Kesselman on Income Splitting
There has been so much discussion of income splitting on this blog that we already have two posts entitled “Income Splitting Redux.” Adding to the mix, the Institute for Research on Public Policy has released a major paper by Jon Kesselman on the subject. He cites my Ottawa Citizen op-ed among many other sources.
I have not fully read and digested Kesselmanâ€™s paper, which opposes labour income splitting, proposes investment income splitting, and suggestsÂ reforms to pension income splitting. On the whole, I think that Kesselman is to be commended for producing a thoughtful paper on complex and important issues. InÂ rejecting splitting of the biggest component of personal income, he largely endorses the position that I and others on the left put forward a year ago.
However, I would tend to quibble with him on investment income splitting. His argument is that many wealthy couples already employ aggressive tax-planning to shift investment income to the lower-earning spouse.Â Overtly allowing such splitting would be fairer to all couples with investment income and reduce the resources devoted to tax planning. However, Kesselman concedes that this approach “likely would entail a net revenue cost for government.”
Existing tax avoidance is too often deployed as an argument for lower rates and looser rules. Instead, tax avoidance should prompt stronger compliance and enforcement provisions.
Kesselmanâ€™s call for even lighter taxation of investment income must be assessed in the broader context of a tax system in which the very rich recipients of such income no longer pay their fair share. Deep corporate tax cuts should allow for somewhat heavier taxation of investment income at the personal level.
A more progressive approach, which might help correct the tax systemâ€™s current inequity, isÂ the Dutch/Danish model of taxing all of a coupleâ€™s investment income (perhaps above a given threshold) in the hands of the higher-earning spouse. However, as Kesselman convincingly demonstrates, there is no theoretically perfect solution to this thornyÂ issue.