When you run a support calculation in Canada, two numbers appear side by side that most people assume should be related: the NDI split (how disposable income is divided after support) and the s.7 apportionment (each party's share of special and extraordinary expenses like childcare, medical, and extracurricular costs).

They never match. Not even close.

Consider a shared-parenting case in Ontario: a payor earning $120,000 and a recipient earning $45,000, with two children and $12,000 a year in childcare. Run that case through the calculator with the childcare properly included, the way the with-child formula requires, and the spousal support range drops from what it would have been if s.7 had been left out. Once support is set, the calculator's own net disposable income figures for both parties are already measured after each side's s.7 share has been subtracted. Exhibit 1 below shows exactly how much the range moves.

Most lawyers accept the resulting split as simply how the framework works. We think it's worth asking three narrower questions: is s.7 actually in the calculation to begin with, does the apportionment rule still make sense once it is, and what happens to that split in the years after the order is made.

First: Is Section 7 Even In the Calculation?

Before you get into whether the s.7 split matches the NDI split, confirm the s.7 expenses are actually in the spousal support calculation at all. The Spousal Support Advisory Guidelines' own Revised User's Guide doesn't hedge on this: leaving s.7 expenses out of the with-child-support formula is "the single most common and most significant mistake under the SSAG." Where the expenses are large, the Guide adds, the error "can be quite serious for the payor," because an s.7 expense that never enters the formula produces a spousal support range that's too high, and it's the payor who overpays every month until someone catches it.

This isn't the apportionment question the rest of this article is about, and it isn't the drift question either. It's the precondition to both. Get this one wrong and the other two are moot: you're arguing about how to split, or how to keep up with, something that was never in the calculation your spousal support range was actually built from.

On Dvorce.ca, an s.7 expense you enter is deducted from each party's individual net disposable income before the spousal support range is drawn. There's no path through the with-child formula that skips it. The calculator supports seven categories of s.7 expense; see below for how the tax treatment differs across them.

Include s.7, and a second question follows: how should it be split between the parties? The default rule, s.7(2), splits it in proportion to income, not evenly. Professor Rollie Thompson, one of the SSAG's own authors, has described exactly when it should move to an even split instead:

That 50/50 s.7 split is opt-in, not the default rule, and it only applies once spousal support has already been set at the 50/50 NDI point. It is a hedge against what happens after the order, covered later in this article, not a description of how s.7 gets apportioned in the ordinary case.

Not All S.7 Expenses Cost the Same, After Tax

The calculator currently recognizes seven categories of s.7 expense: Child Care, Medical/Dental (Uninsured), Health Insurance Premiums, Educational Expenses, Post-Secondary Education, Extracurricular Activities, and Other. The Federal Child Support Guidelines direct that in determining the amount of an s.7 expense, the court must take into account "any subsidies, benefits or income tax deductions or credits relating to the expense" (s.7(3)): the split should run on the net cost, not the sticker price. In practice, how cleanly that net cost can actually be computed varies a lot by expense type.

Child Care has a precise answer. It qualifies for the Income Tax Act's s.63 childcare deduction, and Dvorce.ca computes the exact tax savings by running your tax calculation with and without the deduction and taking the difference, then apportions the net cost, not the gross amount, between the two of you.

The other six categories, and what Dvorce.ca does with each:

• Medical/Dental (Uninsured): qualifies for the federal medical expense tax credit on the part of the expense above a threshold tied to the claimant's own net income (3% of net income, capped at an amount the CRA indexes each year). Dvorce.ca computes that credit for whichever of you paid, by running your tax calculation with and without it, and apportions the net cost.

• Post-Secondary Education: tuition generates a tuition tax credit that belongs to the student first, and the student can transfer up to $5,000 of it to a parent. Only the student's return shows how much, so the expense row has an optional field for the tuition amount transferred to you. Enter it and Dvorce.ca derives the federal tax saving and shares it in proportion to income, the approach the Saskatchewan Court of Appeal ordered in Koback v. Koback, 2013 SKCA 91. Leave it blank and the expense is apportioned at the gross amount.

• Health Insurance Premiums: count as medical expenses for the same federal credit. Dvorce.ca merges them with Medical/Dental into one computed credit for the paying parent, so the threshold is applied once, and apportions the net cost.

• Educational Expenses: relief depends on the specific expense and province; apportioned at the gross amount.

• Extracurricular Activities: generally carries no tax relief; apportioned at the gross amount.

• Other category: a catch-all for anything not covered above, with no dedicated tax treatment; apportioned at the gross amount.

Two things to know about the medical credit. Justice Canada's own worksheet for s.7 expenses (Worksheet 2) names the medical expense credit, alongside the childcare deduction and the transferred tuition amount, as a tax item to net against the expense, which is the basis for deriving it. But we have not found a reported decision that nets this credit on a s.7 medical expense, and the report says so. Provincial medical and tuition credits are not derived yet; the relief in your calculation is the federal credit only.

What that means practically: if a large share of your s.7 expenses is childcare or medical, the number in your range already reflects your real after-tax cost. If it is tuition, enter the transferred amount and it does too. For educational, extracurricular and other expenses, the range is running on the gross amount, worth flagging to your lawyer or mediator if the true net cost, after your own credits, would meaningfully change the split.

A Brief History of How We Got Here

To understand the disconnect, you need to understand that the s.7 apportionment rules and the NDI framework were designed at different times, for different purposes, and were never reconciled.

The Child Support Guidelines (1997)

The Federal Child Support Guidelines came into force in May 1997, replacing the old discretionary regime with a formulaic table-based approach. One of the most significant changes was making child support tax-neutral: unlike the prior system, post-1997 child support is neither deductible by the payor nor taxable to the recipient.

Section 7 governs special and extraordinary expenses: costs that fall outside the table amount because they're significant, necessary, and not adequately covered by the base payment. The rule for sharing them is set out in s.7(2): each party contributes "in proportion to their respective incomes." The statute doesn't prescribe exactly how "income" is defined for this purpose. The text of s.7(2) refers only to "the respective incomes of the spouses." In practice, the SSAG-recommended approach, endorsed by the Revised User's Guide, is to use guideline income adjusted for spousal support, though the court retains discretion to use a different income base where the circumstances warrant it. Where spousal support is paid, it is deducted from the payor's income and added to the recipient's for the purpose of calculating the proportionate share.

The logic for this adjustment is sound in principle: because spousal support is taxable and deductible (unlike child support), it genuinely shifts each party's financial resources. The payor's adjusted income reflects what they actually have left; the recipient's reflects what they've actually received. The s.7 proportions should follow.

The SSAG (2008)

The Spousal Support Advisory Guidelines emerged from a major research project by Professors Carol Rogerson and Rollie Thompson, published in final form in July 2008. They introduced the Low/Mid/High range framework, the with children and without children formulas, and, critically, the concept of individual net disposable income (INDI) as the target measure for spousal support adequacy.

The SSAG authors endorsed using spousal-support-adjusted incomes for s.7 apportionment. At the time the with-children formula was designed, the primary scenario was one where the recipient had primary care of the children. The Revised User's Guide (2016) was explicit about how this should influence the range: cases should "more often resolve in the upper half of the range" because "there is great need in the home of the primary carer for the children." The authors accepted that the formula would often leave the recipient with more than 50% of NDI, and they were comfortable with that: the primary carer typically has greater child-related costs to absorb.

What the framework never addressed was the interaction between that NDI outcome and the s.7 apportionment. If the recipient is meant to have more NDI because of the primary care burden, should they not also carry more of the special expenses that represent part of that burden? The Guidelines never asked that question.

The Structural Disconnect

The SSAG authors deliberately chose INDI, a measure that backs out child support contributions from each party's income, rather than family NDI. They explained why in the 2008 Final Version (Chapter 8): family NDI "bulk[s] up the recipient's income in a somewhat misleading fashion," they wrote, "masking the impact of spousal support upon the recipient parent's individual income." They wanted a cleaner measure of the adult income pool after child obligations are handled.

This is a reasonable design choice for the spousal support formula itself. But it means there are now three income concepts in play: guideline income (used for s.7 apportionment), INDI (used for the SSAG formula), and NDI (used for outcome analysis). These will almost never align, because they incorporate different inputs: taxes, benefits, child support transfers, and deductions are included to varying degrees in each.

The Apportionment Question, After Inclusion

With s.7 properly included, the apportionment rule itself does not change to match the NDI outcome. It stays income-proportionate under s.7(2), on the net cost, with each parent's own tax relief staying with the parent who claimed it. What changes at the 50/50 NDI point is the spousal support amount: the calculator solves for the level of spousal support that puts both parties at exactly 50/50 net disposable income, treating the ordinary proportionate s.7 split as a fixed input to that solve, not as something it rebalances. The NDI share and the s.7 share you see in a report are both measured after that solve, not before it.

Where the Gap Actually Bites

In the demo above, the payor's s.7 contribution at the 50/50 NDI point is $6,599 and the recipient's is $5,401: the recipient's $5,401 is what she pays before the roughly $2,100 she gets back through the s.63 childcare deduction, which shows up as lower tax, and on the $9,911 net cost the payor's share is 66.6%. Change the expense type, the payer, or the income split, and both numbers move. The point isn't that the split always comes out close in gross dollars. It's that whatever it comes out to, both parties' net disposable income figures already have that gross figure subtracted.

The real gap is what happens after the order. Section 7 expenses change: childcare ends, activities change, medical costs come and go. Whether the order fixes each parent's percentage or provides for recalculation, the changed amount moves the NDI split, and spousal support is rarely revisited at the same time, so the split the order was built around drifts away from wherever it started. And most cases never sit at a clean 50/50 NDI point to begin with: the mid-point is still what most practitioners open on, and many settlements land somewhere else in the range entirely. A split that was internally consistent on the day of the order can look very different two years later, not because anyone did the math wrong, but because nobody re-ran it.

Why the 50/50 Point Matters

Thompson's quote above explains why the mid-point default matters here. Defaulting to the formula's mid-point rather than the 50/50 NDI point "has unintended effects, especially noticeable in the British Columbia courts where that trend is too strong," he wrote. A default to the mid-point in high-income shared custody cases can leave the recipient with 58% of NDI, a percentage that is hard to justify in most shared custody cases.

The support amount at the 50/50 NDI point needs no agreement and no order to exist: in shared custody the SSAG range already includes it (SSAG 8.6.3; the Revised User's Guide, s.8(f), says the range "always includes" it), and Dvorce.ca extends the printed range to reach that point where the formula alone would not. The 50/50 s.7 split is the opt-in half: it follows only "if the spouses agree to this 50/50 split of NDI or if a court orders it." So the 50/50 support amount is a location within the range, not a default. Dvorce.ca pre-selects the mid-point for inspection, as most practitioners expect, and shows the 50/50 NDI scenario beside it, one click away. The Revised User's Guide is explicit that there is "no inherent reason in the SSAG to default to the mid-point on amount."

Sequencing matters here too: the 50/50 s.7 split has to be entered first, in the calculator's Agreed S.7 Split (RUG 8(b)) input in the Special Expenses tray. That way, the range the 50/50 NDI point is read from is the one already solved with that split inside it. Reading the 50/50 amount off a range solved under the ordinary proportional split, then agreeing to a 50/50 s.7 split afterward, pulls the NDI split away from 50/50 from day one.

The Primary Care Rationale and Why It's Breaking Down

The SSAG authors had a rationale for the primary care scenario: the parent providing the bulk of daily childcare has costs that don't show up cleanly in any income calculation: the constant availability, the emergency pickups, the foregone professional opportunities, the domestic management that makes two households function. Higher NDI for that parent was justified as partial recognition of costs the formula couldn't fully capture.

This rationale made more sense in the world the Guidelines were designed for: one where most custody arrangements were clearly primary/secondary, where one parent was the default daily caregiver and the other had every-other-weekend access.

That world has substantially changed. Shared parenting arrangements have become significantly more common in practice, and amendments to the Divorce Act enacted in 2019 (Bill C-78), which came into force on March 1, 2021, reflect this shift at the statutory level, though not in the way it is sometimes described.

The amendments replaced the old "maximum contact" principle under s.16(10) with a new s.16(6), which instructs courts to give effect to "the principle that a child should have as much time with each spouse as is consistent with the best interests of the child." Importantly, Parliament explicitly rejected proposals to introduce a presumption of equal parenting time. The Supreme Court of Canada confirmed in Barendregt v. Grebliunas (2022 SCC 22), a relocation case that became the leading authority on the new framework, that the amended language represents a "parenting time factor," not a presumption, and cautioned that it "must not be used to detract from the child-centric nature of the inquiry."

What the amendments do signal is that substantially equal parenting time is now a formally recognized statutory category, with its own distinct legal treatment, including specific burden-of-proof rules for relocation under s.16.93. Where the status quo provides for essentially equal parenting time, the relocating parent bears the burden of proving the move serves the child's best interests. This legislative architecture treats shared parenting as a protected arrangement, not an exception requiring special justification.

The practical effect is that the rationale for giving the recipient more than 50% of net disposable income is increasingly absent. In a shared parenting arrangement, neither parent is absorbing an asymmetric burden of daily care. Both are running full households for the children. Both are incurring equivalent expenses. That is the case for anchoring on the 50/50 NDI point instead of the mid-point in shared custody: with no primary carer to protect, there is no remaining reason for one party to end up above 50% of the household's net disposable income.

What the Calculator Doesn't Automate for You

A few dimensions of this problem worth flagging for practitioners:

The CS set-off layer in shared custody. In shared custody cases, the net child support transfer is already reduced through set-off, and the s.7 share is deducted from INDI on top of that, both before the range is drawn. Neither obligation is uncounted. What a single range figure doesn't show is the two obligations side by side, which is worth asking your report to break out if you want to see the full picture in one place.

Calculation accuracy matters more than it seems. Figures like the eligible dependant credit (available to the recipient, statutorily denied to the payor) and the Canada Child Benefit are real asymmetries, but they can be accounted for in a properly constructed NDI calculation. The problem is that many traditional tools either miss them or calculate them imprecisely, which means the NDI analysis practitioners are relying on may already understate the recipient's real position. A tool that includes these correctly at least gives both parties an accurate picture of the gap.

Thompson's "special s.7 solution" is a hedge against drift, not a substitute for inclusion. Including s.7 in the calculation is not optional and does not depend on anyone invoking anything: it is how the with-child formula is supposed to run, in every case. The 50/50 s.7 split is different. It is an opt-in choice that locks the s.7 split to 50/50 once spousal support is set at the 50/50 NDI point, specifically so that later changes in s.7 expenses don't knock the NDI split off 50/50 without someone having to renegotiate spousal support. Skip inclusion, and there's no NDI split worth protecting in the first place. Get inclusion right but skip the 50/50 s.7 election, and your NDI split is correct on the day of the order and free to drift afterward. The two fixes solve different problems.

What Should Change

Once spousal support is set at the 50/50 NDI point, splitting s.7 50/50 as well is the internally consistent choice, and it's also the drift-proof one. It keeps the NDI split at 50/50 no matter how the s.7 expenses themselves change afterward, without anyone having to go back to court to re-run the numbers. Thompson's SSAG FAQs already support this for shared custody; the gap is that it's opt-in, not automatic, so it only helps the practitioners who know to ask for it. On Dvorce.ca that means entering the 50/50 s.7 split first, in the Agreed S.7 Split (RUG 8(b)) input in the Special Expenses tray, a Case Pass feature, so the 50/50 NDI point comes from the range the calculator then solves with that split already built in. On these numbers, the agreed 50/50 split puts the equal-NDI amount at $12,828 whether the childcare is $12,000 or nothing, because both parties' net disposable incomes move by half the net cost together, so the figure stops depending on the expense at all.

The harder, more useful work is on the accuracy side of s.7(3)'s netting duty: the requirement to apportion expenses on their net cost, after subsidies, benefits, and tax relief, not their gross amount. Dvorce.ca nets out the childcare tax deduction because it's deterministic and computable from the tax calculation itself. The other six categories, medical and dental expenses, tuition, health insurance premiums, educational expenses, extracurricular activities, and the general "other" category, don't have an equally clean answer, and the calculator apportions them at their gross amount today. Closing that gap for the categories where it's computable is a more concrete improvement than reopening the apportionment rule itself.

Making both the split and the drift visible is the first step. If support analysis reports consistently show s.7 contribution shares next to the NDI split they produced, and flag when s.7 expenses have changed since the order was made, practitioners will start to engage with it, and eventually, so will courts.