Might Kenneth Carter roll over in his grave? As Chair of the Royal Commission on Taxation in the Sixties, he might remind us today that their report recommended that the family ought to be the unit of taxation. Their view recognized that family income is a joint product which acknowledges the holding of a job and the care of home and children. They further acknowledged the propensity to divide income artificially thanks to our progressive tax system. However, the Commission’s recommendation was not codified into our tax law.

But now, we fast forward four decades to last Fall, when, out of the blue, the Conservatives floated a miniature version of taxing the family unit…. pension splitting. Commencing in 2007, married pensioners may see a drop in their tax bills, thanks to these new rules. Our last taste of this came in the 80s when couples became able to pool and split their CPP entitlements. However, these provisions are quite different. First, the new splitting rule is a year-by-year election, whereas CPP splitting is a one-time permanent decision. Second, the splitting only goes one way, whereas CPP is pooled and divided equally. Third, the annual amount of split income is totally discretionary, subject to a maximum of one half of the qualifying income. Fourth, the type of income that qualifies for splitting is much broader – essentially income that qualifies for the pension credit for the transferor. This means incomes from work pensions, RRIFs and annuitized RRSPs.

It is the transferor who determines the qualification to split. The transferee requires no qualification other than to be a spouse as defined under tax law. The main benefit, of course, should derive from a difference in the marginal tax brackets of the two parties. However, the age and circumstance of the transferee may contribute further to the value of making the transfer. If the transferee qualifies for the pension credit but otherwise has no other pension income to enjoy it, the transferring-in income will now qualify. A third bonus might include reduced OAS clawback burden for the transferor.

A surprising extra from the mandarins in Finance is that the corresponding proportions of tax withheld on the sources must also transfer. Typically, this would bear more commonly on transfers of work pensions, not RRIFs or annuities, which usually have no tax withheld.

There will be some complexities and confusion with the quarterly installment strategies for each of the parties. To avoid over and under installments by the parties, it may be prudent to do a little proforma tax planning, wherein each would estimate their incomes and pension splitting strategy during the year, and modify their installment payments accordingly.

Whether splitting strategy is addressed during the year or when the tax returns are prepared in the following Spring, ultimately, the determination must be made as to how much to split to achieve an optimal result. The fact patterns for some couples may give an obvious answer – the full one half allowable may be the wisest. Other fact patterns may be less obvious, particularly when both spouses have significant incomes. And remember, this decision must be made every year!

We were sitting around the office one day thinking about this when somebody postulated that the range of choices must be parabolic, meaning that there is one optimum split for every couple’s fact pattern, and further, that computing power should be able to do the heavy lifting for the optimizing search, thus saving endless trial-and-error iteration. Excel spreadsheet technology saved the day with the “solver” add-in plus a few other enhancements to our in-house tax planning application. Armed with this new tool, we ran 30 trials with real client data to test this new tax-saving opportunity.

Now for the surprising results. Every tax practitioner I have spoken with about this new provision has expressed amazement at the revenue give-away by the folks at Finance. I guess some wily veteran practitioners may have been more taciturn, and rightly so. After 30 documented “clinical” trials, we have observed that the average win for our clients is typically very small. Why? First, many taxpayers retired today took planning steps during their working years to improve their tax burdens in retirement, e.g. spousal RRSPs, directed savings to the lower income spouse, etc. Second, the tax bracket structure cuts a broad sweep across middle-range income. Incomes from $37,000 to almost $69,000, while vastly different, are in the same marginal tax bracket. Therefore, splitting within this range may yield zero gain. Third, the increase between two marginal brackets at higher income levels is typically 2%. Thus, splitting, say, $10,000 across two brackets only yields approximately $400. Fourth, some splits gain on tax brackets but trigger extra clawback to the transferee.

Ironically, the very first test case yielded a family tax saving of $6,000, and that was by far the best one. In that case, the transferee spouse had no other income at all and the transferor spouse’s income was quite large.

The solver optimizer turned out to be a lot smarter than we are, when it identified a beneficial “reverse” split from a lower income spouse to the significantly higher income spouse, which would seem counter-intuitive at first. The reason was the clawback. The lower-income transferor crawled down out of clawback territory while the higher-income transferee was already fully clawed back anyway, and thus bore no extra clawback cost from the extra income. The transferor’s gain from reduced clawback was greater than the extra marginal tax rate of the transferee.

It is reasonable to infer that this new Carter-like provision “threw a bone” to Canadian retirees, who last Fall felt their retirement incomes were dealt a blow from the trust unit saga. The Finance Minister probably recalled when a previous government felt the brunt of “Grey Power” when it proposed to remove indexation from universal government pensions.

Single retirees suffering from the trust unit legislation were not thrown the bone, nor were non-retirees, who themselves will instead only have something to look forward to in their retired years.

As another tax preparation season comes upon us next Spring, it will be interesting to see if the tax preparation software developers crank it up a notch and program this solver optimization feature into their product offerings. If they don’t, we will be turning to our in-house tax planning application for help. We think that we have only scratched the surface of understanding the complex tax planning ramifications of pension splitting. We may have to learn the rest with our feet in the fire next Spring.

In closing, remember that Finance Ministers have a habit of coming and going, and policy leanings may depend upon the political stripe currently in power. The lesson is not to depend on this new pension splitting too heavily, at the expense of historical tried-and-true financial planning strategies that have served us over the ages.

See also: Federal Buget 2007 – Pension Splitting

Article originally published: Nov 2007