INTRODUCTION

Historically, federal tax legislation has always been based on the common law.  An example is the deemed disposition of trust assets for tax purposes every 21 years, which is a direct reference to the common law rule against perpetuities for trusts.[1]

Now that in Quebec a trust established by contract by and large corresponds globally to the common law concept of an express trust in terms of its effects, it must be determined whether references to the common law trust in the Income Tax Act[2] are neutral enough to encompass the distinctive features of the civil law trust. Where they are not, the impact of the discrepancies will be analysed and solutions proposed.

Before embarking on the technical study of these differences, we must indicate the approach that will be taken in our analysis of the dissimilarities between the common law and the civil law, although there may be some exceptions.

Amend the tax laws, not the applicable private law

The goal of the exercise is not to amend the tax laws in a way that would make Quebec residents subject to the common law rules, or vice versa. Where the tax treatment diverges, an effort, ideally, should be made to respect the civil law rather than amend it so that it operates like the common law. The principle of respecting the civil law can, on an exceptional basis, be stretched somewhat. One example is the technique applied in paragraph 248(3)(d) for the recognition of certain arrangements such as self-directed RRSPs. That raised some doubts in Quebec's civil law but were acceptable in the other provinces. While the purpose of that paragraph is laudable, its effect is to distort Quebec's civil law insofar as it encourages recourse to common law rules in Quebec rather than making adjustments to the tax provisions so they will work in harmony with the civil law.

Major and unavoidable differences

There are differences between the common law trust and the civil law fiducie that cannot be resolved through tax law. 

Me Jean Charles Hare[3] has identified four main differences between the civil law fiducie and the common law trust:

On the nature of the trust, this author points out the difference between the dual ownership of property held in the common law trust comparatively to the civil law trust patrimony over which no one has ownership.

Cumulation of roles is not permitted by the Civil Code of Québec,[4] since article 1275 C.C.Q. requires the appointment of a trustee who is neither the settlor nor a beneficiary. Although this difference means that an additional trustee must be involved in the trust, a prerequisite that is unnecessary under the common law, it generally does not lead to a different tax treatment for Quebec trusts because of subsection 75(2) I.T.A., among others. In addition, the use of substitution, which tax law treats as a trust, can be considered. With substitution, no trustee other than the institute need be appointed to administer the property.

With respect to the power to appoint, Me Hare writes: [TRANSLATION] "the settlor may give the trustee unlimited power to add or remove trust beneficiaries," which is not permitted under the civil law. This would appear to be a difference that does result in a distinctive tax treatment between Quebec and the other provinces.

Finally, Me Hare addresses the difference in the procedures for the variation and termination of a trust depending on the jurisdiction. Although, in principle, such variations and terminations can be effected on the mere consent of the beneficiaries, those modifications are generally subject to specified requirements enacted by provincial laws. Then, the complexity of those requirements have to be considered when a comparison is made between common law and Quebec civil law, the latter providing for an application to the court in such circumstances.

Other examples of differences in civil law include the prohibition against appointing a corporate trustee other than a trust company,  the absence of a Protector, the absence of an extended case law on trusts and the impossibility of using  a declaration of trust.

This study will not propose changes for most of the above elements since they are a matter of private law and  will evolve along with it. Intervention would be warranted only where a specific difference resulted in a different tax treatment.

Forum shopping

When completed, the current  harmonization will not alter the appeal of finding the provincial private law system with the most advantageous trust legislation. On considering the differences between the civil law trust and the common law trust and the differences among provinces because of their respective trust legislation, some taxpayers will still opt for the law of Quebec, which offers trustees, for example, more security in terms of their personal liability or, on the other hand, the common law trust, which permits the settlor to appoint himself as trustee.[5] However, because harmonization seeks to resolve the differences in tax treatment, at least at the federal level, the choice of jurisdiction should no longer be motivated by such differences.

Avoid any tax difference?

Before proposing amendments to standardize the federal tax treatment of trusts all across Canada, it is essential to ascertain whether it is worthwhile to try to completely eliminate all of the tax advantages or disadvantages arising from different rules from one jurisdiction to another. In our opinion, harmonization should be applied primarily to cases where the legal situation created is essentially the same but, because the concepts are different, their tax treatment is different. This action would then be in line with the prevailing case law favouring the uniform application of tax legislation notwithstanding the applicable private law.

However, harmonization has its limits  because some tax differences are the result of specific rules that are not found in the other legal system. This is true of constructive trust which results from a common law judge's interpretation of a past situation. Although methods can be proposed to limit the tax impact of such differences, constructive trusts and the resulting tax advantages are not available to Quebec residents although less spectacular results may be obtained through the court's acknowledgement of a tacit partnership.[6]

Quebec residents can use substitution to obtain trust tax treatment without the presence of an actual trustee. As it happens, this distinctive feature is the result of the federal government's attempt to adapt its tax legislation to an institution specific to Quebec by applying the tax rules for trusts, a related concept. The result is interesting since the concept of substitution, which had been ignored by Quebec tax specialists, can once more be utilized. Substitution is clearly a concept specific to Quebec, however, and the concomitant tax advantages and disadvantages are not applicable anywhere else in Canada.

When the difference in tax treatment between the two legal systems arises from features specific to one of the systems, it may be impossible to achieve perfect uniformity, although efforts can be made to mitigate the differences.

 Reduce the number of references to concepts that are foreign to the civil law

When faced with legal concepts that are foreign to their jurisdiction, civil law lawyers and notaries need solutions that make the application of the tax rules clearly understandable so that they are not forced to analyse the common law to be able to apply the rules. The concept of "vested indefeasibly" in subsection 70(6) I.T.A. and that of "revert" in subsection 75(2) I.T.A. could be defined, for example. These definitions could be made applicable to the civil law as well as the common law, thus obviating reference to the common law as a complementary source for the provisions. That solution could lead to more differences, however, if the definition of a common-law concept did not correspond to the definition made by that jurisdiction's courts.

Avoid adding to the complexity of the law 

In analysing the proposed solutions, a sustained effort must be made to avoid adding to the complexity of the law. This complexity is a source of differences and unfairness for Canadian taxpayers because of differences in the means available to them to access accurate tax information. To increase the complexity of the law in order to make it horizontally uniform across the whole country would lead to greater disparities in its vertical application across the various classes of taxpayers. 

Impact of taxation on trust agreement

Having set the parameters within which we shall work and in our discussions of problems and solutions, we must be aware of the extent to which tax law influences the law of trusts and the fact that the two areas of law are intrinsically related. That influence is one more reason why it is necessary to harmonize tax legislation whereas, if tax policy were still neutral in terms of deciding whether or not to set up a trust, things would be different.

Far from being neutral, tax rules even impose the framework in which a trust is legally constituted and in so doing occasionally misrepresent the civil law perspective on the issue. Some trusts in fact are directly derived from tax legislation or other public laws.

The following lists some of the elements that by and large derive from tax legislation:

• Choice of settlor (reasons: subsection 75(2), attribution rules, preferred beneficiary election);
• Choice of trustees (reasons: subsection 75(2), Canadian residence) ;
• Choice of property transferred (reason: the resulting tax disposition);
• Choice of beneficiaries (reasons: benefit to shareholder, attribution rules);
• Methods for distributing income and capital (reasons: 21-year rule, presence of a preferred beneficiary or a spouse in a trust exclusively for his or her benefit, requirements to render amounts payable).

Such elements, however, play a fundamental role in a decision to set up a trust, but for tax reasons they must often be articulated differently from what the parties to the trust deed had intended at the outset.

That influence is not peculiar to civil law. The constitution of a common law trust is also subject to tax considerations that must be worked out in terms of the applicable legislation.

Analysis and proposals

Taking the above into consideration, our task is now to analyse the specific elements of the law (other than the concepts of "beneficial" and "legal ownership", which are treated in another study) that raise problems of interpretation when they are applied in a civil law context.

For each tax problem that is identified, one or more solutions will be proposed that could standardize the tax treatment of trusts in Canada while respecting the private law of the provinces.

1.  Problems arising from the interaction between the civil code and the new rules for self-benefit trusts

In 1994, the Civil Code introduced a redefined trust concept that allowed trusts to be used in Quebec to protect assets from creditors or a bad administration, although the nature of this trust is not making unanimity.[7]

To obtain such protection, a person has only to transfer some of his property to the trust in order to take it out of his own patrimony. However, the degree of protection sought will vary depending on the terms of the trust deed. To achieve protection from creditors, the safest trust would appear to be one that allows the trustees, at their discretion, to distribute the trust income and capital among a number of beneficiaries, including the settlor, so that the transferor's interest in the trust would be of no value as long as the discretion of the trustees had not been exercised in his favour.[8] Differences between common law and civil law do not particularly affect the way in which the trust is used. The disadvantage of this kind of trust for tax purposes, however, is that a deemed disposition is triggered when the property is transferred to the trust.

The trust that is the object of the present chapter offers a less greater protection against the creditors but its advantage is that it can receive property without tax consequences.

1.1.  Self-Benefit trust

To avoid a disposition for tax purposes, the trust must be created for the exclusive benefit of the transferor, which in this study will be referred to as a "self-benefit trust" (or fiducie pour soi in French). In this kind of trust, the assets are not directly available to the transferor's creditors but are sheltered in the trust patrimony. If the creditors cannot attack the transfer, which may have been made in violation of their rights,[9] through the other means available to them, they cannot seize the trust assets nor can they cause them to be sold. At most, they can seize the proceeds eventually resulting from the transferor's interest in the trust.

However, Quebec jurists wishing to avail themselves of the self-benefit trust for their clients face significant difficulties because the qualifying criteria, recently included in the I.T.A, that allow tax-free transfers to this kind of trust originate in court decisions from common-law jurisdictions. Therefore, they must be interpreted with the requisite adaptations if they are to be transposed to a civil law context.

A jurist's comment to me illustrates the problem:

            [TRANSLATION]

in this context, it's not only necessary in Quebec law to imitate a transfer of legal ownership that does not involve a transfer of the beneficial ownership, but it must also be ascertained how a life interest in the trust can be disposed of while preserving a reversionary interest on death.

Indeed, although the criteria that the Parliament has elaborated are sufficiently detailed to try to adapt them to the civil law, though not without questions, they derive from concepts foreign to the civil law.

1.2.   I.T.A. Criteria

Two new and similar tax provisions allow for tax-free transfers to trusts by gift or by sale. The proposal is to integrate the first provision to section 73 I.T.A.[10] with the new alter ego trustand the joint spousal or common-law partner trust, both of them are permitted beginning at age 65. These trusts will enjoy the same rollover as the one originally given inter-spousal transfers of capital property.

The second provision is found in subsection 107.4(1) I.T.A. and is one of the situations giving rise to a "qualifying disposition" for a rollover.

The two provisions have roughly the same requirements and, according to the established hierarchy, subsection 107.4(1) I.T.A. applies only if the trust does not qualify under subsection 73(1) I.T.A.,[11] for example, for a transfer of property to the trust by a corporation,[12] or for a transfer of property to the trust that is not capital property.

The criteria for the application of these rollovers are technically as follows:

Like the spousal rollover, subsection 73(1.01) I.T.A. requires that the following conditions be met:

• The settlor's right to receive all of the income during his lifetime;
• No one else may receive any capital before his death.

By comparison, the concept of a qualifying disposition in subsection 107.4(1) L.I.R implicitly includes the following requirements arising from the necessity that there be no change in the beneficial ownership of the property, as this condition is interpreted in the Technical Note accompanying this section:

• the settlor is the sole beneficiary of the trust;
• the settlor is entitled to receive whatever share he requests of the annual income and the capital gains realized from the trust; and
• the trust property returns to the settlor if the trust is ended before his death.^[13]

Unlike section 73 I.T.A., the concept of a qualifying disposition makes it possible to choose whether the rollover will be total or partial.[14]

Two other conditions must be met in order to take advantage of the section 73 I.T.A trust:        

The disposition does not result in a change in the beneficial ownership of the property and there is immediately after the disposition no absolute or contingent right of a person (other than the individual) or partnership as a beneficiary (determined with reference to subsection 104(1.1)) under the particular trust.[15]

With a few minor differences, the same criteria apply to trusts constituted under subsection 107.4 I.T.A.

Subsection 104(1.1) of the Income Tax Act provides that a person is deemed not to have a beneficial right under a trust where the person is beneficially interested in the trust solely because of a right that may arise as a consequence of the law governing the intestacy of an individual or a right that may arise as a consequence of the terms of the will or other testamentary instrument of an individual. The foregoing clarification seems useful given that the law forbids the presence of a potential right of becoming a beneficiary. It would seem appropriate that the possibility of a right to take back by the settlor or by his or her heirs exist and that intestate heirs would be eligible to become potential beneficiaries[16]. However, the reference made to the will leaves us perplexed because the testamentary dispositions will be acknowledged as valid only once the settlor is deceased. This date is quite distant from the date of the initial transfer to the trust. Is the paragraph an indication that the property must necessarily come back to the estate of the settlor ?  If it is the intention of the legislator, he could have been more explicit. A contrario, could we not contend that if unlimited power to appoint beneficiaries is given to the settlor by the trust deed, the appointment which takes effect only at the time of the settlor's death and the termination of the trust, therefore, the condition that no person has a beneficiary right immediately after the transfer, even if this right is just potential, has been fulfilled (without having to examine subsection 104(1.1) I.T.A.) ? It would be difficult to uphold in these circumstances that each and every person is a potential beneficiary. How about the limited power given to the transferor to appoint beneficiaries, but which limited power is broad enough to include all of the people who are naturally potential heirs of the transferor? Thus, because we are of the opinion that the power to appoint beneficiaries stays an option, it will be part of the present analysis as well as testamentary and intestate heirs.

Furthermore, there will be a disposition of the property of such a trust on the day on which the death of the settlor occurs, according to new subsection 104(4)(a.4) I.T.A., rather than every 21 years.

1.3. T he civil law problematics

A practitioner in a civil law environment who is informed of these technical requirements is confronted with at least two questions: 1. What is beneficial ownership? - 2. In a civil law context, what is the mechanism whereby property may be devolved after the settlor's death?

1.3.1.  Concept of beneficial ownership

The concept of beneficial ownership referred to in the Act is dealt in another Department of Justice study. While awaiting the legislative amendments that will result from the latter study, there could be a solution by giving a liberal interpretation to paragraph 248(3)(e) I.T.A., which attempts to harmonize the common law concept of beneficial ownership to the civil law. Arguably, as soon as a person has a beneficiary right in the trust, he has the beneficial ownership of the property in that trust.[17]On the other hand, from the technical interpretations of the Canada Customs and Revenue Agency (the "Agency") reported by Me Jolin,[18] as soon as a person receives a beneficiary right (even a future right) in a trust, that is enough to cause a change in the beneficial ownership and prevent a rollover to a self-benefit trust.[19] Recent publications, however, reveal the extent to which Quebec jurists no longer know what tack to take with regard to this concept.[20] The pending clarifications to the Act will certainly be welcome if they manage to shed light on this debate. For the time being, we shall confine ourselves to noting that, as regards the self-benefit trust, the other requirements in the Act seem to function as a description of what, in the civil law, constitutes no change in the beneficial ownership.

1.3.2.  The mechanism of devolution on the settlor's death

A trust deed usually specifies the person to whom the property will be transmitted on the death of a beneficiary. The various possibilities open to a Quebec jurist under the Income Tax Act when drafting a self-benefit trust have been listed in the legal literature as follows:

            [TRANSLATION]

To comply with the condition that no one but the settlor may be the present or future beneficiary of a self-benefit trust, regardless of age, the trust deed may provide for the happening of one of the following on the death of the settlor-beneficiary:

• make no provision for a beneficiary in the event of the settlor-beneficiary's death;
• provide for who will become the trust beneficiaries in the event of the death of the settlor-beneficiary among the persons mentioned to that effect in the will of the settlor-beneficiary;
• provide for who will become the trust beneficiaries in the event of the death of the settlor-beneficiary among the persons who are the heirs by intestacy of the settlor-beneficiary; or
• provide that all the property of the trust will be deemed to be transferred to the settlor immediately before his death.

The trust deed should not designate the persons who will be beneficiaries of the trust on his death, but this designation may be made in another document, the will. The connection between the two must be explicitly referred to.[21]

Faced with this array of solutions, it would be difficult to argue that this kind of trust cannot be used in Quebec. However, attempts to apply the above solutions will result in various difficulties and the results will vary depending on the particular solution chosen. In the other provinces, however, the use of such trusts does not appear to be as problematic, since the requirements seem to be directly derived from common law concepts such as beneficial ownership, discussed above, and the transfer of a life interest in trust while preserving a reversionary interest on death.

Here are the issues that arise in the civil law with regard to these different variants in the order in which they are listed above:

Remain silent about the identity of the trust beneficiary or beneficiaries on the death of the settlor. (#1)

As we have previously mentioned in our comments concerning subsection 104(1.1) I.T.A., article 1297 of  the Civil Code provides that, where there is no beneficiary, the property devolves to the settlor or his heirs.

This article makes it possible to comply with the requirements of subsection 104(1.1) I.T.A. and leaves it up to the settlor to make a will concerning the trust property, thus giving him complete freedom on the choice of the subsequent beneficiaries. 

The Civil Code offers a tool enabling Quebec jurists to draw up a self-benefit trust in full compliance with the law. However, for this article to be used, there are certain civil law issues that cry out for clarification:

• Does the property return to the deceased's patrimony by the effect of this provision?
• Can the settlor make a legacy by particular title of the trust property?

The two questions address, on the one hand, the quality of the protection obtained for the assets in this manner and, on the other hand, restrictions on testamentary freedom.

The Civil Code is unclear about whether the property returns to the settlor and passes into his patrimony on his death by the effect of article 1297 C.C.Q. If the answer is affirmative, the property becomes available at that time to pay creditors,[22] thus rendering the protection for the assets less effective. In terms of the tax consequences, the inter vivos trust property would be the subject to the deemed disposition on the beneficiary's death and consequently would not receive the spousal rollover.[23] However, if, as we believe, the property goes into the deceased's patrimony, it could then be used to create testamentary trusts. Is this similar to the results achieved by citizens in the other provinces?

On the issue of whether a legacy by particular title of the property from a self-benefit trust would be authorized in this situation, the Civil Code and academic opinion unanimously agree that the Code's use of the term "heirs" [héritiers], generally excludes a legatee by particular title.[24]  On this interpretation, the settlor must take this outcome into account in planning his estate and provide that the trust patrimony will be given to those successors who receive universal legacies or as legatees by universal title.

However, we believe that in a trust context the term "heirs" should be understood in a broad sense as meaning all successors of the settlor, especially since, if the trust property were returned to the settlor during his lifetime, it would be part of the mass of the succession property or estate.  When this issue comes before the courts, it is impossible to imagine that a judge, on the basis of the wording of the Code, would decide that the legacy is not valid and that the trust property should instead be turned over to the persons receiving the residue of his estate in a case where the settlor has left a will clearly providing that the inter vivos trust property is to go to a specific legatee by particular title. Furthermore, article 2456 C.C.Q. is also important. It creates an exception to the definition of "heirs" by specifying that when the term "heirs" is used to designate the beneficiary of an insurance policy it means the person's entire succession generally.

However, this limitation on the settlor's testamentary freedom does not call for a tax-specific solution, although it does clarify how a self-benefit trust fits into the Quebec context.

Provide that on the death of the settlor-beneficiary those named in the will for that purpose will be the beneficiaries of the trust. (#2)

Insofar as the spirit of the tax measure concerning the self-benefit trust does not require that the property return to the settlor's patrimony when the trust ends, the best way to constitute such a trust appears to be to include a specific provision in the trust deed that the property will be handed over to the persons chosen by the settlor under the terms of his will. 

The great advantage of this option lies in the increased protection it offers to the trust property which at no time will pass through the settlor's patrimony and become subject to the claims of the creditors. The tax aspect differs from that of the first option in that it then becomes impossible to create a testamentary trust using the property of the inter vivos trust. It is even necessary to ensure that other trusts are not tainted by that property.[25]

However, it should be asked whether the power of appointment is enough to ensure compliance with paragraph 104(1.1)(a) I.T.A., which among other things permits "a right that may arise as a consequence of the terms of the will or other testamentary instrument of an individual who, at the particular time, is a beneficiary under the trust."

In view of this wording and the general context of the self-benefit trust, the use of the power of appointment raises two questions:

• Does the I.T.A authorize a provision in the trust deed that the settlor will appoint in his will the persons who will receive the trust property on his death without the property passing through his patrimony?
• In civil law, can such a power be drafted to make it unlimited, or should it be restricted to certain classes of persons, in which case the tax impact of such restriction must also be ascertained.^[26]

The government will have to explain its policy on the first issue. The second concerns limitations on the power to appoint under the civil law. This power is the privilege of designating, after the trust has been constituted, the persons who will be its beneficiaries. Professor Brierley describes the power as a right intuitu personae (of a specific person) without patrimonial value.[27]

Article 1282 C.C.Q. would seem favourable to the idea that there is no need to limit the power of appointment to a class of persons where it involves a power retained by the settlor, as opposed to that conferred on the trustee or a third party.  If the Civil Code were interpreted as a tax provision, we could be sure of this conclusion from the wording of the article:

 The settlor may reserve for himself the power to appoint the beneficiaries or determine their shares, or confer it on the trustees or a third person.

…the power to appoint may be exercised by the trustee or the third person only if the class of persons from which he may appoint the beneficiary is clearly determined in the constituting act.

At a recent APFF conference, Me Richard Gauthier said he favoured this interpretation, namely, that the article permits an unlimited power of appointment.[28]

Me Jolin, on the other hand, expressed the view that the power to appoint must be limited:

            [TRANSLATION]

 The problem is that, in a trust constituted in accordance with the Civil Code of Quebec, an unlimited power of appointment is completely invalid.  Moreover, according to article 1282 C.C.Q., only a trustee or a third person can be granted the power to appoint beneficiaries or determine their shares.  It would appear that the Agency has completely overlooked such realities in applying the federal tax system to trusts established under the Civil Code of Quebec. [29]

Since Me Jolin completely overlooks the portion of the article stating that the settlor may reserve this power for himself, we think that his conclusion on the requirement to limit the power of appointment might have been different if he had dealt with that possibility.

When article 1282 C.C.Q. was introduced, the Minister of Justice of Quebec commented briefly that the provision was consistent with the current state of the law.[30]

But what was the state of the law? The case law has consistently affirmed that, in order to be valid, a power of appointment must limit the class of persons in whose favour it can be used.[31] Often cited among these judgments is the decision of the Quebec Court of Appeal in Succession de Brodie. The decision was rendered in a testamentary context.

Professor Brierley[32] has made a comparative study of this judgment and another Ontario decision, also in a testamentary context. The common law jurisdiction had maintained the validity of a general power of appointment, which corresponds to an unlimited power of appointment, whereas the contrary result was reached in Quebec. Professor Brierley explained the result by the differences in the two jurisdictions' concept of the law of property: the common law allowing that a "mere power" in a will to designate beneficiaries took priority over the trust concept, which requires that beneficiaries be designated more specifically, whereas from a civil law standpoint, testamentary freedom must yield to the requirement that the persons who are to benefit from the will must "exist or be identifiable at the time of the testator's death"[33]according to former article 838 C.C.L.C. He explained the result in the civil law in more detail as follows:

The general power of appointment was struck down in Brodie, on the other hand, because the device of a power, like the notion of a trust itself, is perceived to be no more than a modality of gratuitous dispositions of property and, in particular, dispositions by will[…] That general law is to the effect that a disposition of ownership by one person, the owner (the person divesting), necessarily requires that there be certainty in the person to whom the ownership is transferred (the person to be vested). [34]

The Brodie decision was rendered under the old Code, which provided that the constitution of a trust could exist only as a mechanism for the administration of gifts or legacies.[35]  Thus, although the Supreme Court of Canada[36] had already had occasion to rule that ownership of the trust property was vested in the trustee and not the beneficiaries, this did not prevent the Court of Appeal to maintain that in order for the power of appointment to be recognized as valid in the context of a legacy, it must at least define [TRANSLATION] "the class of beneficiaries within which the trustees or legatees will choose.[37]

Article 1282 C.C.Q. would therefore have codified that legal rule. However, it cannot be concluded from this that a power of appointment reserved by the settlor for himself should necessarily receive the same treatment, given that the Code draws a distinction between the two cases and the earlier case law never dealt with this situation. In fact, the traditional concept of the power of appointment is the power left to a third person which must be supervised, whereas when the power is to be exercised by the settlor, it should instead be considered as an unlimited right that the settlor reserves for himself. Furthermore, article 1281 C.C.Q. also provides that the settlor may reserve the right to receive the fruits and revenues or the capital of the trust. In addition, the Civil Code indirectly incorporates an unlimited power of this kind in article 1297 C.C.Q. by providing that the "heirs" of the settlor receive the trust property when there is no beneficiary. In this way, the designation of the beneficiaries is left up to the settlor's will, which is free to designate anyone. Not to admit that the settlor can reserve the power for himself with no limitations would amount to denying the legality of the following clause in a self-benefit trust deed:

            [TRANSLATION]

"On the death of the settlor, the property will be given to the persons who are designated in the settlor's will."

whereas the Code already admits the validity of the following clause:

[TRANSLATION]

"On the death of the settlor, the property will be given to the heirs of the settlor in accordance with article 1297 C.C.Q."

It is true that there is a great difference between these two clauses if one accepts the thesis that the second clause triggers the return of the property to the settlor's estate, while the first clause allows the trust property to pass directly to the persons designated.  The two clauses are equivalent, however, in their lack of precision concerning the identity of those who will receive the property and the scope of the power left to the settlor allowing him to choose any beneficiary.

In the context of the new Code, if we correctly understand Professor Brierley's reasoning, he expresses doubts about whether the settlor may reserve an unlimited power of appointment for himself, although he admits that the Code provides the settlor with an opening in this direction by letting him retain the right to reserve a power of appointment for himself.[38]

Furthermore, in commenting on the new articles of the Civil Code in another paper,[39] Professor Brierley explains that the Code repeats the main themes of the earlier case law and he clarifies the distinction between a trust and the power to appoint. He makes only one allusion to such power's being reserved by the settlor:

            [TRANSLATION]

the second (the power to appoint) implies the power of the holder of the power (trustee or others, even the settlor) to choose the beneficiaries within a determined class… 

Given the fact that a jurist as expert as Professor Brierley has not stated that there have been any changes in the law, which for years has proclaimed that the power of appointment should be limited, together with the impossibility of easily modifying a trust, it is not surprising that some Quebec jurists still do not venture to use the power of appointment without accompanying it with a determination of the class of persons in whose favour it may be exercised.

In deciding how to apply the power of appointment to a self-benefit trust, a Quebec jurist is then faced with two alternatives:

1. Playing the caution card under the civil law and limiting the power of appointment, while at the same time trying to comply with I.T.A. requirements by ensuring that the limitation is as unrestrictive as possible through an authorization to the settlor to choose among all persons having a family relationship to him or a charitable organization. This method ensures greater protection from a trustee in bankruptcy who could be tempted to try to appropriate an unlimited power to himself. However, according to the Agency's technical interpretations, this kind of expanded power of appointment could be seen as an appointment of potential beneficiaries, which would make the trust ineligible for the tax rollover.[40] Me Jolin recently stated, however, that he had secured a verbal agreement from the Agency that such a power of appointment could be used in the context of a self-benefit trust.[41]
2. Or, playing the caution card in tax law and provide for a general and unlimited power of appointment, but this creates the risk that it could be deemed null and article 1297 C.C.Q. would be applied. However, even if the Quebec jurist adopts this solution, he is not at the end of his troubles since, if the return of the property to the settlor's patrimony is a key element to qualify for the rollover to the trust, as the Agency has stated on occasion,[42] by using the unlimited power to circumvent this rule, it might not qualify for that rollover, which could be very expensive.

The following clause, which probably comes from the common law, was examined by the Agency in a technical interpretation[43]:

" on the death of the settlor (" Mr. X"), the trustees of the trust are to deal with the trust property in the same manner as if it formed part of his estate, that is, in accordance with the provisions of his last will, if any, and if none, in accordance with applicable intestacy laws. As an alternative, the trustees are given the power to transfer all or part of the trust property to Mr. X's estate.

This clause is as broad as an unlimited power of appointment. However, according to the Agency, it does not meet the conditions for the rollover expressed in Technical News No. 7,[44] which were the origin of the provisions in section 73 I.T.A. Here are some particularly revealing excerpts from this ruling:

" The effect of the " as if " clause is to provide for other (contingent) beneficiaries of the trust.

(…) the power to appoint beneficiaries would have to be a general power ( …)

Property over which the deceased has a general power which is exercised in the deceased's will is, however, generally available for payment of debts of the deceased."

Although this opinion is based on common law principles, one cannot but be uneasy about the treatment of even an unlimited power of appointment in Quebec.

In fact, the real conclusion of this part is that we do not know how to adapt a general power of appointment to a civil law context and what consequences this would have for the rights of property.

Provide in the trust deed that when the settlor dies, his heirs by intestacy will receive the property. (#3)

In theory, this technique would meet the requirements of paragraph 104(1.1)(b) I.T.A., because the latter paragraph at first glance does not require that the laws of intestacy be applied only where no will exists, in contrast to the Explanatory Notes accompanying section 107.4 I.T.A.

If the trust deed contained a clause, as we proposed in the alternatives listed above, to the effect that the heirs by intestacy are entitled to the trust property, we think this would constitute a step backwards in relation to the other situations that have been analysed, in the sense that the list of subsidiary beneficiaries would be even more rigid and would completely put aside the settlor's discretion. Although the heirs by intestacy are undefined before the settlor's death, the clause would ensure that, if the settlor had a spouse and children on his death, they would share in the property in the proportions dictated by the Civil Code. If there were no longer any spouse, the children would receive everything, and so on. However, if such description of the contingent beneficiaries had been made in the trust deed, the trust would undoubtedly have been disqualified for the tax rollover because of the presence of other future beneficiaries.

Furthermore, this description deprives the settlor of his testamentary freedom over the trust property. This result may create an even greater distance between civil law and common law if it turned out in this latter jurisdiction that the retaining of this testamentary freedom is necessary for the lack of change in beneficial ownership.

We do not  retain this alternative. It is less flexible than the first and presents far more of a risk in the sense that it does not seem to respect the fiscal policy underlying self-benefit trusts. We believe that it is an erroneous interpretation of the bill, which has the object of permitting that one can fall back on the law of intestacy, but only where the settlor does not leave a will. Then, in our opinion, this option is already covered by the option #1, which uses article 1297 C.C.Q.

Transfer the trust capital immediately before the time that is immediately before the death of the settlor. (#4)

A relatively widespread practice in the context of a protective trust,[45] even before the recent amendments, is to have the trust property return to the settlor's patrimony before his death so that the property can pass among the assets of his estate. By doing this, one is not deprived of the possibility of taking advantage of the rollover of the assets to the spouse on death and the possibility of creating testamentary trusts as of that time. The effect of this practice is to advance the time when the property reverts to the settlor by the operation of article 1297 C.C.Q.

To achieve this result, a deeming clause is included in the trust deed providing either that the return of the trust property takes place as a matter of law immediately before the settlor's death or that the trust ends immediately before this time. Thus, on the death of the settlor, the deed would have one notionally return to the moment immediately before the settlor's death for this clause to take effect, whereas at the time this moment occurred the effect did not exist, since the death had yet not occurred. This is nothing more or less than contractually giving retroactive effect to a delivery or termination provision. Although this may be valid as between the parties to the contract, its  effect in the eyes of the tax authorities is debatable. However, in such a context, the argument that the tax authorities are a third person against whom this clause may not be set up is harder to apply than in the context of a counter-letter, where at the beginning of the trust the tax authorities have received a copy of the deed and can take cognizance of the effects provided for on death.

In our opinion, the idea that one can cause or avoid certain tax effects through the use of such a clause is a fundamental issue. In our opinion, the law cannot allow this, because immediately before the death, the delivery of the trust property or the termination of the trust has not occurred. To foresee that it has done so is a fiction. In fact, only on death does the obligation to hand over the property or the end of the trust arise and not before. This was the view of the Federal Court of Appeal in the context of a shareholders' agreement in its recent decision in Nussey Estate.[46]  The judges unanimously refused to recognize the effectiveness of a contractual provision deeming that a share redemption by the company had taken place before the disposition on the shareholder's death when, in his lifetime, the shareholder had not benefited from any such redemption.

In a bench judgment, the Court held that

…the deemed redemption provision in the agreement did not effect a retroactive disposition of the shares the day before A.W. Nussey's death. By its terms, the agreement could only operate once the taxpayer had died, by which time subsection 70(5) of the Act would have applied.

For the same reasons and because, by analogy with this situation, a presumption in the trust deed would make it possible to avoid the deemed disposition of the property on the beneficiary's death, this presumption could not be set up against the tax authorities, since they are not party to this fiction even if they are aware of its existence.

Thus, although option #4 would have the same advantages as option #1 and even more, making it possible to take advantage of the spousal rollover in addition to the possibility of creating testamentary trusts on death, this option would probably not be recognized by the tax authorities, and article 1297 C.C.Q. would apply in any event, as in the first option, but with the risk that the parties concerned would rely on the wording of the trust deed in planning their estates.

Use of this last option is accordingly not recommended.

The following summarizes the consequences of using each option and whether it can be recommended in setting up a self-benefit trust:

1.4. Conclusion

The Department of Finance should clarify which approaches it intends to allow, taking into account the two civil law options: returning the property to the settlor under article 1297 C.C.Q. or using the power of appointment.

2. Vested or vested indefeasibly (Dévolu  ou dévolu irrévocablement)

In trust matters, the fact that a property has vested or vested indefeasibly is essential to meet the conditions for certain kinds of tax relief or tax treatment.[47] Although the concept of devolution is known to the civil law in matters of succession,[48] the term "vested" in the Income Tax Act is translated in the French version not just by the term "dévolu" but also by the term "acquis".[49] It is interpreted through reference to the common law concept of vesting and has a broader meaning than does devolution in civil law.

The concept of indefeasible vesting also refers back to the common law. While its literal meaning seems fairly obvious, its legal meaning is somewhat different, as can be seen from the various definitions for the opposing concept of "defeasible interest", which can mean an interest subject to a condition, an interest in a legacy that has not been paid (subject to being divested) or an interest subject to a power of revocation.[50] Practitioners in a civil law environment must not only understand the nuances involved in the common law concept of indefeasibly in combination with that of vested, but must also adapt these nuances to their own legal system in order to comply with the conditions prescribed by the Income Tax Act. 

Since the topic of indefeasible vesting is treated in a separate research contract, in this study we will content ourselves with indicating the various difficulties of a practical nature occasioned by the use of this concept in civil law trusts.

We may distinguish between two kinds of situations where the concept of devolution (vesting) or acquisition is used in connection with trusts: (1) the devolution of property to the trust patrimony and (2) the devolution of property to a beneficiary while the trust continues to administer it. They both are characterized by their opacity in civil law.

2.1. Indefeasible vesting to the trust

To the first category belongs the rollover on death, for example to a trust created for the exclusive benefit of the spouse under subsection 70(6) I.T.A.[51] Among other things, this subsection requires that the property has vested indefeasibly in the trust.  It is clear on the face of it that such vesting is specific to the settlement of a succession, since subsection 73(1) I.T.A. deals with a similar rollover for an inter vivos transfer without requiring indefeasible vesting.

The only definition of the concept of "indefeasible vesting" to be found in the Income Tax Act is in subsection 248(9.2), which only serves to explain that the property is deemed not to have vested indefeasibly after the death of an individual or a spouse who is a beneficiary of a trust, but does not provide the parameters for vesting.

The main tool for the analysis of the criteria for indefeasible vesting continue to be the Agency's position as reflected in Interpretation Bulletin IT-449R.[52] The adaptation of these criteria to the civil law has been the subject of commentary by the academic community.[53]

Under the Civil Code, the effect of a legacy is established at the time of death and its acceptance only confirms it[54] but, according to the Interpretation Bulletin, indefeasible vesting requires that the property subject to the legacy be clearly identified and concerns the time when the spouse, spouse trust or child of the deceased "obtains a right to absolute ownership of that property in such a manner that such right cannot be defeated by any future event, even though that person may not be entitled to the immediate enjoyment of all the benefits arising from that right."[55]

Thus, the retroactive effect of the civil law or other provincial law[56] is not recognized. According to Me Jolin, one must wait for the heirs to accept[57] and even for the partition of the property before the heir's rights in a property can be certified and it can be claimed that the property has vested indefeasibly in the heir. Even the declaratory effect of partition[58] would not suffice for recognition of the vesting for tax law purposes at a date prior to that of the agreement. However, publication of the declaration of transmission, which renders the liquidator's divestiture official and releases the property for the benefit of the legatees and heirs at the end of his administration[59] is not required for vesting.[60]

Interpretation Bulletin IT-449R contains additional qualifications concerning the indefeasible vesting of shares subject to a buy-sell agreement among shareholders.[61]

As can be seen, notwithstanding the interpretation of indefeasible vesting in a testamentary trust context demonstrates a number of differences concerning the civil law, enough information is available so that a civil law lawyer or notary can be aware of the limits, even if they have been dictated by a common law environment. In an harmonization context of the Act with the private law, it would be appropriate to redefine these limits to make them more inclusive of the civil law. Study of the feasibility of this adaptation could be linked to a review of retroactivity since, among other things, it needs to be determined whether in this context the declaratory effect of the vesting should be recognized or whether the Act should explicitly reject it.

2.2.Vesting in the trust beneficiary

The concept of " vested,"  which is translated by "dévolu" or  "acquis" and used with regard to the beneficiary of a trust is far more nebulous in civil law. It refers to a situation where an interest in a trust is acquired by a beneficiary without the property's having been distributed to the latter. The vesting must sometimes be indefeasible.

2.2.1. Indefeasible vesting in a beneficiary

Paragraph (g) of the definition of "trust" in subsection 108(1) I.T.A. includes an exception for a trust of which all beneficiaries have acquired all their indefeasibly vested rights.  This trust escapes the 21-year deemed disposition rule, but is still subject to the other tax rules for trusts, with the exception of the preferred beneficiary election and section 106 I.T.A.

The technical notes offer a brief explanation of the former version of paragraph 108(1)(g) as follows:[62]

 The definition is amended for 1993 and subsequent taxation years so that the exclusion with respect to unit trusts also applies to trusts under which all interests in which have vested indefeasibly and in which no interest may become effective in the future. The ability of a trust beneficiary to sell his or her interest in the trust or gift the interest during his or her lifetime or on death through the terms of a will is not intended to bring a trust outside the terms of this exclusion. […] This amendment is relevant primarily for those commercial trusts which do not qualify as unit trusts.

As can be seen, apart from the fact that the beneficiary implicitly receives an interest that he may dispose of, these notes are not very helpful in understanding the meaning of indefeasible vesting.

In the section on loans to non-residents, subsection 17(15) I.T.A. specifies that a trust in which all of the interests have indefeasibly vested is synonymous of a non-discretionary trust. However, the scope of this definition is limited to this section.

We must return to Interpretation Bulletin IT-449R in which indefeasible vesting to a beneficiary is explained as follows:

 where [….] the individual's ownership rights cannot be defeated by any future event and no other person has any right whatsoever to an immediate or future benefit from that property or that trust, the property will be considered to vest indefeasibly in that individual.[63]

Paragraph 8(c) of the Bulletin explains that if, under the terms of a will, farm land is directed to be held in a trust for the benefit of the taxpayer's child, to be distributed to the child when the child reaches a specified age and, if the child should die before that age, to be distributed to the child's estate, the property is considered to vest indefeasibly in the child. If the will provides, however, that the land would be distributed to other persons, for example, the taxpayer's grandchildren, if the child should die before attaining the specified age, the land would not vest indefeasibly in the child until the child attained the specified age.

This interpretation seems to be in line with the new subparagraph 108(1)g)(v) I.T.A. permitting to escape the 21-year rule only if the indefeasibly vested interest of a beneficiary should terminate properly, namely "as a consequence of a distribution to the person (or the person's estate) of property of the trust".

However, this reference to the person's estate creates a problem in civil law with respect to personal trusts, at least where the specified age of distribution is later than the date on which the child reaches the age of majority and acquires the capacity to make a will. In fact, under the civil law one cannot generally leave to a beneficiary, by means of a reference to his succession, the freedom to designate replacement beneficiaries[64]. As explained in section 1.3.2 on the power of appointment, it is necessary to specify the classes of persons among whom the power may be exercised.[65] Although the classes may be broad enough to encompass family members, they cannot be equivalent to testamentary freedom since they must usually exclude "friends" and other persons who cannot be described except as members of a class so broad and imprecise that it could include anyone.

Since the power of appointment must be differentiated from proprietorship, it may also be thought that, if an amount has become payable to a beneficiary, he would act as a creditor of the trust for this amount. This debt would then belong to the beneficiary's patrimony and would therefore automatically return on his death to his testamentary succession or succession by intestacy, as the case may be, without passing through the provisions of the trust deed. But it seems to us that the notion of indefeasible vesting does not go so far.

Could a simple reference to the persons who would have received the succession by intestacy of the beneficiary without regard to the latter's would be valid from a tax point of view? To answer this question, we must first find out whether indefeasible vesting in a beneficiary involves a tax requirement that the beneficiary has the same testamentary freedom as if he had been the owner of the property.

In view of all this lack of precision, the first step would be, then, to define the parameters of the indefeasible vesting that takes place in favour of a trust beneficiary. They seem to be well known in the common law, but for a civil law lawyer or notary they mean nothing. Specifically, answers are needed to the following questions:

1. Must the beneficiary have unlimited testamentary freedom over his share of the trust?

Ex: Three children are each entitled to a third of a trust patrimony. The patrimony will be divided into three separate portions, the capital and income from these portions will be appropriated to the exclusive needs of the child for whom each portion is intended and, if the child dies, his portion will be distributed to his heirs. (N.B. This power of appointment could be illegal under the civil law.)

2. Is indefeasible vesting equivalent to granting the beneficiary a financial claim on the trust property that he may transmit to his heirs?

Ex.: A trustee makes a third of the property of the trust patrimony payable, and the trust becomes the debtor, in favour of a child beneficiary and the property that is to be used to pay the debt will be placed in a separate account. No other beneficiary is stipulated for this share. The trustee can distribute all or part of this property at his discretion. On the death of the child, the trustee will distribute the property to his (the child's) succession.

There is only a slight difference between these two situations. We are not even certain that they are really different, since in fact the creditor relationship seems unworkable to us in the context of a personal trust. We can certainly imagine that the trust could become indebted in favour of a beneficiary if an amount has become payable to him, for example when the beneficiary has reached the age of distribution, or when the trustee has exercised his discretion to distribute some income or capital. The writing of a note could then testify of a creditor/debtor relationship for this amount. That debt or note would certainly be part of the property in the beneficiary's succession on his death and would be handed over according to his will or the rules of intestacy.

Except for those situations we do not think that the trust is the debtor of the beneficiary's entire share while retaining the ownership and administration of it. In the Civil Code, the right of the beneficiary in the trust is contingent upon the respect of the conditions enacted by the settlor. On the death of the beneficiary, if his right to receive the capital is not yet opened[66], we believe that his succession will have no right to claim against the trust. The right of the deceased will pass to other beneficiaries according to the trust deed. Furthermore, the case law is in the opinion that the "succession" could not qualified as a replacement beneficiary, because it would constitute an illegal unlimited power of appointment.

So, in civil law, a relation of debtor/creditor for the entire participation of a beneficiary would go beyond the personal trust context. The trustee would no longer be administering the trust patrimony but rather the beneficiaries' patrimony.

Once the answers to these questions have been clarified and the other criteria for indefeasible vesting have been made clearer, if possible, the second step would be to ensure that the criteria are consistent with the civil law, particularly with respect to the power of appointment, which cannot usually be unlimited. If the criteria were found to be inconsistent, they would have to be reviewed and adapted to this system of law, for example by allowing on the death of a beneficiary the use of a large but limited power of appointment, or the use of a disposition referring to the succession by intestacy without the necessity to have a full testamentary freedom.

It should be noted that our comments do not necessarily apply to commercial trusts as it seems that in those trusts, the right of a beneficiary would not terminate upon his death since his participation does not result from a gift or a will.[67] It would then be easier to meet the requirements of an indefeasibly vested right with commercial trusts.

2.2.2. Non indefeasible vesting

In other contexts, the law disregards the indefeasibility requirement and retains only the vesting requirement. Thus, subsection 104(18) I.T.A. allows the taxation of income in the hands of a child even when the income is not paid or made payable to him as long as the right to the income has been vested (or "acquis") to the child.

In the light of the technical interpretations on this subject referring to the common law definition of "vested", at first glance we cannot point out the difference between vesting and indefeasible vesting. Both seem to require a certain right in the trust that can be transferred by the beneficiary himself.[68] 

These interpretations have also shown that the Agency relies on two common law decisions for its interpretation of the concept of "vested", namely, Hashman v. M.N.R.,[69] and Cole Trust v. M.N.R.[70]  These decisions, however, do not offer much enlightenment about the meaning of "vested" since they do not provide a definition but rather apply the concept to specific situations.

We might add another example to the debate: a legacy of an RRSP to a trust under 60(l)(ii)(B) I.T.A. where the trust is for a child under 18. In this situation, the Act requires that the child be the "sole person beneficially interested in amounts payable under the annuity " acquired from the RRSP. Although the English version speaks of "sole person beneficially interested" and not of "vested", this might possibly be a concept equivalent to vesting or, at least, we can assume a contrario, that the fact that an interest is "acquis" could be enough to ensure that the requirement of this provision is met. There is some confusion in Quebec about how it is to be interpreted. For example, it is said that this provision requires a trust whose the sole beneficiary is the child.[71] All that the Act requires, however, is payment to the child of the annuity amount, not the entire trust property. Having just one child as the beneficiary can definitely facilitate taxation of the trust income, but does not solve the question of whether the child's interest can terminate, for example, on his death and be given to his brothers and sisters.

In a civil law context, we would really like to know what the concept of "vested" is all about. It may be that its meaning is not much clearer in common law and that this discomfort stems from Quebec's lack of experience with trusts. It would be advisable, however, for the government to clarify this important concept, use a uniform language when it is appropriate and clarify the differences, if any, that it sees in a right vested in a beneficiary, a right vested indefeasibly and a trust of which one is the sole beneficiary in respect of a certain amount.

Once again, such explanations should, needless to say, take into consideration the limitations posed by the civil law on the application of these definitions.

3. Constructive and resulting trusts[72]

The Civil Code of Quebec does not recognize any trusts but expressly created trusts. The Associate Deputy Minister of the ministère de la Justice has also made it clear that the common law concepts of resulting trust and constructive trust have not been introduced in Quebec's law.[73] 

Tax litigation jurisprudence concerning cases from Quebec has no choice to refuse to apply such trusts.[74]

Nevertheless, reliance on the concepts of resulting trusts and constructive trusts seems to be on the rise in the common law jurisdictions, and such trusts are increasingly used to obtain tax results to the taxpayers' advantage. 

As part of this study of the differences between the common law and civil law jurisdictions in terms of trusts, it is worth taking the trouble to look more closely at these concepts.

3.1. Constructive trusts

A constructive trust (trust by interpretation) arises from the interpretation of a past situation by a common law judge.

The concept of constructive trust is used in two circumstances:

1. As a restitutionary mechanism developed by American and Canadian courts. It corresponds to the civil law doctrine of unjust enrichment.  This kind of remedy is often sought in a matrimonial situation.
2. As a presumption for the protection of the victim in a fraud or misconduct context. This use of the constructive trust originated in Great Britain.

Canadian case law has applied these concepts in two sorts of circumstances. In either case, there may be some interesting tax consequences stemming from the court's decision to recognize a trust rather than make an order transferring the property to the victim on some other basis because the trust so created predates the judgment recognizing the trust. When the existence of a constructive trust is directly argued in a tax case and not in the context of a claim against the person who has been enriched, judges often feel that they are justified in recognizing its effects notwithstanding the fact that the actual creation of the trust falls outside their jurisdiction. Some in fact believe that where the taxpayer's situation meets the conditions where the existence of such a constructive trust could be successfully argued in court, the tax consequences should be defined accordingly.

Let us look at a few examples from the tax case law:

In Karavos v. The Queen,[75] the judge refused to apply the constructive trust concept in a tax case involving restitution. The taxpayer argued that a constructive trust was created by the sale of a building so that part of the gain could be taxed in the hands of her spouse. After analysing the authorities and Supreme Court of Canada decisions, the judge commented as follows:

A constructive trust is a mechanism by virtue of which a Court with equitable jurisdiction can grant redress to an unjustly deprived person. In determining whether unjust enrichment exists and restitution through the invocation of a constructive trust is appropriate a Court may take into account the deprived person's actual financial contributions, (which may properly include the contribution of earnings towards household bills and maintenance), all work performed in relation to the property, both physical and otherwise, and other factors as the performance of housekeeping duties, the raising of children etc. The result is that effectively a Court is required to embark on an examination of the totality of a marital relationship extending over a period of 30 years to determine whether an unjust enrichment occurred and whether it would be appropriately remedied by a declaratory order vesting the claimant with title to property or by granting a monetary award. In my view such an inquiry is inappropriate in an income tax context. The use of a restitutory device to remedy situations of unjust enrichment should not be equated with the determination of a collateral issue necessary in order for this Court to carry out its statutory function, that is, to dismiss or allow an appeal or vacate or vary an assessment.

[…]

Even if the Appellant had established the prerequisites which would enable this Court to find that unjust enrichment existed this appeal could not succeed. In order to determine whether it is appropriate in a given case to invoke the remedy, the Supreme Court proposed a "causal connection" test. This test was considered by Dickson, C.J.C. in Sorochan v. Sorochan where he said:

... It is suggested simply that there should be a "clear link between the contribution and the disputed assets". The question of a connection between the deprivation and the property is further explained as "an issue of fact". That is, courts must ask whether the contribution is "sufficiently substantial and direct" to entitle the plaintiff to an interest in the property in question.

For this appeal the question may be stated as follows: Were Mrs. Karavos' contributions, in a broad sense, sufficiently substantial and direct so as to entitle her to a portion of the profits realized upon the sale of the property? There must be a clear causal connection between the spousal contribution founding the unjust enrichment and the property which is alleged to be the subject of the constructive trust. In my view there is no reasonable connection between the contribution or alleged deprivation and the property. 

Although the facts in this case did not allow the judge to find a trust, it is interesting to note that the definition concerns striking a balance in a couple's domestic relations.

The other component of the constructive trust, the "protective" aspect has a lesser foothold in Canadian case law, according to authors Brown and Rajan:[76]

The history of the constructive trust has created uncertainty about its current doctrinal basis.  Although it has been argued that the current position in Canada is that the constructive trust is to be regarded as a remedy and not a substantive institution, it is not clear that a Court would refuse to impose a non-remedial constructive trust on the basis of the British institutional notion of the trust as a cause of action in appropriate circumstances. […] the specific nature of the constructive trust may have important tax consequences since at least one Tax Court decision appears to have distinguished between the remedial and non-remedial constructive trust when providing tax relief.

The decision mentioned is Fletcher v. M.N.R.[77] In that case, the Tax Court of Canada expressed the opinion that the term "trust" in the Act includes all kinds of trusts, even a constructive trust arising from the misconduct of a person acting under a power of attorney.[78] In the result, this case allowed Mr. Fletcher, a non-resident, to plead with success that he had realized a capital loss under section 115 I.T.A. as a result of the disposition of his interest in a Canadian trust,[79] a trust that was a constructive trust created by his agent's misconduct.

Thus, although it did not have the jurisdiction to declare the existence of a constructive trust and require the payment of compensation, the Tax Court of Canada did not deprive itself of the power to recognize the tax consequences of such a trust. This court based its intervention on the need to identify the owner of the beneficial ownership in a property in order to assess the taxes correctly.[80]

The Agency recognized that a constructive trust is a trust in a tax context at a Round Table[81] held after the Fletcher decision. In response to a number of questions, in particular whether it was a bare trust that should be disregarded from a taxation point of view, the Agency replied:

It is our view that a constructive trust will be a trust that is subject to the application of all relevant provisions of the Act.  In these trusts, the constructive trustee usually has the control of the trust property within the meaning of subsection 104(1) of the Act.  Such a trust would be deemed to have been created at the time the property or control thereof is acquired by the ultimate constructive trustee, or later on, as the circumstances may require. Normally, the time of the Court judgment would not be relevant to this determination.  In our opinion, a constructive trust would be a personal trust within the meaning of the definition contained in subsection 248(1) of the Act.

Another example of the application of the constructive trust concept for tax purposes can be found in Anderson Estate v. The Queen.[82] Real property of the deceased had escaped the deemed disposition rules on his death on the basis that, at that time, he no longer held the beneficial ownership of it, even though the title was in his name. After his death it was recognized that the property belonged to his sister-in-law who had lived with him for 57 years and had never been paid for her work. This finding made it unnecessary to consider whether a trust existed since the court relied on her beneficial ownership of the property. From a civil law viewpoint, it is difficult to understand the nuance between recognizing a constructive trust of the remedial kind and deeming the beneficial ownership to have already been transferred because of the work performed by a family member, but the result seems to be the same, a result that is not available in Quebec.

Liability for tax under section 160 I.T.A. was also avoided in the Savoie[83] case on the basis of a constructive trust, which was argued along with resulting trust. The judge held that it was legitimate to argue the existence of the trust for the first time in the context of a tax case when it had to be determined who had the ownership of the property; therefore, the recognition of such trusts is not exclusively reserved to courts hearing a case between two spouses.[84]

3.2. Resulting trusts

A resulting trust (fiducie par déduction) is described summarily by Professor Waters:[85]

 … a resulting trust arises whenever legal or equitable title to property is in one party's name, but that party, because he is a fiduciary or gave no value for the property, is under an obligation to return it to the original title owner, or to the person who did give the value for it. 

A resulting trust requires at least two elements: a common intention, expressed or not, that the property be held in trust, and facts supporting this intention or from which this intention may be deduced.[86]

Like the constructive trust and often in conjunction with it, resulting trust has been argued on a number of occasions in tax cases from common law jurisdictions.

One example is Holizki[87] in which the Federal Court of Appeal confirmed the trial decision finding that the husband, who owned 99% of the shares of the family business, had always held a part in a resulting trust for the benefit of his wife. Once again, the facts that led to this result were the usual situation where a couple pools their property, the husband is in business and subsequently incorporates his business in his own name. His wife's involvement, among other things, had consisted in guaranteeing loans for the business, working in it and making up any cash shortfalls from her nursing income. It should be noted that during the years the business was in operation the income was all taxed in the husband's hands.

The judge summarized their situation as follows: "There was no express trust agreement and no discussion between Mervin and Maureen that he was holding any property in trust for her. Both Mervin and Maureen testified that it was just "understood" that the business belonged to both of them." [88].

The consensus between the spouses regarding their common property, despite the fact that the title to the shares was in the husband's name alone, allowed the couple to claim that, although the husband had subsequently transferred 49% of the shares to his wife, the transfer should escape to the attribution rules since the ownership had never been transferred. By applying the concept of resulting trust, the wife had always been the owner of these shares. And the court found in their favour.

A similar decision was rendered in Disbrowe v. The Queen.[89]

A Quebec jurist might be somewhat sceptical about this line of cases after all the efforts put forth by the Department of Finance to prevent spouses under a community of property regime in Quebec from using their co-ownership under that regime to split income![90]

This jurist still remembers the words of Judge Dubé in a Federal Court - Trial Division decision in 1989 refusing to recognize a capital gain split between spouses married under Quebec's community of property regime, despite the spouses' co-ownership of the property:[91]

…it would be quite unfair for taxpayers in one province to be favoured by provincial legislation dealing with the application of the Act, which should affect all Canadian taxpayers equally.

3.3. What makes such trusts so attractive from a tax standpoint

The majority of the decisions involving both resulting or constructing trusts and taxation is connected with transactions between spouses or other related persons. The consequences vary and can include avoidance of the attribution rules, of the deemed disposition on death or of the application of section 160 I.T.A. When such trusts are found to exist, the result is that the property in question has not in fact been transferred since it is deemed to have always been held in trust for the benefit of the person who suffered the impoverishment as a result of the relationship that has been established, whether conjugal or other.

As said earlier, other situations between unrelated persons involving misconduct or unjust enrichment can also give rise to such trusts.[92] These situations appear less problematic, however, since they do not open the door to abuse given the opposing interests of the persons in question. From a tax perspective, however, if it must be considered that a trust existed in a tax context, as in the Fletcher decision,[93] it would be normal then to reassess the years that were not statute-barred and, among other things, take into account the 21-year deemed disposition rule. Oddly enough, as the authors Brown and Rajan[94] have noted, those decisions where tax relief was claimed in this context have not reconsidered the taxpayers' whole tax file on the basis of the trust's existence.

Opinion on the issue of the retroactive tax impact of resulting or constructive trusts is not unanimous.[95]  There is disagreement as to whether the existence of such trusts and thus the beneficial ownership of the person claiming the gain begins at the time of the event giving rise to his entitlement, on the acquisition of the property to which the trust applies, in the course of the events giving rise to an unjust enrichment, at the time a claim was instituted on this basis or when the judgment recognizing the trust is rendered.[96]

It is nonetheless interesting to note that, at the tax level, once the existence of a constructive or resulting trust was recognized, none of the recent decisions was decided on the basis that the trust did not exist until judgment had been rendered by the Court.

3.4. Solutions for Quebec

In view of all this fiscal uncertainty, what should be retained and what should be asked to ensure fairness for Quebec?

Where the differences begin to be significant in the tax treatment accorded in civil law and common law jurisdictions is with unjust enrichment claims in the context of a constructive or resulting trust[97] and in matrimonial matters in particular. The civil law also offers a remedy based on enrichment in the same circumstances^[98] or on the compensatory allowance between married persons[99] but, as far as we know, whatever compensation has been obtained has never had a retroactive tax impact.

One author has summarized the treatment of unjust enrichment as follows:

            [TRANSLATION]

Arising from the same equitable principle and applying the same conditions for its exercise, the case law from Quebec and the common law provinces has evolved interrelatedly, with the common law provinces incorporating principles taken from Quebec law and vice versa.

[…] It must be emphasized, however, that in terms of compensation the common law solutions could not be fully incorporated into our Quebec law, which does not recognize the constructive trusts or trusts by interpretation that are relied on by the common law courts in awarding in some cases ownership of the other spouse's property.[100]

The only cases admitting of a retroactive tax effect in Quebec are those involving claims for restitution based on a tacit partnership. In Beaudoin-Daigneault v. Richard, the Supreme Court of Canada recognized that there could have been a tacit partnership between common law spouses who were partners in a farm with the result that the husband's signature on the initial contract of purchase was merely that of a mandatary of the partnership.[101] For its part, the ministère du revenu du Québec, in disallowing certain losses, pleaded the existence of a tacit partnership between spouses.[102] The remedy is not confined to the civil law; taxpayers in common law provinces have also pleaded the partnership concept.[103] However, a number of factors limit the use of this remedy and make it less popular in Quebec and elsewhere than probably common law constructive or resulting trusts.[104]

It should be pointed out that, although the use of the common law constructive and resulting trusts and even beneficial ownership[105] has been developed in tax case law over the last decade to good effect in matrimonial cases, their use is still not applicable to Quebec.

What solutions could be considered to deal with the unfairness?

• refuse to recognize such trusts as trusts for tax purposes;
• recognize the family, or at least the couple, as a tax unit;
• prevent resulting or constructive trusts from having a retroactive tax effect by enacting measures similar to those adopted for Quebec's community property regime; and
• recognize throughout Canada that a reasonable consideration, corresponding to the payment of a spouse's entitlement on the basis of enrichment, is valuable consideration.

3.4.1. The government could reverse its position of recognizing the existence of a constructive trust or even a resulting trust as a trust for tax purposes;

Like the bare trust[106], which is deemed not to be a trust, it could be provided that this kind of trust as well is not recognized for the purposes of tax legislation.[107]

And if this draconian solution is not suitable in circumstances where such trusts result from the misconduct or fraud of an unrelated person, an exception could be made for those cases and the trust could even be recognized as dating from the commission of the wrongful act, provided that the act were not statute barred. Since this kind of situation is relatively uncommon in Quebec, this exception might have no impact since, although in some cases as in the facts of Fletcher[108] the tax result might be advantageous, the victim could also find himself having to be responsible for taxes on income from a property that he did not own at the time.

However, we do not think that this solution would fully cover matrimonial cases, since common law courts sometimes base their decisions on the concept of beneficial ownership without necessarily always having to refer to the constructive or resulting trust.[109]  Thus, it is not so much the trust's tax status that matters as the fact that there is no transfer by the spouse who pays the debt resulting from an unjust enrichment situation.

3.4.2. The family as a tax unit (or eliminating the attribution rules between spouses)

The main inequities resulting from the fact that resulting and constructive trusts are not applicable in Quebec are that Quebec residents are not allowed to split capital gains or investment income from the portion of the property that is intended to redress the financial imbalance between the spouses, and the fact that sufficient consideration cannot be argued in the context of section 160 I.T.A. in relation to the settlement of rights arising from their life in common.

This is only the tip of the iceberg for family tax policy. The Department of Finance still refuses to recognize the family as a single tax entity, yet there are more and more limitations that reduce the deductions and tax credits that are based on the income of the entire family.

Tax policy thus favours recognition of the family unit when tax expenditures can be reduced. For example, two persons who work for an equivalent income and manage their wealth independently but live together as a couple will have to share the principal residence exemption if each owns a residence. Furthermore, a mother with a modest income because of her family responsibilities will have her child tax benefits reduced if her common law spouse (conjoint de fait) has a high income. Since it is implicitly assumed that this mother has access to her spouse's property to raise her children, it would be logical to encourage the legal transfer of that property to her name without the application of the attribution rules. Especially since it can be anticipated that, even with this tax advantage, such transfers will in fact enjoy limited popularity…

Recognition of the family, or at least the couple, as a tax unit would be a sort of bow to family unity by eliminating the attribution rules between spouses and reducing tax complexity, especially since the attribution rules are difficult to monitor and poorly understood.[110]

This solution would do away with tax cases based on the effects of a constructive or resulting trust, beneficial ownership or a tacit partnership as they relate to the attribution rules. Thus it would help make the tax system more neutral as far as the choice of the most appropriate remedy goes. It might also promote shared ownership of property by couples during their union rather than on the breakdown of their relationship when such discussions become very painful.[111] To obtain the tax split, the title should have to be transferred to the spouse.

This solution would also have the advantage of placing all spouses on an equal footing: those who equalize their spouse's property throughout their relationship and those who refuse or neglect to do so and find themselves at the end of their relationship having to make court-ordered adjustments.

It would not solve the thorny issue of section 160 I.T.A.  In this regard, an exception could be added to the Act similar to the one granted on marriage breakdown for cases where a reasonable division of property is made during the union. Otherwise, it might be thought that the Act favoured marriage breakdown when the couple is in financial difficulty! 

On the other hand, if the couple were recognized as a tax unit for all purposes of the Act, all of the property belonging to the couple might be viewed as a way to guarantee payment of taxes…

3.4.3. Prevent resulting and constructive trusts from having a retroactive tax effect by enacting measures similar to those adopted for Quebec's community property regime.

Like denying the existence of a trust for tax purposes, this solution is certainly very simple to implement in that it does not require extensive legislative amendment and would not generate additional tax expenditures. On reflection, however, it might just be a case of rearranging the deck chairs on the Titanic.

In fact, we believe that adopting such a presumption would lead to increased reliance on another argument, having more or less the same effect, but this time all across Canada, to get around the attribution rules and section 160,[112] i.e., the sufficient consideration rule. In an appropriate context, the argument would consist in claiming that the interspousal transfer was made for equivalent value, i.e., in settlement of all or part of the debt pursuant to the spousal relationship.

In our opinion, litigation involving constructive and resulting trusts cannot find a fair solution applicable to the country as a whole unless such situations are viewed in their entirety. Merely adding a presumption to the Act that these particular kinds of trust do not exist or are of no effect for tax purposes in order to prevent the income splitting by which their use might give rise could open the door to arguments about the market value of the consideration. Needless to say, it would be better to define the new remedy clearly in a way that encompasses and clarifies throughout Canada the question of the effect of such trusts as well as the questions concerning the payment of household expenses and spousal compensation.

3.4.4. Recognize everywhere in Canada that a reasonable consideration, corresponding to the payment of a spouse's entitlement based on enrichment, constitutes a valuable consideration.

According to the foregoing, instead of reappraising the tax system as it relates to resulting or constructive trusts, tax policy could solve the issue of fairness between Quebec and the other provinces by clearly establishing the principle that the payment of marriage debts is a transfer for consideration everywhere in Canada provided that the transfer is genuine and reasonable in the circumstances. This solution is similar to eliminating the attribution rules between spouses, with the difference that the amendment would be restricted to transfers resulting from situations involving financial rebalancing.

It should be noted in this respect that, even without such official recognition, the reasonable consideration argument is quietly making inroads. We earlier proposed that this argument could be used in determining the compensatory allowance during marriage. Our position was that the debt owned by the creditor-spouse in respect of this allowance could, if it were settled on the occasion of a transfer of a property, constitute sufficient consideration so that the exception in subsection 74.5(1) I.T.A. could be used -- to the extent that the subsection 73(1) I.T.A. rollover was not used.[113] This solution was also argued to avoid the application of section 160 I.T.A. in Savoie. The judge in that case found sufficient consideration, in view of the wife's rights, as follows:

Her relinquishment, upon the transfer of the property, of her inchoate right to apply in a provincial superior court for a declaration that she held a 50 % beneficial interest in the property would have a value equal to the 50 % interest in the property and would constitute the consideration that paragraph 160(1)(e) requires to be taken into account in determining her liability under that paragraph.[114]