Fraser Forum

William Watson: A thousand points of tax

Printer-friendly version

Whatever your politics, this was quite a week for Canadian economists.

Our president-elect, that is the president-elect of the Canadian Economics Association (CEA), was named to the federal cabinet. On Oct. 19, Jean-Yves Duclos (pictured above), until recently chair of the department at Université Laval, was elected MP in the Quebec City riding of Québec. And on Nov. 4, Prime Minister Justin Trudeau named him to the cabinet as minister for the new portfolio of Families, Children and Social Development.

Not many professors have leapt from the faculty lounge to the federal cabinet. Stéphane Dion, the new minister of foreign affairs, did in 1996. So did Pierre Trudeau, in 1965. But to my knowledge, no economist has done it and certainly not a president or president-elect of the CEA.  

In my Financial Post column this week, I described a couple of Duclos’ articles including the lead article of the latest issue of the Canadian Journal of Economics, which appeared just a day before Duclos became minister. Here I want to mention my favourite Duclos diagram, which comes from Duclos’ 2006 Innis Lecture to the Canadian Economics Association.

Let me describe it. On the horizontal axis is “market income,” which is income people earn in the labour and capital markets before governments tax their income or give them income transfers. The range is $0 to $100,000. On the vertical axis is the “marginal tax rate,” the rate at which an extra dollar of income is taxed. The range there is 0 to 100 per cent. What does the graph look like? It looks like the Milky Way. It’s not well-behaved at all. It’s a genuine scatter—a scatter cloud, really—with hundreds of points.

How come?

Duclos and two colleagues did the very dirty work of figuring out exactly which implicit tax rates were involved in each of the thicket of social programs and income transfers offered by both the federal and Quebec governments in 2002, and how these implicit tax rates varied according to people’s and families’ ages, social status, number of dependents, and so on and so on.

An implicit tax is implicit only in the sense that it doesn’t officially appear in the income tax code. But if a program’s benefits are income-tested, and after a certain point, start to be phased out as recipients earn more income on their own, the effect is exactly the same as if it were a formal income tax. If you earn an extra dollar, you lose a certain percentage benefit so your net income from the effort is less than the full dollar. “Implicit” doesn’t reduce the economic impact. A tax by any other name stings as much.  

The reason there’s a galaxy’s worth of points in the scatter is that there are lots and lots of programs, family sizes, ages, ability status, and so on. What’s striking is how many of the data points are to the graph’s left (in lower income ranges) but nevertheless are above 50 per cent, with a non-significant number being above 90 per cent. The highest marginal tax rates are paid by the people with the lowest incomes (many of whom actually “pay” negative average rates—that is, they receive net transfers from government).

The saving grace is that some of these points don’t represent very many people. Still, in 2002 2.5 per cent of Quebecers faced marginal tax rates of 100 per cent or more. And the median marginal tax rate was over 35 per cent.

It’s not clear what, if anything, the new Trudeau government plans to do about high marginal effective tax rates. In fact, its campaign promise to income-test child tax benefits will add another such tax rate to the system—another point to what is already “a thousand points of tax,” as it were.

But it’s comforting to realize that at least one person now sitting at the federal cabinet table has an acute understanding of what a marginal effective tax rate is and how much damage even something implicit can do.


Subscribe to the Fraser Institute

Get the latest news from the Fraser Institute on the latest research studies, news and events.