SUPREME COURT OF CANADA

Citation: Contino v. Leonelli‑Contino, [2005] 3 S.C.R. 217, 2005 SCC 63

Date: 20051110

Docket: 30100

Between:

Joanne Leonelli‑Contino

Appellant

and

Joseph Contino

Respondent

Coram: McLachlin C.J. and Major, Bastarache, Binnie, LeBel, Deschamps, Fish, Abella and Charron JJ.

Reasons for Judgment:

(paras. 1 to 83)

Dissenting reasons:

(paras. 84 to 157)

Bastarache J. (McLachlin C.J. and Major, Binnie, LeBel, Deschamps, Abella and Charron JJ. concurring)

Fish J.

Contino v. Leonelli‑Contino, [2005] 3 S.C.R. 217, 2005 SCC 63

Joanne Leonelli‑Contino Appellant

v.

Joseph Contino Respondent

Indexed as: Contino v. Leonelli‑Contino

Neutral citation: 2005 SCC 63.

File No.: 30100.

2005: January 14; 2005: November 10.

Present: McLachlin C.J. and Major, Bastarache, Binnie, LeBel, Deschamps, Fish, Abella and Charron JJ.

on appeal from the court of appeal for ontario

Family law — Maintenance — Federal Child Support Guidelines — Shared custody — Proper approach to application of s. 9 of Federal Child Support Guidelines, SOR/97‑175.

The appellant mother and the respondent father entered into a separation agreement in 1992. The separation agreement provided for joint custody of their son, whose daily residence was to be with his mother, and for the payment by the father of $500 per month in child support. In 1998, this amount was raised to $563. Three years later, the father applied for a reduction in the amount of child support pursuant to s. 9 of the Federal Child Support Guidelines because the child was now in his physical custody 50 percent of the time. Both parties filed their 1998, 1999 and 2000 tax returns as well as their respective financial statements. At the time, the mother’s income was about $68,000 and the father’s around $87,000. Both parties attributed 50 percent of their fixed and variable expenses to the child. The mother and father respectively assigned $1,916.95 and $1,814 of their total expenses to the child. The motions judge granted the motion and reduced the amount of child support to $100 per month. The Divisional Court set aside the decision and ordered the father to pay the full Table amount of $688 per month. The Court of Appeal reduced the monthly amount payable by the father to $399.61. It used the simple set‑off amount as a starting point for determining the support amount (s. 9(a)) and adjusted the set‑off amount by applying a multiplier of 67.6 percent to account for the mother’s fixed costs (s. 9(b)) and by taking the actual situation of the parents and the child into account (s. 9(c)).

Held (Fish J. dissenting): The appeal should be allowed.

Per McLachlin C.J. and Major, Bastarache, Binnie, LeBel, Deschamps, Abella and Charron JJ.: Section 9 of the Guidelines expressly provides for a particular regime in cases of shared custody, and this implies a departure from the payor/recipient model that comes under s. 3. Section 9 requires a court to determine the amount of child support in accordance with the three listed factors once the 40 percent threshold is met. The specific language of s. 9 warrants emphasis on flexibility and fairness to ensure that the economic reality and particular circumstances of each family are properly accounted for. The three factors structure the exercise of the discretion and none of them should prevail. The weight given to each factor will vary according to the particular facts of each case. Under s. 9, there is no presumption in favour of awarding at least the Guidelines amount under s. 3. Nor is there a presumption in favour of reducing the parent’s child support obligation downward from the Guidelines amount, as it is possible that, after a careful review of all of the factors in s. 9, a court will come to the conclusion that the Guidelines amount is the proper amount of child support. [19‑31] [39]

Under s. 9(a), a court is required to take the financial situations of both parents into account, but the provision does not include a conclusive formula to determine how the Table amounts are to be considered or accounted for. The simple set‑off amount is the preferable starting point for the s. 9 analysis, but it must be followed by an examination of the continuing ability of the recipient parent to meet the needs of the child, especially in light of the fact that many costs are fixed. Where both parents are making effective contributions, it is necessary to verify how each parent’s actual contribution compares to the Table amount that is provided for each of them when considered payor parents. This will provide the judge with better insight when deciding whether the adjustments to be made to the set‑off amount are based on the actual sharing of child‑related expenses. The court retains the discretion to modify the set‑off amount where, considering the financial realities of the parents, it would lead to a significant variation in the standard of living experienced by the children as they move from one household to the other. [40‑51]

Section 9(b) does not refer merely to the expenses assumed by the payor parent as a result of the increase in access time from less than 40 percent to more than 40 percent. This paragraph recognizes that the total cost of raising children may be greater in shared custody situations than in sole custody situations. Given that some applications under s. 9 are not meant to obtain a variation order but constitute a first order and that the Table amounts in the Guidelines do not assume that the payor parent pays for any expense for the child, the court will consider all of the payor parent’s costs. The court will examine the budgets and actual expenditures of both parents in addressing the needs of the children and determine whether shared custody has resulted in increased costs globally. These expenses will be apportioned between the parents in accordance with their respective incomes. [52‑53]

Lastly, s. 9(c) vests the court with a broad discretion to analyse the resources and needs of both the parents and the children. It is important to keep in mind the objectives of the Guidelines, requiring a fair standard of support for the child and fair contributions from both parents. The court will look at the standard of living of the child in each household and the ability of each parent to absorb the costs required to maintain the appropriate standard of living in the circumstances. Financial statements and/or child expense budgets are necessary for a proper evaluation of s. 9(c). There is no need to resort to s. 10 and s. 7 of the Guidelines either to increase or to reduce support, since the court has full discretion under s. 9(c) to consider “other circumstances” and order the payment of any amount above or below the Table amounts. It may be that s. 10 would find application in an extraordinary situation, but that is certainly not the case here. [68‑72]

It is important that the parties lead evidence relating to ss. 9(b) and 9(c), and courts should demand information from the parties when the evidence is deficient. A court should neither make “common sense” assumptions about costs incurred by the payor parent, nor apply a multiplier to account for the fixed costs of the recipient parent. [56-57]

Taking into consideration all the factors prescribed in s. 9 and applying them to the particular factual context of this case, the mother should be awarded the sum of $500 per month in child support. [73‑80]

Per Fish J. (dissenting): Support orders under s. 9 of the Guidelines are discretionary by design, but constrained by principle and subject to the overriding requirement of fitness. In determining an appropriate award, all the governing factors set out in s. 9 must be considered. “Appropriate” does not mean mathematically or methodologically ascertainable with precision. It means within an acceptable range that is in each case determined by applying in a principled manner the s. 9 factors to the proven facts and particular circumstances of the matter. An appropriate support order in this case should ensure insofar as possible that the child of the parties enjoys a standard of living that is reasonably comparable to his standard of living before the divorce and does not vary markedly in material respects moving from one household to the other. The method for achieving this outcome should be evidence‑based. [92] [103] [105]

Section 9(a) of the Guidelines requires the court to take the Table amounts into account in fixing child support for shared custody arrangements. The simple set‑off of the Table amounts for sole custody may be a convenient starting point in a global consideration of all the factors that must be weighed under s. 9. The purpose of s. 9(b) is to ensure that the increased costs of shared custody are properly reflected in the support order. This relates essentially to the duplication of fixed costs and to other expenses that result from the exigencies of shared custody. The extent of the duplication of fixed costs will generally be apparent from the budgets submitted by the parties. However, where there is no evidentiary basis for taking into account the increased costs of shared custody arrangements, a court should not resort to multipliers, but instead can reopen the hearing for that purpose. Section 9(c) is the appropriate place for an apportionment of certain expenditures according to the respective incomes of the parents, including the duplicated expenses identified pursuant to s. 9(b). Finally, having applied all the s. 9 factors and in spite of the broad language of s. 9(c), a trial judge may still not have arrived at a just award. In such cases, s. 10(1) of the Guidelines allows a court to “award an amount of child support that is different from the amount determined under any of sections 3 to 5, 8 or 9”. [114-18] [125] [130-31]

In this case, the set‑off amount is $128 per month, and only two types of expenditures should be equalized between the parents: the duplications and other incremental costs inherent in shared custody, which s. 9(b) requires a court to consider, and the variable child care costs that might otherwise be shared by the parents. Furthermore, there are two non‑numerical factors under s. 9(c) that must be taken into account: the disparity between the net assets of the parents and, since this case involves a modification and not an initial support order, the arrangement in effect prior to shared custody. The monthly child support previously paid by the father is an important consideration in the circumstances of this case because the mother incurred fixed costs that were in part a function of the support she was receiving at the time. However, the support previously paid creates no entitlement to continued support at the same level where the parties have moved to shared custody. The Court of Appeal awarded a support order of $399.61. This award lies within the acceptable range. The Court of Appeal set out the basic principles correctly. Its unfortunate observation as to the permissible use in some circumstances of a “stock multiplier” had no bearing on its conclusion, and the limited effect of its resort to a multiplicative factor is adequately compensated by the other factors. [129] [132] [137] [145] [150] [154]

Cases Cited

By Bastarache J.

Distinguished: Francis v. Baker, 1999 659 (SCC), [1999] 3 S.C.R. 250; referred to: Green v. Green (2000), 187 D.L.R. (4th) 37, 2000 BCCA 310; Rizzo & Rizzo Shoes Ltd. (Re), 1998 837 (SCC), [1998] 1 S.C.R. 27; Chartier v. Chartier, 1999 707 (SCC), [1999] 1 S.C.R. 242; Jamieson v. Jamieson, [2003] N.B.J. No. 67 (QL), 2003 NBQB 74; Berry v. Hart (2003), 233 D.L.R. (4th) 1, 2003 BCCA 659; Fletcher v. Keilty (2004), 269 N.B.R. (2d) 302, 2004 NBCA 34; Slade v. Slade (2001), 195 D.L.R. (4th) 108, 2001 NFCA 2; Dean v. Brown (2002), 209 N.S.R. (2d) 70, 2002 NSCA 124; Hill v. Hill (2003), 213 N.S.R. (2d) 185, 2003 NSCA 33; Cabot v. Mikkelson (2004), 242 D.L.R. (4th) 279, 2004 MBCA 107; Dennis v. Wilson (1997), 1997 1933 (ON CA), 104 O.A.C. 250; Wylie v. Leclair (2003), 2003 49737 (ON CA), 64 O.R. (3d) 782; E. (C.R.H.) v. E. (F.G.) (2004), 29 B.C.L.R. (4th) 43, 2004 BCCA 297; Luedke v. Luedke (2004), 198 B.C.A.C. 293, 2004 BCCA 327; Gieni v. Gieni (2002), 29 R.F.L. (5th) 60, 2002 SKCA 87; Middleton v. MacPherson (1997), 1997 11517 (AB KB), 204 A.R. 37; Moran v. Cook (2000), 2000 22542 (ON SC), 9 R.F.L. (5th) 352; Harrison v. Harrison (2001), 2001 60967 (ON SC), 14 R.F.L. (5th) 321; Paras v. Paras, 1970 370 (ON CA), [1971] 1 O.R. 130.

By Fish J. (dissenting)

Paras v. Paras, 1970 370 (ON CA), [1971] 1 O.R. 130.

Statutes and Regulations Cited

Divorce Act, R.S.C. 1985, c. 3 (2nd Supp.), ss. 15(8) [now repealed], 17(8) [now repealed], 26.1(2) [ad. 1997, c. 1, s. 11].

Federal Child Support Guidelines, SOR/97‑175, ss. 1, 3, 4, 5, 7, 8, 9, 10.

Authors Cited

Canada. Department of Justice. Child Support Team. Research Report. Formula for the Table of Amounts Contained in the Federal Child Support Guidelines: A Technical Report. Ottawa: The Department, 1997.

Canada. Department of Justice. Children Come First: A Report to Parliament Reviewing the Provisions and Operation of the Federal Child Support Guidelines, vol. 2. Ottawa: The Department, 2002.

Colman, Gene C. “Contino v. Leonelli-Contino — A Critical Analysis of the Ontario Court of Appeal Interpretation of Section 9 of the Child Support Guidelines” (2004), 22 C.F.L.Q. 63.

Federal/Provincial/Territorial Family Law Committee. Child Support: Public Discussion Paper. Ottawa: The Committee, 1991.

Finnie, Ross, Carolina Giliberti and Daniel Stripinis. An Overview of the Research Program to Develop a Canadian Child Support Formula. Ottawa: Department of Justice, 1995.

MacDonald, James C., and Ann C. Wilton. Child Support Guidelines: Law and Practice, 2nd ed., vol. 1. Toronto: Carswell, 1998 (loose‑leaf updated 2004, release 4).

McLeod, James G. “The Proposed Child Support Guideline Package: The Scope of Judicial Discretion”, in Federal Child Support Guidelines: Reference Manual. Ottawa: Department of Justice, 1997, F‑1.

Melli, Marygold S. “Guideline Review: Child Support and Time Sharing by Parents” (1999), 33 Fam. L.Q. 219.

Melli, Marygold S., and Patricia R. Brown. “The Economics of Shared Custody: Developing an Equitable Formula for Dual Residence” (1994), 31 Houst. L. Rev. 543.

Millar, Paul, and Anne H. Gauthier. “What Were They Thinking? The Development of Child Support Guidelines in Canada” (2002), 17 C.J.L.S. 139.

Payne, Julien D., and Marilyn A. Payne. Child Support Guidelines in Canada 2004. Toronto: Irwin Law, 2004.

Rogerson, Carol. “Child Support Under the Guidelines in Cases of Split and Shared Custody” (1998), 15 Can. J. Fam. L. 11.

Sullivan, Ruth. Sullivan and Driedger on the Construction of Statutes, 4th ed. Markham, Ont.: Butterworths, 2002.

Thompson, D. A. Rollie. “Annotation to E. (C.R.H.) v. E. (F.G.)”, 2004 BCCA 297, 2004 CarswellBC 1157.

Thompson, D. A. Rollie. “Case Comment: Contino v. Leonelli‑Contino” (2004), 42 R.F.L. (5th) 326.

Wensley, Kim Hart. “Shared Custody — Section 9 of the Federal Child Support Guidelines: Formulaic? Pure Discretion? Structured Discretion?” (2004), 23 C.F.L.Q. 63.

APPEAL from a judgment of the Ontario Court of Appeal (O’Connor A.C.J.O., Weiler and Rosenberg JJ.A.) (2003), 2003 30327 (ON CA), 67 O.R. (3d) 703, 232 D.L.R. (4th) 654, 42 R.F.L. (5th) 295, 178 O.A.C. 281, [2003] O.J. No. 4128 (QL), allowing an appeal from a judgment of the Divisional Court (Carnwath, E. Macdonald and Czutrin JJ.) (2002), 2002 2774 (ON SCDC), 62 O.R. (3d) 295, 166 O.A.C. 172, 30 R.F.L. (5th) 266, [2002] O.J. No. 4620 (QL), allowing an appeal from a decision of Rogers J. and setting aside her order reducing the father’s child support payments. Appeal allowed, Fish J. dissenting.

Deidre D. Smith, Susan E. Milne and Gary Joseph, for the appellant.

Thomas G. Bastedo, Q.C., and Samantha Chousky, for the respondent.

The judgment of McLachlin C.J. and Major, Bastarache, Binnie, LeBel, Deschamps, Abella and Charron JJ. was delivered by

Bastarache J. —

1. Introduction

[1] When the federal government decided to adopt in 1997 the Federal Child Support Guidelines, SOR/97-175 (“Guidelines”), its first decision was to choose between different formulae and design a system that would be adapted to the Canadian context. The formulae that were considered with greatest attention were the four in use in the United States: (1) the Income‑Shares Model, where the child should receive the same amount of the parental income, in proportion to each parent’s income, as before the separation; (2) the Delaware or Melson Formula, where basic needs are met before determining how the child is to share the remaining parental income; (3) the Flat Percentage of Income Model, where it is assumed that each parent will spend the same percentage of his or her income on the child and the non‑custodial parent’s share is fixed by regulation; and (4) the Income Equalization Model which is designed to equalize the standards of living of custodial and non‑custodial parents so that the child will experience the lowest reduction in standard of living possible (Federal/Provincial/Territorial Family Law Committee, Child Support: Public Discussion Paper (1991), at pp. 10‑11).

[2] The government decided to adopt a unique formula in the case of split custody; that is the situation where each spouse has custody of one or more children. It is best described as the revised fixed percentage. It is included in ss. 3 and 8 of the Guidelines (see Appendix) and has the features of the flat percentage formula, but uses a specific set of underlying principles to arrive at percentages that vary according to income level. The formula produces a schedule of payment amounts taking into consideration tax consequences. It provides for some add-ons with respect to special expenses (R. Finnie, C. Giliberti and D. Stripinis, An Overview of the Research Program to Develop a Canadian Child Support Formula (1995), at pp. 27‑28).

[3] When dealing with shared custody, however, the formula used in ss. 3 and 8 was not retained. New categories of custodial arrangements were created under s. 9 which states:

9. Where a spouse exercises a right of access to, or has physical custody of, a child for not less than 40 per cent of the time over the course of a year, the amount of the child support order must be determined by taking into account

(a) the amounts set out in the applicable tables for each of the spouses;

(b) the increased costs of shared custody arrangements; and

(c) the conditions, means, needs and other circumstances of each spouse and of any child for whom support is sought.

These shared custodial arrangements required the application of an entirely different formula, one that is not designed with the same guiding principles. Guidelines amounts applicable to the former non‑custodial parent or to the highest income earner in the case of a first application cannot therefore be considered to be presumptively applicable. Shared custody arrangements are not a simple variation of the general regime, they constitute by themselves a complete system.

[4] The application of the factors under s. 9 of the Guidelines have proven to pose serious difficulties. The problems have been addressed in terms of fairness. As mentioned by professor C. Rogerson in her article “Child Support Under the Guidelines in Cases of Split and Shared Custody” (1998), 15 Can. J. Fam. L. 11, at p. 20:

Pushing in favour of some adjustment is a concern for fair and consistent treatment of payors who incur increased expenses during the time they spend with the child. There are two dimensions to the fairness claim. The first is fairness between the payor and the support recipient, who is arguably being relieved of some costs assumed by the payor. The second is fair and consistent treatment of the payor as compared to payors at the same income level who may not be spending any money directly on their children apart from the payment of child support.

But then adjustments are hard to evaluate. More time spent with a child may not involve increased spending or significant savings for the other parent. Where there is a significant disparity of incomes, a new formula can mean a drastic change in the amount of support for the lower‑income parent, who was previously the custodial parent, and exacerbate the differences in standard of living in the two households. There is also a concern that shared custody can entail more cost in duplication of services and leave less money for support.

[5] Against this backdrop, the role of the Court is to interpret the Guidelines as drafted by Parliament. Section 9 is labelled “Shared custody”. Forty percent or more time spent with physical access to the child triggers the application of the three factors in s. 9. We are not concerned in this case with the difficulties sometimes encountered in determining whether the threshold has been met, but with the quantum of support to be awarded once it is. The Court is being asked to decide whether the s. 9 award can be greater than the Guidelines amount; whether the Guidelines amounts are presumptively applicable; whether all three factors in s. 9 are to be given equal weight; whether “increased costs” under s. 9 refers to increased costs of the previously non‑custodial parent or increased costs resulting from the shared custodial arrangement; whether a multiplier can be used in the absence of evidence of increased costs; and how actual needs, conditions and means are taken into account in deciding on a deviation from the Guidelines amounts. These questions must be approached in the context of the particular facts of this case, to which I now turn.

1.1 Overview of the Facts

[6] The mother and the father were married on October 30, 1982, and their only child, Christopher, was born on March 26, 1986. After they separated, they entered into an agreement, dated May 25, 1992, that provided for joint custody of Christopher. His daily residence was to be with the mother, free and liberal access being given to the father. The agreement further provided that the father would pay child support in the amount of $500 per month, subject to an annual cost of living increase. No annual increases were ever paid by the father, and in 1998, the mother applied for the Guidelines amount of support. By minutes of settlement signed in July of 1998, the father agreed to pay $563 per month in child support, based on his annual income of $68,712, adjusted annually in accordance with the Guidelines. Again, child support was not adjusted, although the father’s income rose to $83,527.58 in 1999. The parties also shared equally in Christopher’s orthodontic expenses, although the mother’s income of $53,292 was less than that of the father.

[7] In 2000, the mother began taking a course on Tuesday nights and asked if the father would switch nights with her so that Christopher would be with the father on Tuesdays instead of Thursdays. The father stated that he would take Christopher on both days for the duration of the course.

[8] As a result of having Christopher with him for an additional night each week, in September of 2000, the father requested a reduction in child support based on the s. 9 shared custody provisions, but the mother refused. In March 2001, the father applied to vary the amount of child support on the grounds that the extra night resulted in Christopher being with him 50 percent of the time. Both parties filed their 1998, 1999 and 2000 tax returns as well as their respective financial statements. Both parties attributed 50 percent of their fixed and variable expenses such as mortgage, taxes and groceries to Christopher and both claimed additional expenses for him. The mother and father respectively assigned $1,916.95 and $1,814 of their total expenses to Christopher. The mother stated that monthly expenses for him for clothes, school fees and activities for him totalled $275 per month, while the father claimed $120 per month in variable expenses. The mother also invested $153.84 monthly in a Registered Education Savings Plan (“RESP”) for Christopher’s benefit. Both these variable expenses and the amount invested monthly in the RESP were included in both parents’ total expenses.

1.2 Judicial History

1.2.1 Ontario Superior Court of Justice (Unified Family Court)

[9] Rogers J. endorsed the record that in recognition of the shared custody regime, pursuant to s. 9, there would be a reduction in the Guidelines amount. She ordered a retroactive reduction in support to $100 per month, commencing in September 2000, requiring the mother to repay the overpayment at a rate of $50 per month. Further, she ordered the parties to share equally in s. 7 expenses (special or extraordinary expenses). The parties were entitled to claim the dependent deduction for Christopher in alternate years. The father was awarded costs in the amount of $3,800.

[10] A transcription of the proceedings revealed that there had been an earlier finding that Christopher currently lives with his father 50 percent of the time as it was conceded by the mother at a pre-hearing conference. Rogers J. permitted the father to file recent copies of his tax returns but did not permit the mother, who was self‑represented, to question any of the figures. Rogers J. fixed the mother’s income at $68,082, inclusive of a $7,000 bonus. The father’s income was found to be $87,000. She found that the Table amount of child support for the father was $688 per month, while for the mother, it was $560 per month. She determined the difference in their respective child support obligations to be $128, half of which was $64, an amount that she then grossed up by 50 percent to produce a figure of $96. Rogers J. raised this to $100 per month for ease of calculation.

[11] Rogers J. indicated that she would not hear evidence from the mother concerning the financial hardship she experienced as a result of a move that she made for Christopher’s benefit, as this was not in the affidavit material filed before her. She remarked:

I can only hear about what’s already in the materials. So go from what the evidence is that you’ve introduced and you may wish to tell me how that relates to this but essentially there’s been a finding that there’s a fifty percent regime here. We’re just doing math today. Okay? [Emphasis added.]

1.2.2 Ontario Superior Court (Divisional Court) (2002), [2002 2774 (ON SCDC)](https://www.canlii.org/en/on/onscdc/doc/2002/2002canlii2774/2002canlii2774.html), 62 O.R. (3d) 295

[12] The Divisional Court found that the motions judge had failed to consider whether there had been any increased costs as a result of the shared custody regime, as required by s. 9(b), and that she had also failed to consider the factors listed in s. 9(c). Without this analysis, the court stated that it was difficult to determine how the support amount was calculated. The court also observed that there was a lack of judicial consensus on how support should be calculated under s. 9 and that the proper procedure was also an issue. The court reviewed the relevant portions of the Divorce Act, R.S.C. 1985, c. 3 (2nd Supp.), noting that the court was required to order the Guidelines amount except with respect to five discretionary situations: children over the age of majority (s. 3(2)); high income earners (s. 4); step‑parents (s. 5); undue hardship (s. 10); and shared custody (s. 9). The court noted the principles enunciated by this Court in Francis v. Baker, 1999 659 (SCC), [1999] 3 S.C.R. 250, setting out the proper approach to the application of the Guidelines. Acknowledging that Francis v. Baker dealt with the proper approach to the exercise of the court’s discretion to deviate from the Guidelines amount where the payor’s income was over $150,000, the court determined that the same principles should apply to the other permitted deviations to ensure a consistent approach. The court held that in considering an application for deviation under any statutory exception, a court must first:

(a) make a presumption in favour of the Guidelines amount;

(b) impose an onus on the party seeking a deviation to establish on “clear and compelling evidence” that the deviation is in the child’s best interest;

(c) consider all the statutory factors noted in the section establishing a permitted deviation without providing pre‑eminence to any factor;

(d) deny any application for a deviation based merely upon invocation of the discretionary provision;

(e) focus on the child’s actual circumstances and not perceived parental fairness considerations, such as balancing of parental means. [para. 16]

[13] The Divisional Court presented a three-step analysis for applications for deviation from the Guidelines amount, beginning with a determination of whether the applicant exercises a right of access for no less than 40 percent of the time over the course of the year. Secondly, the court must find that the presumption in favour of the Guidelines amount has been rebutted. This is satisfied if the applicant has discharged the onus of establishing on clear and compelling evidence that the decision is in the child’s best interests. The court stated that there is no right of deviation merely upon invocation of the discretionary provision of s. 9. The court must consider all the statutory factors without providing pre‑eminence to any factor. The focus of determination must be the child’s actual circumstances, not perceived parental fairness, such as balancing parental means. Finally, once the court makes the second finding and the presumption is rebutted, the court considers s. 9(a), (b) and (c) in the exercise of its discretion. In addition, the court adopted the view of the British Columbia Court of Appeal in Green v. Green (2000), 187 D.L.R. (4th) 37, 2000 BCCA 310, and rejected the formulaic approach to s. 9. It granted the appeal and increased support to $688 per month retroactive to September 1, 2000. The court was of the view that there was no clear and compelling evidence that the deviation was in the child’s best interest and that the motions judge was clearly wrong in law to depart from the Guidelines amount.

1.2.3 Ontario Court of Appeal (2003), [2003 30327 (ON CA)](https://www.canlii.org/en/on/onca/doc/2003/2003canlii30327/2003canlii30327.html), 67 O.R. (3d) 703

[14] Weiler and Rosenberg JJ.A., for the court, found that the Divisional Court had erred in its interpretation of s. 9 of the Guidelines. They held that once the applicant parent shows that the 40 percent threshold has been reached, there is no discretion at this early stage; the court must apply s. 9 and the presumptive Table amounts no longer apply. The Court of Appeal was also of the view that the motions judge had erred in adopting a formulaic approach. Recognizing that the courts in the country were generally divided on the proper approach to s. 9, the Court of Appeal indicated a preference for a structured discretionary approach in order to add a dimension of predictability and objectivity. Although the Court of Appeal expressed concern about the lack of information before the motions judge concerning the increased costs of shared custody, it was satisfied that there was sufficient evidence in the record to determine the appropriate quantum under s. 9 without falling back on a strict formulaic approach.

[15] According to the Court of Appeal, Francis v. Baker could not be taken out of context and did not apply to s. 9. The court found that the Divisional Court erred in failing to consider the significant differences in wording contained in the various discretionary sections. In particular, the court found that the Divisional Court’s analysis introduced a presumption not called for by the language of the section, one that is also contrary to the principles of statutory interpretation. The Divisional Court was wrong in finding that there is no right of deviation merely upon invocation of the discretionary provision of s. 9. According to the Court of Appeal, where the 40 percent threshold has been met, the provision establishes the manner of calculating the amount of child support. The language of s. 9 requires the court to determine child support by taking into account the three factors contained in that section: paras. (a), (b), and (c).

[16] The Court of Appeal explained that under para. (a), calculation of the simple set‑off amount was a useful starting point in determining the amount of support. The set‑off amount must then be adjusted in accordance with paras. (b) and (c). Under s. 9(b), the court found that it must take into account the increased costs of the shared custody arrangement. The paragraph recognizes that the increased access for one parent does not result in a dollar for dollar reduction for the other parent and that each parent must bear some of the burden of the additional costs. The court acknowledged that the appropriate method for accounting for the increased costs of shared custody is through the examination of evidence, but in the absence of such evidence, the use of a multiplier is a useful tool to recognize the custodial parent’s fixed costs, and to augment the figure arrived at after a simple set‑off. The Court of Appeal was of the view that under s. 9(b), the court could make common sense assumptions that the non‑custodial parent would have additional costs for variable items such as food, entertainment and transportation, but that evidence would have to be adduced if the parent wanted the court to consider any increase in fixed costs such as those associated with larger accommodations. Finally, according to the Court of Appeal, para. (c) gives the court a broad discretion to take into account the actual situation of the parents and children in ordering support. The court indicated that one goal should be to ensure that the child enjoys a comparable standard of living in both households.

[17] Weiler and Rosenberg JJ.A. then applied these factors to the circumstances of the case. They started with a simple set‑off and chose a multiplier of 67.6 percent because the mother’s accommodations costs were 67.6 percent of her fixed costs for the child. This produced a figure of $215. The court noted that the father had not provided any evidence of increased costs for Christopher, so it made the common sense assumption that variable costs for food and entertainment had risen pursuant to s. 9(c). The court went on to consider the spending patterns of both parents for the child, noting that the mother’s expenses were $403.41 and the father’s $270 per month, for a total of $673.41. The court found that the ratio of the father’s salary to the mother’s salary was 55:45, and that the father should therefore pay 55 percent of $673.41 or $370, the mother being responsible for $303. As the father was only paying $270 per month, the court found that he should pay the mother an additional $100 per month, to be added to the set‑off amount of $215. The mother’s monthly RESP payment of $153.84 for Christopher was then factored in by the court, and divided on the same ratio. This resulted in the addition of $84.61 to the $315, for a total of $399.61 in support to be paid by the father. As there was no evidence of financial hardship before the court, no increase under s. 10 was ordered.

[18] The court determined that the motions judge had erred in making the reduced support award retroactive to the time the father first requested a reduction because the father had never increased support in accordance with the cost of living as required by the agreement of the parties.

2. Analysis

2.1 Interpretation of Section 9 of the Guidelines

[19] In order to determine the correct interpretation to be given to s. 9 of the Guidelines, it is necessary to examine the words of the provision in their entire context and in their grammatical and ordinary sense harmoniously with the scheme of the Guidelines, the object of the Guidelines, and the intention of Parliament (see, e.g., Rizzo & Rizzo Shoes Ltd. (Re), 1998 837 (SCC), [1998] 1 S.C.R. 27, at para. 21; Francis v. Baker, at para. 34; Chartier v. Chartier, 1999 707 (SCC), [1999] 1 S.C.R. 242).

[20] Before turning to the heart of this case, it is important to point out what is in essence an issue of semantics. Parties and courts across the country have inconsistently referred to the parents under s. 9 as the “custodial” parent, “non-custodial” parent, “payor” parent and “recipient” parent. There is no perfect terminology. However, it is clear that in a shared physical custody arrangement, given the nature of child support, one cannot ignore that a transfer of money from one parent to the other will almost always occur. Thus, for sake of clarity, I will use the concepts of “payor” parent and “recipient” parent.

[21] For ease of reference, I again reproduce s. 9 of the Guidelines:

9. Where a spouse exercises a right of access to, or has physical custody of, a child for not less than 40 per cent of the time over the course of a year, the amount of the child support order must be determined by taking into account

(a) the amounts set out in the applicable tables for each of the spouses;

(b) the increased costs of shared custody arrangements; and

(c) the conditions, means, needs and other circumstances of each spouse and of any child for whom support is sought.

[22] The mother submits that there is a presumption in favour of the Guidelines that applies to the exercise of all discretionary powers, including those found in s. 9. According to her, the onus is on the party seeking a deviation to establish on “clear and compelling evidence” that the deviation is in the child’s best interest. She relies on this Court’s decision in Francis v. Baker. The same approach was taken by the Divisional Court. I cannot accept her argument.

[23] In Francis v. Baker, I held that, under s. 4 of the Guidelines (see Appendix), which deals with the situation of high income earners (income over $150,000), there is a presumption in favour of the Guidelines amount. Guideline figures can only be increased or reduced if the party seeking such a deviation has rebutted the presumption of appropriateness. No right of deviation exists merely by pleading the discretionary provision. As earlier noted, s. 9 however expressly provides for a particular regime in cases of shared custody. This implies a departure from the payor/recipient model that comes under s. 3. In fact, s. 3 recognizes that the calculations under that provision will not apply where “otherwise provided under these Guidelines”.

[24] While ss. 3(2), 4, 5 and 10 (see Appendix) provide a framework establishing a structured discretion, each provision incorporates distinct factors which are absent in s. 9. Sections 3(2) and 4 specifically prescribe that the amount in the Guidelines is mandatory unless the court considers that there are reasons to find that it is inappropriate. Section 9 does not contain such a presumption. As submitted by the father, if the drafters of the Guidelines had intended this approach, they would have used the same words to provide for direction in all of the relevant sections. In fact, the wording of s. 9 is imperative. The court “must” determine the amount of child support in accordance with the three listed factors once the 40 percent threshold is met. There is no discretion as to when the section is to be applied: discretion exists only in relation to the quantification of child support (J. D. Payne and M. A. Payne, Child Support Guidelines in Canada 2004 (2004), at p. 254).

[25] Given the presumption of consistent expression, it is possible to infer an intended difference in meaning from the use of different words or a different form of expression: R. Sullivan, Sullivan and Driedger on the Construction of Statutes (4th ed. 2002), at p. 164. Consequently, the reliance by the mother on the presumption enunciated in Francis v. Baker cannot stand and the Court of Appeal was correct in distinguishing the decision.

[26] Furthermore, s. 9(a) refers only to the Table amounts, not the Guidelines amounts, thus precluding consideration under that paragraph of all of the discretionary factors that are allowed under the Guidelines for departure from the Table amounts, as in the case of s. 7 add‑ons. This suggests that s. 9(a) is relatively narrow in its scope and cannot form the exclusive basis for an award of support under s. 9 unless paras. (b) and (c) are taken into consideration (see Rogerson, at pp. 57-58; K. H. Wensley, “Shared Custody — Section 9 of the Federal Child Support Guidelines: Formulaic? Pure Discretion? Structured Discretion?” (2004), 23 C.F.L.Q. 63).

[27] The three factors structure the exercise of the discretion. These criteria are conjunctive: none of them should prevail (see Wensley, at p. 90; Payne and Payne, at p. 254; Jamieson v. Jamieson, [2003] N.B.J. No. 67 (QL), 2003 NBQB 74, at para. 24). Consideration should be given to the overall situation of shared custody and the costs related to the arrangement while paying attention to the needs, resources and situation of parents and any child. This will allow sufficient flexibility to ensure that the economic reality and particular circumstances of each family are properly accounted for. It is meant to ensure a fair level of child support.

[28] The mother argues that, for shared custody, the Guidelines provide limited discretion to deviate from the amounts determined under s. 3, and then only upon consideration of the child’s actual circumstances. For the reasons mentioned above, I must reject this argument. That is not the nature of the discretion provided in s. 9.

[29] In the Court of Appeal decision, Weiler and Rosenberg JJ.A. write, at para. 40:

It seems to us that the error by the Divisional Court was in treating the s. 3 calculation as the “Guidelines amount”. The court made the same error in saying that “[t]here is no right of deviation merely upon invocation of the discretionary provision of s. 9.” To the contrary, where the 40 per cent threshold in s. 9 has been met, that provision establishes the manner of calculating the amount of child support, a calculation quite different from that set out in s. 3 or s. 4. As we read s. 9, especially the use of the phrase “must be determined” in the opening paragraph, deviation from the amount determined under s. 3 or s. 4 is required and the court must follow the process set out in s. 9. We appreciate that this results in more uncertainty and introduces more subjectivity in the calculation of child support orders, but that is an inevitable consequence of the wording of s. 9.

[30] These comments may lead some parents to think that there should be an automatic reduction in the amount of child support in a case such as this one. In my opinion, there is only an automatic deviation from the method used under s. 3, but not necessarily from the amount of child support. As submitted by the mother, it is quite possible that after a careful review of all of the factors in s. 9, a trial judge will come to the conclusion that the Guidelines amount will remain the proper amount of child support (see, e.g., Berry v. Hart (2003), 233 D.L.R. (4th) 1, 2003 BCCA 659).

[31] Thus, not only is there no presumption in favour of awarding at least the Guidelines amount under s. 3, there is no presumption in favour of reducing the parent’s child support obligation downward from the Guidelines amount (Wensley, at pp. 89-90).

[32] The underlying principle of the Guidelines is that “spouses have a joint financial obligation to maintain the children of the marriage in accordance with their relative abilities to contribute to the performance of that obligation” (Divorce Act, s. 26.1(2) (see Appendix)). The Guidelines reflect this principle through these stated objectives (Guidelines, s. 1):

(a) to establish a fair standard of support for children that ensures that they continue to benefit from the financial means of both spouses after separation;

(b) to reduce conflict and tension between spouses by making the calculation of child support orders more objective;

(c) to improve the efficiency of the legal process by giving courts and spouses guidance in setting the levels of child support orders and encouraging settlement; and

(d) to ensure consistent treatment of spouses and children who are in similar circumstances.

[33] These objectives create a palpable tension in the Guidelines (Rogerson, at pp. 59 and 93). At para. 40 in Francis v. Baker, I wrote that the proper construction of a provision “requires that the objectives of predictability, consistency and efficiency on the one hand, be balanced with those of fairness, flexibility and recognition of the actual ‘condition[s], means, needs and other circumstances of the children’ on the other”. Like s. 4 in that case, s. 9 must here be interpreted with these objectives in mind. Parliament, in adopting s. 9, deliberately chose to emphasize the objectives of fairness, flexibility and recognition of the actual conditions, means, needs and other circumstances of each spouse and of any child for whom support is sought, even if to the detriment of predictability, consistency and efficiency to some degree. The legislator recognized in s. 9 that there is a wide range of situations of shared custody depicting the reality of different families. The British Columbia Court of Appeal, in the same vein, observed that there is a myriad of fact patterns which come under the application of s. 9: Green v. Green, at para. 34.

[34] The mother submitted to the Court a detailed account of the legislative history of Bill C‑41 which provided for the necessary amendments to the Divorce Act (1985) and the attendant regulations which became the Guidelines. Many of the policy concerns raised by the mother are not disputed in this case and may, in my opinion, divert the Court from the real legal issue under scrutiny.

[35] The mother argues that the Department of Justice made clear to the Standing Senate Committee that the Table amounts assumed that an access parent was paying substantial costs for the benefit of the child when exercising access. She points out that there was a last‑minute amendment that lowered the threshold of shared custody to 40 percent, instead of 49.9 percent, with no accompanying amendment to the Table. The mother relies upon this proposition to assert that the Table as enacted takes into consideration expenses incurred by the non-custodial parent having custody up to 49.9 percent of the time. Thus, as I understand her argument, there would be less need for a reduction in the level of child support to the recipient parent once the 40 percent threshold is reached.

[36] I am of the opinion that this view is inconsistent with the statute. I agree with the father that the formula used to establish the standard Table amounts assumed that all of the expenditures for the children are met by the recipient parent and no account for any child‑related expenditures is incurred by the payor parent at any level of access (see Canada, Department of Justice, Formula for the Table of Amounts Contained in the Federal Child Support Guidelines: A Technical Report (1997), at p. 2; Wensley, at pp. 83‑85; G. C. Colman, “Contino v. Leonelli-Contino — A Critical Analysis of the Ontario Court of Appeal Interpretation of Section 9 of the Child Support Guidelines” (2004), 22 C.F.L.Q. 63, at pp. 71-74; P. Millar and A. H. Gauthier, “What Were They Thinking? The Development of Child Support Guidelines in Canada” (2002), 17 C.J.L.S. 139, at pp. 149 and 155-56).

2.2 Factors Under Section 9

[37] The framework of s. 9 requires a two‑part determination: first, establishing that the 40 percent threshold has been met; and second, where it has been met, determining the appropriate amount of support.

[38] With respect to the second part of the determination, the litigious issue in the case at bar, courts across the country have struggled to develop an interpretation of s. 9 that is consistent with the Guidelines’ objectives. While the approaches vary widely, they can be divided in two categories. One approach, similar to the approach used by the motions judge, can be described as the “formulaic approach”. The other approach, which may be described as the “discretionary approach”, eschews the use of formulae.

[39] The specific language of s. 9 warrants emphasis on flexibility and fairness. The discretion bestowed on courts to determine the child support amount in shared custody arrangement calls for the acknowledgment of the overall situation of the parents (conditions and means) and the needs of the children. The weight of each factor under s. 9 will vary according to the particular facts of each case. I will now consider each of the three s. 9 factors.

2.2.1 Section 9(a) — Amounts Set Out in the Applicable Tables for Each of the Spouses

[40] The first factor requires that the court take into account the financial situations of both parents (instead of the sole income of the spouse against whom the order is sought, as in s. 3). It is important to highlight the fact that the final and fully considered version of s. 9 does not include a conclusive formula to determine how the Table amounts are to be considered or accounted for.

[41] The Court of Appeal, while it agreed that the use of a formula is not explicitly required in the section, concluded that the set‑off approach in s. 8 could be a useful starting point to bring consistency and objectivity to the determination, especially in cases where there is limited information and the incomes of the parties are not widely divergent. I agree, but would caution against deciding these issues without proper information. I would particularly caution against a rigid application of the set-off which can entail, in the case of a variation order, a drastic change in support, dubbed the “cliff effect” by commentators (M. S. Melli and P. R. Brown, “The Economics of Shared Custody: Developing an Equitable Formula for Dual Residence” (1994), 31 Houst. L. Rev. 543, at p. 565; Rogerson, at p. 74; Wensley, at p. 70), that may not be warranted when a close examination of the financial situation of the parents and standard of living in both households is considered. The value of the set‑off is in finding a starting point for a reasonable solution taking into account the separate financial contribution from each parent. A court will depart from the set-off amount or make adjustments to it if it is inappropriate in light of the factors considered under ss. 9(b) and 9(c). The set-off amount must therefore be followed by an examination of the continuing ability of the recipient parent to meet the needs of the child, especially in light of the fact that many costs are fixed. As mentioned by numerous commentators, this is a problem in many cases where there is a great discrepancy in the incomes of the parents (see Rogerson, at p. 64). It is also a problem in cases where one parent actually incurs a higher share of the costs than the other (taking responsibility for clothing or activities for instance). I would also note that the 40 percent threshold itself should be irrelevant to this evaluation; the cliff effect is not merely a result of the threshold; it is a result of the different methodology.

[42] The Court of Appeal (as well as the father) summarized a number of applications of the set‑off approach adopted by Canadian courts (see Fletcher v. Keilty (2004), 269 N.B.R. (2d) 302, 2004 NBCA 34; Slade v. Slade (2001), 195 D.L.R. (4th) 108, 2001 NFCA 2; Dean v. Brown (2002), 209 N.S.R. (2d) 70, 2002 NSCA 124; Hill v. Hill (2003), 213 N.S.R. (2d) 185, 2003 NSCA 33; Cabot v. Mikkelson (2004), 242 D.L.R. (4th) 279, 2004 MBCA 107; Dennis v. Wilson (1997), 1997 1933 (ON CA), 104 O.A.C. 250; Wylie v. Leclair (2003), 2003 49737 (ON CA), 64 O.R. (3d) 782 (C.A.); Green v. Green; Berry v. Hart; E. (C.R.H.) v. E. (F.G.) (2004), 29 B.C.L.R. (4th) 43, 2004 BCCA 297; Luedke v. Luedke (2004), 198 B.C.A.C. 293, 2004 BCCA 327; Gieni v. Gieni (2002), 29 R.F.L. (5th) 60, 2002 SKCA 87; see also Children Come First: A Report to Parliament Reviewing the Provisions and Operation of the Federal Child Support Guidelines (2002), vol. 2, at pp. 68‑70; J. C. MacDonald and A. C. Wilton, Child Support Guidelines: Law and Practice (2nd ed. (loose-leaf)), vol. 1, at pp. 9-11 to 9-16).

[43] The three main applications of the set‑off formula adopted by the courts are:

1. Simple (or straight) set‑off: The support payment is calculated by determining the Table amount for each of the parents as though each was seeking child support from the other. The amount payable is the difference between the two amounts (see, e.g., Middleton v. MacPherson (1997), 1997 11517 (AB KB), 204 A.R. 37 (Q.B.); Luedke v. Luedke).

2. Pro‑rated set‑off: The Table amount for each of the parents is reduced by the percentage of time the child spends with each parent. The recipient parent’s amount of time with the children is multiplied by the payor’s Guidelines amount and the payor parent’s amount of time with the children is multiplied by the recipient parent’s Guidelines amount. These two pro‑rated amounts are then set‑off against one another (see, e.g., Moran v. Cook (2000), 2000 22542 (ON SC), 9 R.F.L. (5th) 352 (Ont. S.C.J.); Harrison v. Harrison (2001), 2001 60967 (ON SC), 14 R.F.L. (5th) 321 (Ont. S.C.J.); E. (C.R.H.) v. E. (F.G.)). A variation of this approach is the “straight pro-rate” which takes the percentage of time the recipient parent has custody of the children multiplied by the Guidelines amount for the payor parent.

3. Set‑off plus multiplier: The set‑off amount (simple set-off or pro-rated set-off) is increased by a multiplier (usually 1.5), based on the assumption that a portion of the recipient parent’s costs are fixed, and therefore, unaffected by the increased time the child spends with the other parent.

[44] I agree with the father and the Court of Appeal that the simple set‑off is preferable to the pro-rated set-off as a starting point for the s. 9 analysis in view of the language used by the legislator in para. (a). The pro‑rated set‑off was criticized by a number of scholars and courts, including the Court of Appeal in the case at bar, who refused to apply it as it disproportionately impacts on the lower income spouse (see Green v. Green, at para. 32). Professor Rogerson, at pp. 74-75, explains the “cliff effect” created by this approach:

The method of pro‑rating the table amounts to reflect custodial time that was adopted in Hunter, and which has been legislated as part of the shared custody formulas in some American jurisdictions, creates what commentators have called the “cliff” problem. A 1% increase in access, from 39% to 40%, can result in a 40% decrease in the table amount of support to be paid. This is both a conceptual and a practical problem. On the conceptual level, in a sole custody situation, the full table amount is paid even though the custodial parent may have the child only 61% of the time, and the access parent may have the child the remaining 39% of the time. It seems illogical that as a result of a 1% increase in access by the other parent, which allows him or her to reach the shared custody threshold, the custodial parent should be deprived of support for the full 40% of the time the child spends with the other parent. On a practical level, the “cliff” effect will make a custodial parent reluctant to allow an access parent even small amounts of extra time because they will carry such dramatic financial consequences.

[45] I will address the problems related to the multiplier later in these reasons.

[46] However, even the simple set‑off is incomplete. The Court of Appeal was of the view that it is based on certain underlying assumptions and that it will need to be revised in accordance with paras. (b) and (c) in cases where these assumptions do not accord with the reality of shared custody:

(1) The Table amount is the maximum that a parent with that income can [presumably] afford to pay for support (apart from extraordinary expenses);

(2) The ordinary needs of the child equal the Table amount;

(3) The parents each spend an equal amount for the care of the child; and

(4) The child’s needs are reduced by virtue of a shared custody arrangement on a dollar for dollar basis. [para. 67]

[47] I have not been able to identify the source of these assumptions. With respect, I would question whether the Table amount is the maximum that a parent can afford. I would rather think it is the average expenditure of parents with the referable income (see notes 5 and 6 of Schedule I of the Guidelines). This would accord with the methodology for estimating Table amounts and be more consistent with the discretion given to order amounts higher than those found in the Table in special circumstances (see Guidelines, s. 4 and s. 9; also J. D. McLeod, “The Proposed Child Support Guideline Package: The Scope of Judicial Discretion”, in Federal Child Support Guidelines: Reference Manual (1997), p. F-27). The only possible source for the assumption that the Table amount is the maximum that a payor parent with the referable income can afford to pay for support, is contained in the 1997 research report of the Department of Justice Child Support Team entitled Formula for the Table of Amounts Contained in the Federal Child Support Guidelines: A Technical Report. Under the heading “Underlying Principles and Assumptions”, the report states:

The objective of the formula that generates the child support tables is simply to find a means of calculating an amount to be transferred from the paying parent to the receiving parent. The transferred sum should maximise the amount available to be spent on the children while still allowing an adequate reserve for the self support of the paying parent. Several assumptions have been incorporated into the model. First, it is assumed that within the principal residence of the children, the parent and the children will share the same standard of living. A second assumption presumes that if the incomes of the parents are equal, it is fair and equitable that each should contribute equally to the financial support of the children, regardless of the extent of their contribution to the nurturing of the children. [Emphasis added; p. 1.]

I do not consider this statement to be particularly relevant, as maximising the amount available for the children is not the same as establishing Table amounts which indicate the maximum that a payor parent can afford to spend on child support. The first constitutes a general purpose, while the latter entails a specific methodology in calculating the Table amounts. In identifying the actual assumptions which were incorporated into the model used to generate the tables, the report does not state that the Table amounts were meant to indicate the maximum that a payor parent with the referable income can afford to spend on support.

[48] I would also question the second assumption; in my view, the Table is not based on a consideration of specific needs of children, but on the average expenditure of parents and consideration of the ability to pay of parents (see McLeod, at pp. F-8 and F-9). The third assumption is easily accepted: the set-off assumes each parent has an equal share of variable expenses. This said, I do not think it is necessary to identify a situation contrary to that which would correspond to underlying assumptions in order to apply s. 9(b) and (c). What is important is that the set-off does not take into account actual spending patterns as they relate to variable costs or the fact that fixed costs of the recipient parent are not reduced by the increased spending of the payor parent.

[49] Hence, the simple set-off serves as the starting point, but it cannot be the end of the inquiry. It has no presumptive value. Its true value is in bringing the court to focus first on the fact that both parents must make a contribution and that fixed and variable costs of each of them have to be measured before making adjustments to take into account increased costs attributable to joint custody and further adjustments needed to ensure that the final outcome is fair in light of the conditions, means, needs and other circumstances of each spouse and child for whom support is sought. Full consideration must be given to these last two factors (see Payne, at p. 263). The cliff effect is only resolved if the court covers and regards the other criteria set out in paras. (b) and (c) as equally important elements to determine the child support.

[50] It should be noted here that the Table amounts are an estimate of the amount that is notionally being paid by the non‑custodial parent; where both parents are making an effective contribution, it is therefore necessary to verify how their actual contribution compares to the Table amount that is provided for each of them when considered payor parents. This will provide the judge with better insight when deciding whether the adjustments to be made to the set-off amount are based on the actual sharing of child-related expenses.

[51] This is where discretion comes into play. The court retains the discretion to modify the set-off amount where, considering the financial realities of the parents, it would lead to a significant variation in the standard of living experienced by the children as they move from one household to another, something which Parliament did not intend. As I said in Francis v. Baker, one of the overall objectives of the Guidelines is, to the extent possible, to avoid great disparities between households. It is also necessary to compare the situation of the parents while living under one roof with the situation that avails for each of them when the order pursuant to s. 9 is sought. As far as possible, the child should not suffer a noticeable decline in his or her standard of living. Still, it is not a discretion that is meant to set aside all rules and predictability. The court must not return to a time when there was no real method for determining child support (Paras v. Paras, 1970 370 (ON CA), [1971] 1 O.R. 130 (C.A.)).

2.2.2 Section 9(b) — Increased Costs of Shared Custody Arrangements

[52] What should the courts examine under this heading? Section 9(b) does not refer merely to the expenses assumed by the payor parent as a result of the increase in access time from less than 40 percent to more than 40 percent, as argued in this Court. This cannot be for at least two reasons. First, it would be irreconcilable with the fact that some applications under s. 9 are not meant to obtain a variation of a support order, but constitute a first order (see Payne, at p. 261). Second, as mentioned earlier, the Table amounts in the Guidelines do not assume that the payor parent pays for the housing, food, or any other expense for the child. The Tables are based on the amount needed to provide a reasonable standard of living for a single custodial parent (see Formula for the Table of Amounts Contained in the Federal Child Support Guidelines: A Technical Report, at p. 2). This Court cannot be blind to this reality and must simply conclude that s. 9(b) recognizes that the total cost of raising children in shared custody situations may be greater than in situations where there is sole custody: Slade v. Slade, at para. 17; see also Colman, at pp. 71-74; Wensley, at pp. 83-85. Consequently, all of the payor parent’s costs should be considered under s. 9(b). This does not mean that the payor parent is in effect spending more money on the child than he or she was before shared custody was accomplished. As I discuss later in these reasons, it means that the court will generally be called upon to examine the budgets and actual expenditures of both parents in addressing the needs of the children and to determine whether shared custody has in effect resulted in increased costs globally. Increased costs would normally result from duplication resulting from the fact that the child is effectively being given two homes.

[53] A change in the actual amount of time a payor parent spends with a child will therefore give rise under s. 9(b) to an inquiry in order to determine what are, in effect, the additional costs incurred by the payor as a result of the change in the custodial arrangement. I say this because not all increases in costs will result directly from the actual amount of time spent with the child. One parent can simply assume a larger share of responsibilities, for school supplies or sports activities for example. For these reasons, the court will be called upon to examine the budgets and actual child care expenses of each parent. These expenses will be apportioned between the parents in accordance with their respective incomes.

2.2.3 Section 9(c) — Conditions, Means, Needs and Other Circumstances

[54] It is clear then that not every dollar spent by a parent in exercising access over the 40 percent threshold results in a dollar saved by the recipient parent: Green v. Green, at para. 27. Professor Rogerson refers to this at pp. 20-21:

On the other hand, allowing such an adjustment raises many concerns. Increased time spent with a child does not necessarily entail increased spending on the child. Furthermore, dollars spent by an access or secondary custodial parent do not necessarily translate into a dollar for dollar reduction in expenditures by the primary custodial parent, many of whose major child‑related costs are fixed — such as housing and transportation; any savings will typically be only with respect to a small category of expenditures for food and entertainment. Particularly in cases where there is a significant disparity in income between the parents, reductions in the basic amount of child support may undermine a lower‑income custodial parent’s ability to make adequate provision for the child or children, and will certainly exacerbate the differences in standard of living between the two parental homes.

Indeed, irrespective of the residential arrangement, it is possible to presume, in the absence of evidence to the contrary, that the recipient parent’s fixed costs have remained unchanged and that his or her variable costs have been reduced only modestly by the increased access. Thus, when no evidence is adduced, the court should recognize the status quo regarding the recipient parent.

[55] The analysis should be contextual and remain focussed on the particular facts of each case. For example, an application that represents a variation of a prior support arrangement, will usually raise different considerations from a s. 9 application where no prior order or agreement exists. In the former case, the recipient parent, when he or she first got custody, may have validly incurred expenses based on legitimate expectations about how much child support would be provided. These expenses should be taken into consideration and a court should have proper regard to the fixed costs of the recipient parent.

[56] Moreover, as asserted by Prowse J.A. in Green v. Green, at para. 35, it is important that the parties lead evidence relating to s. 9(b) and (c). This evidence has often been lacking, with the result that the courts have been forced either to make assumptions about increased costs (as was done by the Court of Appeal in the present case), or to dismiss the application under s. 9 for lack of an evidentiary foundation.

[57] In my opinion, courts should demand information from the parties when it is deficient. Three main options have been discussed and applied by appellate courts:

(1) Rely on the parties’ financial statements and child expense budgets which provide a fairly reliable source of information;

(2) Adjourn the motion to provide additional evidence (see, e.g., Cabot v. Mikkelson, at para. 43);

(3) Make “common sense” assumptions about costs incurred by the payor parent and apply a multiplier to account for the fixed costs of the recipient parent.

The third option is not acceptable, as I will explain below.

[58] In the present case, the Court of Appeal relied on “common sense” assumptions. The Court found that the father must have incurred additional variable costs for such items as food and entertainment. The Court of Appeal should have considered the total additional costs attributable to the situation of shared custody under s. 9(b), the evidence adduced permitting this, and should not have simply assumed what more additional costs would be.

[59] The Court of Appeal also resorted to the multiplier. The assumption behind the multiplier is that 50 percent of the recipient parent’s costs are fixed and, therefore, unaffected by the time the children spend with the payor parent. The multiplier operates to obviate the necessity of the parties calling evidence of the increased costs associated with children living for substantial periods of time in two households. While this formula takes into consideration the increased costs of shared custody, it does so in a somewhat inflexible fashion: Green v. Green, at para. 33.

[60] In the opinion of the appellate court, the use of such a method recognizes the concerns raised by the commentators and the courts that dollars spent on increased access or shared custody do not necessarily lead to a reduction in expenditures for the recipient parent. The use of a multiplier also furthers two of the objectives of the Guidelines, predictability and consistency, in calculating support. According to the Court of Appeal, used with discretion, a multiplier can provide a mechanism for recognizing the relative inflexibility of some of the recipient parent’s costs. In the absence of evidence concerning fixed costs of the recipient parent, the most common multiplier is 50 percent, which is applied to the set‑off amount. The amount of the multiplier ought, however, according to the Court of Appeal, to be adjusted depending on the circumstances.

[61] The father argues that, as currently drafted, the Guidelines do not support the use of a multiplier. He submits that while it may be correctly assumed that there are increased costs associated with a shared custody arrangement, it is not possible to simply assume the amount of that increase. I agree.

[62] Multipliers are controversial in the jurisprudence and commentaries. They have been characterized as unfair and discriminatory (Wensley, at pp. 83-85; Colman, at p. 77). One cannot help observing that, in the case at bar, the Court of Appeal applied a multiplier to account for the fixed costs of the mother, but it required the father to prove his additional or increased costs of custody. There is no basis for such asymmetrical treatment of each parent’s fixed and duplicated housing costs: D. A. R. Thompson, “Case Comment: Contino v. Leonelli‑Contino” (2004), 42 R.F.L. (5th) 326, at p. 331. In fact, it seemingly ignores the fact that the initial set-off takes into account the fixed costs of both parents.

[63] The British Columbia Court of Appeal commented on the use of a multiplier in Green v. Green, at para. 35 and concluded to its inapplicability:

In order to apply s. 9 however, it is important that the parties lead evidence relating to s. 9(b) and (c); that is, of “the increased costs of shared custody arrangements” and “the conditions, means, needs and other circumstances of each spouse [parent] and of any child for whom support is sought”. This evidence has often been lacking, with the result that the courts have been forced either to make assumptions about increased costs, or to refuse the application under s. 9 for lack of an evidentiary foundation. The latter option is particularly unsatisfactory in cases of in‑person litigants who often have little idea about the nature of the evidence which is required. A carefully crafted, standard form, “fill‑in‑the‑blanks” affidavit attached to the Guidelines may be of some assistance in that regard. Having said that, I recognize that it is not always easy for an access parent to demonstrate precisely what costs have increased as a result of increased access, and by how much. That problem highlights the attraction for some of a multiplier approach, since it has the benefit of simplicity and ease of application for in‑person litigants, in particular. In the American jurisdictions which use a multiplier, it is incorporated into the relevant child support guidelines. I do not interpret our Guidelines, as currently drafted, as justifying such an approach. [Emphasis added.]

[64] In Slade v. Slade, at para. 19, the Newfoundland Court of Appeal also considered the application of a multiplier:

There are other approaches. For example, a multiplier of 1.5 has been applied to the set‑off figure on the assumption that 50% of the custodial parent’s costs are fixed and therefore unaffected by the time the children are with the other parent. The multiplier is commonly used in United States jurisdictions which have sought to avoid the necessity of calling evidence of actual increased costs associated with shared custody. Section 9 would indicate that in Canada the examination of the actual circumstances was chosen as the appropriate method. However, as Justice Prowse said in Green: “This evidence has often been lacking, with the result that the courts have been forced either to make assumptions about increased costs, or to refuse the application under s. 9 for lack of an evidentiary foundation” (para. 35). Justice Prowse indicated that she did not view s. 9 as justifying the multiplier approach. However, she was unwilling to say that there was only one approach possible under the section. The multiplier has been used in a number of cases in this Province and, to be frank, when specific evidence regarding the increased cost of shared custody is not before the trial judge the multiplier would appear to be a method of providing rough justice; though I would agree with Justice Prowse, the research to support the multiplier of 1.5 has not been done in Canada. [Emphasis added.]

[65] D. A. R. Thompson in his Annotation to E. (C.R.H.) v. E. (F.G.), 2004 BCCA 297, 2004 CarswellBC 1157, highlights some of the more egregious problems related with the use of a multiplier:

The problems with the Contino analysis are entirely in steps (2) and (3). A multiplier for increased costs is already built into the straight set‑off. Multipliers are not needed so long as you steer clear of any use of the pro‑rated set‑off, as the Ontario Court of Appeal directed (unlike the B.C. Court of Appeal). A multiplier on top of a straight set‑off of Table amounts is clearly double‑counting. The Contino court’s use of actual spending under the third step leads to further double‑counting — possibly even some triple‑counting — of expenses for children.

See also Thompson, “Case Comment: Contino v. Leonelli‑Contino”, at pp. 329‑31.

[66] M. S. Melli in “Guideline Review: Child Support and Time Sharing by Parents” (1999), 33 Fam. L.Q. 219, at p. 232, observes that the multiplier is an instrument that may be difficult to use given its consequences:

. . . using a multiplier on the child support amount also has the effect of producing a lesser reduction in the payment by the nonresidential parent. When that parent is the lower income parent, the result may be as detrimental to the child as a decrease to the other parent. Furthermore, the principal costs of shared time, the duplication of facilities are borne by the nonresidential parent. Some of these costs are already factored into the child support formula as costs of visitation. To increase the child support award across the board to both parents makes the nonresidential parent pay twice for certain costs and seem structured to discourage time sharing by parents. For these reasons, guideline reviewers should investigate carefully proposals to use a multiplier.

[67] It is of primary importance to note that, so far, the research to support the multiplier of 1.5, or any other multiplier for that matter, has not been done in Canada: Slade v. Slade, at para. 19. Even the Department of Justice in its Children Come First Report, does not recommend the use of a multiplier in the absence of available research in Canada to show how much shared custody increases cost: “Without empirical evidence on the relative proportion of fixed and ‘shiftable’ costs, the Department of Justice cannot support the use of a multiplier as a presumption in shared custody cases” (p. 74).

[68] Section 9(c) vests the court a broad discretion for conducting an analysis of the resources and needs of both the parents and the children. As mentioned earlier, this suggests that the Table amounts used in the simple set‑off are not presumptively applicable and that the assumptions they hold must be verified against the facts, since all three factors must be applied. Here again, it will be important to keep in mind the objectives of the Guidelines mentioned earlier, requiring a fair standard of support for the child and fair contributions from both parents. The court will be especially concerned here with the standard of living of the child in each household and the ability of each parent to absord the costs required to maintain the appropriate standard of living in the circumstances.

[69] The Court of Appeal enumerates a number of factors to be considered under this subsection:

1. Actual spending patterns of the parents;

2. Ability of each parent to bear the increased costs of shared custody (which entails consideration of assets, liabilities, income levels and income disparities); and

3. Standard of living for the children in each household.

[70] The actual spending patterns of the parents have already been considered under s. 9(b). These factors are helpful, the last one being particularly useful for the exercise of discretion in a predictable manner. As I indicated above, financial statements and/or child expenses budgets are necessary for a proper evaluation of s. 9(c).

[71] Moreover, given the broad discretion of the court conferred by s. 9(c), a claim by a parent for special or extraordinary expenses falling within s. 7 of the Guidelines (see Appendix) can be examined directly in s. 9 with consideration of all the other factors (see Slade v. Slade, at paras. 26-30). Section 9(c) is conspicuously broader than s. 7.

[72] The Court of Appeal, when reversing the decision of the Divisional Court, posited that a reduction in support under s. 9 will sometimes result in undue hardship to the recipient parent and that in such cases the court will need to consider the provisions of s. 10(1) of the Guidelines. In my opinion, there is no need to resort to s. 10, either to increase or to reduce support, since the court has full discretion under s. 9(c) to consider “other circumstances” and order the payment of any amount, above or below the Table amounts (see “Case Comment: Contino v. Leonelli-Contino”, at p. 332). It is not that “other circumstances” of each spouse and “hardship” are equivalent terms, it is that the discretion of the court, properly exercised, should not result in hardship. It may be that s. 10 would find application in an extraordinary situation, but that is certainly not the case here.

2.3 Application of Section 9 to the Facts

[73] Now that the principles which regulate the appropriateness of child support awards under s. 9 of the Guidelines have been clarified, I will turn to the facts of this particular case.

[74] In this case, the motions judge granted the motion and reduced the $563 in monthly child support that the father had been paying to $100. As the Court of Appeal stated, “[s]he applied a mathematical formula similar to the formula found in s. 8 of the Guidelines for split custody and essentially set‑off the father’s and mother’s Table amounts. She then required the father to pay the difference, with a very minor adjustment” (para. 1).

[75] It seems very clear from the reasons for judgment of Rogers J. that she did not exercise her discretion properly, having relied on a mathematical analysis which is at odds with the approach for determining child support under s. 9 of the Guidelines. In her order, Rogers J. mentioned:

There is now a shared custody regime and there is going to be a reduction from the guideline amount. I have gone through the math with Mr. Cooper. I thank him for his mathematical skills and or those of his calculator. It seems as though the number is something around $96.

[76] The motions judge’s failure to provide adequate reasons for this conclusion is apparent from her order. For its part, the Divisional Court held that there was no right of deviation from the s. 3 presumptive amount merely upon passing the 40 percent threshold; this too, is an erroneous approach.

[77] The Court of Appeal fell into error by assuming, rather than applying evidence of the additional costs attributable to the shared custody, and by using a multiplier. The simple set‑off verified against the budgets submitted by the parties under s. 9(c) is however acceptable in the absence of other evidence, since it leads to an examination of the actual capacity of each party to contribute to the expenses and consideration of the standard of living of both households. The Court of Appeal exercised its discretion to order the total variable expenses to be shared in proportion to respective incomes. In my analysis, I will confirm this but will give further attention to the actual ability of each parent to absorb increased costs.

[78] The record before this Court contains sufficient evidence, i.e., affidavits and financial statements from both parents, in order to set the correct support payment without having to order a new trial. I nevertheless would caution against making awards without having taken steps to obtain a complete record. As determined by the motions judge, the Table amount for the father on an income of $87,315 is $688 per month; the Table amount for the mother on an income of $68,082 is $560. The set-off amount is $128. The father’s child expense budget reveals monthly expenditures attributable to the child of $1,814 and the mother’s child expense budget reveals monthly expenditures attributable to the child of $1,916.95. These budgets were accepted at trial and should not be questioned here. They establish that expenditures are not the same for both parents, and that there is in fact a large amount of duplication with regard to fixed costs. Both of these factors point to the need for significant adjustments to the set-off amounts. The second step in the analysis consists of looking at the ratio of income between the parties of 56:44 (in the interest of precision and exactitude, I have slightly modified the ratio used by the Court of Appeal); the father ought to be responsible for 56 percent of the total child related expenditures, $2,089.33, and the mother ought to be responsible for 44 percent of the total child-related expenditures, $1,641.62. Already contributing $1,814, the father would be required to pay the mother the sum of $275.33. In addition, attention must be given to the fact that the father’s net worth is $255,750 and the mother’s $190,651; this is consistent with the means and conditions test in s. 9(c) and will be dealt with later.

[79] The set-off amount under s. 9(a) is $128, but, as I have just noted, other circumstances and the evidence presented under s. 9(c) requires that it be adjusted. Based only on the sharing of child-related expenditures apportioned against the income of the parents, the father would be required to pay the mother a sum of $275.33 per month. Furthermore, examining all the costs of both parents, I have found no evidence that the fixed or variable costs of the mother decreased in any way following the shared custody arrangement; on the other hand, there is no evidence that the extra time devoted by the father or, more generally, the change brought to the custodial arrangement, has resulted in any increase in the father’s actual expenses. Because this s. 9 application represents a variation from a long-standing financial status quo upon which the mother incurred valid expenses on behalf of this child, these realities are important considerations. As mentioned earlier, the means and conditions test in s. 9(c) requires that I also consider the difference in net worth between the parents, which is $65,099, and the general ability of each parent to absorb increased costs.

[80] This means that I must now consider the impact of a new support order on the standard of living of the child under s. 9(c). I cannot ignore the fact that, in this case, I am dealing with a variation order and not a first time order. Up until this litigation, by way of settlement, for a number of years, the mother was receiving over $500 from the father (an amount that was not adjusted in 1999 even though the father’s income rose to $83,527.58). Finally, while the motions judge refused to consider this fact, it is clear from the record that the mother moved to a new house in 2000 because she believed it was in the child’s best interest, in the reasonable expectation that she would continue to receive $563 a month or more from the father. This expense, which was not challenged as inappropriate by the father, has to be considered part of the contextual analysis which includes consideration of the financial conditions and means of the mother. The purchase of the new home created some financial difficulties for the mother since she had to collapse a significant amount of her RRSPs (and consequently pay income tax on what she cashed in). She was legitimately relying on the support payment she was receiving from the father pursuant to the earlier arrangements made between them. He could not have ignored that. In light of these factors, I have come to the conclusion that the child support must be set at $500 per month. I see no reason to question the view of the Court of Appeal that the facts of this case do not substantiate a retroactive order.

[81] I would like to add a word of caution regarding the use of budgets at the appellate level in light of the approach adopted by Fish J. It is important, in my view, that all financial information be presented for a proper application of s. 9, but details concerning the appropriateness of individual items should be decided by the trial judge, on the basis of representations made by the parties. I, with all due respect, find it incongruous for this Court to question the unchallenged appropriateness of taking into account the cost of one car, for each household, or the fact that the father wishes to bring his son to restaurants on a regular basis. What we have here is a set of facts that have been accepted and are just one part of the general framework for the application of s. 9. I do not believe a line by line questioning of expenses and reapportionment of those expenses is fairer to the parties or more appropriate just because it involves a more moderate change of the set-off amount of $128.

3. Conclusion

[82] The determination of an equitable division of the costs of support for children in shared custody situations is a difficult matter; it is not amenable to simple solutions. Any attempt to apply strict formulae will fail to recognize the reality of various families. A contextual approach which takes into account all three factors enunciated by Parliament in s. 9 of the Guidelines must be applied.

[83] The appeal is allowed and the amount of support to be paid by the father to the mother is set at $500 per month. Like the justices of the Court of Appeal, I am of the view that success was divided at all levels and that the difficulties associated with the interpretation of s. 9 of the Guidelines makes this case a proper one to order that both parties bear their own costs throughout.

The following are the reasons delivered by

Fish J. (dissenting) —

I

[84] Child support provides an economic safety net for its intended beneficiaries — the children, not the parents — of failed family relationships. It serves to “break the fall” by affording the children a reasonable standard of living commensurate with the combined resources of their parents.

[85] And, in cases of shared custody, child support seeks to secure for the child, insofar as possible, a similar standard of living in the two households concerned. Children may prefer one household to the other, but that cannot be made to depend on the respective means or resources of their parents.

[86] Where a child resides with one parent less than 40 percent of the time, that parent is deemed, by legislative fiction, to incur no child-related expenses at all. The appropriate level of child support is fixed with mathematical precision by the Federal Child Support Guidelines, SOR/97-175 (“Guidelines”), (or a corresponding provincial grid).

[87] That sort of social policy decision is a matter for Parliament. And Parliament has spoken.

[88] In cases of shared custody, where the parents each have access to or custody of the child at least 40 percent of the time, Parliament has spoken as well. There, instead of imposing a mathematical calculus or grid, Parliament has left to the courts the determination of child support in accordance with s. 9 of the Guidelines.

[89] Section 9 sets out in general terms the factors that judges must take into account. But it does not assign them relative weight or indicate — even in general terms — how the factors bear on one another. Still less does s. 9 contemplate, with respect to the shared custody arrangements that are governed exclusively by its terms, a single “correct” award.

[90] Rather, s. 9 requires the court in each instance to determine the amount of child support “by taking into account”: (a) the Table amounts that would apply if it were a case of sole custody; (b) the increased costs of shared custody arrangements; and (c) “the conditions, means, needs and other circumstances of each spouse and of any child for whom support is sought”.

[91] Applying s. 9 to the evidence in this case, the motions judge ordered Christopher’s father to pay Christopher’s mother $50 per month for Christopher’s support ($100 less $50 in retirement of overpayment); the Divisional Court increased the amount to $688 ((2002), 2002 2774 (ON SCDC), 62 O.R. (3d) 295); and the Court of Appeal fixed support at $399.61 ((2003), 2003 30327 (ON CA), 67 O.R. (3d) 703). As I have already mentioned, s. 9 does not command a single result. But neither does it authorize such vastly divergent awards.

[92] In short, s. 9 support orders are discretionary by design, but constrained by principle and subject to the overriding requirement of fitness. Support orders in shared custody cases must fall within the boundaries drawn by Parliament in setting out the governing factors under s. 9 of the Guidelines. All of these factors must be considered by the courts in determining an appropriate award. “Appropriate” does not mean mathematically or methodologically ascertainable with precision. It means within an acceptable range that is in each case determined by applying in a principled manner the factors set out in s.—9 to the proven facts and particular circumstances of the matter.

[93] Of the widely divergent awards in the courts below, only the amount fixed in the Court of Appeal falls within that range. It appears to me reasonable on its face and consistent with the governing principles. I would therefore dismiss this appeal.

II

[94] The predictability ensured with mathematical precision by the Guidelines for sole custody cases has in some quarters created an understandable but futile expectation of like certainty with respect to shared custody arrangements.

[95] Predictability and precision in sole custody awards result from the mandatory application of Table amounts included in the Guidelines. These Table amounts are built on Statistics Canada’s “40/30” Equivalence Scale (see: Canada, Department of Justice, Formula for the Table of Amounts Contained in the Federal Child Support Guidelines: A Technical Report (1997), at p. 3). The Equivalence Scale, in turn, is based on empirical research that showed a 40 percent increase in cost to a household upon adding a second member (child or adult), and 30 percent more for each additional member (both percentages being approximate). In this model, the additional members reside in the household 100 percent of the time.

[96] In shared custody arrangements, the child joins two households, each for at least 40 percent of the time. Support orders are in those circumstances governed by s. 9 of the Guidelines. Section 9(a) requires that the Table amounts applicable to single custody be taken into account as one of a series of factors. There is no meaningful way to graft those Table amounts onto shared custody support orders. They are not meant to be either added to or substracted from the ratio of expenses incurred by either parent at the time the support order is made.

[97] Likewise, s. 9(b) of the Guidelines assumes that two households are more expensive than one due to duplicated fixed costs and the loss of economies of scale. But this assumption adds nothing of concrete assistance in fixing an appropriate amount of support.

[98] I understand very well that the Guidelines call for objectivity, predictability and efficiency. With respect to shared custody, however, Parliament has departed from the strict table-based model that leads to relative certainty. The focus of s. 9 is on fitness, not formulas. This approach relies on the wisdom and experience of trial judges — the responsibility is primarily theirs — to determine fair awards by applying to the facts as they find them the mandatory considerations set out in s. 9.

[99] Parliament has thus favoured judicial discretion, exercised in a principled manner, over the relative certainty of mathematically driven determinations. This legislative choice leaves no statutory vacuum. It signifies Parliament’s confidence that judges will exercise wisely and in accordance with the governing criteria the discretion vested in them by s. 9 of the Guidelines, ensuring a sufficiently reliable and reasonably predictable result.

[100] In short, s. 9 requires judges to fix child support in shared custody arrangements without the benefits and constraints of a mandatory grid. This deliberate rejection by Parliament of a precise formula or methodology does not authorize the courts to conjure one up from the void.

III

[101] For many years Canadian courts took guidance from Kelly J.A.’s finding in Paras that “the objective of maintenance should be, as far as possible, to continue the availability to the children of the same standard of living as that which they would have enjoyed had the family break-up not occurred”: see Paras v. Paras, 1970 370 (ON CA), [1971] 1 O.R. 130 (C.A.), at p. 134.

[102] Although the method for determining child support was fundamentally altered by the adoption of the Guidelines in 1997, Kelly J.A.’s words continue to resonate. Section 26.1(2) of the Divorce Act, R.S.C. 1985, c. 3 (2nd Supp.) (am. S.C. l997, c. 1), incorporates the former ss. 15(8) and 17(8), which have been seen as a reflection of Kelly J.A.’s observations in Paras: see J. D. Payne and M. A. Payne, Child Support Guidelines in Canada 2004 (2004), at p. 10.

[103] In my view, an appropriate support order in this case should ensure insofar as possible that Christopher, the child of the parties, enjoys a standard of living that is reasonably comparable to his standard of living before the divorce and does not vary markedly in material respects moving from one household to the other. The proper pursuit of this objective will of course bear in mind, as I indicated earlier, that child support aims to alleviate the economic impact of divorce on the child; it must not be turned into a ramrod for equalizing the incomes of former spouses.

[104] Like Kelly J.A. (Paras, at pp. 134-35), I understand that the resources of both parents will frequently be inadequate after their physical separation to maintain their previous standard of living, under a single roof. Where the two conflict, it is the child’s standard of living and not theirs that is to be favoured.

[105] The method for achieving this outcome should be evidence-based. In this respect I am essentially in agreement with my colleague Bastarache J. and the Court of Appeal in this case. Both have stressed the importance of an evidence-based approach.

[106] My colleague indicates that “[t]he record before this Court contains sufficient evidence, i.e., affidavits and financial statements from both parents, in order to set the correct support payment without having to order a new trial” (para. 78). This should not be understood as a general rule that the parents’ budgets are “all the necessary evidence” for deciding cases under s. 9. Trial judges have found that budgets alone often present an incomplete or misleading snapshot of the real situation. Upon examining the budgets and considering any supplementary information obtained at the hearing, trial judges are authorized for this reason to adjourn the proceedings should they think this necessary to complete the record.

IV

[107] The Tables included in the Guidelines assume that no child care expenditures are made by a parent with whom a child spends less than 40 percent of their time. That assumption vanishes when s. 9 is triggered.

[108] Where access by one of the parents hovers just below 40 percent, concerns will arise that crossing the threshold can produce a precipitate drop in child support. To avoid this “cliff effect”, the custodial parent (the parent who has sole custody) may be tempted to cling for selfish reasons to its financial advantages — for example, by hiring more babysitters to prevent the other parent from reaching the 40 percent level that brings with it shared custody. The access parent, on the other hand, may be tempted to express an enhanced interest in spending more time with the child, driven more by pecuniary than by parenting considerations.

[109] Bastarache J., like the Court of Appeal, discusses the cliff effect in relation to the meaning of “increased costs” of shared custody arrangements under s. 9(b). My colleague considers that “increased costs” means all costs incurred by the payor parent; the Court of Appeal, on the other hand, found that the “increased costs” are the incremental costs incurred after the 40 percent threshold is crossed. As we shall see, it appears to me preferable to deal with s. 9(b) simply as a binding reminder of the duplications and other incremental costs inherent in shared custody, which the set-off Table amounts under s. 9(a) do not take into account.

[110] In the present case, to the credit of both parents, there was no such evident tug of war. Both parents appear to have kept the interests of their child paramount throughout.

[111] When the mother decided to take a course that would prevent her from caring for Christopher on Tuesdays, she offered to let him spend the evening with his father rather than arrange for a relative to take him. As she explained before the motions judge, “I always tried to do whatever was in my power to ensure that Christopher has a close relationship with his father”.

[112] Nor is there any suggestion of machination in the father’s offer to take Christopher both Tuesdays and Thursdays rather than exchanging one day for the other. He did not bring his application for modification on the heels of this change. After the mother’s course ended, Christopher remained with his father, at least partially at Christopher’s request. Only at this point did the father seek an adjustment in child support. He then waited another half-year before pressing his claim in the courts.

[113] Care must nonetheless be taken not to interpret and apply s. 9 in a manner that creates a cliff effect in this case — or aggravates the risk of a cliff effect in others.

V

[114] Simple set-off of the Table amounts for sole custody may be a convenient way to begin the process of fixing an appropriate amount of support and, in this way, serve as a “starting point”. It is not, however — and should not be thought of — as a preliminary amount to be increased or decreased, or to have added to it or subtracted from it, other amounts determined on examination of the factors set out in paras. (b) and (c) of s. 9.

[115] I agree with Bastarache J. that the Court of Appeal erred in its apparent endorsement of “stock multipliers”. The expression “stock multiplier” refers here to an abstract or predetermined figure that bears no relation to the specific facts of the case.

[116] As I understand it, the Court of Appeal endorsed the use of a multiplier of 1.5 (sometimes called the “Colorado multiplier”, in reference to the jurisdiction in which the figure apparently originated) where there is no evidentiary basis for taking into account, as required by s. 9(b), the increased costs of shared custody arrangements. In my view, stock multipliers of that sort should not be used at all. The better approach was suggested by the Court of Appeal itself: In the absence of evidence required to make a fact-based, case-specific determination, the trial court can reopen the hearing for that purpose.

[117] I think it fair to emphasize, however, that the Court of Appeal did not in fact resort to a stock multiplier in this case. It applied instead a multiplicative factor based on the evidence in the record. To the extent that this factor may have been imperfectly established or applied, its impact on the result was limited and, in any event, offset by the countervailing effect of the other factors I shall examine below.

[118] Section 9(a) requires the court to take the Table amounts into account in fixing child support for shared custody arrangements. For shared custody arrangements, simple set-off of the Table amounts credits each parent with what that parent would receive if he or she had sole custody. It may be thought, in that way, to adequately compensate both parents. But that compensation is more apparent than real, since it disregards the duplications and other incremental costs inherent in shared custody. The purpose of s. 9(b), in my view, is to ensure that these costs are properly reflected in every support order made under s. 9.

[119] The phrase “increased costs of shared custody arrangements” in s. 9(b) must be viewed in this light and I do not find it particularly helpful to focus on either the starting point applied by the Court of Appeal or the starting point preferred by Bastarache J. With respect, I think it preferable to regard s. 9(b) as a binding reminder that setting-off the Table amounts is, like other “starting points”, the beginning but not the end of the exercise.

[120] Here, the ultimate burden of the exercise is to fashion a support order with the interests of the child foremost in mind. It must, however, be patterned on the facts and not made of whole cloth.

[121] It may be useful to observe that the cliff effect can only arise on a modification of child support. A cliff exists where there is a large vertical change over a small horizontal change. On a first order, there is no risk that a small change in access or custody will cause a precipitous decline in support. There is simply a judicial determination in accordance with the applicable statutory provisions.

[122] On the other hand, where there is a previous support order based on sole custody, it may create a kind of momentum that the change to shared custody will not be entirely capable of arresting. That is particularly so in this case, where concrete and irreversible financial decisions were made by the mother in reliance on the amount of child support then being paid to her by the father — such as collapsing her RRSPs and buying a new home.

[123] Christopher’s mother thus sacrificed her own financial well-being and her savings for the future so that Christopher could have the best possible life now. The father does not dispute that the mother’s move to Woodbridge was in Christopher’s best interest. The move to the new home was made in November 2000 while the father was still paying $563 per month though the parties had in fact agreed to a shared custody arrangement. It could not therefore have escaped him that the purchase was likely made in partial reliance on his continuing support.

[124] This is one of the relevant factors under s. 9(c), all of which are well set out, indicatively and not exhaustively, by Bastarache J. And I agree with my colleague that commensurate importance should be attached in this case to the respective resources of the parents and to the situation created consensually before they moved to shared custody. I shall return to this branch of the matter in discussing an appropriate amount of support.

[125] Having applied all the factors and in spite of the broad language of s. 9(c), a trial judge may still not have arrived at a just award. When all the line normally on the bobbin is played out, there remains one last reserve. Section 10(1) allows a court to “award an amount of child support that is different from the amount determined under any of sections 3 to 5, 8 or 9”. “Undue hardship” is the reserve chute when the Guidelines’ basic fabric proves insufficient to break the fall.

[126] Although the discretion of the court is made very broad by the “other circumstances” already contemplated in s. 9(c), that cannot be thought to preclude the application of s. 10, first and foremost because Parliament has otherwise decreed — expressly.

[127] Section 10 is more than “other circumstances”: It more solidly grounds a parent’s claim, and it gives a trial judge specific guidance in adjudicating it. Even if “other circumstances” allows for the same result, a judge would have to begin construction on a bare lot, with none of the foundation and superstructure that s. 10

already puts in place.

VI

[128] In my respectful view, the support orders made by the motions judge and the Divisional Court in this case both result from an understandable but mistaken quest for certainty and simplicity. No formula can be devised that will at once respect the words of Parliament and achieve the predictability of a universally applicable calculus. The plain fact of the matter is that s. 9 of the Guidelines contemplates a judicial assessment of a variety of considerations — some competing and some complementary.

[129] The first consideration relates to the Guidelines Table amounts. For the reasons explained, they cannot simply be grafted onto shared custody arrangements. Yet s. 9(a) requires that they be taken into account. I have suggested that this might conveniently be done by setting-off the Table amounts that govern sole custody and treating the result as a “starting point” — the first step in a global consideration of all of the factors that must be weighed under s. 9. Here, the set-off amount is $128 per month.

[130] The second consideration, made mandatory by s. 9(b), relates to the increased costs of shared custody. On my view of the matter, this relates essentially to the duplication of fixed costs and to other expenses that result from the exigencies of shared custody — housing, toys, clothes, books and supplies in both places, and so forth. The inquiry is essentially evidence‑driven. The extent of the duplication of fixed costs will generally be apparent from the budgets submitted by the parties, as tested and clarified at the hearing.

[131] The third consideration relates to s. 9(c), which requires the court to take into account “the conditions, means, needs and other circumstances of each spouse and of any child for whom support is sought”. This is the appropriate place, in my opinion, for an apportionment of certain expenditures according to the respective incomes of the parents, including the duplicated expenses identified pursuant to s. 9(b). Section 9(b) flags the need to take into account the increased costs of shared custody; s. 9(c) by referring to the respective means of the parents suggests how this should be done.

[132] Since this case involves a modification and not an initial support order, s. 9(c) requires us to bear in mind the support order in effect immediately prior to the shared custody arrangement. And we are bound as well, of course, to consider the disparity in assets and indebtedness of the parents. These considerations do not yield firm numbers — they in fact yield no numbers at all — but they do provide guidance in fixing an appropriate amount of support.

[133] In apportioning the parents’ expenditures between them in accordance with their incomes, we have only the budgets to rely on in this case. Understandably, the budgets attribute entirely to Christopher the expenditures that relate to him alone: for example, his clothing, school lunches, summer camp, books, tuition, school activities, and gifts from Christopher to other children.

[134] Both parents attribute to Christopher 50 percent of virtually all of their other “total living expenses”. Christopher is thus deemed, arbitrarily it seems, to share equally with his father and mother the cost of their charitable donations, chiropractic and physical therapy, and car insurance, and in the father’s case car loan payments and meals outside the home as well. Most of these expenditures appear to have little or no connection to Christopher.

[135] An additional difficulty arises where the household is not limited to the parent and child alone. In this case, for example, the Court of Appeal noted that the father has remarried (para. 11). It is therefore unclear why 50 percent of any — let alone all — household expenditures should be attributed to Christopher.

[136] However illogical, this 50 percent apportionment was accepted by the parties and relied on by the courts throughout. At this stage, I therefore feel bound to do likewise, albeit reluctantly.

[137] Unlike Bastarache J., however, I would not simply take the global monthly expenditure from the parents’ budgets. Only two categories or types of expenditures should in my view be equalized between the parents: first the duplications and other incremental costs inherent in shared custody, which s. 9(b) requires us to consider; second, the variable child care costs that might otherwise be shared by the parents.

[138] To begin with, I would thus apportion only duplicated fixed expenditures of the parents (expenditures made by both parents where the expenditure by one does not decrease the expenditure of the other). And I would consider here only the actual, proven and reasonably necessary increased costs inherent in shared custody.

[139] In this regard, it appears from their budgets that the mother spends $889.49 monthly under the “housing” head of her budget, and the father spends $738.50. The total of the two is $1627.99 which, apportioned according to the ratio of their incomes (56:44), invites a monthly equalization payment of $173.17 by the father to the mother.

[140] The second group of expenditures to be apportioned covers variable costs and disbursements that may be unevenly distributed between the parents. These expenditures should be shared fairly rather than on the basis of who writes the cheque.

[141] For example, one parent may spend more on clothing for the child and on gifts from the child to others because the child prefers to go shopping with that parent. Similarly, one parent may spend more than the other on certain fees (activities, books, tuition, special projects, field trips and summer camp) either because that parent has greater enthusiasm for these activities or is the one who typically enrolls the child. In neither case should the parent who initially covers these costs have to bear them alone.

[142] The kinds of expenditures that lend themselves to this unfairly shared burden are those that the parents spend entirely on Christopher. Variable expenses attributed at 50 percent — for example, for meals outside the home, which are claimed by the father in this case — stand on a different footing, since they benefit both the parent and the child and are generally incurred by both parents in accordance with their respective lifestyles.

[143] On the evidence before us, the mother spends $428.84 per month on Christopher’s clothing, school fees, school lunches, school activities, summer camp, gifts from the child to others, and an RESP. The father’s expenditures of this sort (he does not contribute to the RESP) amount to $120 per month. Applying the ratio of incomes as a corrective therefore requires the father to pay $187.35 to the mother, in addition to the $173.17 mentioned earlier (in dealing with duplicated fixed costs), or a total of $360.52 to be paid monthly by Christopher’s father to Christopher’s mother.

[144] I note in passing that the respective Table amounts in this case correspond, almost exactly, to the 56:44 ratio of the parents’ incomes: the $688 payable by the father under the Guidelines is about 56 percent of $1248 (the sum of the Table amounts), while the $560 payable by the mother equals about 44 percent. In this way, the simple set-off under s. 9(a) and the ratio method under s. 9(c) are essentially the same mathematical operation applied to different amounts.

[145] Now, with the figure of $360.52 in mind — but making no final determination yet — there are the two non-numerical factors under s. 9(c) that have still to be considered. They must be weighed judicially, not arithmetically. The first is the situation prior to shared custody, and the second is the disparity between the net worth of the parents.

[146] For the reasons mentioned above, the mother’s reliance on the $563 per month in child support previously paid to her by the father is a significant factor in this case. To the father’s knowledge if not with his actual agreement, the mother purchased a home, cashed in RRSPs and paid tax on them all in the expectation that she would continue to receive the same or a similar amount in the future. That expectation, though understandable, creates no entitlement to continued support at the same level notwithstanding a move to shared custody.

[147] The second non-numerical factor to be considered is the disparity between the assets and liabilities of the parents. This factor is relevant for two reasons. First, because assets may create for the parents an opportunity to generate income, increasing their “means” and their ability to provide for the child. Second, a heavy debt load may require that a larger proportion of the parent’s income be put in the service of those debts, decreasing that parent’s ability to contribute to child care.

[148] The disparity in total income involves different considerations that have already been taken into account. As a discrete consideration under s. 9(c), income disparity is relevant to a fair apportionment of total child care expenses and duplicated fixed costs. To make of the relatively moderate earnings disparity in this case an additional, independent factor would undermine the notion that child support is not intended to be a mechanism for equalizing the parents’ incomes.

[149] It is, of course, relatively easy to ventilate in this way the various factors that must be considered in making a s. 9 support order, and far more difficult to assess their combined or cumulative effect. Yet that is the judicial determination required under s. 9 of the Guidelines where, as here, the parties are bound by a shared custody arrangement.

[150] The support previously paid by the access parent to the parent having sole custody is not presumptively applicable to shared custody, even where the fixed costs of the parties are not significantly affected by the change. As I mentioned earlier, however, it is an important consideration in the circumstances of this case. Here, the support previously paid was $563 per month. The mother incurred fixed costs that were in part a function of the support she was receiving at the time. This, I repeat, must be taken into account, though it creates no entitlement to the same level of support where the parties have moved to shared custody, which is governed by a fundamentally different provision of the Guidelines.

[151] Bearing in mind all of the circumstances, I believe that the situation prior to shared custody and the disparity in the net worth of the parties together militate in favour of a support order in the $500 range. Other important factors, however, point to a significantly lower amount: notably, the $360.52 derived by apportioning the relevant child care costs according to the respective incomes of the parents and, more strikingly still, the set-off amount of $128, which must also be taken into account pursuant to s. 9(a) of the Guidelines. These are both mandatory considerations in determining child support in cases of shared custody. To attribute little or no importance to them is to impermissibly ignore the plain words of s. 9, read as a whole.

VII

[152] For all of the foregoing reasons and with respect for those who are of a different view, I would dismiss the appeal.

[153] The motions judge ordered Christopher’s father to pay Christopher’s mother support for their child that amounts in effect to $50 per month. The Divisional Court substituted an award of $688. I agree with the Court of Appeal that the motions Judge and the Divisional Court both erred in their application of s. 9 of the Guidelines and both reached unacceptable results.

[154] The Court of Appeal substituted a support order of $399.61. This award lies within the acceptable range that is in each case determined by applying to the facts, in a principled manner, all of the relevant factors set out by Parliament in s. 9 of the Guidelines. Except as indicated, I believe the Court of Appeal set out the basic principles correctly. Its unfortunate observation as to the permissible use in some circumstances of a “stock multiplier” of 1.5 had no bearing at all on its conclusion, and the limited effect of its resort to a multiplicative factor of 1.67 is adequately compensated by the other factors I have mentioned.

[155] I have followed a somewhat different path in pursuing the same quest as the Court of Appeal. But I have arrived at a similar result.

[156] Trial judges are required to exercise their discretion in fashioning s. 9 support orders, and there is no single methodology that best ensures an appropriate result. If the route I have chosen resembles a trial-like exercise, this should be taken as guidance meant to be helpful by example, and not just by exposition.

[157] Finally, I agree that there should be no order as to costs.

APPENDIX

Divorce Act, R.S.C. 1985, c. 3 (2nd Supp.)

26.1 . . .

(2) [Principle] The guidelines shall be based on the principle that spouses have a joint financial obligation to maintain the children of the marriage in accordance with their relative abilities to contribute to the performance of that obligation.

Federal Child Support Guidelines, SOR/97-175

1. [Objectives] The objectives of these Guidelines are

(a) to establish a fair standard of support for children that ensures that they continue to benefit from the financial means of both spouses after separation;

(b) to reduce conflict and tension between spouses by making the calculation of child support orders more objective;

(c) to improve the efficiency of the legal process by giving courts and spouses guidance in setting the levels of child support orders and encouraging settlement; and

(d) to ensure consistent treatment of spouses and children who are in similar circumstances.

3. (1) [Presumptive rule] Unless otherwise provided under these Guidelines, the amount of a child support order for children under the age of majority is

(a) the amount set out in the applicable table, according to the number of children under the age of majority to whom the order relates and the income of the spouse against whom the order is sought; and

(b) the amount, if any, determined under section 7.

(2) [Child the age of majority or over] Unless otherwise provided under these Guidelines, where a child to whom a child support order relates is the age of majority or over, the amount of the child support order is

(a) the amount determined by applying these Guidelines as if the child were under the age of majority; or

(b) if the court considers that approach to be inappropriate, the amount that it considers appropriate, having regard to the condition, means, needs and other circumstances of the child and the financial ability of each spouse to contribute to the support of the child.

4. [Incomes over $150,000] Where the income of the spouse against whom a child support order is sought is over $150,000, the amount of a child support order is

(a) the amount determined under section 3; or

(b) if the court considers that amount to be inappropriate,

(i) in respect of the first $150,000 of the spouse’s income, the amount set out in the applicable table for the number of children under the age of majority to whom the order relates;

(ii) in respect of the balance of the spouse’s income, the amount that the court considers appropriate, having regard to the condition, means, needs and other circumstances of the children who are entitled to support and the financial ability of each spouse to contribute to the support of the children; and

(iii) the amount, if any, determined under section 7.

5. [Spouse in place of a parent] Where the spouse against whom a child support order is sought stands in the place of a parent for a child, the amount of a child support order is, in respect of that spouse, such amount as the court considers appropriate, having regard to these Guidelines and any other parent’s legal duty to support the child.

6. (1) [Special or extraordinary expenses] In a child support order the court may, on either spouse’s request, provide for an amount to cover all or any portion of the following expenses, which expenses may be estimated, taking into account the necessity of the expense in relation to the child’s best interests and the reasonableness of the expense in relation to the means of the spouses and those of the child and to the family’s spending pattern prior to the separation:

(a) child care expenses incurred as a result of the custodial parent’s employment, illness, disability or education or training for employment;

(b) that portion of the medical and dental insurance premiums attributable to the child;

(c) health‑related expenses that exceed insurance reimbursement by at least $100 annually, including orthodontic treatment, professional counselling provided by a psychologist, social worker, psychiatrist or any other person, physiotherapy, occupational therapy, speech therapy and prescription drugs, hearing aids, glasses and contact lenses;

(d) extraordinary expenses for primary or secondary school education or for any other educational programs that meet the child’s particular needs;

(e) expenses for post‑secondary education; and

(f) extraordinary expenses for extracurricular activities.

(2) [Sharing of expense] The guiding principle in determining the amount of an expense referred to in subsection (1) is that the expense is shared by the spouses in proportion to their respective incomes after deducting from the expense, the contribution, if any, from the child.

8. [Split custody] Where each spouse has custody of one or more children, the amount of a child support order is the difference between the amount that each spouse would otherwise pay if a child support order were sought against each of the spouses.

9. [Shared custody] Where a spouse exercises a right of access to, or has physical custody of, a child for not less than 40 per cent of the time over the course of a year, the amount of the child support order must be determined by taking into account

(a) the amounts set out in the applicable tables for each of the spouses;

(b) the increased costs of shared custody arrangements; and

(c) the conditions, means, needs and other circumstances of each spouse and of any child for whom support is sought.

10. (1) [Undue hardship] On either spouse’s application, a court may award an amount of child support that is different from the amount determined under any of sections 3 to 5, 8 or 9 if the court finds that the spouse making the request, or a child in respect of whom the request is made, would otherwise suffer undue hardship.

Appeal allowed, Fish J. dissenting.

Solicitors for the appellant: MacDonald & Partners, Toronto.

Solicitors for the respondent: Bastedo Stewart Smith, Toronto.