G. (L.) v. B. (G.), [1995] 3 S.C.R. 370

L.G. Appellant

v.

G.B. Respondent

Indexed as: G. (L.) v. B. (G.)

File No.: 23629.

1995: March 2; 1995: September 21.

Present: La Forest, L'Heureux‑Dubé, Sopinka, Gonthier, Cory, McLachlin and Iacobucci JJ.

on appeal from the court of appeal for quebec

Divorce ‑‑ Support ‑‑ Variation ‑‑ De facto union ‑‑ Separation agreement incorporated in divorce judgment providing for payment of support to wife ‑‑ Agreement concluded by parties at time when to husband's knowledge wife was seeing friend with whom she now cohabits ‑‑ Whether change sufficient to justify varying support ‑‑ To what extent court's discretion limited by agreement ‑‑ Divorce Act, R.S.C., 1985, c. 3 (2nd Supp.), ss. 15, 17.

Divorce ‑‑ Support ‑‑ Variation ‑‑ Separation agreement incorporated in divorce judgment providing for payment of child support ‑‑ Agreement concluded by parties when child was a minor, whereas he is now married and independent ‑‑ Whether change sufficient to justify cancelling child support ‑‑ Divorce Act, R.S.C., 1985, c. 3 (2nd Supp.), ss. 15, 17.

The parties were married in 1960 and divorced in 1986. The divorce judgment, rendered pursuant to the 1985 Divorce Act, ratified an agreement on corollary relief entered into between the parties. That agreement provided inter alia that the respondent would pay the appellant $2,600 a month as spousal support and $100 a month for the minor child. This spousal support would not be reduced unless the appellant earned wages of more than $15,000 a year, in which case it would be reduced by the amount of the excess. At the time this agreement was entered into the appellant was seeing a friend with whom she had been cohabiting since May 1989. In July of that year the respondent, pursuant to s. 17 of the 1985 Divorce Act, filed an application to vary the corollary relief. In this application he sought (1) cancellation of the child support, as the child had been living with him since the divorce; (2) a declaration that the appellant was financially self‑sufficient, and (3) cancellation of the spousal support payable to the appellant. The evidence disclosed that during the marriage the appellant looked after the house and their three children. She also contributed to the activities of the respondent's business. However the appellant, who is now 53 years old, is still not in the labour market. Her new companion pays her between $1,000 and $1,300 a month and lent her $45,000 for the purchase of a condominium. They share all other common expenses. The Superior Court dismissed the applications to quash the support orders. The Court of Appeal, relying on the "presumption of self‑sufficiency", according to which the recipient of spousal support who is living in a de facto union has the burden of showing that, notwithstanding the cohabitation, she still needs the support, reduced the support payable to the appellant to $1,250 a month. It also cancelled the child support from the date on which the application to vary the corollary relief was filed.

Held: The appeal should be allowed.

Per Sopinka, Cory, McLachlin and Iacobucci JJ.: Variation of support orders is governed by s. 17 of the 1985 Divorce Act and the test developed in Willick applies. For a support order to be varied under s. 17, there must be a material change of circumstances, that is, a change that, if known at the time, would likely have resulted in different terms. If the matter which is relied on as constituting a change was known at the relevant time it thus cannot be relied on as the basis for variation. Here, the trial judge found that, at the time of the agreement, the respondent knew that the appellant was "seeing" a third party and that it was foreseeable that they would cohabit. In view of this finding, the trial judge correctly concluded that there was no material change of circumstances. There was no basis in fact or law for the Court of Appeal to reverse this conclusion. This is not an appropriate case to review the application of Pelech, Richardson and Caron to the support provisions of the 1985 Divorce Act.

Since the child is now married and independent, the conditions for variation exist with respect to the support payable for him. This support order should be cancelled from the date of the hearing of this appeal.

Per La Forest, L'Heureux‑Dubé and Gonthier JJ.: Since the question at issue must be considered in light of the principles stated in the 1985 Divorce Act, the criteria set out in Pelech, Richardson and Caron, which were rendered in the context of the provisions and philosophy of the 1968 Divorce Act, are inapplicable. The 1985 Act moved away from the tendency favouring a "clean break" which a majority of courts had adopted under the 1968 Act and emphasized the substantive rather than formal equality of the spouses in the marriage and at the time of the divorce. The 1985 Act thus rejected the presumption of economic self‑sufficiency and substituted for it a number of criteria that would take into account the advantages and disadvantages to spouses accruing from the marriage or its breakdown. Without completely departing from the objective of economic self‑sufficiency, the 1985 Act underlined that this objective can only be pursued "in so far as practicable".

For the initial support order to be varied under the 1985 Divorce Act, there must be a sufficient change between the parties to justify variation, as provided by s. 17(4) of that Act. The test set out in Willick in connection with an application to vary support for children applies mutatis mutandis to the variation of spousal support. Once the sufficiency of a change has been established, the court must next determine the extent of the variation and to do so it generally has to make an assessment of the entirety of the present circumstances of the parties. If they have concluded an agreement it must assess the agreement in light of the factors and objectives that govern spousal support under ss. 15(5), 15(7), 15(8), 17(4), 17(7) and 17(8). At the time of divorce the agreement is only one of the factors listed in s. 15(5) that should be taken into account in assessing inter alia the duration and amount of spousal support. The four objectives mentioned in s. 15(7) must also be considered. Additionally, s. 15(5) makes no mention of the necessity of a causal connection. In so far as a duty of support results from marriage, the notion of causality can only be dealt with in accordance with the various principles and objectives set out in the 1985 Act. Although s. 17, which governs variation orders, restates the general provisions applicable to a support order without specifically mentioning the obligation to take into account agreements concluded between the parties, it cannot be concluded that such agreements should be ignored when applications to vary support orders are made, especially when they were intended to be a final settlement and were ratified by the original support order, an order which must be taken into account. The weight to be given to these agreements will depend, first, on the extent to which the agreement reflects the principles and objectives stated in s. 17, and second, on the scope and nature of the change which has occurred, taking into account all the circumstances of the parties. The more the agreement or support order takes into account the various objectives of the 1985 Act, especially that of promoting an equitable distribution of the economic consequences of the marriage and its breakdown, the more likely it will be to influence the outcome of the variation application. Under the 1985 Act courts accordingly retain a discretionary power the exercise of which will depend on the particular facts of each case and which will be exercised in accordance with the factors and objectives mentioned in the Act. The existence of an agreement, final or otherwise, should not have the effect of precluding such an analysis.

In the case at bar the evidence discloses no change that would be sufficient to justify variation of the support order respecting the wife. The fact that the appellant is living with a companion certainly does not mean that she can be presumed to be financially independent. While the 1985 Divorce Act deals with the economic self‑sufficiency of the former spouses, it is only as one of the many objectives that must be met by the support order. What is more, this concept of economic independence is only favoured "in so far as practicable". Further, if the support order provides that spousal support will cease as soon as the former spouse attains economic self‑sufficiency, such self‑sufficiency is not to be presumed: it must be proven. Such evidence has not been presented. First, the agreement, made and ratified by the support order initially rendered, was drawn up when the appellant was seeing the person who foreseeably would become her companion in the not too distant future. Second, the possibility that the appellant would become financially self‑sufficient and rejoin the labour market, in view of her age, her absence from the labour market for many years and, presumably, the lack of any adequate professional training in a competitive labour market, was entirely illusory, quite apart from the fact that she was unable to accumulate retirement pension and enjoy other benefits associated with employment. In short, given the current Act and the evidence, including the initial support order and the agreement entered into between the parties, which was intended to be final, the Court of Appeal was not justified in concluding that the appellant had attained adequate financial self‑sufficiency to relieve the respondent of part of his support obligation as stipulated in the support order. Most importantly, it could not start from a presumption of financial self‑sufficiency which is not supported by either the 1985 Act or its interpretation.

So far as the child support paid to the former wife is concerned, since the child is now married and self‑supporting, which was not the case before the trial judge and in the Court of Appeal, the nature and magnitude of this change are such that the child support should be cancelled, from the date the appeal at bar was heard.

Cases Cited

By Sopinka J.

Applied: Willick v. Willick, 1994 CanLII 28 (SCC), [1994] 3 S.C.R. 670; referred to: Pelech v. Pelech, 1987 CanLII 57 (SCC), [1987] 1 S.C.R. 801; Richardson v. Richardson, 1987 CanLII 58 (SCC), [1987] 1 S.C.R. 857; Caron v. Caron, 1987 CanLII 59 (SCC), [1987] 1 S.C.R. 892.

By L'Heureux‑Dubé J.

Applied: Willick v. Willick, 1994 CanLII 28 (SCC), [1994] 3 S.C.R. 670; Moge v. Moge, 1992 CanLII 25 (SCC), [1992] 3 S.C.R. 813; not followed: Pelech v. Pelech, 1987 CanLII 57 (SCC), [1987] 1 S.C.R. 801; Richardson v. Richardson, 1987 CanLII 58 (SCC), [1987] 1 S.C.R. 857; Caron v. Caron, 1987 CanLII 59 (SCC), [1987] 1 S.C.R. 892; referred to: Corkum v. Corkum (1988), 1988 CanLII 8613 (ON SC), 14 R.F.L. (3d) 275; Snyder v. Snyder (1987), 1987 CanLII 8409 (NS CA), 10 R.F.L. (3d) 144; Schroeder v. Schroeder (1987), 1987 CanLII 5322 (MB QB), 11 R.F.L. (3d) 413; Williams v. Williams (1988), 1988 CanLII 5470 (NL SC), 13 R.F.L. (3d) 321; Willms v. Willms (1988), 1988 CanLII 4625 (ON CA), 14 R.F.L. (3d) 162; Droit de la famille‑‑382 (1988), 1988 CanLII 1033 (QC CA), 16 R.F.L. (3d) 379; Brody v. Brody (1990), 1990 CanLII 5876 (AB QB), 25 R.F.L. (3d) 319; Publicover v. Publicover (1987), 1987 CanLII 8376 (NS FC), 9 R.F.L. (3d) 308; Lynk v. Lynk (1989), 1989 CanLII 194 (NS CA), 21 R.F.L. (3d) 337; Doncaster v. Doncaster (1989), 1989 CanLII 4720 (SK CA), 21 R.F.L. (3d) 357; Fisher v. Fisher (1989), 1989 CanLII 4358 (ON SC), 22 R.F.L. (3d) 225; Story v. Story (1989), 1989 CanLII 5317 (BC CA), 23 R.F.L. (3d) 225; Linton v. Linton (1990), 1990 CanLII 2597 (ON CA), 1 O.R. (3d) 1; Romanoff v. Romanoff (1992), 1992 CanLII 13007 (MB QB), 41 R.F.L. (3d) 433; Droit de la famille‑‑1567, 1992 CanLII 3052 (QC CA), [1992] R.J.Q. 931; Droit de la famille‑‑1688, 1992 CanLII 3653 (QC CA), [1992] R.J.Q. 2797; Bush v. Bush (1989), 1989 CanLII 8812 (ON SC), 21 R.F.L. (3d) 298; Messier v. Delage, 1983 CanLII 31 (SCC), [1983] 2 S.C.R. 401; Brockie v. Brockie (1987), 1987 CanLII 141 (MB CA), 8 R.F.L. (3d) 302; Droit de la famille‑‑333, [1987] R.J.Q. 294.

Statutes and Regulations Cited

Divorce Act, R.S.C. 1970, c. D‑8 [formerly S.C. 1967‑68, c. 24], s. 11(1), (2).

Divorce Act, R.S.C., 1985, c. 3 (2nd Supp.), ss. 15(5), (7), (8), 17(1), (4), (7), (8), (10).

Authors Cited

Bailey, Martha J. "Pelech, Caron, and Richardson" (1989‑90), 3 C.J.W.L. 615.

Bala, Nicholas. "Domestic Contracts in Ontario and the Supreme Court Trilogy: `A Deal is a Deal'" (1988), 13 Queen's L.J. 1.

Canada. Law Reform Commission. Family Law. Ottawa: The Commission, 1976.

Canada. Department of Justice. Divorce Law In Canada: Proposals For Change. Ottawa: Department of Justice, 1984.

Davies, Christine. Family Law in Canada. Toronto: Carswell, 1984.

Duff, David G. "The Supreme Court and the New Family Law: Working through the Pelech Trilogy" (1988), 46 U.T. Fac. L. Rev. 542.

Durnford, John W., and Stephen J. Toope. "Spousal Support in Family Law and Alimony in the Law of Taxation" (1994), 42 Can. Tax J. 1.

Goubau, Dominique. "Une nouvelle ère pour la pension alimentaire entre ex‑conjoints au Canada" (1993), 72 Can. Bar Rev. 279.

Heeney, Thomas A. "The Application of Pelech to the Variation of an Ongoing Support Order: Respecting the Intention of the Parties" (1989), 5 C.F.L.Q. 217.

McDermid, D. R. "The Causal Connection Conundrum" (1989), 5 C.F.L.Q. 107.

McLeod, James G. Annotation (1987), 1987 CanLII 57 (SCC), 7 R.F.L. (3d) 225.

Neave, Marcia. "Resolving the dilemma of difference: A critique of `The Role of Private Ordering in Family Law'" (1994), 44 U.T.L.J. 97.

Payne, Julien D. "Further Reflections on Spousal and Child Support After Pelech, Caron and Richardson" (1989), 20 R.G.D. 477.

Payne, Julien D. "Spousal and Child Support After Moge, Willick and Levesque" (1995), 12 C.F.L.Q. 261.

Proudfoot, Patricia, and Karen Jewell. "Restricting Application of the Causal Connection Test: Story v. Story" (1990), 9 Can. J. Fam. L. 143.

Rogerson, Carol J. "Judicial Interpretation of the Spousal and Child Support Provisions of the Divorce Act, 1985 (Part I)" (1990‑91), 7 C.F.L.Q. 155.

Rogerson, Carol J. "The Causal Connection Test in Spousal Support Law" (1989), 8 Can. J. Fam. L. 95.

APPEAL from a judgment of the Quebec Court of Appeal, [1993] Q.J. No. 673, J.E. 93‑880 (sub nom. Droit de la famille‑‑1783), allowing in part the respondent's appeal from a decision of the Superior Court rendered on May 20, 1990, allowing in part his application to vary a support order. Appeal allowed.

Simon Lahaie, for the appellant.

George Artinian, for the respondent.

The reasons of La Forest, L'Heureux-Dubé and Gonthier JJ. were delivered by

[1] L'Heureux‑Dubé J. ‑‑ This appeal concerns the variation of a support order made in a divorce judgment rendered pursuant to the Divorce Act, R.S.C., 1985, c. 3 (2nd Supp.) (formerly S.C. 1986, c. 4). That judgment ratified an agreement on corollary relief entered into between the parties. The question at issue is to what extent the court's discretion is limited by such an agreement.

I. Facts

[2] The appellant L.G. and the respondent G.B., 53 and 56 years old respectively, were married in 1960. Three children were born of the marriage. The two eldest, D. and B., had reached the age of majority when the divorce occurred. The youngest, F., was 12 years old and at boarding school at the time. The parties lived together for over 25 years during which the appellant looked after the maintenance and education of their three children, did the housework and also contributed to the activities of the respondent's business. The divorce took place on November 18, 1986.

[3] The corollary relief agreement, signed between the parties beforehand, each with the assistance of independent legal counsel, contained the following stipulations. The parties would have joint custody of their minor son, the appellant having physical custody and the respondent visiting rights on the conditions specified therein. The respondent would pay the appellant $2,600 a month as spousal support and $100 a month for the minor child. The support was subject to indexing under art. 638 of the Civil Code of Quebec, S.Q. 1980, c. 39. The respondent would also pay all the schooling, clothing, recreational and school meal expenses of their minor son. Any tax deduction for the dependant minor child would benefit the respondent.

[4] The agreement further provided that the spousal support payable to the appellant would not be reduced unless she earned wages of more than $15,000 a year. Any employment income over $15,000 would reduce the spousal support by the amount of the excess.

[5] On the question of the division of property, the agreement stated that the appellant would retain the ownership of the furniture but that, by February 1, 1987 at the latest, she would leave the marital home and transfer its ownership to the respondent. In the meantime, and after the appellant's departure, the respondent would pay all real estate taxes, water rates and other charges of a similar nature affecting the marital residence, as well as major repairs, fire and liability insurance and mortgage. The respondent further agreed to pay as compensation for enrichment, in full and final settlement of the provisions of the marriage contract and in final settlement of any amount which may otherwise be owed to the appellant, excluding support, the sum of $120,000 without interest, payable in four equal instalments of $30,000 on February 1, 1987 and on December 1 of 1987, 1988 and 1989.

[6] At the time this agreement was entered into, the appellant was seeing a friend with whom she had been cohabiting since May 1989. The respondent had also begun cohabiting with a new companion.

[7] In July 1989, the respondent, pursuant to s. 17 of the 1985 Divorce Act, filed an application to vary the corollary relief, seeking (i) legal custody of the child, a minor at the time (now married), (ii) cancellation of the child support of $100 a month, (iii) a declaration that the appellant was financially self‑sufficient, and (iv) cancellation of the spousal support payable to the appellant.

[8] The evidence disclosed that the appellant, who was 45 years old at the time of divorce, is still not in the labour market and that her new companion pays her between $1,000 and $1,300 a month and lent her $45,000 for the purchase of a condominium. They share all other common amenities.

[9] The Quebec Superior Court allowed the respondent's application in part and, on an appeal by the respondent, the Court of Appeal allowed the appeal in part: hence the appeal at bar.

II. Judgments

Superior Court

[10] With respect to the corollary relief relating to the parties' son, then living with his father, Benoit J. varied custody only to the extent of awarding physical custody to the respondent, and upheld the order for $100 a month payable to the appellant for the child.

[11] The trial judge stated that the question of spousal support must be looked at in light of all the evidence as well as the agreement concluded between the parties. After noting the respondent's admission as to his capacity to pay as well as the fact that there had been no unfavourable change in his financial situation since the divorce, he concluded:

[translation] The [respondent] has not set any term to his obligation to pay the indexable spousal support nor to the [appellant's] right to receive this alimony and to be able to earn an indexable salary of $15,000.00 without a reduction of the alimony. Clearly, the [respondent] cannot be permitted to alter the agreement which was freely entered into four years ago. However, the Court may later on conclude that the [appellant] has lost her entitlement to the present alimony because she has failed to discharge her responsibilities.

[12] Considering the effect of the appellant's new circumstances, the judge concluded that her cohabitation with a third party was foreseeable at the time of divorce and the spousal support in the agreement would ordinarily have taken this into account. In spite of this, there was no stipulation reducing or cancelling the support in such circumstances. Furthermore, as the relationship offered no guarantee of any permanence, it could not secure the appellant's financial independence.

[13] The application was therefore allowed in part, solely as to the physical custody of the then minor child.

Court of Appeal (per Vallerand, Baudouin and Fish JJ.A.), J.E. 93‑880

[14] Fish J.A., writing for the court, concluded that the appellant's new union [translation] "[had] achieved sufficient stability for . . . the support obligations imposed on the ex‑husband at the time of the divorce to be reassessed". The Court of Appeal considered that the agreement entered into between the parties was not a final settlement of all obligations pertaining to the marriage and its breakdown. Relying on the "presumption of self‑sufficiency", according to which the recipient of spousal support who is living in a de facto union has the burden of showing that, notwithstanding the cohabitation, she still needs the support, the Court of Appeal reduced the amount payable to the appellant to $1,250 a month as of April 20, 1993, the date of judgment. It cancelled the child support of $100 a month retroactive to July 24, 1989. Only these two aspects of the support order are at issue before us.

III. Applicable Legislation

[15] The case at bar is governed by the 1985 Divorce Act. Although ss. 17(1), 17(4) and 17(7) are directly applicable here, ss. 15(5) and 15(7) and 17(8) of the Act are also relevant:

15. . . .

(5) In making an order under this section, the court shall take into consideration the condition, means, needs and other circumstances of each spouse and of any child of the marriage for whom support is sought, including

(a) the length of time the spouses cohabited;

(b) the functions performed by the spouse during cohabitation; and

(c) any order, agreement or arrangement relating to support of the spouse or child.

(7) An order made under this section that provides for the support of a spouse should

(a) recognize any economic advantages or disadvantages to the spouses arising from the marriage or its breakdown;

(b) apportion between the spouses any financial consequences arising from the care of any child of the marriage over and above the obligation apportioned between the spouses pursuant to subsection (8);

(c) relieve any economic hardship of the spouses arising from the breakdown of the marriage; and

(d) in so far as practicable, promote the economic self‑sufficiency of each spouse within a reasonable period of time.

17. (1) A court of competent jurisdiction may make an order varying, rescinding or suspending, prospectively or retroactively,

(a) a support order or any provision thereof on application by either or both former spouses; or

(b) a custody order or any provision thereof on application by either or both former spouses or by any other person.

(4) Before the court makes a variation order in respect of a support order, the court shall satisfy itself that there has been a change in the condition, means, needs or other circumstances of either former spouse or of any child of the marriage for whom support is or was sought occurring since the making of the support order or the last variation order made in respect of that order, as the case may be, and, in making the variation order, the court shall take into consideration that change.

(7) A variation order varying a support order that provides for the support of a former spouse should

(a) recognize any economic advantages or disadvantages to the former spouses arising from the marriage or its breakdown;

(b) apportion between the former spouses any financial consequences arising from the care of any child of the marriage over and above the obligation apportioned between the former spouses pursuant to subsection (8);

(c) relieve any economic hardship of the former spouses arising from the breakdown of the marriage; and

(d) in so far as practicable, promote the economic self‑sufficiency of each former spouse within a reasonable period of time.

(8) A variation order varying a support order that provides for the support of a child of the marriage should

(a) recognize that the former spouses have a joint financial obligation to maintain the child; and

(b) apportion that obligation between the former spouses according to their relative abilities to contribute to the performance of the obligation.

IV. Analysis

[16] The real question at the heart of this case concerns the effect of an agreement or consent order as to corollary financial relief, on a subsequent application to vary spousal support pursuant to the 1985 Divorce Act. Each party relied on Pelech v. Pelech, 1987 CanLII 57 (SCC), [1987] 1 S.C.R. 801, Richardson v. Richardson, 1987 CanLII 58 (SCC), [1987] 1 S.C.R. 857, and Caron v. Caron, 1987 CanLII 59 (SCC), [1987] 1 S.C.R. 892 ("the trilogy").

[17] In particular, the appellant alleged that the agreement entered into between her and her former husband is final and that, in view of the trilogy, the Court of Appeal erred in varying the support order based on that agreement. For his part, the respondent maintained that the agreement in question, unlike those which were the subject of the trilogy, is not a final one and, in any case, a fundamental change (having no causal connection with the marriage) has taken place in the situation of the appellant and their son which provides a basis for the remedy sought, in accordance with case law developed subsequent to the trilogy.

[18] This leads me to review the trilogy, which was rendered in the context of the provisions and philosophy of the Divorce Act, S.C. 1967‑68, c. 24 (later R.S.C. 1970, c. D‑8).

A. The 1968 Divorce Act

[19] The relevant provisions of the 1968 Divorce Act dealing with support orders and their variation (as reproduced in the R.S.C. 1970) read as follows:

11. (1) Upon granting a decree nisi of divorce, the court may, if it thinks it fit and just to do so having regard to the conduct of the parties and the condition, means and other circumstances of each of them, make one or more of the following orders, namely:

(a) an order requiring the husband to secure or to pay such lump sum or periodic sums as the court thinks reasonable for the maintenance of

(i) the wife,

(ii) the children of the marriage, or

(iii) the wife and the children of the marriage;

(b) an order requiring the wife to secure or to pay such lump sum or periodic sums as the court thinks reasonable for the maintenance of

(i) the husband,

(ii) the children of the marriage, or

(iii) the husband and the children of the marriage; and

(c) an order providing for the custody, care and upbringing of the children of the marriage.

(2) An order made pursuant to this section may be varied from time to time or rescinded by the court that made the order if it thinks it fit and just to do so having regard to the conduct of the parties since the making of the order or any change in the condition, means or other circumstances of either of them. [Emphasis added.]

[20] It should be borne in mind here, as Professor C. Davies (Family Law in Canada (1984), at p. 328) has pointed out, that before the 1968 Act was enacted, divorce laws in Canada were far from uniform.

[21] The main purpose of the 1968 Act was, first, to standardize divorce throughout Canada and to provide additional grounds for divorce. Further, the statute entitled courts to make corollary orders for support and custody upon granting divorce. Support orders had to take into account the conduct as well as the condition, means and other circumstances of the parties. The Law Reform Commission of Canada described the law of support in the following terms:

Before that time [the 1968 Act], the right to maintenance on divorce could only be lost as a result of a judicial determination, based on known, settled and pre‑existing rules of law, that the claimant spouse had committed a matrimonial offence. This was arbitrary, but certain. The 1968 Act changed the law to allow the court to award maintenance in any event, but the result has been a maintenance rule that is both arbitrary and uncertain. The Act now requires that the award be based on the court's evaluation of conduct in addition to a consideration of the spouses' condition, means and circumstances. This means that the financial implications of a maintenance claimant's marital economic experience are always subject to the uncertainty of a behavioural evaluation according to whatever criteria a judge may find compelling. The proper standard of conduct is not defined by law, nor is the nature of the relationship between conduct and financial rights. Both these matters are, according to one appellate court decision, "within the entire and absolute discretion" of the trial judge. These inherently subjective standards lack the certainty that is essential if justice is to be done in determining the economic consequences of marriage breakdown, where the outcome will often represent the fruits of the labour of the spouses' adult lifetimes.

(Family Law (1976), at pp. 39‑40.)

[22] What was not spelled out in the 1968 Act was quickly made up by the courts, which adopted the concept of a "clean break", namely the marked tendency of the courts to assume that economic self‑sufficiency of either spouse could and should be achieved as soon as possible after the divorce. In this context, spousal support could only be of a temporary nature, i.e. a means of enabling its recipient, generally the wife, to go into or return to the labour market, thus allowing the parties to rebuild their separate lives.

[23] This is the context in which Pelech, Richardson and Caron were decided. Although the judgments were rendered in 1987, after the coming into force of the current Divorce Act, they dealt with consensual support orders made pursuant to the 1968 Act. Pelech and Caron involved motions to vary support orders based on s. 11(2) of that Act, while Richardson concerned a support order based on s. 11(1). In all three cases a separation agreement was entered into between the parties.

[24] According to Wilson J., writing for the majority in Pelech (at pp. 851‑53):

Absent some causal connection between the changed circumstances and the marriage, it seems to me that parties who have declared their relationship at an end should be taken at their word. They made the decision to marry and they made the decision to terminate their marriage. Their decisions should be respected. They should thereafter be free to make new lives for themselves without an ongoing contingent liability for future misfortunes which may befall the other.

Where parties, instead of resorting to litigation, have acted in a mature and responsible fashion to settle their financial affairs in a final way and their settlement is not vulnerable to attack on any other basis, it should not, in my view, be undermined by courts concluding with the benefit of hindsight that they should have done it differently.

[25] This restates the gist of the philosophy underlying the 1968 Act as formulated by a majority of the courts of that time.

[26] La Forest J., dissenting in Richardson, supra, took a completely different approach at pp. 878‑79:

To allow separation agreements the kind of compelling weight argued for in this case is effectively to rewrite the Act so as to provide that where such an agreement exists, the trial judge's discretion is solely to vary the agreement in those cases only where radical or, to use the expression employed in some cases, catastrophic changes have occurred since it was made.

I am not, as I mentioned earlier, saying that a separation agreement is not an important fact to be considered by the judge in exercising his or her discretion. Such an agreement falls within the factors "conduct of the parties" or "other circumstances" mentioned in s. 11(1). It is obviously better for the parties to settle their affairs whenever possible. As well, settlement by the parties assists the courts in performing their function; it can save time and costs. Still, the latter considerations should not be used as an excuse to displace the court's function. Consequently, as many courts have stated, while such an agreement should not lightly be disturbed, at the end of the day, the judge cannot be bound by it.

[27] As discussed in Moge v. Moge, 1992 CanLII 25 (SCC), [1992] 3 S.C.R. 813, both the case law and academic writers were not long to react to the trilogy. At the outset, Professor J. G. McLeod held the view that the rules set out in the trilogy should be applied to any non‑consensual support order, even in the absence of a final agreement:

The reasons [of Wilson J.] are also likely to affect the granting of support in the absence of a settlement agreement. The reasons of Wilson J. in Pelech, Richardson and Caron confirm a basic support model. In order to obtain support, a claimant must prove:

(1)need;

(2)that the need arises for a legally acceptable reason; and

(3) that the need/inability is causally connected to the marriage. [Emphasis added.]

(Annotation (1987), 1987 CanLII 57 (SCC), 7 R.F.L. (3d) 225, at p. 232.)

See also D. R. McDermid in "The Causal Connection Conundrum" (1989), 5 C.F.L.Q. 107, at p. 119.

[28] Some courts adopted this approach, not without some hesitation in certain cases (Corkum v. Corkum (1988), 1988 CanLII 8613 (ON SC), 14 R.F.L. (3d) 275 (Ont. H.C.)), but more resolutely in others: Snyder v. Snyder (1987), 1987 CanLII 8409 (NS CA), 10 R.F.L. (3d) 144 (N.S.C.A.); Schroeder v. Schroeder (1987), 1987 CanLII 5322 (MB QB), 11 R.F.L. (3d) 413 (Man. Q.B.); Williams v. Williams (1988), 1988 CanLII 5470 (NL SC), 13 R.F.L. (3d) 321 (Nfld. S.C.); Willms v. Willms (1988), 1988 CanLII 4625 (ON CA), 14 R.F.L. (3d) 162 (Ont. C.A.); Droit de la famille‑‑382 (1988), 1988 CanLII 1033 (QC CA), 16 R.F.L. (3d) 379 (Que. C.A.), and Brody v. Brody (1990), 1990 CanLII 5876 (AB QB), 25 R.F.L. (3d) 319 (Alta. Q.B.).

[29] Other courts confined the trilogy to consensual situations: Publicover v. Publicover (1987), 1987 CanLII 8376 (NS FC), 9 R.F.L. (3d) 308 (N.S. Fam. Ct.); Lynk v. Lynk (1989), 1989 CanLII 194 (NS CA), 21 R.F.L. (3d) 337 (N.S.C.A.); Doncaster v. Doncaster (1989), 1989 CanLII 4720 (SK CA), 21 R.F.L. (3d) 357 (Sask. C.A.); Fisher v. Fisher (1989), 1989 CanLII 4358 (ON SC), 22 R.F.L. (3d) 225 (Ont. Div. Ct.); Story v. Story (1989), 1989 CanLII 5317 (BC CA), 23 R.F.L. (3d) 225 (B.C.C.A.); Linton v. Linton (1990), 1990 CanLII 2597 (ON CA), 1 O.R. (3d) 1 (C.A.); Romanoff v. Romanoff (1992), 1992 CanLII 13007 (MB QB), 41 R.F.L. (3d) 433 (Man. Q.B.); Droit de la famille‑‑1567, 1992 CanLII 3052 (QC CA), [1992] R.J.Q. 931 (C.A.), and Droit de la famille‑‑1688, 1992 CanLII 3653 (QC CA), [1992] R.J.Q. 2797 (C.A.).

[30] In Bush v. Bush (1989), 1989 CanLII 8812 (ON SC), 21 R.F.L. (3d) 298 (Ont. U.F.C.), Judge Steinberg summarized the state of the law on the trilogy as follows (at pp. 302‑3):

The trilogy has been widely reviewed and judicially interpreted in differing ways and the present state of the law of spousal maintenance is somewhat unclear.

Some courts have concluded that the causal connection test should be given a broad application in all spousal support applications, whether or not the actions are by way of an initial application or application to vary, whether or not there were prior existing agreements regarding maintenance as between the spouses, and whether or not the proceedings are under the Divorce Act, 1985, or the Family Law Act. Other courts have attempted to apply the ratio in the trilogy in a much narrower manner. My view tends towards the latter approach.

It should be noted that the decisions rendered in the trilogy were decided under the now repealed Divorce Act, 1968. The present application is under the Divorce Act, 1985, which contains criteria for the granting or varying of maintenance orders, which did not exist under the old legislation. In that context, I agree with the views of my colleague Judge Mendes da Costa that the principles in the trilogy should not be applied so as to "read into legislation words that are not there": see Andreeff v. Andreeff, Ontario Unified Family Court, 9th May 1989, at p. 11 (now reported 1989 CanLII 4115 (ON SC), 20 R.F.L. (3d) 277 at 285).

[31] One must say that, prior to the trilogy, the state of the law was no clearer. Describing the effect of the more or less incoherent approach taken by courts Chouinard J., in Messier v. Delage, 1983 CanLII 31 (SCC), [1983] 2 S.C.R. 401, said the following (at p. 409):

I cannot state the matter any better than Judge Rosalie S. Abella of the Provincial Court, Family Division, for the judicial district of York in Toronto, did in an article entitled "Economic Adjustment On Marriage Breakdown: Support", (1981) 4 Family Law Review 1. She wrote the following at p. 1:

To try to find a comprehensive philosophy in the avalanche of jurisprudence which is triggered by the Divorce Act (RSC 1970 c D‑8) and the various provincial statutes is to recognize that the law in its present state is a Rubik's cube for which no one yet has written the Solution Book. The result is a patchwork of often conflicting theories and approaches.

[32] For its part, academic commentary has generally been very critical of the trilogy. The commentator T. Heeney objected to the broad interpretation given to it:

In the trilogy, the Supreme Court of Canada has more or less stated that what is done should not be undone. Finality is the pervasive judicial objective that runs through Pelech, Caron and Richardson. . . . Rarely, however, has a case been more misunderstood than Pelech, and rarely have courts been so willing to extract principles out of context, and apply them to fact situations where they simply do not fit. The quest for finality has led judges to cast in stone agreements that were never intended to be permanent and inflexible, and to impose constraints of finality on ongoing maintenance relationships, where finality is impossible. [Footnotes omitted.]

("The Application of Pelech to the Variation of an Ongoing Support Order: Respecting the Intention of the Parties" (1989), 5 C.F.L.Q. 217, at p. 217.)

[33] Similarly, Professor C. J. Rogerson criticized the effect of the causal connection as contemplated in Pelech:

The current causal connection test . . . is more aligned with the clean break model of spousal support. Just as the clean break theory arbitrarily deems self‑sufficiency to exist, the causal link theory arbitrarily deems the causal link to have been broken, with the result that the claimant spouse is deemed responsible for his or her own support. Instead of presuming, until strong evidence to the contrary is led, a causal connection between a spouse's inability to meet his or her needs and what went on during the marital relationship, the causal link test presumes the opposite.

("The Causal Connection Test in Spousal Support Law" (1989), 8 Can. J. Fam. L. 95, at p. 122.)

[34] Professor N. Bala, for his part, cautions:

. . . while the promotion of finality is desirable, this should not be used as a justification for precluding the judicial overriding of unfair agreements. Rather than discouraging the parties from entering into such agreements, knowledge that the courts may intervene to set aside unfair agreements should encourage the parties to initially enter into agreements which are fair. [Emphasis added.]

("Domestic Contracts in Ontario and the Supreme Court Trilogy: `A Deal is a Deal'" (1988), 13 Queen's L.J. 1, at p. 61.)

See also, inter alia, D. G. Duff, "The Supreme Court and the New Family Law: Working through the Pelech Trilogy" (1988), 46 U.T. Fac. L. Rev. 542; J. D. Payne, "Further Reflections on Spousal and Child Support After Pelech, Caron and Richardson" (1989), 20 R.G.D. 477; M. J. Bailey, "Pelech, Caron, and Richardson" (1989‑90), 3 C.J.W.L. 615; P. Proudfoot and K. Jewell, "Restricting Application of the Causal Connection Test: Story v. Story" (1990), 9 Can. J. Fam. L. 143; C. J. Rogerson, "Judicial Interpretation of the Spousal and Child Support Provisions of the Divorce Act, 1985 (Part I)" (1990‑91), 7 C.F.L.Q. 155; M. Neave, "Resolving the dilemma of difference: A critique of `The Role of Private Ordering in Family Law'" (1994), 44 U.T.L.J. 97.

[35] I must refer here to the dissent by La Forest J., which Professors J. W. Durnford and S. J. Toope underline as follows:

La Forest J penned a powerful dissenting opinion in which he challenged the analysis in Pelech by measuring its effects in a case such as Richardson. According to La Forest J, the majority's emphasis upon protecting the sanctity of spousal contracts was misconceived, as was its supposed promotion of equal autonomy. The relevant legislation expressly allowed for judicial discretion in weighing the factors that should be considered in awarding support. The existence of a contract was only one such factor, albeit an important one. More fundamentally, La Forest J questioned the model of rational choice implicit in the majority reasons in Pelech and Richardson. He noted that divorce is one of the most stressful occasions in any person's life and that many people do "very unwise things, things that are anything but mature and sensible, even when they consult legal counsel." Agreements should not be treated as sacrosanct in this emotionally fraught context. [Emphasis added; footnotes omitted.]

("Spousal Support in Family Law and Alimony in the Law of Taxation" (1994), 42 Can. Tax J. 1, at pp. 17‑18.)

[36] Finally, many writers such as Professor Rogerson have pointed out that the discussion prompted by the philosophy underlying the trilogy questions the very concept of marriage and its economic consequences:

What must be realized is that the debate about causal connection is really a debate about the purposes of spousal support in general, a debate which ultimately raises questions about our understanding of marriage as an institution. As Rosalie Abella has argued:

The problem (of the non‑establishment of a uniform philosophy of support) really lies with an inability to agree on what the purposes of economic adjustments on divorce or separation should be. And this, not surprisingly, derives from an inability to agree on what the purpose of marriage should be. [Footnotes omitted.]

("The Causal Connection Test in Spousal Support Law", supra, at p. 103.)

[37] This is precisely what the 1985 Act was to accomplish by, on the one hand, moving away from the tendency favouring a "clean break" which a majority of courts had adopted under the 1968 Act and its implicit presumption of economic self‑sufficiency, and, on the other, by emphasizing the substantive rather than formal equality of the spouses in the marriage and at the time of the divorce.

B. The 1985 Divorce Act

[38] In response to frequent criticism and rapid changes in society, in 1976, the Law Reform Commission of Canada tabled its report on family law, supra, an exhaustive study proposing a sweeping revision of the 1968 Divorce Act based on a different approach (at p. 1):

Changes we have experienced socially, as well as changes in the composition, structure, expectations and thwarted hopes of families and their members have at best led to palliative accommodations by the law to social pressures, such as making divorce generally available, but hardly to a re‑examination of the image of the family the law reflects. This image may by now be so far removed from reality that the law and its institutions may weaken rather than strengthen family life, especially in crisis situations.

[39] The following observation by the Law Reform Commission is wholly relevant to the questions at issue here (at pp. 17‑18):

When a marriage has broken down, there is much that the law ought to do to assist the persons affected by the radical changes in circumstances to adjust to the new situation and to protect those who have relied to their detriment on the expectation that marriage would be permanent. We suggest a shift in legal policy towards a process that focuses on the social and economic implications of marriage breakdown for the spouses and their children, premised on finding fair and constructive solutions to the problems resulting from the ending of this most important human relationship.

[40] The federal Department of Justice later issued a series of proposals based on this report, which eventually served as a model for the 1985 Divorce Act. As regards support orders, the Department of Justice recommended the following approach:

The objectives of maintenance should be clearly spelled out in the Divorce Act to ensure that, as far as possible, the economic consequences of divorce are settled fairly and consistently in divorce cases. These objectives should establish general guidelines to be used in determining the nature and value of maintenance awards, ruling out such considerations as the behaviour of the spouses.

(Canada, Department of Justice, Divorce Law In Canada: Proposals For Change (1984), at p. 22.)

These principles are to be found in ss. 15 and 17 of the 1985 Act, which I have earlier cited in part.

[41] The current Act, therefore, adopted as its underlying philosophy a partnership in marriage and, at the time of a divorce, an equitable division of its economic consequences between the spouses. It thus rejected the presumption of economic self‑sufficiency and substituted for it a number of criteria that would take into account the advantages and disadvantages to spouses accruing from the marriage or its breakdown. It did not, however, completely depart from the objective of economic self‑sufficiency, although it underlined that this objective can only be pursued "in so far as practicable".

[42] Paradoxically, after the enactment of the current Act, courts continued to rely on the pattern established under the 1968 Act, namely the clean break model of self‑sufficiency as a primary objective, i.e. economic self‑sufficiency of the recipient of alimony, generally the wife, as soon as possible after the divorce. This tendency, which actually became more marked after 1985, was attributed to the Supreme Court's judgments in Messier v. Delage, supra, and the trilogy, and, in particular, to the causal connection which it advocated, although these cases were decided under the 1968 Act.

[43] Vigorous criticism from academic writers (see, for example, Rogerson, "Judicial Interpretation of the Spousal and Child Support Provisions of the Divorce Act, 1985 (Part I)", supra; Payne, "Further Reflections on Spousal and Child Support After Pelech, Caron and Richardson", supra; D. Goubau, "Une nouvelle ère pour la pension alimentaire entre ex‑conjoints au Canada" (1993), 72 Can. Bar Rev. 279), and the reaction of certain courts, in particular the Quebec Court of Appeal (Droit de la famille‑‑1688, supra, per LeBel J.A.), were quick to challenge this trend, which they contended was at odds with the very language of the 1985 Act and its underlying philosophy. In Moge v. Moge, supra, the Supreme Court sought to put an end to the uncertainty that prevailed by adopting an interpretation of the 1985 Act consistent with its language and underlying philosophy, especially as regards support obligations and their subsequent variation.

[44] That judgment clearly rejected the concept of formal equality which had previously prevailed and, with it, the main assumption at the heart of the economic self‑sufficiency model:

All four of the objectives defined in the Act must be taken into account when spousal support is claimed or an order for spousal support is sought to be varied. No single objective is paramount. The fact that one of the objectives, such as economic self‑sufficiency, has been attained does not necessarily dispose of the matter.

Many proponents of the deemed self‑sufficiency model effectively elevate it to the pre‑eminent objective in determining the right to, quantum and duration of spousal support. In my opinion, this approach is not consonant with proper principles of statutory interpretation. The objective of self‑sufficiency is only one of several objectives enumerated in the section and, given the manner in which Parliament has set out those objectives, I see no indication that any one is to be given priority. Parliament, in my opinion, intended that support reflect the diverse dynamics of many unique marital relationships.

It is also imperative to realize that the objective of self‑sufficiency is tempered by the caveat that it is to be made a goal only "in so far as practicable". This qualification militates against the kind of "sink or swim" stance upon which the deemed self‑sufficiency model is premised.

In the result, I am respectfully of the view that the support model of self‑sufficiency . . . cannot be supported as a matter of statutory interpretation, considering in particular the diversity of objectives set out in the Act.

(Moge, supra, at pp. 852, 853 and 857-58.)

[45] This is the background against which this appeal comes to the Court. The specific point at issue is whether the criteria set out in the trilogy should continue to be applied under the 1985 Act.

[46] In my opinion, the answer must be no and, in this connection, I agree with the analysis of Professor Payne in "Spousal and Child Support After Moge, Willick and Levesque" (1995), 12 C.F.L.Q. 261, at p. 271:

In my opinion, Moge v. Moge clearly demonstrates that we must turn to the Divorce Act to ascertain its application in all divorce cases where spousal support is sought, including those where a separation agreement has been negotiated to regulate spousal support. Under subsection 15(5) of the Divorce Act, an agreement is only one of many factors that must be considered in determining the right to, amount and duration of spousal support. Furthermore, all four of the objectives of spousal support orders under subsection 15(7) must be considered; their application is not excluded simply because a final agreement or settlement has been negotiated. There is no requirement or causal connection built into subsection 15(5) of the Divorce Act, although the notion of causal connection is recognized in subsections 15(7) and 17(7) of the Divorce Act. It is doubtful whether the principles defined in Pelech, Caron and Richardson, which were enunciated in the context of the Divorce Act, 1968, can survive. Although the Supreme Court of Canada trilogy was intended to provide more definitive guidelines that legal practitioners could apply with respect to the right to, amount and duration of spousal support, it has generated more confusion than precision and for that reason, if for no other, is unlikely to survive when the Supreme Court of Canada reviews the question. [Emphasis added; footnotes omitted.]

See also Goubau, supra, at p. 300.

[47] The question here at issue must, therefore, be considered in light of the 1985 Act and the criteria which flow from it, as interpreted in Moge, supra. First, it must be determined whether a sufficient change has occurred between the parties to require the court's intervention. Second, the court must assess the effect of the agreement in light of the factors and objectives that govern spousal support under ss. 15(5), 15(7), 15(8), 17(4), 17(7) and 17(8) of the Act.

C. Variation of Support Orders

[48] Variation of support orders is governed by s. 17 of the current Act. Section 17(4) states that before it varies a support order "the court shall satisfy itself that there has been a change in the condition, means, needs or other circumstances of either former spouse or of any child of the marriage". This provision applies to any variation of a support order, whether in favour of a spouse or of the children. Consequently, the test developed in Willick v. Willick, 1994 CanLII 28 (SCC), [1994] 3 S.C.R. 670, in connection with support for children applies mutatis mutandis to the variation of spousal support. To begin with, the judge must determine whether there has been a change in the parties' situation since the last support order.

(1) Change

[49] What sort of change is appropriate? Willick, supra, explained what is meant by "change". Sopinka J. said the following (at p. 688):

In deciding whether the conditions for variation exist, it is common ground that the change must be a material change of circumstances. This means a change, such that, if known at the time, would likely have resulted in different terms. The corollary to this is that if the matter which is relied on as constituting a change was known at the relevant time it cannot be relied on as the basis for variation.

[50] In that same decision I made the following observations (at pp. 733‑34):

In my view, having regard for the wording of s. 17(4) of the Act, the preliminary threshold test ensures that . . . support orders will not be reassessed by courts anytime a change, however minimal, occurs in the circumstances of the parties or their children. This approach recognizes the value in some degree of certainty and stability between the parties. Parties must be encouraged to settle their difficulties without coming before the courts on each and every occasion. Nonetheless, the threshold test cannot be applied properly unless the sufficiency of the change in circumstances is evaluated against the backdrop of the particular facts of the case at hand. It is important to point out that the Act does not qualify "change" but merely states that "the court shall satisfy itself that there has been a change". . . . [Emphasis in original.]

[51] As to what change is sufficient, I went on to say (at p. 734):

To begin with, "sufficiency" of the "change" must be defined in terms of the parties' overall financial situation. Moreover, the fact that a change was objectively foreseeable does not necessarily mean that it was contemplated by the parties. Finally, although any change which is not contemplated may be considered by a judge to be sufficient, it is obvious that not every change will justify variation. Most importantly, however, and notwithstanding the above observations, while the onus of proving the sufficiency of the change in condition, means, needs or other circumstances rests upon the applicant . . ., the diversity of possible scenarios in family law dictates that courts maintain a flexible standard of judicial discretion which does not artificially limit the adaptability of the Divorce Act provisions. [Emphasis in original.]

(2) Agreement

[52] The analysis which the court must undertake to determine the extent of the variation, once the sufficiency of a change has been established, was discussed in Willick, supra, where I noted in this connection (at pp. 734‑35):

Once a sufficient change that will justify variation has been identified, the court must next determine the extent to which it will reconsider the circumstances underlying, and the basis for, the support order itself. For the reasons below, I believe that it is artificial for a court to restrict its analysis strictly to the change which has justified variation. Moreover, while a variation hearing is neither an appeal nor a trial de novo, where the alleged change or changes are of such a nature or magnitude as to make the original order irrelevant or no longer appropriate, then an assessment of the entirety of the present circumstances of the parties . . . is in order.

[53] Section 15(5)(c) of the 1985 Act, which governs the initial support order, specifically provides, with respect to agreements entered into between spouses in contemplation of their divorce, that "[i]n making an order under this section, the court shall take into consideration . . . any order, agreement or arrangement relating to support of the spouse or child". An agreement is only one of the factors listed in s. 15(5) that should be taken into account in assessing inter alia the duration and amount of spousal support (Brockie v. Brockie (1987), 1987 CanLII 141 (MB CA), 8 R.F.L. (3d) 302 (Man. C.A.)). In addition, the four objectives mentioned in s. 15(7) must also be considered.

[54] In s. 15(5), no mention is made of a necessary causal connection: it differs in this respect from s. 17(10), which does not apply here. In so far as a duty of support results from marriage (or from cohabitation under certain provincial statutes), the notion of causality can only be dealt with in accordance with the various principles and objectives set out in the 1985 Act, as Osborne J.A. noted in Linton, supra, at pp. 23 and 28-29:

Although there are many cases in which causal connection principles have been applied without much, if any, reference to the Divorce Act (1985) in my opinion causal connection principles are only relevant to the extent that a causal connection requirement or objective may be taken to emerge from the Divorce Act (1985). . . .

To the extent that causal connection principles can be found in s. 15(7)(a) of the Divorce Act, I do not think that those principles should be used to provide for a clean break or deemed self‑sufficiency approach to support. . . . Although I acknowledge that self‑sufficiency may lead to a clean break, it must be remembered that the objective of self‑sufficiency, as referred to in s. 15(7)(d), is qualified. Thus, the clean break approach to support must be viewed as having a limited legislative underpinning.

[55] Section 17, which governs variation orders, restates for its part the general provisions applicable to a support order without specifically mentioning the obligation to take into account agreements concluded between the parties. One should not conclude, however, that such agreements should be ignored when applications to vary support orders are made, especially when they were intended to be a final settlement, and were ratified by the original support order, an order which must be taken into account. As Wilson J. said in Pelech (at p. 850):

People should be encouraged to take responsibility for their own lives and their own decisions. This should be the overriding policy consideration.

[56] Having said that, while it is true that the parties should be encouraged to reach an agreement on the economic consequences resulting from their divorce rather than going to the courts, such agreements are only one factor, "albeit an important one", which must be considered in the exercise of the judge's discretionary power (Durnford and Toope, supra). The weight to be given to agreements will depend, first, on the extent to which the agreement reflects the principles and objectives stated in s. 17 of the 1985 Act and, second, on the scope and nature of the change which has occurred, taking into account all the circumstances of the parties. The more the agreement or support order takes into account the various objectives of the Act, especially that of promoting an equitable distribution of the economic consequences of the marriage and its breakdown, the more likely it will be to influence the outcome of the variation application. In drafting future agreements, counsel would be well advised to articulate the bases on which both spousal and child support covenants have been negotiated.

[57] I should mention in passing that, in such an equitable distribution, the spouses may agree on various ways of dividing their assets depending on their financial situation at the time of divorce. Thus, spousal support for an unlimited time, such as that contemplated by the parties here, may well be compensatory in nature rather than simply needs based, especially when the parties do not have the resources to make a lump sum payment at the time of divorce, which may also, in some cases, confer some tax advantages. In the absence of any such mention in the agreement, the parties' intention in this regard cannot be presumed. However, it is clear here that nowhere in the agreement is there a mention of any attempt to dispose of the financial resources of the parties in accordance with the criteria laid down by the 1985 Act. No mention is made, for example, of the advantages and disadvantages resulting to the former spouses from the marriage or from its failure. It would thus be open to conclude that spousal support, unlimited as to time, could have been set taking these matters into account. Those are facts which the trial judge cannot ignore on an application to vary where an agreement has been entered into between the parties at the time of divorce or later. In this connection, as I noted earlier, it would clearly be useful for the parties to mention the various factors and objectives they took into account in their agreement to share the economic consequences of the marriage and its breakdown.

[58] Having said that, the fact remains that, under the 1985 Divorce Act, courts retain a discretionary power the exercise of which will depend on the particular facts of each case and which will be exercised in accordance with the factors and objectives mentioned in the 1985 Act. The existence of an agreement, final or otherwise, should not have the effect of precluding such an analysis.

[59] This is the background against which the appeal at bar must now be decided.

V. Application to the Facts

[60] In light of the foregoing I consider that Benoit J. correctly applied the 1985 Act and made no error in his analysis of the evidence.

[61] In particular, the trial judge took into account, as one factor among others, the agreement concluded between the parties at the time of the initial support order which one now wants to vary. He went on to consider whether there was evidence of a change since the support order, the nature and magnitude of which would justify the variation. After a careful review of the evidence, he came to the conclusion that such evidence had not been presented and he dismissed the respondent's application.

[62] The respondent had relied on two events which, in his opinion, justified intervention by the Court. The first concerned the parties' son, F., who was a minor at the time and living with his father. Given that F. is now married and self‑supporting, which was not the case before the trial judge and in the Court of Appeal, the nature and magnitude of the alleged change are such that the child support of $100 a month should be cancelled, as it was by the Court of Appeal.

[63] The other change alleged is the fact that, since 1989, the appellant has been living with the person she was seeing at the time of the divorce. The trial judge analysed the evidence in this regard and concluded:

[translation] The fact that [the appellant] shares an apartment with a third person is a new fact. The question that must be answered is whether this new fact is a change that can affect the support order. At the time the consent was signed [the appellant] was seeing this third person, to [the respondent's] knowledge. It was in the ordinary course of events foreseeable that after the divorce, especially in the circumstances, [the appellant] would live with the third party or some other third party. The hypocritical maintenance of separate residences should not be encouraged in any case. Although the event was foreseeable, [the respondent] did not see fit at the time to stipulate that entitlement to the alimony would cease, or even that it would be reduced, if the respondent lived with someone else. As [the appellant] has not remarried, she cannot exercise any right against her companion and he has no obligation to her. Additionally, there is no guarantee that the relationship between [the appellant] and the third party will be permanent. . . . For the moment the situation should not cause the respondent to lose her right to alimony.

[64] On the same evidence, the Court of Appeal came to a different conclusion solely on the basis of what it called the "presumption of self‑sufficiency". If I understand this concept fully, as the Court of Appeal applied it to the circumstances of the case at bar, the fact of living together creates a presumption that the recipient of spousal support has become financially self‑sufficient and that, accordingly, at least prima facie, the payer of the support is released in whole or in part from his or her support obligation.

[65] There is nothing in the 1985 Act that supports such a presumption. While the Act deals with the economic self‑sufficiency of the former spouses, it is only as one of the many objectives that must be met by the support order (s. 17(7)(d)). What is more, this concept of economic independence is only favoured "in so far as practicable".

[66] If such a "self‑sufficiency presumption" may have existed due to the interpretation given to the 1968 Act by doctrine and jurisprudence, it is no longer possible under the 1985 Act. The main decision cited by the Court of Appeal in support of the rule on which it relied, Droit de la famille‑‑333, [1987] R.J.Q. 294, in fact concerned a support order rendered pursuant to the former Act. As I mentioned earlier, such an interpretation is no longer possible under the current Act for the reasons that I have discussed above.

[67] Clearly, if the support order provides that spousal support will cease as soon as the former spouse attains economic self‑sufficiency, as this concept is expressed in the 1985 Act, and this is a question of fact in each case, then the support obligation will end. However, such economic self‑sufficiency is not to be presumed: it must be proven.

[68] In the case at bar such evidence has not been presented. First, the agreement, made and ratified by the support order initially rendered, was drawn up when the appellant was seeing the person who foreseeably, would become her companion in the not too distant future. Second, the possibility that the appellant would become financially self‑sufficient and rejoin the labour market, in view of her age, 53, her absence from the labour market for many years and, presumably, the lack of any adequate professional training in a competitive labour market, was entirely illusory, quite apart from the fact that she was unable to accumulate retirement pension and enjoy other benefits associated with employment. Despite the fact that there was no evidence in this regard, the parties were undoubtedly well aware of the situation since the support obligation undertaken by the respondent was not limited in time, but only as to the amount and, for that matter, conditionally.

[69] Finally, the fact that the appellant is living with a companion certainly does not mean that she can be presumed to be financially independent, no more than the fact that if the respondent had been in the same circumstances as the appellant at the time of the divorce, he could have been regarded as financially independent merely because he was living with a companion who was in the labour market and sharing common expenses, as well as giving him a gift and a loan. In short, the evidence discloses no change that would be sufficient to justify variation of the support order rendered four years earlier when the divorce was granted.

[70] In these circumstances, given the current Act and the evidence in the case at bar, including the initial support order and the agreement entered into between the parties, which, in my view, was intended to be final, the Court of Appeal was not justified in concluding that the appellant had attained adequate financial self‑sufficiency to relieve the respondent of part of his support obligation as stipulated in the support order. Most importantly, it could not start from a presumption of financial self‑sufficiency which is not supported either by the 1985 Act or the interpretation given to it (Moge, supra). I note that the Court of Appeal's judgment differs from an earlier judgment of the same court in which LeBel J.A wrote:

[translation] Reading the provisions of the [1985] Divorce Act confirms both the complexity and diversity of the objectives sought both when the initial support is granted and when a variation order is made. Making the attaining of financial self‑sufficiency as quickly as possible and a clean break in prior relations the central objective of this legislation is an oversimplification.

(Droit de la famille‑‑1688, supra, at p. 2804.)

[71] In light of the foregoing, in view of the error of principle it made and the deference owed by an appellate court to the findings of a trial judge, the intervention of the Court of Appeal was inappropriate.

VI. Conclusion

[72] For these reasons I would allow the appeal, reverse the judgment of the Court of Appeal and restore the order made by the trial judge, except as to the support order payable for the parties' son, which I would cancel from the date the appeal at bar was heard, the whole with costs to the appellant in all courts.

The judgment of Sopinka, Cory, McLachlin and Iacobucci JJ. was delivered by

[73] Sopinka J. -- I have read the reasons prepared by my colleague L'Heureux-Dubé J. and agree with the disposition of the appeal which she proposes. I also agree with her reasons wherein she states that variation of support orders is governed by s. 17 of the current Divorce Act, R.S.C., 1985, c. 3 (2nd Supp.), and that the test developed in Willick v. Willick, 1994 CanLII 28 (SCC), [1994] 3 S.C.R. 670, applies. In that case, a majority agreed with the following statement of the threshold test for determining a change of circumstances (at p. 688):

In deciding whether the conditions for variation exist, it is common ground that the change must be a material change of circumstances. This means a change, such that, if known at the time, would likely have resulted in different terms. The corollary to this is that if the matter which is relied on as constituting a change was known at the relevant time it cannot be relied on as the basis for variation.

[74] Here, the trial judge found as a fact that at the time of the agreement the respondent knew that the appellant was "seeing" the third party and that it was foreseeable that they would cohabit. In view of this finding, the trial judge correctly concluded that there was no material change of circumstances. There was no basis in fact or law for the Court of Appeal to reverse this conclusion.

[75] While I fully agree that this Court, in an appropriate case, will have to review the application of the trilogy (Pelech v. Pelech, 1987 CanLII 57 (SCC), [1987] 1 S.C.R. 801, Richardson v. Richardson, 1987 CanLII 58 (SCC), [1987] 1 S.C.R. 857, and Caron v. Caron, 1987 CanLII 59 (SCC), [1987] 1 S.C.R. 892) to the support provisions of the 1985 Divorce Act, in my view this is not the appropriate case. Apart from being unnecessary to this decision, the case was not presented on this basis. No mention of the trilogy is made in the reasons of the trial judge. The Court of Appeal, while mentioning the trilogy, found the cases unapplicable and reversed the trial judge on the basis of a presumption of self-sufficiency based on the appellant's new relationship. I agree with L'Heureux-Dubé J. that the Court of Appeal erred in this respect. In this Court, any reference to the trilogy was at best at the periphery of the main argument.

[76] I would, however, dispose of the appeal as proposed by my colleague.

Appeal allowed with costs.

Solicitors for the appellant: Alarie, Legault & Associés, Montreal.

Solicitors for the respondent: Martineau, Walker, Montreal.