpretation. Unlike Megaw LJ, Sach LJ recognizes the clear distinction between conceptual and evidential certainty.[90] He does not refuse the need for conceptual certainty[91]while simultaneously not limiting it to evidential certainty.[92]Quite simply, Penner explains that it is a question of fact whether someone ‘is’ or ‘is not’ a member of a class. The availability of evidence does not disregard the trust instrument. This fit ‘perfectly’[93] to the criteria of ‘relatives’ as meaning ‘a descendent from a common ancestor’.[94]
As discussed above, discretionary trusts are wholly different from fixed trusts because they profess different requirements provided by the settlor. As such, the list test is better suited for a fixed trust where the people are limited and can be ascertained, particularly in the case of familial situations. Conversely, despite sharing a subtle difference with powers, a discretionary trust by itself, does not segregate itself completely from a power. A discretionary trust and a mere power share a lot more in common than they do apart, as such the test that was originally taken for powers from Re Gulbenkian can be easily be adopted for a discretionary trust, as expressed by Lord Wilberforce.
* This paper was submitted as part of the author’s LL. B programme at BPP University (UK). Some alterations have been made to the paper to include further comprehensive details.
** Moshiuzzaman, is final year law student pursuing his LL. B from BPP University (UK) via a distance learning programme from London College of Legal Studies (South), Dhaka (Bangladesh).
[1] Apart from charitable trusts; A.J. Oakley, Parker and Mellows: The Modern Law of Trusts (9th edition, Sweet & Maxwell 2008) p – 58, para – 3-029
[2] Knight v Knight (1840) 3 Beav 148 at 173 as per Lord Longdale; Jill E Martin, Hanbury & Martin Modern Equity (12th edition, Sweet & Maxwell Publication 2015) p 79
[3] Jill E Martin, Hanbury & Martin Modern Equity (12th edition, Sweet & Maxwell Publication 2015) p 79
[4] In Morice v Bishop of Durham (1805) 10 Ves 522, it was held that a trust must be enforceable by the court to be valid; Alastair Hudson, Equity and Trusts (6th edition, Routledge-Cavendish Publications 2010) p 116
[5] McPhail v Doulton [1971] AC 424 at 451
[6] [1971] AC 424
[7] Graham Virgo, The Principles of Equity & Trusts (2nd edition, OUP 2016) p – 102, para – 4.4.4
[8] Jill E Martin, Hanbury & Martin Modern Equity (12th edition, Sweet & Maxwell Publication 2015) p 89 paras – 4-010
[9] Alastair Hudson, Equity and Trusts (6th edition, Routledge-Cavendish Publications 2010) p 124
[10] Graham Virgo, The Principles of Equity & Trusts (2nd edition, OUP 2016) p – 102, para – 4.4.4
[11] Inland Revenue Commissioners v Broadway Cottages Trusts [1955] Ch 20 at 29 as per Lord Jenkins.
[12] Paul Matthews, A Heresy and a Half in Certainty of Objects [1984] Conv 22 at 27
[13] ibid
[14] This test was formulated by Lord Denning MR in Re Gulbenkian [1968] Ch 126 at 134
[15] Alastair Hudson, Equity and Trusts (6th edition, Routledge-Cavendish Publications 2010) p 126
[16] Inland Revenue Commissioners v Broadway Cottages Trusts [1955] Ch 20
[17] See supra note 6
[18] McPhail v Doulton [1971] AC 424 at 447 as per Lord Wilberforce
[19] Alastair Hudson, Equity and Trusts (6th edition, Routledge-Cavendish Publications 2010) p 156
[20] McPhail v Doulton [1970] 2 All ER 228 at 235 as per Lord Hodson (dissenting)
[21] In re Gulbenkian’s Settlement [1970] AC 508
[22] This is or is not test.
[23] McPhail v Doulton [1971] AC 424 at 450 as per Lord Wilberforce
[24] ibid
[25] Morice v Bishop of Durham (1805) 10 Ves 522 as per Jenkins LJ
[26] McPhail v Doulton [1971] AC 424 at 451 as per Lord Wilberforce
[27] ibid
[28] ibid n.26
[29] McPhail v Doulton [1971] AC 424 at 454 as per Lord Wilberforce
[30] ibid
[31] McPhail v Doulton [1971] AC 424 at 455 as per Lord Wilberforce
[32] This cannot be delegated to third parties or another trustee. As stated per s.11 of Trustee Act 2000; Michael Haley and Lara McMurtry, Equity & Trusts (3rd edition, Sweet & Maxwell 2011) p 425, para – 12.34
[33] Freme v Clement (1881) 18 Ch D 499; Philip H. Pettit, Equity and the Law of Trusts (12th edition, OUP 2012) p 33
[34] Philip H. Pettit, Equity and the Law of Trusts (12th edition, OUP 2012)
[35] F.E. Spearing, Discretionary Trusts & Power: Distinction without a difference, 7 Trent LJ (1983) p 108
[36] Ibid n.34 p 33
[37] Ibid n.34
[38] F.E. Spearing, Discretionary Trusts & Power: Distinction without a difference, 7 Trent LJ (1983) p 107
[39] Dennis R. Klinck, McPhail v Doulton and Certainty of Objects: A Semantic Criticism, 20 Ottawa L. Rev. (1988) at p 379-380
[40] Morice v Bishop of Durham (1805) 10 Ves 522 at [539]-[540] as per Lord Eldon L.C.
[41] ibid as per Jenkins LJ
[42] A ‘power of appointment’ will only be valid if the objects – beneficiaries of it can be ascertained with certainty, as in re Gulbenkian’s Settlement [1970] AC 508
[43] F.E. Spearing, Discretionary Trusts & Power: Distinction without a difference, 7 Trent LJ (1983) p 108
[44] Ibid
[45] Ibid n.43
[46] Re Manisty’s Settlement [1974] Ch 17
[47] F.E. Spearing, Discretionary Trusts & Power: Distinction without a difference, 7 Trent LJ (1983) p 108
[48] McPhail v Doulton [1971] AC 424 at 449 as per Lord Wilberforce
[49] F.E. Spearing, Discretionary Trusts & Power: Distinction without a difference, 7 Trent LJ (1983) p 109
[50] Re Bibby (J) & Sons Trust Deed, Davies v I.R.C. (1952) W.N. 402; F.E. Spearing, Discretionary Trusts & Power: Distinction without a difference, 7 Trent LJ (1983) p 109
[51] [1970] AC 508 (HL)
[52] Benjamin Donnell, McPhail v Doulton and the Fundamental Distinctions between Trusts and Powers 4 N.E.L Rev. (2016) p 52
[53] Ibid
[54] John Hopkins, ‘Trust – Discretionary Trusts – Certainty of Objects’ [1970] 28 CLJ 206, 207
[55] [1968] Ch 126 at 133B as per Lord Denning M.R. sitting in the Court of Appeal.
[56] Ibid n.51
[57] [1970] AC 508 (HL) at 518
[58] Often referred to as trustees in trust instruments.
[59] ibid n.51
[60] ibid n.51
[61] [1971] AC 424 (HL) at 448
[62] The case concerned discretionary trusts.
[63] ibid n.51
[64] McPhail v Doulton [1971] AC 424 at 451 as per Lord Wilberforce
[65] The application of the ‘Is or is not’ test in Re Baden (No 2) [1972] EWCA Civ 10 by the three judges Sachs, Stamps and Megaw LJJ demonstrate how wide and distinct the test can be depending on the interpreter.
[66] Benjamin Donnell, McPhail v Doulton and the Fundamental Distinctions between Trusts and Powers 4 N.E.L Rev. (2016) p 53
[67] John Hopkins, ‘Certain Uncertainties of Trust and Powers’ [1971] 29 CLJ 68, 69; Benjamin Donnell, McPhail v Doulton and the Fundamental Distinctions between Trusts and Powers 4 N.E.L Rev. (2016) p 53
[68] George Palmer, ‘Private Trusts for Indefinite Beneficiaries’ (1873) 71 Mich L Rev 359, 366
[69] Benjamin Donnell, McPhail v Doulton and the Fundamental Distinctions between Trusts and Powers 4 N.E.L Rev. (2016) p 53
[70] McPhail v Doulton [1971] AC 424 at 452 as per Lord Wilberforce
[71] Benjamin Donnell, McPhail v Doulton and the Fundamental Distinctions between Trusts and Powers 4 N.E.L Rev. (2016) p 54
[72] Warbuton v Warbuton 4 Bro.P.C. 1
[73] This is contrasting to the decision reached in Richardson v Chapman 7 Bro.P.C. 318 as noted by Lord Wilberforce in McPhail v Doulton [1971] AC 424 at 451
[74] Benjamin Donnell, McPhail v Doulton and the Fundamental Distinctions between Trusts and Powers 4 N.E.L Rev. (2016) p 54
[75] [1973] Ch 9
[76] Michael Haley and Lara McMurty, Equity & Trusts (4th edition, Sweet & Maxwell 2014) p 72 paras – 2.51
[77] Ibid
[78] Dependents were not regarded as uncertain – Parliaments and other deeds had described the individuals as financially dependent upon other for their support; J E Penner, The Law of Trusts (10th edition, OUP 2016) p 213, para – 7.63
[79] J E Penner, The Law of Trusts (10th edition, OUP 2016) p 213, para – 7.63
[80] Ibid
[81] Ibid n.79
[82] J E Penner, The Law of Trusts (10th edition, OUP 2016) p 213, para – 7.66
[83] Ibid p 214, para – 7.66
[84] Re Baden No 2 [1973] Ch 9 at [27A]
[85] Harding v Glyn [1739] 1 Atk. 469
[86] Re Baden No 2 [1973] Ch 9 at [29A]-[29B]
[87] Re Baden No 2 [1973] Ch 9 at [24C]-[24D]
[88] J E Penner, The Law of Trusts (10th edition, OUP 2016) p 213, para – 7.65
[89] Ibid
[90] J E Penner, The Law of Trusts (10th edition, OUP 2016) p 213, para – 7.63
[91] Re Baden No 2 [1973] Ch 9 at [19H]-[20A]
[92] Re Baden No 2 [1973] Ch 9 at [19H]
[93] J E Penner, The Law of Trusts (10th edition, OUP 2016) p 213, para – 7.64
[94] Ibid