When a Payor's Income Swings, What Number Sets Support?

The situation

Winston worked as a paramedic in Ancaster, a job with a fixed salary that barely moved from one year to the next. His common-law partner of nine years, Rania, ran her own electrical contracting business — some years she cleared close to $130,000 after a run of new-build subdivision work, and other years, when commercial projects dried up, her income fell closer to $60,000. They had lived together, unmarried, raising their son Samir, and had built a household around that unevenness: heavier spending in the good years, tighter budgets in the lean ones, a shared mortgaged home they both put money into as they could.

When the relationship ended, they agreed quickly that Samir would live primarily with Winston, with Rania having him several nights a week. What they could not agree on was what child support Rania should pay. Under Ontario's child support guidelines, the paying parent's support obligation is based on their annual income — but Rania's income was not one number. It was a jagged line, and each of them had a different idea of where on that line to draw support from.

The problem

Winston came to us after several months of informal back-and-forth had gone nowhere. Rania was proposing to base support on her most recent tax year, which happened to be one of her weakest — commercial electrical work had slowed across the region and her reported income was around $58,000. Winston pointed to the two years before that, when her income had been closer to $115,000 and $128,000, and argued support should reflect what she actually earned in a typical year, not the one convenient low point.

Both had a fair argument, and that was the real problem. For a salaried employee, the guideline income is usually just last year's total income figure from their tax return — straightforward. For someone self-employed with genuinely variable income, the guidelines allow a court, or parties negotiating an agreement, to average income over recent years rather than freezing on a single one, particularly when a business has cyclical or unpredictable earnings. But averaging is not automatic, and there is real judgment in how many years to average and how to treat unusual spikes or dips that will not repeat. Rania's low year was not fabricated or timed to the negotiation — the slowdown was real — but using it alone would have set support well below what Samir's household actually needed and below what Rania herself expected to earn again the following year.

There was a second layer to the dispute. Because Winston and Rania had never married, some of what people assume applies to married spouses — automatic property division rules under the Family Law Act, for instance — did not apply to them the same way, since those rules are specifically limited to married spouses. Child support itself, however, is owed regardless of whether parents were married, so that part of the disagreement was squarely about the numbers, not the relationship status. Winston also raised, early on, whether he might be owed spousal support given how long they had lived together and the income gap between a paramedic's steady salary and Rania's better years — but after reviewing their situation, we told him plainly that a claim built mainly on his own steadier income against her fluctuating one was unlikely to succeed and would only add cost and delay to the more pressing issue, which was Samir's support.

What we did

1. Pulled three years of financial disclosure, not one. We asked Rania, through her own advisor, for notices of assessment and full business financial statements for the three most recent tax years, plus her year-to-date figures for the current year. A single year, in either direction, was never going to be a reliable picture of what she actually earned.
2. Built an averaging proposal grounded in the actual pattern. Looking at the three years together, her income averaged out to roughly $100,000 — closer to the middle years than to either the peak or the recent slump. We proposed averaging her three most recent years for the purposes of setting support, which is a recognised approach for self-employed payors with fluctuating income, rather than anchoring to whichever single year favoured one side.
3. Priced in the slowdown honestly, without ignoring it. We did not argue the low year should be disregarded. Instead we proposed weighing it into the average rather than throwing it out, and flagged that if the slowdown continued, that would show up in next year's average too — the formula would move with her real income over time rather than locking in a number that quickly went stale.
4. Negotiated directly with Rania's advisor rather than heading to court. Litigating a contested income determination in family court over a few thousand dollars a year in support would have cost both of them more, in time and money, than the dispute itself was worth. We opened with the averaging proposal and the supporting figures and kept the conversation focused on the math rather than the history between them.
5. Built in a yearly review instead of a one-time fix. Because Rania's income was genuinely likely to keep swinging, we proposed the parties exchange income documents each spring and recalculate the three-year average annually, adjusting the monthly support figure accordingly — so neither side would need to renegotiate from scratch, or go back to court, every time her business had an unusual year.
6. Set out the agreement in a written separation agreement with independent legal advice on both sides. A support arrangement based on an averaging formula only holds up if both people understood exactly how it works and both had their own advisor confirm that before signing — so if a dispute arose later, no one could argue they had not understood what they agreed to.

The outcome

Rania's advisor pushed back initially, wanting to average only her two most recent years, which would have kept the number lower. After two rounds of negotiation, both sides settled on the three-year average, with an agreement that the annual review would use a rolling three-year window each time — so the lean year would always be balanced by stronger years on either side of it, and no single bad or unusually good year could swing the number too far in either direction.

The monthly support figure that came out of the average landed meaningfully higher than what Rania had first proposed, and noticeably lower than what Winston had first asked for — which is what a fair reading of three years of real numbers should produce. Neither of them got the number they opened with, and both told us, separately, that they could live with where it landed. That is the honest shape of most negotiated support outcomes involving variable income: not a clean win for either side, but a number that tracks reality closely enough that it does not need to be fought over again every year.

The yearly review clause did its job the following spring, when Rania's income ticked back up to around $95,000 for the year — the rolling average adjusted the support figure upward by a modest amount, and both sides accepted the recalculation without needing to involve us again. That is precisely the outcome the averaging structure was built for.

What you can learn from this