Spousal Support Advisory Guidelines: The Revised User's Guide

19 Retirement

Retirement cases are now more common, thanks to the “boomers” retiring. Retirement raises a host of tricky issues, usually at the stage of variation or review. In this edition of the User’s Guide, we decided to gather these retirement issues in one place, unlike previous editions. In the SSAG itself, retirement only merited a single paragraph under “Exceptions”. Obviously, this is a growth area for SSAG analysis.

First there are threshold issues. In a variation case, does retirement amount to a “material change” or not? Is retirement a ground of review? In most instances, retirement means a reduction in the payor’s income, but the “early retirement” cases ask whether income should be imputed to the retiring payor, either the former employment income or some other income post-retirement.

Once we get past those threshold issues, before applying the SSAG formulas, there may still be another issue, the “double-dipping” pension issue under Boston v. Boston, 2001 SCC 43. Boston is treated by the SSAG as a property-based “exception”, recognizing the interplay between the division of a pension as family property and the pension’s use as income for support purposes.

Finally, former spouses and partners will eventually have to draw down their capital for current needs, a challenge for income-based guidelines.

For an excellent earlier analysis of the retirement case law, see Marie Gordon, “Back to Boston: Spousal Support After Retirement” (2009), 28 Canadian Family Law Quarterly 125.

(a) Early retirement

When the payor retires “early” and seeks a reduction in spousal support, there will be close scrutiny of the decision to retire. When will a retirement be described as “early”? The courts are not always clear. For our purposes, an “early” retirement is either a retirement on a reduced pension or a retirement on a full or unreduced pension before 65 years of age, in the absence of health issues or other special circumstances. If the court sees the early retirement as “voluntary” and not necessary or reasonable, then it is likely that spousal support will not be changed. As Gordon notes, many of these cases involve longer marriages and significant compensatory claims: above at 151-166.

Where there is an application to vary, the court might take the view that the early retirement does not amount to a “material change” and leave spousal support unchanged: Cossette v. Cossette, 2015 ONSC 2678 (Div.Ct.); Sangster v. Sangster, 2014 NBCA 14; Walts v. Walts, 2013 ONSC 6787; MacLanders v. MacLanders, 2012 BCCA 482; Jordan v. Jordan, 2011 BCCA 518; Marshall v. Marshall, 2011 ONSC 5972; Francis v. Logan, 2008 BCSC 1028; Gajdzik v. Gajdzik, 2008 BCSC 160; Moffatt v. Moffatt, [2003] O.J. No. 3912 (S.C.J.) (income maintained for 5 years until normal retirement date). The Ontario Divisional Court also found no “material change” in another early retirement case, but its authority is undermined by a number of errors in its reasons: Hickey v. Princ, 2015 ONSC 5596.

Early retirement will be accepted where justified by health issues: LeMoine v. LeMoine, [1997] N.B.J. No. 31 (C.A.). Or by economic uncertainty and lay-offs: Beck v. Beckett, 2011 ONCA 559. Retirement at a younger age from the Armed Forces has produced conflicting results: Powell v. Levesque, 2014 BCCA 33 (justified, support varied), but Sangster v. Sangster, 2014 NBCA 14 (not justified, no material change).

In other cases, the courts have found the retirement decision itself to be reasonable, but then a part-time employment income is imputed to the early retiree. In Donovan v. Donovan, 2000 MBCA 80, the imputed employment earnings plus his pension brought the retired police officer back to his pre-retirement income level. A similar result obtained in Rothschild v. Sardelis, 2015 ONSC 5572, where the payor’s employment income was imputed, but only for the one additional year he should have worked before retiring. When the early retiree’s income is imputed at or close to the pre-retirement level, the line between this “imputing” approach and the above “no material change” approach disappears in practical terms.

In Stephen v. Stephen, 2004 SKQB 386, a part-time income was imputed even where the RCMP officer retired early because of stress and various physical ailments, leaving the payor with an income in-between his pension and his former full-time income. To similar effect, see LeBlanc v. LeBlanc, 2013 NBCA 22 and Beck v. Beckett, 2011 ONCA 559. On imputing income in such cases, see Rollie Thompson, “Slackers, Shirkers and Career-Changers: Imputing Income for Under/Unemployment” (2006), 26 Canadian Family Law Quarterly 135, esp. 158-160.

(b) Does retirement constitute a basis for a change in spousal support?

Retirement may be included as an explicit ground of review in an order or agreement, especially where the retirement is likely to occur in the near future as a “genuine and material uncertainty”, to use the language of Leskun. In these cases, there is no need to prove a “material change”. Given the uncertain and confused treatment of “material change” by some courts, lawyers have frequently used a review clause to avoid that debate. For a court-ordered review at retirement, see Vaughan v. Vaughan, 2014 NBCA 6.

Alternatively, retirement may be explicitly included in the definition of material change in an order or agreement as is permitted by L.M.P., e.g. Slongo v. Slongo, 2015 ONSC 2093.

If the order or agreement is silent on retirement, retirement is usually a “material change”, although some cases continue to apply the wrong test of “foreseeable” from Miglin, rather than the correct test, i.e. was retirement  considered or taken into account in the previous order (see discussion above under “Variation and Review”). For a discussion of the appropriate test, see Rollie Thompson, “To Vary, To Review, Perchance to Change: Changing Spousal Support” (2012), 31 Canadian Family Law Quarterly 355. Some courts still confusingly find retirement is not a material change because it is “foreseeable”, e.g. Hickey v. Princ, 2015 ONSC 5596 (Div.Ct.).

(c) Previously-divided pensions: double-dipping, “Boston” and the SSAG

Years later, lawyers and judges still struggle with the practical implications of Boston. In the SSAG, Boston is recognized as an “exception”, where the use of the SSAG formulas must be modified in assessing the amount of spousal support.

Boston contained three important “retirement” holdings. First, where a pension has been divided as property and a court is addressing spousal support, “to avoid double recovery, the court should, where practicable, focus on that part of the payor’s income and assets that have not been part of the equalization or division of matrimonial assets, when the payee’s continuing need for support is shown” (para 64). Second, the recipient has an obligation to use her or his share of the divided property to generate income and to make a contribution to self-sufficiency and, if not, then income can be imputed to the recipient. The recipient in effect has to create her or his own “pension” (para 54). Third, the majority in Boston recognized that double recovery cannot always be avoided, noting that economic hardship and need may justify an exception to the general rule. For good reviews of Boston, see Gordon, above, as well as Carol Rogerson, “Developments in Family Law: The 2000-2001 Term” (2001), 15 S.C.L.R. (2d) 307 at 329-354.

Addressing Boston issues requires a step-by-step approach. The SSAG starting point in assessing spousal support is usually the full incomes of both parties. First, Boston creates an exception to this general approach. The payor thus bears the burden of convincing the court to use an income less than his or her Guidelines income under this exception. By way of actuarial or other evidence, the payor must prove to the court that some portion of his or her current pension income has already been divided as property. Second, if the payor does meet this initial burden of proving the “double-dipping” exception, then the burden shifts to the recipient to convince the court that the hardship or need exception applies under Boston.

Boston was decided in 2001, before the advent of the Spousal Support Advisory Guidelines. In Boston itself, the Supreme Court ultimately deferred, without much explanation, to the motions judge’s reduction of spousal support from $3,433 to $950 per month. The move towards income-based guidelines with the SSAG in 2005 led to courts trying to find a formulaic method to apply the “no double-dipping” approach from Boston.

Here’s what we had to say in the March 2010 User’s Guide about Boston and double-dipping:

The Advisory Guidelines do not change the law from Boston v. Boston, [2001] 2 S.C.R. 413, which governs “double-dipping” in respect of pension division and spousal support. That law can best be understood as relating to entitlement, entitlement to share in the already-divided portion of the payor’s pension income. Under the Advisory Guidelines, Boston is recognized as the basis of an “exception” (SSAG 12.6.3) …

Boston articulates a general rule against “double-dipping”, i.e. that spousal support should not be paid out of pension income from a pension that has already been divided as part of the property division between the spouses. But Boston then goes on to recognize exceptions to the rule, exceptions which have been discussed at length in the appeal cases of Meiklejohn v. Meiklejohn, [2001] O.J. No. 3911 (C.A.); Chamberlain v. Chamberlain, 2003 NBCA 34, 36 R.F.L. (5th) 241; and Cymbalisty v. Cymbalisty, 2003 MBCA 138, 44 R.F.L. (5th) 27. The most common exceptions to the rule against “double-dipping” are based upon hardship and need. Boston applies in cases where there has not been an in specie or statutory division of the pension, but instead the recipient spouse has received other assets or a lump sum in lieu of the pension. To apply the Boston rule against double-dipping, a court needs evidence of the prior valuation and division of the pension, to determine which portion of the payor’s current income has been divided. In some cases the divided portion will be quite small relative to the undivided portion of the payor’s pension income and Boston will not have any impact: Leepart v Leepart, 2009 CarswellSask 54, 2009 SKQB 47.

Where a pension is divided at source when it is paid out, as is the case under British Columbia or Nova Scotia legislation, then the problems of Boston can usually be avoided, e.g. Trewern v. Trewern, [2009] B.C.J. No. 343, 2009 BCSC 236. In these cases, both spouses simply include the pension payments in their income and the previously divided portions of the pension effectively cancel each other out.

The application of Boston within the context of the Advisory Guidelines raises complex issues. The Boston exception under the Advisory Guidelines recognizes that some adjustment may need to be made to the application of the SSAG to avoid “double-dipping”, but determining when and how that adjustment is to be made raises difficult issues, in part because of the “fuzziness” of Boston itself.

In two Ontario cases, courts have made a very formulaic adjustment to the Spousal Support Advisory Guidelines to avoid “double-dipping” that may not accurately reflect the Boston ruling. In each case, the court reduced the payor’s income by the amount of the divided pension and then calculated the without child support formula range on the reduced payor income: Hurst v. Hurst, [2008] O.J. No. 3800 (S.C.J.) and Gammon v. Gammon, [2008] O.J. No. 4252, 2008 CarswellOnt 6349 (S.C.J.). In both of these cases, the previously-divided pension was a small part of the payor’s total income, so the problems were not so obvious. As well, in Hurst this method of adjustment was dictated by the parties’ agreement. In other cases, however, this formulaic adjustment of income may be too mechanical and rigid, and may lead to inappropriate results. It not only risks by-passing the analysis of the exceptions that are built into Boston, but may also lead to arbitrary results under the Advisory Guidelines.

To take a simple example, assume the spouses have been married for 20 years and the wife has an income from non-pension sources of $10,000 per year. The payor husband has retired with an annual income of $50,000, of which $30,000 represents the previously-divided pension. If the above Boston adjustment is used, then the SSAG range would be $250-$333/mo. Treating the payor as a person living on $20,000 a year, which is the “floor” for spousal support, could lead to an award at the low end of the range, or even to the elimination of spousal support. This would ignore the payor’s base income of $30,000 upon which he can live.

In some recent SSAG cases, courts have taken the full income of the payor into account in calculating the range, relying upon the hardship and need exceptions to the “double-dipping” rule: see Scott v. Scott, [2009] O.J. No. 5279 (S.C.J.) and Jenkins v. Jenkins, [2009] M.J. No. 271, 2009 MBQB 189.

Another aspect of the Boston “double-dipping” rule is the requirement that the recipient spouse convert into income the assets received and “traded off” against the payor’s pension in the property division.

In two recent cases, the payor spouse took early retirement and then argued that the recipient wife in her early ‘fifties should be required to access her share of the pension or have income imputed for SSAG purposes, but the courts rejected this “reverse Boston” argument: Szczerbaniwicz v. Szcerbaniewicz, [2010] B.C.J. No. 562, 2010 BCSC 421; and Swales v. Swales, [2010] A.J. No. 297, 2010 ABQB 187.

Despite these cautions about a simple formulaic attempt to apply Boston, the most common method for adjusting the SSAG formulas to reflect the rule against double-dipping continues to be to plug into the formula only the payor’s income generated by the undivided portion of the pension, plus any other non-pension income: Elliston v. Elliston, 2015 BCCA 274; Murphy v. Murphy, 2015 BCSC 408; Pascall v. Mbolekwa, 2015 ONSC 7444; MacQuarrie v. MacQuarrie, 2012 PECA 3; Stephenson v. Stephenson, 2012 ONSC 1867 (Div.Ct.); Landry v. Mallette, 2014 ONSC 5111. An amount is then determined within that lower SSAG range.

Here we have to go back to the basics of Boston, some of which are less than clear. We will try to disentangle the three main holdings in Boston, and relate them to the SSAG.

First, the “no double dipping” approach. In determining spousal support after a pension division, the focus should be upon the undivided portion of the pension received by the payor, said the Court. The complications mostly arise for defined benefit pensions. In most cases, the support recipient has received her or his share of the pension by way of a lump sum paid out of the pension plan and into a locked-in retirement investment, or in other offsetting non-pension assets or by an equalization payment. Eventually, the recipient will have to convert these funds into some form of recurring income for retirement, a “pension” of some kind in the words of Boston. It is a very difficult task to impute what the recipient’s income should be from the divided assets, or when that income should commence (as most recipients are younger than the payors). It is a simpler task, although not a “simple” task, to exclude the already-divided pension income from the payor’s income and then to determine spousal support using that reduced payor income and the actual recipient income. And, as the payor spouse has the necessary information available, the law places the burden upon the payor to prove what portion of the pension income has already been divided as property.

Second, as a practical matter the recipient spouse will have to generate a “pension” from her or his assets eventually. Remember that in Boston the wife had received the large matrimonial home and some other assets, and had increased her assets over time. There is much discussion of this issue in the Supreme Court majority’s reasons. But this can inadvertently lead to another kind of “double-counting”, revealed in Boston itself and discussed in Rogerson, above. If the payor’s income is reduced, taking out all the previously-divided portion of the pension, and then the recipient’s income is increased to include investment income on any assets derived from the property division (or, even worse, to include some estimate of the annuity that could be generated from the assets), then there is double-counting working against the recipient. To be balanced as between the parties, you must take the divided pension income (or its equivalent in investment terms) out of both sides or out of neither. This point is not clearly stated in Boston.

Third, the hardship or need exception within Boston. It is not always clear from Boston why this “exception” is limited to need, as has been the subsequent interpretation of the decision. There is much discussion in the case about “hardship” and “need”, but little recognition of compensatory claims post-retirement. At first glance, a compensatory claim would seem to have as strong a claim, if not stronger, for an exception after retirement, especially after a long traditional marriage. Given the way that compensatory and non-compensatory factors are intertwined after a long marriage, it may not matter that the exception is treated as one for non-compensatory support. In the end, the Supreme Court did not apply the need exception on the facts of Boston. Further, the Court never defined its view of “need” in such cases.

The formulaic approach applied to avoid double-dipping often produces intuitively appealing outcomes on the facts. The ingredient common to all the cases cited above is that the bulk of the pension was NOT divided, leaving most of the pension income available as the payor’s “income” for spousal support purposes. In these cases, the undivided pension income of the payor fell in the range of 62 per cent of the full pension income (MacQuarrie, Stephension) to 76 per cent (Pascall) or 81 per cent (Murphy). In Elliston, the husband was still working while collecting his military pension, such that his income available for support was 74 to 81 per cent if his full pension income was considered (the husband’s pre-cohabitation service portion seems to have been forgotten). The other common element in these cases is that all were lengthy relationships, ranging from 16 to 21 years, such that the without child support formula divided a substantial percentage of the remaining income by way of support to the recipient.

If the income from the undivided portion of the pension is below $20,000, it is important NOT to treat the payor as having an actual income below $20,000 and NOT apply the SSAG provisions about the floor: Brisson v. Brisson, 2012 BCCA 396. This important point was ignored in Stephenson v. Stephenson, 2012 ONSC 1867 (Div.Ct.) and Rothschild v. Sardelis, 2015 ONSC 5572.

Turning to the hardship/need exception, in the majority of post-Boston cases, this “exception” has swallowed the “rule”. For some recent cases, see: Hickey v. Princ, 2015 ONSC 5596 (Div. Ct.); Senek v. Senek, 2014 MBCA 67 (needs-based exception applied on appeal); and Landry v. Mallette, 2014 ONSC 5111(needs-based exception does not apply because wife’s needs being met by income of new partner). For two cases not quite as clear about the exception, see Flieger v. Adams, 2012 NBCA 39, affirming 2011 NBQB 237 and Dishman v. Dishman, 2010 ONSC 5239. To complicate matters further, many of the “Boston hardship/needs exception” cases involve disability or illness issues, as is true of Hickey v. Princ and Landry v. Mallette.

Where courts use the SSAG to apply this “exception” formulaically, the full pension income of the payor is usually used to calculate the SSAG range, without much explanation: e.g. Smith v. Werstine, 2014 ONSC 5319. The full pension income will generate a higher range, which moves the amount in the direction that is desired. Again, it is not obvious that this formulaic approach offers the appropriate outcome in every case, as was pointed out in Slongo v. Slongo, 2015 ONSC 2093. We would suggest that calculations be done for alternative incomes, as discussed under “Determining Income” above. First, do a calculation on the full pension income and, second, do another for only the undivided portion of the payor’s pension, before locating an amount under the “exception” within Boston. The hardship and need of the recipient will have an impact upon the location of that amount.

Further, it should be remembered that “need” in spousal support law should not be treated as bare or subsistence need, but as a relative concept, tied to the marital standard of living. The longer the marriage, with “merger over time”, the stronger the non-compensatory claim to the marital living standard. This view of need should inform the hardship/need exception under Boston.

In the end, under the SSAG, Boston requires courts to exercise more flexibility, both in applying the general “rule against double-dipping” and in applying the rather large need exception to that “rule”. The double-dipping “rule” of Boston is itself an exception from the ordinary calculation of spousal support. In many cases, the formulaic calculation (seen in cases like Stephenson) may produce a tolerable result, but only where the relationship is lengthy and the bulk of the pension is not divided (either because there has been a significant period of pension contribution post-separation or because of other sources of payor income). Similarly, the formulaic use of the payor’s full income under the hardship/need exception may generate reasonable outcomes over a range of cases, but not all.

(d) Termination of spousal support

Retirement is one of the reasons that spousal support ends. Retirement cases will usually involve the application of the without child support formula. In a long marriage, where one or both parties have pensions that are equalized, retirement may mean both have similar assets and incomes and thus no basis for continuing spousal support in most cases. If neither has an employment pension and both spouses look to their divided CPP/RRQ public pensions and OAS (Old Age Security) and sometimes GIS (Guaranteed Income Supplement), again their incomes will likely be similarly low, e.g. Arbou v. Robichaud, 2012 NBQB 16.

In between those two extremes, there will likely be continuing income disparities, although often reduced in size after retirement. In some of these cases, the payor’s actual post-retirement income will drop below the “floor” of $20,000 gross per year: Whittick v. Whittick, 2014 BCSC 1597; Heywood v. Heywood, 2013 ONSC 58; and A.M.R. v. B.E.R., 2005 PESCTD 62. These cases were discussed above under “Ceilings and Floors”. There are exceptions where support is continued after retirement and despite a payor income below the floor, but only in long marriages where the recipient has little or no income at all: Pratt v. Pratt, 2008 NBQB 94 (wife on social assistance, husband $14,116/yr, support only $300/mo.) and M.(W.M.) v. M.(H.S.), 2007 BCSC 1629 (wife’s income zero, husband $17,800/yr, support of $600/mo, low SSAG).

In many cases, by the time the payor reaches retirement, the couple is nearing the end of the duration of spousal support anyway and the drop in payor income at retirement provides the ground for termination, e.g. Powell v. Levesque, 2014 BCCA 33 (8-year relationship, recipient disabled, payor retires on full pension from Armed Forces at 44 with health problems of her own, support paid 12 years, terminated). However, spousal support can continue past retirement, as Boston reminded us. For cases of long or late marriages (“rule of 65”), where support is indefinite, retirement and the payor’s drop in income will often create the grounds for termination or the imposition of a time limit.

(e) Living off capital and income-based guidelines

Eventually, as we get old enough, we all have to “live off our capital”, to drawn down our capital resources to pay for our current needs, especially those without pensions. RRSPs have to be converted into RRIFs (Registered Retirement Income Funds) or annuities. Businesses and farms have to be sold. Interest from investments becomes insufficient to fund daily needs. As Leskun v. Leskun, 2006 SCC 25 reminded us, capital is part of “means” and can be the basis for paying spousal support.

This poses a problem for income-based guidelines like the SSAG. In effect, there are two steps to the SSAG analysis at this advanced stage, assuming entitlement: first, what income should be imputed to the spouses as reasonable withdrawals from capital in addition to whatever current income the spouses may earn; and, second, the formula calculation for amount, under the without child support formula. Or, if older spouses have roughly similar assets after the division of property, a court can terminate spousal support and leave each spouse to manage their own capital to meet their needs, as occurred in Puiu v. Puiu, 2011 BCCA 480 (34-year traditional marriage, separated 2005, husband 66, wife 61, neither working).

Date modified: