Noelle Heen-Lune Rathee v. Rajinder Singh Rathee

COURT FILE NO.: FS-19-00014536-0000

DATE: 2023-12-12

SUPERIOR COURT OF JUSTICE - ONTARIO

RE: NOELLE HEEN-LUNE RATHEE, Appellant

AND:

RAJINDER SINGH RATHEE, Respondent

BEFORE: R. L. Akazaki J.

COUNSEL: Lorne Wolfson, for the Applicant Ida Sherkat, for the Applicant Ilana Zylberman Dembo, for the Respondent Batya Berman, for the Respondent

HEARD: October 3, 2023

REASONS FOR JUDGEMENT

AKAZAKI J.

EXECUTIVE SUMMARY

[1] This was an appeal of a family law arbitrator’s spousal support award. The Respondent husband is an ophthalmologist with a thriving practice. His annual income doubled to $1,552,000 during his seven-year marriage to the Applicant wife. During that time, she left her managerial position at a pharmaceutical company to look after their daughter. He put her on the payroll of his professional corporation to run the household spending through its books and lower his income tax. In the summer of 2013, he consulted five family lawyers to help him plan leaving his wife. A month or two later, he found a new house, moved out his home office paraphernalia, and told his wife that he was moving out. She was blindsided.

[2] Before he left, the husband shopped online for furniture and even asked her what she thought about specific pieces. He tried to calm her distress by promising to discharge all his financial support obligations through the legal process. In the meantime, he expected her to remain in the matrimonial home and to continue looking after their five-year-old. He would keep his wife on the income-splitting payroll of his professional corporation to reimburse her for household expenses.

[3] For a while, he returned to the house often. He participated in marriage counselling and led his wife to believe that they could get back together. He even took his wife and daughter on holidays. But by 2019, he wanted to move on. He started to pay child support. The wife then agreed to a mediation-arbitration procedure with a senior family law practitioner. In 2020, he removed his wife from the payroll. Mediation failed.

[4] They proceeded to arbitration. The main dispute consisted of the wife’s pursuit of equalization of net family property, ongoing child support (monthly table amounts and “s. 7” expenses), spousal support of $29,689 per month within the range calculated under the Spousal Support Advisory Guidelines (“SSAGs”), and retroactive child and spousal support in the amount of $3,000,000 for nine years of unpaid monthly instalments to the date of the hearing.

[5] The husband did not dispute liability for child support. However, he contended the monthly amount should be lower and that all his “status-quo" income-splitting payments and paid child support should offset his retroactive child support obligations. He also took the position that he owed the wife no spousal support. He contended she could resume the career she had left after their child was born. He also argued that she failed to give effective notice of the claim until the start of proceedings in 2019. In addition to denying any spousal support obligation, he wanted to back-charge her rent totaling $362,688. After claiming this and other credits on his side of the support ledger, he took the position that she owed him $350,373, and that this sum should be clawed back against her share of the net family property.

[6] The arbitrator equalized the parties’ net family property by requiring the husband to pay $682,382, including prejudgment interest. He found the husband had underpaid child support during the 2013-2022 period by $106,397, after applying all the maintenance and child support payments. He dismissed the claim for occupation rent and made some directions regarding life insurance and the sale of the matrimonial home. He then fixed a sum of $250,000 for all past and future spousal support, including prejudgment interest, and awarded $171,330 in costs to the husband as the successful party in the dispute.

[7] As I will calculate later, the lump sum, net of interest, represented monthly spousal support obligations during the 2013-2022 period of $2,238.17 despite his monthly earnings of well over $100,000. The wife appealed the spousal support and costs awards. Her main objection to the support award was that the arbitrator gave no reasons for departing so radically from the income-based range in the SSAGs. The SSAGs, she contended, would have called for support from the high-earning husband that was many, many times greater, and for a duration for as long as the daughter’s graduation from high school.

[8] On appeal, the husband’s counsel agreed that the arbitrator did not explain how he arrived at the $250,000 sum. His counsel’s DivorceMate figures for the expected range of monthly spousal support did not differ much from the wife’s calculations. The defence of the award consisted of a “reverse engineering” of the $250,000 lump sum figure based on two assumptions. First, he assumed the arbitrator imputed a significant discretionary discount of income above the “ceiling” annual income of $350,000. Second, he assumed the arbitrator accepted his defence that no spousal support was claimable before 2019, yielding a duration of support between three and four years. Both assumptions were contradicted by express findings.

[9] The arbitrator imputed the husband’s full income, as proven at the hearing, for child and spousal support. The arbitrator also rejected the husband’s delayed notice argument. (Even if the arbitrator had limited the duration to three or four years, the husband would have owed spousal support of between $900,000 and $1,250,000 even at the low end of the husband’s SSAGs range.) The arbitrator award of retroactive child support from 2013 to 2021 and acceptance of the husband’s offset for the same period was also consistent with his rejection of the delayed notice argument.

[10] The arbitrator’s actual rationale for his award, in paragraphs 153-57 of his decision, was that a lump sum as described in Davis v. Crawford, 2011 ONCA 294, would satisfy the wife’s entitlement to retroactive and prospective spousal support. The lump sum described in the Davis case referred to a present value of prospective periodic support that would allow a clean break between the parties if the payor could afford to pay. The principle did not allow the arbitrator to award of a lump sum to reduce the actual value of a $30,000 per month obligation over nine years to $250,000.

[11] Were it not for this fatal legal error and unreasonable result, the arbitral reasons should have provided the factual findings and data inputs for a spousal support calculation of the monthly support under the with-child-support SSAGs formula starting in 2013:

a) the parties’ actual or imputed incomes;

b) the use of the total incomes for support purposes, without grounds for discretionary reduction of the amounts above $350,000;

c) quantum of child support and allocation of private school tuition;

d) the husband’s election to apply all of his payments as credit toward unpaid child support, meaning he had paid no spousal support;

e) delay not affecting the wife’s entitlement; and

f) the wife’s reliance on her husband’s promise to pay spousal support once it was determined in the legal process.

[12] The formula based on the above inputs inescapably resulted in accrued spousal support in the order of $3,000,000 or more. For reasons known only to the wife and her counsel, she reduced her support demand from $3,000,000 to $2,000,000 in the closing submission at the arbitration hearing. Since the husband was permitted to use all his prior payments to reduce accrued child support to $106,387, this proved a costly waiver of $1,000,000. After deducting the arrears in child support from the $2,000,000, the court is able to substitute the $250,000 lump sum with the amount of $1,893,603 in spousal support arrears. The error in principle therefore resulted in an unreasonable outcome and therefore requires the court to set the spousal support award aside.

[13] The costs award must also be set aside. In case I am incorrect in my disposition of the main issue on appeal, I will state the reasons why the wife was the successful party in the arbitration, even if the award of spousal support was much less than what she had reason to expect.

[14] The balance of these reasons will be organized in the following order:

• Relevant arbitral findings and reasons for the awards under appeal.

• Appellate jurisdiction and reasonableness as the applicable standard of review on questions of law.

• Adequacy of the arbitrator’s reasons.

• The wife’s entitlement to spousal support from 2013 to 2022.

• The arbitrator’s rejection of the husband’s arguments for limiting the duration of the award on the grounds of delayed notice, and additional grounds for rejecting them, based on promissory estoppel.

• The range of monthly spousal support the arbitrator ought to have used to determine the periodic entitlement under the SSAGs formula.

• The arbitrators’ error of law in applying the Davis criteria regarding lump sums to pick an arbitrarily low global amount of spousal support instead of adding up the monthly payments.

• The wife’s self-imposed $2,000,000 cap for child support and spousal support, limiting the spousal support claim to $1,893,603 after deducting $106,397 in accumulated child support arrears.

• The costs decision was untenable, in any event of the appeal.

• Conclusion summarizing the disposition of the appeal.

ARBITRAL FINDINGS AND REASONS FOR THE SPOUSAL SUPPORT AND COSTS AWARDS

[15] The property division and equalization aspects of the award were not in contention. I will not refer to them except later when I deal with the appeal from the costs award. However, the issues pertaining to occupation rent, child support and spousal support were connected.

[16] The starting point for examining the award for spousal support is the arbitrator’s rejection of the wife’s demands of a combined lump sum for retroactive child and spousal support. At paragraph 126 of his decision, the arbitrator “categorically” rejected the wife’s “ask” of a $2,000,000 lump sum award of spousal and child support as “completely excessive, punitive and unreasonable.”

[17] I pause to observe that a rough calculation of the spousal support arrears was hidden in plain sight. The arbitrator had, in paragraph 123, just found that the husband’s assets had grown by $5,600,000 during the same post-separation period, as a result of the parties’ agreement to continue the status quo. In paragraph 129, he calculated the child support in the amount of $945,978. As I will describe later, the SSAGs with-child-support formula for spousal support was 40% to 46% of the net income differential after deduction for child support (SSAGs, p. 33). Assuming the increase of capital was the accumulation of net disposable income, this meant the range of spousal support would have been $1,861,680 to $2,140,850, and that the “ask” was right at the bottom of the range.

[18] Ordinarily, the separation date is the first reference point for a calculation of entitlement. Where one spouse’s income far exceeds the other’s, the law presumes an immediate division of income. Thus, for the sake of argument, the wife should have expected a monthly spousal support payment from her husband based on his 2013 income of $1,552,000 (subject to discretion over $350,000) and her income of $86,209. Using the calculations by the husband’s counsel with average figures in the ensuing period, the husband would have started to owe an amount in the order of $30,000 per month.

[19] In paragraph 129, the arbitrator allowed the husband to apply all the “status quo” payments to reduce his retroactive child support obligations from $945,798 to $106,397. Beyond this offset, the arbitrator did seem aware that the husband’s demand for a $350,373 reimbursement of support overpayments was smoke and mirrors. In particular, the husband:

1. used the uncharacterized “status quo” payments to offset retrospective child support obligations, relying on the priority of child support over spousal support under s. 15.3 of the Divorce Act, R.S.C., 1985, c. 3 (2nd Supp.);

2. used the wife’s occupation of the home and the same “status quo” payments as the basis for denying her claim to spousal support as being separate from child support, on the basis that such occupation allowed her to be self-sufficient and that was at the time she and the daughter left the home; and

3. charged his wife occupation rent, for the use of the home that he had already sought credit as ersatz support.

[20] At paragraph 112, the arbitrator noted the husband’s demand for $350,373.26 by asking that he “monetize Ms. Rathee’s benefit in remaining in the home from October 2013 until now.” Earlier, at paragraphs 40-48, the arbitrator had denied the occupation rent claim on the basis that (a) the husband had “conduced to the wife’s remaining in the home as part of maintaining the status quo on separation,” (b) that it provided the child the stable home that both parents wanted, and (c) both parties benefited from its escalation in value during the interval. In paragraph 123, the arbitrator appeared to have rejected the husband’s attempt to monetize the occupation of the home and to have applied a more standard family law analysis:

123. … Dr. Rathee’s assets grew from separation until the hearing before me from $4.7 million to $10.3 million while Ms. Rathee’s grew from $2.4 million to $2.9 million, again, speaking to Dr. Rathee’s advantages from the marriage if not also the marriage breakdown as Ms. Rathee assumed the bulk of child-rearing.

[21] I pause to observe that the above findings appeared out of place in the section dealing with the issue of child support. Nevertheless, based on the findings in paragraph 123, most family law jurists would have concluded that the compensatory grounds for spousal support under s. 15.2 of the Divorce Act were very strong during the 2013-2022 period and should have extended the duration of support until the child completed high school. However, at paragraph 94, the arbitrator had already concluded, in another non sequitur: “On the evidence before me and having regard to the relevant statutory considerations, I have concluded that the wife’s spousal support entitlement is at or near the end.”

[22] After dealing with allocation of s. 7 expenses and security for support, the arbitrator turned to his award of spousal support starting in paragraph 145. After considering the wife’s demand, he rejected the husband’s argument that it should be disallowed for the period before 2019 because of her delay in pursuing support. He then concluded, at paragraphs 153-157, that a lump sum would “satisfy Ms. Rathee’s retroactive and prospective entitlement, to both compensatory and non-compensatory support.” He did not provide a calculation or other explanation for the $250,000 figure. Instead, he cited the Davis criteria for awarding a lump sum in lieu of periodic support as his rationale for arriving at the $250,000 figure and then concluded:

156. I find the lump sum suggested by Mr. Wolfson excessive and unreasonable based on the evidence. I also find the amount suggested by Ms. Zylberman Dembo inadequate given the roles assumed in the marriage and Ms. Rathee’s career interruption needing more redress than Dr. Rathee acknowledges.

157. In consideration of all the objectives and factors of a spousal support award, including the parties’ means, needs and financial circumstances, I find that a lump sum payment of $250,000 [per fn. 47: including prejudgment interest] will fully compensate Ms. Rathee for any disadvantage arising from the marriage and marriage breakdown and the roles and responsibilities assumed during the marriage, especially considering its relatively short duration and the delay in pursuing her claims.

[23] The arbitrator released supplementary reasons to clarify various issues raised by the parties, but nothing further regarding his spousal support award. He provided a calculation of prejudgment interest on the equalization payment, but not for the retroactive portion of spousal support.

[24] The arbitrator’s ruling regarding costs focused on various issues in the arbitration. He concluded that the husband prevailed, or at least prevented the full extent of the wife’s claims. At paragraph 40 of the costs award, he concluded that amounts sought by both parties in their bills of costs were reasonable.

JURISDICTION AND STANDARD OF REVIEW

[25] The jurisdiction to hear the appeal from a family arbitration lies to this court under s. 45(6)(b) of the Arbitration Act, 1991, S.O. 1991, c. 17. The types of questions that are subject to appellate review depend on the provisions of the arbitration agreement. The standard of review depends on several factors beyond the requirement that questions of fact and mixed fact and law require deference to the arbitrator and questions of law do not. The general reluctance to disturb arbitral findings stems from the parties’ contractual choice to have their dispute resolved privately, but there is a countervailing contractual intention to allow a full appeal where one is provided in the agreement. Where the issue is spousal support, there is a specialized standard of appellate review resembling a “reasonableness” standard used in administrative law.

[26] If the arbitration agreement did not confer a right of appeal, an appeal would have been permitted only with leave of the court under s. 45(1). The general reluctance of courts to interfere with arbitral awards stems from the case law regarding leave applications. The implied agreement of the parties was to choose private arbitration as a speedy and final resolution of disputes. The default rule under s. 45(1) therefore requires courts to be cautious about granting leave to appeal only where questions of law can be “clearly perceived and delineated”: Elk Valley Coal Partnership v. Westshore Terminals Ltd. (2008), 2008 BCCA 154, 291 D.L.R. (4th) 645, at para. 17. When leave is granted, appeals restricted to questions of law require that the appellate court hold the arbitrator to a reasonableness standard and not a correctness standard, unless the question of law entails constitutional law or importance to the legal system. The rationale for this additional deference is that the parties have chosen their own decision-maker based on subject-matter expertise and are therefore accorded deference similar to that of an administrative tribunal in interpreting the law applicable to a specialized area: Sattva Capital Corp. v. Creston Moly Corp., 2014 SCC 53, [2014] 2 S.C.R. 633, at para. 105.

[27] If the agreement provides a right of appeal, under s. 45(2) and (3) the parties may appeal an arbitration award on questions of law, fact, or of mixed fact and law. In the mediation-arbitration agreement in this matter, art. 14.1 provided appeals as of right on all three types of questions.

[28] The express agreement of the parties to allow a direct right of appeal on the three types of questions denotes the intention of the parties to be governed by a standard of appeal equivalent to the appeal of a trial judgment to a provincial court of appeal. The Superior Court should not interfere with the parties’ freedom to contract: Schickedanz v. Wagema Holdings Ltd., 2022 ONSC 5315, [2022] O.J. No. 4206, at paras. 18-19, relying on TELUS Communications Inc. v. Wellman, 2019 SCC 19, [2019] 2 S.C.R. 144, at para. 131. The appellant wife thus argued that the arbitrator’s decision deserved “as much deference on appeal as does a decision of a trial judge,” and relied on Rosenberg v. Yanofsky, 2019 ONSC 6886, [2019] O.J. No. 6057, at paras. 7-8.

[29] Tracing the judicial annotations back from Rosenberg, the origin of the above statement regarding equivalence with appeals from trial decisions was Fisher v. Fisher, 2008 ONCA 11, 88 O.R. (3d) 241, at para. 25, which referred to the specific experience of the trial judge in that case. In Davis, at para. 82, the appellate court assumed that an “experienced family law judge” would have considered tax implications of a lump sum spousal support award even though the issue was not spelled out in the decision. The idea that an appellate court’s deference could depend on the trial judge’s experience appears to stretch the role of casuistry underlying case-based reasoning in the common law. Would it have been open for the court in Davis to assume that a judge without family-law experience had not considered the tax implications? I do not read Davis to mean there should be a sliding scale of deference based on the known experience level of the trial judge or arbitrator, or that seasoned adjudicators should be accorded a free pass in articulating reasons. I therefore do not accord the arbitrator with experiential deference apart from the parties’ choice to choose a distinguished senior member of the family bar.

[30] Both parties relied on the leading case of Housen v. Nikolaisen, 2002 SCC 33, [2002] 2 S.C.R. 235, in which the Supreme Court of Canada set out the level of deference to lower courts and tribunals based on the category of question being put to the appellate court, which I will rephrase for present purposes.

a. On pure questions of law, the standard is correctness, and an appellate court can substitute the opinion of the arbitrator with its own.

b. Findings of fact cannot be reversed unless the arbitrator has made a “palpable and overriding error.”

c. On questions of mixed fact and law the standard is also “palpable and overriding error,” unless the arbitrator made some extricable error in legal principle. That legal error is subject to a standard of correctness.

[31] Housen only sets out general principles applicable to all appeals. Family law cases have attracted more specialized treatment. In Hickey v. Hickey, 1999 CanLII 691 (SCC), [1999] 2 S.C.R. 518, at para. 11, the Supreme Court has stated that “appeal courts should not overturn support orders unless the reasons disclose an error in principle, a significant misapprehension of the evidence, or unless the award is clearly wrong.” This appellate standard captures the discretionary element of support orders.

[32] The tendency in appeals from family law judgments to superimpose on the general Housen standard a subject-matter deference stems from the fact that most family law disputes are statutory. In turn, family law statutes confer considerable discretion on courts to determine most issues. The Divorce Act, s. 2(1), and the Family Law Act, s. 1(1), confer jurisdiction on specific courts, sometimes overlapping, and sometimes depending on the marital status of the disputants. Whenever statutes confer jurisdiction over subject matter, even the correctness standard imports some level of difference, at least for the purposes of interpreting the statute. In the arbitral context, the analogy to administrative law has led to the adoption of a reasonableness standard similar to judicial review: Sattva, at para. 105.

[33] A reasonableness standard widens the range of legal outcomes by allowing the adjudicator a “range of acceptable and rational solutions”: Dunsmuir v. New Brunswick, 2008 SCC 9, [2008] 1 S.C.R. 190, at para. 47, as brought into the reasoning in Sattva, at para. 104. The emergence of this deferential standard of review for questions of law stems from the principle of administrative law that tribunals should be permitted to interpret their home statutes and regulations in accordance with their subject-matter expertise. Since Sattva, the Supreme Court has required the reasonableness standard to be applied in all appeals and judicial reviews of specialized tribunals, except with respect to questions that extend beyond the tribunal’s remit: Canada (Minister of Citizenship and Immigration) v. Vavilov, 2019 SCC 65, [2019] 4 S.C.R. 653, at paras. 55-72. Perhaps the reasonableness standard is best captured by the idea that such an appeal should not be a “treasure hunt for error” but rather a tracing of the “overarching logic” as bearing a reasonable conclusion from the evidence: Vavilov, at para. 102. The appeal in this case did not involve hunting for error but searching for the actual logic behind the award. As I observed in paragraph 17 above, the elements for estimating a cumulative spousal support figure under the SSAGs formula for the 2013-2022 period were in plain sight, but the arbitrator did not use them.

[34] In the case of spousal support, the discretion applies to the determination of appropriate quantum of monthly support within a range under the SSAGs, based on the facts of the case. In Fisher, at paras. 102-03, the Court of Appeal held that the SSAGs “assist in informing an appellate standard of review.” The rationale for stipulating this special standard of appellate review is founded on judicial acceptance of the SSAGs as a normative instrument. The SSAGs defines the logic for spousal support, and deviation from it requires reasonable explanation.

[35] The SSAGs were a project of the Canadian Department of Justice to make spousal support awards more predictable and easier for parties to settle without the necessity of significant legal expenses. It also allowed generalist courts to account for regional and provincial differences in economics and taxation, without expertise in those fields. The true starting point for spousal support is s. 15.2 of the Divorce Act, notably subsections 15.2(4), (5), and (6):

Factors

(4) In making an order under subsection (1) or an interim order under subsection (2), the court shall take into consideration the condition, means, needs and other circumstances of each spouse, including

(a) the length of time the spouses cohabited;

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

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

Spousal misconduct

(5) In making an order under subsection (1) or an interim order under subsection (2), the court shall not take into consideration any misconduct of a spouse in relation to the marriage.

Objectives of spousal support order

(6) An order made under subsection (1) or an interim order under subsection (2) 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 any obligation for the support of any child of the marriage;

(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.

[36] These factual and public-policy components require case-specific data inputs regarding the parties’ marriage and breakup, as well as objectives concerning gender equality, recognition of traditional work in the home, gendered consequences of motherhood, economic autonomy and post-separation childcare. To impose some level of normative control over judicial discretion, Canadian courts have made wide use of the SSAGs formulas and commercially available computer software programs that simplify the legal work.

[37] Unlike the child support guidelines, the SSAGs are not binding legal authority. However, the failure to apply the SSAGs has come to be a marker of unsound reasoning. An award must disclose adequate reasons justifying departure from the SSAGs, because without the SSAGs, “it is very difficult to establish a principled basis for arriving at a figure for spousal support”: Slongo v. Slongo, 2017 ONCA 272, 137 O.R. (3d) 654, at para. 105. If principled reasons are provided, departure from the SSAGs ranges does not in itself constitute grounds for appeal. In Politis v. Politis, 2021 ONCA 541, 158 O.R. (3d) 230, at paras. 28-29, the Court of Appeal for Ontario stated:

[28] The appellant is correct in pointing out that the ranges generated by the Spousal Support Advisory Guidelines … formulas are the presumptive starting point for awarding support. … While not binding, the SSAGs should not be lightly departed from. ... Any departure requires adequate explanation. ... That being said, while the SSAGs formula offers a valuable tool in assessing a reasonable amount of spousal support, there are complicating factors that must be considered….

[29] Where, in my view, the appellant’s position errs is in equating the principled guidance offered in the SSAGs as a whole with the values generated by the short-hand formulas. Those formulas are intended to be used as tools only and, according to the SSAGs themselves, cannot be applied automatically in every case.

(bold added, and citations removed.)

[38] In this appeal, the wife has not appealed any findings of fact. Nor has she appealed the award based on an error of law in determining the specific legal consequence of facts in relation to the law of spousal support, i.e. a question of mixed fact and law as described in Canada (Director of Investigation and Research) v. Southam Inc., 1997 CanLII 385 (SCC), [1997] 1 S.C.R. 748, at para. 35. Had the arbitrator applied the SSAGs formula or gave reasons to depart from it, and had he adjusted the income above the $350,000 ceiling, these would have entailed questions of mixed fact and law. On the other hand, an appeal from an award of spousal support disconnected from the SSAGs entails a question of law based on a reasonableness standard. This standard applies rather than the correctness standard, because the SSAGs define the expected range of legal outcomes, and the courts expect judges and arbitrators to give individualized and fact-specific reasons either to apply the SSAGs in a particular way or to depart from it.

[39] The wife contends that the arbitrator either gave no cogent reasons explaining the award or failed to consider the SSAGs, either to calculate support within an appropriate range or to depart from the range. As I explain in the next section, the inadequacy of reasons is a question of law, but that ground of appeal fails if the decision contained sufficient findings to apply the SSAGs in coming to a reasonable range from which to make an award. If the award did not fall into that range and did not provide any reasons for departing from the SSAGs, this would constitute a separate error of law.

[40] Finally, the appeal from the costs award is also subject to guiding principles regarding the standard of review. Obviously, if the appeal is granted and the success is reversed, so too will the costs. However, that is not automatic in this case, because the wife has appealed only one part of the arbitral award. The court will have to determine the effect of reversal of the support award on the overall success or failure of the wife in the dispute.

[41] If the appeal is not granted, appellate courts will not lightly interfere with costs awards. They will require an error in principle or obvious error because of the highly discretionary nature of such awards: Hamilton v. Open Window Bakery Ltd., 2004 SCC 9, [2004] 1 S.C.R. 303, at para. 27. (Incidentally, the appellant wife did not seek leave to appeal from the costs award, and the husband did not raise the absence of a request for leave. Where the parties have agreed on a simple appeal process, as they did here, the appellate court cannot read in a leave requirement: Schickedanz, at paras. 19-25.)

ADEQUACY OF REASONS

[42] Before turning to the SSAGs analysis as applied to the arbitral award, I will address the adequacy of the arbitrator’s reasons in determining the lump sum. In making the $250,000 lump sum award, the wife cited as a ground for appeal the conclusory nature of the arbitrator’s decision to reject her demand for retroactive and prospective spousal support. The inadequacy of reasons is a unique ground of appeal because, in the extreme instance where neither the foundation for the decision nor the error can be gleaned on appellate review, a re-hearing is the only remedy: Waterway Houseboats Ltd. v. British Columbia, 2020 BCCA 378, [2020] B.C.J. No. 2146, at para. 379, quoting from Bedwell v. McGill, 2008 BCCA 6, [2008] B.C.J. No. 22:

The question of whether a trial judge's reasons are adequate is a threshold issue. If the reasons are not adequate to permit proper appellate review, the appeal must be allowed and a new trial ordered. Inadequate reasons do not enable the appellate court to make its own findings of fact and conclusions of law based on those findings. In addition, the issue of adequacy of reasons is different than the issue of whether the failure of the trial judge to address critical evidence constitutes a palpable and overriding error.

[43] Appeals are from orders, not the reasons: Heston-Cook v. Schneider, 2015 ONCA 10, [2015] O.J. No. 120, at para. 12. Nevertheless, the appellate court has no original jurisdiction to conduct a re-hearing, to cure the absence of arbitral findings of fact or subordinate determinations of fact and law. An arbitrator must state the reasons for an award: Arbitration Act, 1991, S.O. 1991, c. 17, s. 38. A breach of that duty undermines the order by frustrating the rights of parties to appeal the order: R. v. Sheppard, 2002 SCC 26, [2002] 1 S.C.R. 869, at para. 66. Conclusory statements fail to reveal factual and legal analysis and are therefore inadequate: Gibson v. Insurance Corporation of British Columbia, 2008 BCCA 217, [2008] B.C.J. No. 898, at paras. 20-23.

[44] Nevertheless, the duty is not meant to be onerous. A brief and intelligible narrative whose internal logic is sound will suffice in most instances, even if the appellate court needs to draw inferences. The appellate court cannot intervene simply because the lower court or tribunal encountered problems expressing itself: Hill v. Hamilton-Wentworth Regional Police Services Board, 2007 SCC 41, [2007] 3 S.C.R. 129, at para 101.

[45] The duty to give reasons could be relaxed even further where the parties have freely contracted to proceed by way of a streamlined mediation-arbitration procedure by the same neutral decision-maker instead of choosing to proceed with a different one for the arbitration phase. This choice entails a combined package of services and not two separate packages. It implicitly tolerates influence on the arbitrator of factors and positions learned during the confidential mediation phase that cannot be relied on formally in decision-making. Indeed, the value to the parties of the arbitrator’s knowledge of privileged information from the mediation phase extends beyond litigation economy. Each party derives some expectation of reward for taking a reasonable position in the arbitral phase, because what is reasonable will be informed by the parties’ positions during the mediation. See: Akazaki, “Overcoming Bias and Promoting Mediation-Arbitration in Canadian Civil Litigation,” 53 Advoc. Q. 349 (2022-23), at 362.

[46] The advantage to the parties of the hybrid procedure is that the trier of fact and law has already been actively involved in the proceedings, and thus has a more intimate knowledge of the justice of a case than a traditional arbitrator or judge outfitted only with formal pleadings or admissible evidence. In McClintock v. Karam, 2015 ONSC 1024, 124 O.R. (3d) 616, at para. 70, Gray J. made the following observation about the hybrid process:

[70] If the mediator/arbitrator must move to the arbitration phase, it cannot be expected that he or she can entirely cleanse the mind of everything learned during the mediation phase, and of every tentative conclusion considered, or even reached, during the mediation phase. However, at a bare minimum the parties are entitled to expect that the mediator/arbitrator will be open to persuasion, and will not have reached firm views or conclusions.

[47] The traditional origins of mediation-arbitration in final-offer arbitration (“FOA”) date back to the 1980s. FOA structures the arbitral phase as a continuation of mediation instead of a de novo adjudicative process: Akazaki, op. cit., pp. 351-52. Thus, parties’ agreement to mediation-arbitration entails acceptance of the negotiation phase as informing intangible aspects of the adjudication such as credibility and reasonable conduct. However, the arbitrator cannot accept or reject a claim based only on an intangible appreciation of the parties they gained from the outset of the combined process. In this instance, the spousal support award is only defensible if supported by the evidence introduced during the arbitral phase and guided by law. Thus, in this case, it would have been permissible for the arbitrator to exercise his discretion to some extent based on his knowledge of the parties’ bargaining positions, but it was not open for him to be governed by such factors without regard to the factual and legal case unveiled during the formal arbitration.

[48] There is a further distinction between mediation-arbitration and FOA. FOA is principally interest-based dispute resolution (e.g., an employee wants more personal benefits out of a contract; an employer wants more work or better performance). Mediation-arbitration starts as interest-driven negotiation and transitions into a rights-based adjudication. The arbitrator cannot, as he did at paragraph 126 of his award, “categorically reject” the wife’s demand “as being completely excessive, punitive and unreasonable” without stating the range of spousal support to which she could legally claim. In particular, he first needed to determine the range of legal entitlement as guided by the SSAGs before commenting on its adequacy or its excessiveness.

[49] The wife argued that the arbitral award cannot stand because of the failure to apply the SSAGs. She contended that the $250,000 award represented a marked departure from the formula ranges for which the arbitrator did not offer any explanation. The husband’s position is that the award can find support in the SSAGs and therefore required no explanation. Had the wife harboured such a concern, there was an opportunity to receive clarification which the parties did in fact use but not on this issue.

[50] At paragraphs 115-122 of Halliwell v. Halliwell, 2017 ONCA 349, 138 O.R. (3d) 671, the Court of Appeal held that the trial judge’s failure to apply entitlement considerations to the application of the SSAGs formula was an extricable error in principle. One side’s assumption that the arbitrator carried out the proper SSAGs analysis falls short of reasons demonstrating that he had, in fact, carried it out. Unlike the tax issue in Davis, the SSAGs formula is the centerpiece of Canadian spousal support jurisprudence.

[51] The arbitral award mentioned the SSAGs three times, albeit twice in the same context. In paragraph 91, it referred to the wife’s contention that even shorter marriages can still call for long post-separation spousal support awards according to the with-child-support formula. In footnote 18 of the same paragraph, the arbitrator referred to s. 8.5.5, a section devoted to “The problem of short marriages with young children.” In paragraph 117, the arbitrator did find that the status quo payments were “not necessarily consistent or commensurate with his child and spousal support obligations as found in the Child Support Guidelines and Spousal Support Advisory Guidelines.” If zero payment of spousal support was “not necessarily consistent or commensurate” with the SSAGs, the arbitrator did not explain how $250,000 was consistent or commensurate with it. I do not construe such references to the SSAGs, without more, as evidence that the arbitrator actually used the with-child-support formula to derive a range of monthly spousal support amounts in any of the years between 2013 and 2022. He thus did not perform the foundational task, either to establish the range, or to give reasons why the range was inappropriate.

[52] The husband’s counsel attempted to piece together a SSAGs analysis through various parts of the award and references to the record before the arbitrator. Despite this effort, I am compelled to reach a conclusion similar to paragraphs 115-22 of Halliwell, in which Gillese J.A. wrote that the trial judge found threshold entitlement to spousal support, cited relevant case law and legislation, imputed payor income, addressed other issues, but ultimately did not tie in the entitlement factors to the location within the SSAGs range (or not within the range) that the quantum of support should be determined.

[53] The next question is whether the record contained sufficient information for this court to perform the SSAGs analysis. If not, the only viable remedy would be to remit the question to another arbitrator for re-hearing: Schulstad v. Schulstad, 2017 ONCA 95, [2017] O.J. No. 513, at paras. 61-65. The question of sufficiency of information to perform a SSAGs analysis is hampered to some extent by the absence in the award of a quantum of monthly support, a duration of past or retroactive support, and duration of prospective support beyond the finding that it was at or near the end. Prejudgment interest, pursuant to the Courts of Justice Act, R.S.O. 1990, c. C.43, s. 128(1), must also be “calculated.” In paragraphs 30-37 of his Supplementary Arbitration Award, he did provide interest calculations for the net family property equalization. He provided no similar calculation of interest for the retroactive spousal support, and he did not identify how much was retroactive and how much was prospective.

[54] In Davis, at para. 82, the appellate court was able to infer steps such as income tax implications of a 15-year monthly support obligation calculated at $1,000, in the course of arriving at a lump sum award of $135,000. There, the court held, “Although a more detailed explanation of the figures the trial judge used in making her calculations would have been desirable, in the end, we cannot say the award was clearly wrong.” Here, the reader of the arbitral reasons has no means of knowing how much of the $250,000 award constituted past support and how much was prospective support, let alone the arithmetical inputs for each.

[55] Reliance on the length of the marriage (para. 157) was at odds with the guidance provided by the SSAGs for with-child-support spousal support. The SSAGs formula prioritizes the continuing primary parenting role of the wife over duration of the marriage and recognizes that her years of sacrifice resulting from the marriage are still ahead of her. Given the arbitrator’s finding at paragraph 123, that the wife continued such sacrifice during the 2013-2022 period, the compensatory basis of spousal support should have been similar to a with-child-support spousal claim for a sixteen-year marriage. Reliance on the length of marriage as “relatively short” was therefore an unreasonable reversal of the priority given to the effect of child-rearing.

[56] Further, the arbitrator’s reliance on the wife’s delay in pursuing her claims (para. 157) contradicted his express finding that the delay did not affect her entitlement to spousal support (para. 152). The absence of reasons explaining the contradiction gives rise to an apprehension that he simply misdirected himself.

[57] These deficits and inconsistencies in the arbitrator’s reasons do not, however, mean the award was so inadequate as to preclude appellate review. For the reasons to fall below that threshold standard, the arbitrator would have had to fail to make findings on principal inputs for the SSAGs formula: Schulstad, at paras. 61-62. Here, the data would include the parties’ income, duration of marriage, spousal roles during marriage, circumstances of breakup, and child support. While this means I must use the factual data in the award to test whether $250,000 came within a reasonable range of outcomes, it does not mean that I am required to remit the quantification for a re-hearing.

ENTITLEMENT AND NINE-YEAR DURATION OF SPOUSAL SUPPORT

[58] The husband did not appeal the award. He therefore did not contest on appeal the arbitrator’s determination at paragraph 153 that the wife had satisfied the threshold question of entitlement to retroactive and prospective spousal support, on both a compensatory and a non-compensatory basis. In paragraphs 145-57 of his award, the arbitrator returned to the threshold issue of entitlement, but only to reject the husband’s argument that she was not entitled to spousal support on account of delay and lack of hardship. One can glean from paragraphs 152, 153 and 157 that the arbitrator had accepted the wife’s argument for compensatory spousal support based on her role during the marriage as the child’s primary caregiver, and for non-compensatory spousal support based on the financial dependence and disparity of income. I concur with the arbitrator that, based on his factual narrative of the history of the marriage, the wife’s right to spousal support was based both on need and on compensation. Moreover, the conduct of the husband in planning and announcing the breakup, as well as his plan to keep the wife dependent on him financially, perpetuated the compensatory basis of the spousal support for years after he moved out.

[59] Nevertheless, it is important to restate the rationale for entitlement, as the Court of Appeal for Ontario described in Halliwell, at paras. 108-09, borrowing heavily from the SSAGs (underlining mine):

[108] The application of the SSAGs formulas, whether under or above the ceiling, requires a preliminary consideration of entitlement. The entitlement question then informs the approach to be taken in applying the SSAGs.

[109] As stated at s. 3.2.2. of the SSAGs:

The Advisory Guidelines do not deal with entitlement. . . . The Advisory Guidelines were drafted on the assumption that the current law of spousal support, post- Bracklow, continues to offer a very expansive basis for entitlement to spousal support. Effectively any significant income disparity generates an entitlement to some support, leaving amount and duration as the main issues to be determined in spousal support cases. . . . The basis of entitlement is important, not only as a threshold issue, but also to determine location within the formula ranges or to justify departure from the ranges as an exception.

[110] It is important to note that s. 3.2.2 recognizes that entitlement plays two different important roles in determining spousal support. First, entitlement is a threshold issue. Second, entitlement determines location within the formula ranges or to justify departure from the ranges.

[60] In paragraph 88, the arbitrator found that the wife had considerable employable skills that she could use within a short period of time. This appeared to be the sole factor considered in his conclusion at paragraph 94 that the wife’s support entitlement was “at or near the end.” This conclusory statement could have been justified, if the husband had discharged all his statutory support obligations, both to his child and to his wife, to foster the recipient spouse’s self-sufficiency. Since the wife’s employment qualifications had not changed since 2013, one deduces from the arbitrator’s reasoning that the husband’s non-payment of spousal support and underpayment of child support during the nine years delayed her self-sufficiency. If “within a short period of time” meant an interval starting in 2013, that would have been wholly inconsistent with the finding that the entitlement was “at or near the end.” The prolonged dependency (or delay in self-sufficiency) was consistent with the arbitral findings of the husband’s 2013 promise to discharge all his support obligations in the family law process (paragraphs 20-21). The advantage to the husband of avoiding those same responsibilities did not escape the arbitrator’s notice:

128. I agree with Ms. Rathee, however, that Dr. Rathee was either aware or ought to have been that he would owe not only an equalization payment of some significance but also child and spousal support. Not having made payment(s), it enabled him to amass capital and enjoy his income free from constraint.

[61] Similarly, at paragraph 152, the arbitrator rejected the husband’s argument that the wife’s entitlement only started in 2019. In the absence of another chronological reference, the entitlement could only have started in 2013.

[62] The basic SSAGs formula, the official commentary and the academic explanation emphasize that the child support recipient’s compensatory entitlement to spousal support, once established, starts immediately on separation and is left open-ended because the recipient parent’s self-sufficiency is held back by the ongoing childcare responsibility. Starting with the basic formula, the effect of childcare on the duration of spousal support is a defining element (underlining mine):

3.3.4 The with child support formula

In cases where there are dependent children, the with child support formula applies. The distinctive treatment of marriages with dependent children and concurrent child support obligations is justified by both theoretical and practical considerations and is reflected in current case law.

On the theoretical front, marriages with dependent children raise strong compensatory claims based on the economic disadvantages flowing from assumption of primary responsibility for child care, not only during the marriage, but also after separation. We have identified this aspect of the compensatory principle as it operates in cases involving dependent children as the parental partnership principle, and have drawn on this concept in structuring the with child support formula. For marriages with dependent children, length of marriage is not the most important determinant of support outcomes as compared to post-separation child-care responsibilities.

On the practical front, child support must be calculated first and given priority over spousal support. As well, the differential tax treatment of child and spousal support must be taken into account, complicating the calculations. The with child support formula thus works with computer software calculations of net disposable incomes.

Under the basic with child support formula:

• Spousal support is an amount that will leave the recipient spouse with between 40 and 46 percent of the spouses’ net incomes after child support has been taken out. (We refer to the spouses’ net income after child support has been taken out as Individual Net Disposable Income or INDI).

• The approach to duration under this formula is more complex and flexible than under the without child support formula; orders are initially indefinite in form (duration not specified) but the formula also establishes durational ranges which are intended to structure the process of review and variation and which limit the cumulative duration of awards under this formula. These durational limits rely upon both length of marriage and the ages of the children.

[63] The Revised User’s Guide to the SSAGs explains the practical effect of the above guidance: “The real issue, at both interim and initial stages, will be the location of an amount for spousal support within the SSAG range.” The co-author of the SSAGs and Revised User’s Guide further described the rationale for discounting the importance of the length of marriage and emphasizing the compensatory nature of the duration of support where one party has primary care of the child (underlining mine):

The scenario is familiar, mentioned earlier: young husband and wife, together for three years, they have two-year-old twins, she is home, he earns a sizeable income, they separate and she continues as the primary parent for the twins. The husband, and his lawyer, will see a three-year marriage and a limited spousal support obligation. But the bulk of the disadvantage is not behind the wife, but in front of her. The age of the children may complicate her return to the paid labour market and, once she does return to employment, her parenting responsibilities will likely continue to limit her earning capacity for a lengthy period of time.

In decided cases, judges consistently ignore or underestimate the compensatory disadvantage going forward. Too often, we see judges ordering short time limits at first instance, keyed to the length of the relationship, rather than the care of the children, a result utterly inconsistent with Moge. This still happens, despite the range for duration under the SSAG for such cases, using the age of children test in shorter marriages, with the lower end tied to the last child commencing full-time school and the upper end fixed by the end of high school.

D.A. Rollie Thompson, ‘Ideas of Spousal Support Entitlement’ (2014), 34 Can. Fam. L.Q. 1 (p. 18 of WestlawNext publication)

[64] A plain reading of the SSAGs with-child-support formula, aided by the interpretive resources available to family law practitioners, indicates that the wife’s entitlement to spousal support begins from the moment her husband left her to look after their child primarily on her own. We also know the arbitrator determined the entitlement was at or near the end. Based on the standard of appellate review, I am reluctant to disturb that finding, even though it seemed to have been based on the single factor of the wife’s employability in pharmaceutical sales.

[65] The SSAGs range of duration of support for the with-child-support formula is between the time the youngest child enters full-time school and the time the child finishes high school. “This test will typically apply to marriages of less than ten years” (SSAGs, pp. “x” and 84). Since the child was in grade nine at the time of the arbitration, a round number of nine years could have been a reasonable measure of the length of entitlement, based on the arbitrator’s conclusion regarding duration of the entitlement and, at paragraph 153, regarding her entitlement to both retroactive and prospective spousal support on both a compensatory and non-compensatory basis. It would have been helpful if the arbitrator had stated this reasoning, but the outcome cannot be considered unreasonable.

[66] The award thus contained the factual basis for the period of entitlement as starting in 2013 and ending sometime in 2022.

REJECTION OF HUSBAND’S DELAY ARGUMENT, HUSBAND ESTOPPED

[67] The arbitrator’s findings therefore should have led to a straightforward conclusion that the wife was entitled to spousal support on the grounds of need and compensation for a period of nine years after separation. In fact, in paragraph 152, he rejected the husband’s three arguments against awarding retroactive spousal support, viz. that (1) she had failed to provide notice of the claim until the 2019 start of mediation, (2) he had not engaged in blameworthy conduct and understood he was fulfilling all his support obligations, and (3) the wife had not experienced hardship and he would endure hardship if he were ordered to pay retroactive support. Paragraph 152 stated:

152. In considering the cases of S. (D.B.) v G. (S.R.), Kerr v. Baranow and others, I find the reasons for Ms. Rathee’s delay relatively explicable, especially in the later years, and not affecting her entitlement to spousal support, as opposed [to] the fixing an amount that appropriately and equitably balances the countervailing interests of the parties.

[68] Previously, he had also rejected the delay argument at paragraphs 127-28. While attributing some responsibility to the wife for failing to drive the matter forward more quickly, he held that the husband was or ought to have been aware that he would owe spousal support. “Not having made payment(s),” the arbitrator concluded, “it enabled him to amass capital and enjoy his income free from constraint.”

[69] Against the context of the cited case law, his finding that the wife’s delay did not affect her entitlement meant the arbitrator rejected the husband’s two arguments based on delay. The subordinate clause in paragraph 152 starting with the word “as opposed” also meant he also rejected the husband’s argument based on relative hardship. Kerr v. Baranow, 2011 SCC 10, [2011] 1 S.C.R. 269, paras. 211-12, harmonized the law of retroactive spousal support with the notice rule for child support in D.B.S. v. S.R.G, 2006 SCC 37, [2006] 2 S.C.R. 231. In paras. 123-25 of D.B.S., Bastarache J.’s majority opinion put forth a general rule that, in most circumstances, retroactive support should not be awarded for a period prior to three years before notice. However, the payor spouse “should not be permitted to profit from his/her wrongdoing.” In applying these principles to spousal support, the Supreme Court at paras. 213-16 in Kerr found the British Columbia Court of Appeal had erred by failing to consider all the circumstances of the recipient spouse and by attaching “such serious consequences” to her delay in pursuing support.

[70] I also add that, since the arbitrator calculated retroactive child support for the entire post-separation period, he was also correct in finding there was no principled reason to limit spousal support based on the Kerr and D.B.S. rationale.

[71] Although he did not comment on it, the arbitrator’s rejection of the arguments for limiting the duration of support was also consistent with are more compelling reason, namely the husband’s 2013 promise to pay the support he owed in a “financial settling-up” within the legal process. At paragraph 21, the arbitrator made it clear that the promise would be fulfilled when the retroactive support would be agreed or adjudicated:

21. I also find as a fact that Dr. Rathee to his credit, meant this to mean that his financial

obligations, including any retroactive support, would be assessed and either agreed or adjudicated and he would discharge them correspondingly. I am also satisfied and find that Ms. Rathee had every reason to rely on these representation(s).

[72] During the appeal hearing, husband’s counsel questioned the arbitral finding and said the husband never made such a promise. However, the evidence for the arbitrator’s finding appeared at pp. 78-81 and pp. 100-05, which included portions of the husband’s arbitration hearing transcript (Appeal Book and Compendium, pp. 1580-83 and 1602-07). In particular, the husband had consulted five lawyers in May and June of 2013 about his rights and obligations on separation. He had understood that, beyond support for the children, their relative incomes and length of marriage made it likely that he would have to pay spousal support. He also understood that, due to his high income, his financial obligations could be “quite significant.” On cross-examination, at pp. 104-05 of the transcript, he then described a July 2013 conversation in which he and his wife agreed to continue their existing family finance arrangements and later “figure out” what should later ensue:

Q. And what did you mean when you say “we’ll figure it out?”

A. I’m – what I meant, specifically – I’m glad you ask -- is we need the help of legal counsel and an understanding of what the obligations are, specifically, based on our -- based on my financial statements and income, to know more.

Q. And -- and other than the sort of general understanding that you have acknowledged earlier, would it be fair to say that neither you nor Noelle, at that point in time, had any real understanding as to how much she would be entitled to for an equalization payment, for child support, for spousal support? Would that be a fair statement? [emphasis added].

MS. ZYLBERMAN DEMBO: I -- I object. That's not a fair question to be putting to the witness.

THE ARBITRATOR: I don't agree. If Dr. Rathee can answer it, he should answer it.

A. The answer to the question, I – I don't think either one of us -- I can't speak for Noelle, but I certainly wasn't aware of the exact numbers in terms of the obligation. I was not aware. I knew they were numbers that were significant. I couldn't tell you what they were.

[73] The period after the husband’s promise of a deferred reckoning of spousal support was also material to the retroactive support obligation. The arbitrator, at paragraph 16 of the decision, recited the wife’s evidence that she was “blindsided” by the husband’s announcement that he was leaving her. He then concluded at paragraph 18 that the husband’s behaviour during the ensuing period would have raised a reasonable prospect of cohabitation, such that the husband’s date of separation should be used for equalization of property (October 2013 instead of July 2013).

[74] The parties participated in marriage counselling. For the next two years, at least, the husband spent a considerable amount of time in the matrimonial home. The parties and their child went on vacations together. During that time, in the wife’s words, they functioned “like a family.”

[75] The act of “lulling to sleep” has been described as a form of estoppel or ratification: Dominion Gresham Guarantee and Casualty Company v. Bank of Montreal, 1930 CanLII 318 (UK JCPC), [1930] 4 D.L.R 689 (UK JCPC), at 693. More importantly, the husband did not renege on his promise to pay spousal support until he took the position in the 2019-2022 mediation-arbitration that the wife should have been satisfied with the uncharacterized support that he applied to retroactive child support. The husband’s promise and the post-separation conduct both barred him from later resiling from the spousal support obligation and from citing delay or lack of notice as grounds for reducing his liability.

[76] Maracle v. Travellers Indemnity Co. of Canada, 1991 CanLII 58 (SCC), [1991] 2 S.C.R. 50, remains the leading case on the doctrine of promissory estoppel. Promissory estoppel is an exception to the principle that estoppels operate only as shields (defences to claims) and not swords (claims), in that the estoppel precludes a defendant from relying on a defence. It would be more precise to say that promissory estoppel does not a really operate as a sword, but rather as a claimant’s shield against the defendant’s use of a shield. At p. 59 S.C.R., Justice Sopinka described the necessary elements of a promissory estoppel, in the context of a defendant’s ability to raise a limitation period defence:

The principles of promissory estoppel require that the promissor, by words or conduct, intend to affect legal relations. Accordingly, an admission of liability which is to be taken as a promise not to rely on the limitation period must be such that the trier of fact can infer from it that it was so intended. There must be words or conduct from which it can be inferred that the admission was to apply whether the case was settled or not, and that the only issue between the parties, should litigation ensue, is the issue of quantum. Whether this inference can be drawn is an issue of fact. If this finding is in favour of the plaintiff and the effect of the admission in the circumstances led the plaintiff to miss the limitation period, the elements of promissory estoppel have been established.

[77] If such an estoppel can preclude reliance on a hard limitation period, it would apply with equal if not greater force to the more malleable judge-made notice and delay rules in D.B.S. and Kerr. The arbitrator’s findings, at paragraphs 20-21, clearly constituted words and conduct by the husband admitting to spousal support liabilities that he would honour, and that the only issue left for the legal process to resolve was the issue of quantum. It therefore does not matter whether the wife stood on her rights during the 2013-19 period. The husband’s assurance effectively reversed the onus on himself. If the financial status quo pending settlement was racking up too much retroactive spousal support, he ought to have placed the wife on notice earlier that he wished to limit or cap his liability. He did not do so.

[78] Had the wife sued him for spousal support at the time of the separation in 2013, the likely result would have been an interim order, and the only real issue on that point would have been the quantum in accordance with the SSAGs. The only party suffering prejudice as a result of the delay interval was the wife.

[79] Before leaving the estoppel issue, I observe that appellate courts generally refrain from considering issues that were not clearly raised at first instance: Perez v. Governing Council of the Salvation army in Canada (1998), 1998 CanLII 7197 (ON CA), 42 O.R. (3d) 229 (ON CA). Nevertheless, it is hard to limit the ambit of appellate review to issues defined in the original proceeding when none were so defined. The appeal book contained no pleadings or arbitration statements apart from the effective relief sought by the parties and the parties’ affidavits. There appears to have been no opportunity to plead the 2019 notice defence, or an estoppel to the same 2019 notice defence based on prior assurances that the support issues will all be settled in the legal process. The mediation-arbitration process the parties chose, for cost and efficiency reasons, may have obviated the necessity for formal pleadings apart from the statements of relief. Perhaps the parties had raised various issues during the mediation. Article 9.2 of the mediation-arbitration agreement left it for the arbitrator to set the procedure, subject to consultation with the parties.

[80] The husband might argue that it would be unfair to impose an estoppel preventing him from denying liability for spousal support during the 2013-2019 interval. However, he freely admitted the promise in his testimony and the arbitrator found that the wife could rely on it. The only thing missing from the piece was a label in the arbitrator’s reasons calling the reliance on the promise an estoppel from denying liability for spousal support during that period.

[81] Finally, the various instances in the decision where the arbitrator found that the husband knew the spousal support would be subject to a settling-up and had agreed to the deferral from 2013 amounted either to a deemed or effective notice of her claim, if for no reason other than his acknowledgment. The significance of this fact, by operation of the Kerr principles or of a promissory estoppel, is that all the support from the 2013 agreement forward was prospective support to which the wife was by law entitled, and arrears is more suitable in describing the unpaid spousal support during that period than retroactive. The following language from the judgment of Lang J.A. in MacKinnon v. MacKinnon, 2005 CanLII 13191 (ON CA) (cited in Kerr at para. 211), could very well have replaced paragraphs 153-57 of the arbitral decision to lead the arbitrator to a reasonable award of spousal support (underlining mine):

[18] In fairness to the trial judge … the question of spousal support for the two-year period pending the trial was apparently characterized by both counsel at trial, as it was on appeal, as a claim for retroactive support. In my opinion, however, this case does not raise a question of retroactive support.

[22] Absent any unusual reason arising from the factors and objectives set out in the Divorce Act, an applicant who requests financial disclosure in preparation for the negotiation or litigation of a support claim, and who then proceeds reasonably to a disposition of the claim, presumptively is entitled to prospective support from the date of notice that a support claim is being pursued.

[25] During the two years in issue, the husband had the financial means to pay spousal support. He knew that spousal support would be decided at trial. In such circumstances, there is no reason to relieve the husband from his responsibility to pay spousal support.

SSAGs RANGE OF MONTHLY SPOUSAL SUPPORT

[82] After rejecting the husband’s argument based on delay and notice, as stated in Kerr and D.B.S., there was no reason to preclude the wife’s entitlement to spousal support for the same length of time as the child support, i.e., from 2013. The fact that the lump sum included prejudgment interest requires the appellate court to consider how the award, net of interest, translated into a monthly support obligation. Using the specific prejudgment interest rate(s) and the 106.5-month duration used by the arbitrator for the property equalization payment, one could calculate a principal monthly support figure of $238,364. This results in a $2,238.17 monthly support obligation for a payor earning over $100,000 a month. Considered as nine years of monthly support payments, the $250,000 figure was clearly wrong.

[83] The $2,238.17 figure was about a tenth of the monthly support granted in Halliwell. The duration of spousal support in Halliwell was determined by the absence of child support and a much longer marriage. However, the immediate obligation at the time of separation would not have been very different from the instant case. The formulas prioritize different factors, but not so much as to reduce the millionaire support payor’s obligation to a non-earning support recipient to that of a middle-income payor. In Halliwell, the Court of Appeal for Ontario recalculated the support by reducing a $1,000,000 income to $675,000 after imputing the recipient spouse’s expected investment income from a $3,000,000 equalization payment. The result was a recalculation of the $28,978 monthly payment to $21,000. In contrast to Halliwell, the arbitrator found at paragraph 95 of the decision that the husband’s full income, as supported by the expert income valuator, should be used for support purposes.

[84] The basic elements of a spousal support calculation were in the arbitral award, starting with the parties’ income at separation and the ensuing period. At paragraph 95, the arbitrator made findings based on the income valuator’s evidence, for both child and spousal support. Notably, he did not discount income above the $350,000 SSAGs discretionary ceiling (as the Court of Appeal did in Halliwell). There is no valid reason for the appellate court to question these income findings, combined with the absence of stated reasons to depart from the full number above $350,000. In paragraph 99, he excluded capital gains in 2020 and preferred one expert’s calculation over the other. At footnote 23 of the decision, he incorporated the wife’s income figures by reference to her submissions.

[85] Stepping back from the arbitrator’s findings in paragraph 95 of the award, I note that there are generally two approaches to high-income cases involving the ceiling. Where the payor income is marginally above $350,000, a “minimum plus” approach entails a calculation of support at $350,000, with an adjustment upwards while considering child support. Where the payor income is well above the ceiling, the formulas are not automatic but do provide an appropriate method of determining spousal support for the income above the ceiling. Although the discretion above the ceiling requires individualized and fact-specific analysis, ignoring the SSAGs formula is not an option: Hathaway v. Hathaway, 2014 BCCA 310, [2014] B.C.J. No. 1996, at paras. 46-48.

[86] Returning to paragraph 95 of the arbitral decision, the specific finding that the full incomes ought to be used for child and spousal support was not qualified by any reason to dial it back to a figure above the $150,000 ceiling for table child support under s. 4 of the Federal Child Support Guidelines (SOR/97-175) or the $350,000 ceiling for spousal under the SSAGs. The average of the husband’s annual income for the 2013-2021 period amounted to $1,178,400. The average for the wife was $130,660. At the appeal hearing, the husband’s counsel presented various DivorceMate calculations, based on incomes of $1,253,000 for the husband and $136,816 for the wife, with 90/10 sharing of the s. 7 expense for private school tuition of $31,050. When I ran the numbers, my monthly figures were lower than his counsel’s:

Low Mid High

My figures: $24,637 $27,688 $30,739

Husband’s figures: $26,531 $29,721 $32,906

[87] Calculated over the period between separation and the arbitration, it should have been obvious that a rather orthodox SSAGs calculation, based on the with-child-support formula, should have produced lumps sums at least ten times more than the $250,000 awarded.

[88] The flaw in my calculation in using average income figures to forego having to calculate the child support annually is that the income differentials fluctuated marginally, and prejudgment interest might be off because the husband’s income figures were, for accounting and other reasons, higher immediately after 2013 than during the middle period. I do not view this as a real issue because the wife asked for a combined child and spousal support award of $2,000,000; and because the SSAGs ranges are wide enough that the arithmetic cannot be reduced to an exact calculation. On the issue of the “ask,” as the arbitrator called it at paragraph 126, the award of retroactive child support of $106,397 meant the highest he could have awarded for retroactive spousal support was $1,893,603.

[89] While the result here is unaffected by the place in the range used to calculate the spousal support payments, and the arbitrator clearly did not feel compelled to choose from within the range, the tendency in the case law to view compensation as the main purpose of spousal support would suggest that the higher end of the range should have been the right choice: See D. A. Rollie Thompson, “Ideas of Spousal Support Entitlement” (2014), 34 CFLQ 1, at p. 9 of Carswell ed.

[90] The with-child-support formula under the SSAGs recognizes, on the one hand, that some of the child support will alleviate the wife’s needs; but, on the other, that the wife’s obligation to perpetuate the pre-separation living standard for the child (especially compared to the payor’s) will require more income-sharing by the payor spouse. Such is the nature of the compensatory element of the formula. In Chutter v. Chutter, 2008 BCCA 507, [2008] B.C.J. No. 2398, at para. 79, Rowles J.A. of the British Columbia Court of Appeal stated:

[79] In my view, the trial judge’s emphasis on his finding that the assets the appellant received in the division of property can provide her with a standard of living comparable to that which she enjoyed during the marriage overlooks the principles underlying compensatory support where one spouse has been disadvantaged by the marriage or its breakdown relative to the other spouse. As this Court held in Tedham, at para. 60, a spousal support award based on compensatory principles should continue until compensation is achieved, even if the recipient spouse has achieved a degree of self-sufficiency. In arriving at that conclusion, this Court cited Allaire, in which the Ontario Court of Appeal rejected the husband’s argument that any disadvantages flowing from the marriage must defer to the fact that the wife, at the time of trial, was nevertheless able to earn a reasonable income (para. 21).

[91] Similarly, the Court of Appeal for Ontario’s decision in Mason v. Mason, 2016 ONCA 725, 132 O.R. (3d) 641at para. 199, is instructive in choosing a figure within the SSAGs range in accordance with various factors, instead of picking a number:

[199] The Spousal Support Advisory Guidelines: The Revised User’s Guide provides that courts should avoid the tendency to “default” to the mid-range amount of spousal support. Section 9 of the 2016 Revised User’s Guide explicitly states, “[t]he mid-point of the SSAG ranges for amount should NOT be treated as the default outcome.” In determining the appropriate quantum of support within the range, a court is required to consider the support factors and objectives found in the Divorce Act and the Family Law Act. The SSAGs also provide a number of factors to consider while choosing a location within the range, including the strength of the recipient’s compensatory claim, the recipient’s need, property division and debts, and the payor’s needs and ability to pay.

[92] Despite the husband’s submission that the wife had a business sideline, the arbitrator did not mention it in the award. Whatever prospective spousal support the wife may have been entitled to receive, the equalization payment cannot be annuitized into a stream of income for the 2013-2022 period because the husband had not paid it. Because the husband paid no spousal support during that period but, rather, expected the wife to be content with a status quo that also underpaid child support, the wife’s claim for spousal support should have been determined at the higher end of the range. Indeed, the finding that she was entitled to spousal support based on need and compensation put the monthly amount at the higher end.

[93] If I am wrong in my ruling below that the wife’s claim on appeal is capped at $2,000,000 less the award of child support arrears, her spousal support entitlement can be calculated on the high end at $30,739 monthly over nine years (106.5 months of arrears plus 1.5 months after the arbitrator’s calculation), in the amount of $3,319,812, plus prejudgment interest of $290,541.

ARBITRATOR’S MISUSE OF LUMP SUM CRITERIA IN DAVIS v. CRAWFORD

[94] Even the lower end of the SSAGs ranges, calculated over nine years, should have produced a calculation of retroactive spousal support in the order of ten times more than the $250,000 award. In paragraphs 153-57, after rejecting the husband’s arguments for limiting the duration of retroactivity, and there being no cited reason to reduce the income base or the monthly amount, the arbitrator decided to apply the principles behind lump sum awards as stated in Davis and picked the number. In justifying the lump sum as a good idea, he identified five relevant factors from the Davis case: the husband’s ability to pay it, terminating the financial nexus between the parties, the wife’s need for capital, the adequacy of periodic payments, and “Whether appropriate to satisfy an award of retroactive spousal support.” He appears to have extracted these factors, especially the last one, from paras. 65-68 of Davis (underlining mine):

[65] These statutory provisions make it clear that ability to pay is an important consideration in making an award of spousal support, including lump sum spousal support.

[66] Most importantly, a court considering an award of lump sum spousal support must weigh the perceived advantages of making a lump sum award in the particular case against any presenting disadvantages of making such an order.

[67] The advantages of making such an award will be highly variable and case-specific. They can include but are not limited to terminating ongoing contact or ties between the spouses for any number of reasons (for example, short-term marriage; domestic violence; second marriage with no children, etc.); providing capital to meet an immediate need on the part of a dependant spouse; ensuring adequate support will be paid in circumstances where there is a real risk of non-payment of periodic support, a lack of proper financial disclosure or where the payor has the ability to pay lump sum but not periodic support; and satisfying immediately an award of retroactive spousal support.

[68] Similarly, the disadvantages of such an award can include the real possibility that the means and needs of the parties will change over time, leading to the need for a variation; the fact that the parties will be effectively deprived of the right to apply for a variation of the lump sum award; and the difficulties inherent in calculating an appropriate award of lump sum spousal support where lump sum support is awarded in place of ongoing indefinite periodic support.

[95] In Davis, the Court of Appeal upheld the award of a lump sum of $135,000 based on a $1,000 monthly payment over 15 years, using a 4 percent discount interest rate. The court rejected the appellant’s submission, based on Mannarino v. Mannarino, 1992 CanLII 14022 (ON CA) that lump sums should only be made “where there is a real risk that periodic payments would not be made” (Davis, para. 51). The court nevertheless concluded that “most spousal support orders will be in the form of periodic payments” and a lump sum was considered “as a substitute for an award of periodic support” (Davis, paras. 70 and 76). Irrespective of whether ordered periodically or in a lump sum, the payor’s obligation must first be determined by reference to the ordinary rules, including the all-important use of the SSAGs formula (Davis, paras. 75-76; and Parton v. Parton, 2018 BCCA 273, at para. 47).

[96] The phrase I underlined in para. 67 is the only place in Davis where the Court of Appeal referred to retroactive spousal support. The discussion, devoted to the use of lump sums instead of future periodic payments in cases where payors had the ability to pay them, focused on the desirability of effecting a clean financial break between the former spouses. The court clearly meant to say that a lump sum in lieu of periodic payments could be paid at the same time as a lump sum of retroactive support (as opposed to instalments). This meaning is buttressed by the word “immediately” as a qualifier of “satisfying.” Isolated from this context, the arbitrator could perhaps have construed para. 67 as extending the authority under s. 15.2(1) of the Divorce Act to award a lump sum as applying to retroactive awards. The problem with this reading is that retroactive awards are already lump sums in the sense of being a settling-up of the unpaid monthly instalments. The monthly amounts are determined by the SSAGs formula, so the only judicial discretion to be exercised is to limit the duration of pre-notice retroactivity as described in para. 211 of Kerr, as discussed previously.

[97] The reference to the use of lump sum awards as an opportunity to satisfy immediately a retroactive award does not, however, detract from the mandatory consultation of the SSAGs. In the SSAGs, lump sums are described only as a restructuring of periodic payments instances where the payor can afford to make the payment, and “the global amount of support remains the same” (SSAGs, pp. 103-04 and 162). This last point is crucial. Davis only dealt with the circumstances for restructuring the periodic payments into an immediate payment without altering the economic value of the recipient spouse’s entitlement. Indeed, at para. 68, the court noted the obvious difficulty of awarding a lump sum where the duration of support is indefinite. The value of a lump sum was therefore never intended to be anything other than the economic equivalent of a stream of future payments over a duration of support that was either certain or capable of being estimated.

[98] Here, the arbitrator found the delay in proceedings did not affect the wife’s entitlement. Therefore, there was no reason to choose a lump sum over any other form of payment for the amounts owed between 2013 and the 2022 arbitration. The Davis factors for choosing a lump sum over a periodic payment were only relevant to the short time left in her entitlement after the arbitral hearing, as tentatively described in paragraph 94 of his decision. His use of the Davis factors as an alternate method of calculating the support and changing the global amount from a straight calculation of monthly support multiplied by nine years was an error of law that led to a wholly unreasonable result. In so doing, he effectively wiped out nine years of retroactive support, without any tenable grounds.

[99] The flaw in the arbitrator’s reasoning operated a fortiori if one considers the support the husband agreed in 2013 to pay as arrears of prospective, rather than retroactive support, as described in MacKinnon. As illustrated in the trial judge’s calculation in Davis, the calculation of a lump sum by applying a discount rate of interest (cf. rule 53.09 of the Rules of Civil Procedure, R.R.O. 1990, Reg. 194) justifies a lump sum of a future stream being less than the arithmetic sum. In contrast, there can be no justification for reducing the value of payments that the husband ought to have been made as prospective spousal support, ex ante counting from the 2013 agreement.

EFFECT OF THE WIFE’S $2,000,000 SELF-IMPOSED COMBINED CAP

[100] As previously stated, calculation of retroactive spousal support between 2013 and 2022, when combined with the award of retroactive child support, would result in a combined award well over the $2,000,000 claimed in the wife’s closing submissions at the arbitration. Our courts have dealt with cases where civil juries have awarded damages more than the amounts that plaintiffs have pleaded, and trial judges have allowed amendments to the claim: M.B. v. 2014052 Ontario Ltd. (Deluxe Windows of Canada), 2012 ONCA 135, 109 O.R. (3d) 351, at para 72. I have not been able to locate a case dealing with an amount on appeal that is greater than the amount claimed at first instance.

[101] I am mindful of the amount claimed in the actual claim for retroactive child support and spousal support: it was $3,000,000. In retrospect, that was a fair estimate, and the wife had originally pleaded it. Nevertheless, the abandonment of $1,000,000 of claim would have changed the landscape significantly. It is not necessary to speculate how it affected the husband’s litigation strategy and the arbitrator’s reasoning. While the reduction did not appear to affect the arbitrator’s categorical rejection of the demand as excessive, etc., one must consider it from the perspective of the husband and his legal counsel. Just as the wife was entitled to rely on the husband’s promise settle the spousal support later, the husband was entitled to rely on her intentional relinquishment of her right to claim more than $2,000,000 in combined unpaid child and spousal support. Moreover, detrimental reliance is not required in the case of waiver: Cronnox Inc. v. Lloyd’s Underwriters, 2018 ONSC 6437, [2018] O.J. No. 5671, at paras. 77-79.

[102] I therefore reduce the amount of the retroactive amount of spousal support to the maximum amount of jeopardy the husband faced at the final hearing of the arbitration. I therefore substitute the amount of $250,000 with $2,000,000, less the portion of $106,397 already awarded for unpaid child support. The difference is $1,893,603.

COSTS DECISION

[103] Based on my decision of the lump sum award for spousal support, the costs award of the arbitrator cannot stand. Although spousal support was only one issue in the arbitration, the financial change in outcome would leave no doubt that the wife would have been the prevailing party overall.

[104] In the event my decision is overturned on further appeal, I will deal with the costs award in isolation, i.e., as if the arbitrator’s decision were not set aside.

[105] In paragraphs 5-18 of his costs award, the arbitrator appears to have touched on all the applicable jurisprudential considerations. Nevertheless, he outlined them in a haphazard manner, informed by a belief that “costs awards are typically complex, especially in cases of any serious contention.” This belief appears to have carried itself forward into the analysis at paragraphs 32-49. That analysis consisted of four segments. First, he found the financial issues to be relatively complex. Second, he found the husband’s conduct, including his resiling from a tentative agreement during the mediation portion, to be troubling and warranting some penalty but falling short of bad faith. Third, he considered the reasonableness of the lawyers’ fees. Fourth (although not labelled as such), he considered the parties’ settlement offers. After a comparison of the award with the parties’ pre-hearing offers to settle and of the offers against each other, the arbitrator held that the husband was “the more successful party.”

[106] At paragraph 46 of his costs award, the arbitrator appears to have been persuaded that, in terms of an overall bottom line, the wife could have been better off had she accepted the husband’s offer to settle.

[107] The arbitrator outlined his thoughts regarding these four steps before concluding that the husband was the more successful party and therefore entitled to partial indemnity costs of $206,330, less a reduction of $35,000 to reflect the husband’s misconduct and procedural failings during earlier parts of the proceeding. In isolation, the penalty for factors including the husband’s conduct during the mediation stage of the proceeding appears to be an error in principle warranting quantification in the whole amount of $206,330. Offers to settle are unique in being able to be revealed from behind the without-prejudice arras. For mediations to be effective, nothing said or done, in good faith or bad, can later become the subject of prejudice during the arbitration or litigation. This principle was also memorialized in large part in art. 7.3 of the parties’ med-arb agreement.

[108] Rule 24 of the Family Law Rules, O.Reg. 114/99, deals with entitlement in subrules (1) to (11). Subrules (12), (12.1), and (12.2) deal with the amount of costs. These two logical segments reflect a simplified approach to costs, consistent with the overall scheme of the regulation to make family court procedures accessible to a large proportion of self-represented lay litigants: Coats, et al., Ontario Family Law Practice, 2024 (Toronto, LexisNexis Canada Inc., 2023), p. 1789.

[109] By over-complicating the costs rule, the arbitrator made extricable errors of law.

[110] The first segment of rule 24 segment starts with subrule 24(1)’s primary presumption that “a successful party” is entitled to costs and concludes with various exemptions and discretionary departures from the presumption.

[111] Subrule 24(4) allows discretion to award costs against a successful party or to deny that party costs if the party behaved unreasonably. Subrule 24(5) allows consideration of whether a party made an offer to settle in the inquiry of how reasonable or unreasonable the conduct was. The arbitrator appears to have embarked on a comparative analysis of the parties’ offers to conclude that the husband was the “more successful party.” For example, the equalization award was between the parties’ offers, but much closer to the husband’s offer. The arbitrator ruled that the husband prevailed on the issue. Had the wife accepted the offer, she would have been worse off than the award. The costs rules reward parties for beating their offers, not for the proximity of their offers to the result in relation to the other parties’ offer. This was a clear an extricable misinterpretation of the rule. Since both parties made bona fide offers to settle, subrule 24(5) was irrelevant. The rules do not permit a meta-analysis comparing offers with offers. The comparison must be between offers and outcomes.

[112] There is a separate rule, subrule 18(14), that can reverse the fortune of a costs award based on comparisons of results to offers. There, a successful offeror can move the needle indicating the measure of success or failure by making an offer that would have put the adversary in a better position than the result at the end of the hearing. However, these costs consequences arise from the date of the offer, and the presumptive application of subrule 24(1) for legal expenses up to that date remain unaffected.

[113] Finally, I return to the husband’s promise to discharge all his legal obligations, and the fact that by maintaining the financial status quo he was able to thrive financially while his wife remained in a comfortable yet dependent and static position. At the arbitration, he took a hard position, denied spousal support obligations, underpaid child support, and tried to claw back from the property settlement by charging rent on the property the wife also owned. He was the net payor in every category of the statutory family law dispute. Since there was no comprehensive offer to settle from him that would have put the wife in a better position than the result overall, the wife has to have been found to be entitled to costs even though she was dissatisfied with the spousal support ruling.

CONCLUSION

[114] The appeal is allowed. The arbitrator’s spousal support award of $250,000 inclusive of prejudgment interest is hereby set aside and substituted with an award of $1,893,603. With prejudgment interest at 2% over 106.5 months of accumulating payments and an effective rate of 1%, the prejudgment interest as of the arbitration would have been $168,057. Post-judgment interest of 3% per annum shall apply from the date of the award, i.e., from August 20, 2022.

[115] I set aside the arbitral costs award and substitute it with an award to the wife, in the amount of $276,591 fixed by the arbitrator. Post-judgment interest of 4% per annum shall apply from the date of the costs award, i.e., from January 17, 2023.

[116] I encourage the parties to settle their costs of the appeal. If required, the wife can serve and file a bill of costs and costs submissions not exceeding two pages, within 20 days hereof. The husband will have a further 20 days to respond with submissions not exceeding two pages. These materials shall be served and filed and sent to my attention via my judicial assistant.

Akazaki J.

Released: December 12, 2023