Canada’s bloated 1 per cent

Statistics Canada’s release on the escalating incomes of the top 1 per cent gained a lot of media coverage — and also provoked some very defensive reactions by major organs of the Canadian media.

This included an almost rabid column by Financial Post editor Terence Corcoran accusing Statistics Canada of engaging in class warfare and, in a McCarthyite manner, personally attacking some of their existing and former employees.     (I guess he doesn’t read the New York Times much, where that well-known communist Warren Buffett was quoted as saying “There’s class warfare, all right, but it’s my class, the rich class, that’s making war, and we’re winning.”)

The Globe also weighed in with an editorial urging us to “Hug the 1 per cent“ claiming they get a bad rap and that they are a net benefit to Canada because they pay more than their share in income taxes. The Globe editorialists forgot about the other taxes Canadians pay.  When all those other taxes (sales, payroll, property) are considered, the top 1 per cent actually pay a lower share of their income in taxes than all other income groups, including the poorest 10 per cent, as the CCPA’s comprehensive analysis demonstrated.

Another notable reaction was that the income of the top 1 per cent, with a median of $283,400, isn’t that high. (Of course, that raises the question: if $283K isn’t that high, how low does that mean the median of $28K for the 99% is?).  If anything Statistics Canada downplayed the incomes of the top 1 per cent and our growing income gap. For instance, the average mean (as opposed to median) income of the top 1 per cent was actually $429,600 in 2010. The real incomes of the top 1 per cent almost doubled (up by 96%) in the 25 years from $251,250 in 1982 to $492,900 in 2007 (in 2010$).

Meanwhile the average real income of the bottom 90% increased by only 5% over this quarter decade from $27,600 in 1982 to $29,300 in 2010.

There’s a wealth of other data associated with this release now freely available for analysis on the Statscan website, with lots of detail by city and region.

For instance, the data show that the average income tax rate paid by the top 1 per cent declined from 40.6% in 1997 down to 33.3% in 2010. If these top 1 per cent taxpayers (slightly more than 250,000 individuals) were taxed at the same average rate as they were in 1997, our governments would have an extra $8 billion a year in revenue.

By province, if the average income tax rate on the top 1% in Ontario in 2010 was the same as in the late 1990s, they would have paid an additional $2.9 billion in taxes. In Quebec, the top 1% would have paid an extra $1.2 billion. If Alberta’s richest 1% paid the same rate of income tax in 2010 as they had in 1997, they would have paid an extra $2 billion in taxes and in BC if the top 1 per cent paid the same rate of tax in 2010 as they did in 1997, it would have provided an extra $1.3 billion in revenue.

The drop in average taxes paid by the top 1 per cent wasn’t because of a cut in the top federal marginal income tax rate, which has been 29% since 1988, but was caused both by an expansion of tax loopholes (with lower rates on capital gains and stock options) and by provincial cuts to top income tax rates, such as Alberta’s shift to a 10% flat tax rate.

And in these times of continuing fiscal concern, these are the some ways we can regain these revenues: by closing tax loopholes and raising provincial top income tax rates. And if the top 1 per cent paid their fair share of tax, they might get a little more love from the rest of us.


  • Just for those interested, I want to place emphasis on the income tax file that is used by statcan for income derivation purposes. I actually worked on that file many years ago, and we actually had to exclude very a hand full of the top incomes because it was skewing the data and rather than making adjustments and investigating the reasons for the increases, we just excluded a handful of large millionaire remitters.

    Also know that there is a serious amount of under reporting at the top end of the income range when it comes to taxes- I believe some of that was uncovered and documented by Paul Martin’s large review of the tax system in the 90’s.

    And top that up with a large amount of off shoring of wealth and you have some very serious breakdowns in the prevailing notions of democracy and taxes.


  • Let’s start with the fact that in 2010 the top 1% earned 10.6% of the income but paid 21.2% of income tax. That makes the idea that the top 1% don’t pay their “fair share” of taxes a little dubious.

    The Marc Lee paper is very interesting but uses a definition of income designed to get the desired answer. In particular, CPP premiums are counted as taxes (news to the people at the CLC pushing for enhanced CPP), inheritance is counted as income (quite a surprise to the executor that paid any capital gains tax due on death), and income includes imputed rent on housing (most people would be surprised by the idea that their house pays them).

    Let’s continue with the fact that the income share of the top 1% is down from 2006. That makes the adjective “escalating” look at little odd.

    Additionally, Table 1 in the Statscan release shows that only 52.7% of the people in the top 1% category were there five years ago. This has been under-reported (in this blog post as elsewhere). Any notion that “The 1%” can be reified as some kind of permanent aristocracy is clearly pure nonsense, when the data shows the half-life of the category is five years.

    In other news, static estimates of additional revenue that would allegedly be raised by higher tax rates are, um, not robust. See Table 1 at:


    to see some elasticity of taxable income figures specific to Canada.

  • It is true my paper uses a broad concept of income, consistent with the literature on tax incidence, and better reflecting “command over resources”. Those items and others should be counted as income, precisely because they predominantly benefit high income earners.

  • SC: thanks for your inside info and confirmation of some of what I suspected with these tax files under-reporting incomes at the top end, as well of course of the problem of tax avoidance through offshoring/ tax havens.

    RCP: thank you also for your comments.

    I don’t think many people would consider a per capita head or poll tax to be a model of fairness in this century; most probably instead consider levels of income (or wealth) as a base for notions of fairness.

    I see Marc has already responded to some of your points about the concept of income in his study.

    Yes, the share of the 1% declined after the financial crisis, but I expect it will have risen again since. I didn’t write about the issue of mobility, but Miles Corak has done a lot of work on that: I believe he’s found increasing correlation, but that’s a whole different issue not that relevant to my point.

    The issue of elasticity of taxable income also involves considerable discussion. Very briefly:

    1) as noted above by SC, the top income figures appear to be underreported;

    2) there’s a wide range of estimates for ETI and the Finance figures for top incomes, while politically opportune, appear to be on the high end of those: Their calculations were done during a period when high incomes escalated and top income tax rates were cut deeply, but it is questionable how much the latter caused the former (and even if it did, see (5) below). I expect that there were a lot of other economic factors that influenced these results that they didn’t account for in their econometric equations: anyone who has worked in that field knows how this is done;

    3) much of the ETI can reflect tax shifting behaviour, rather than real economic factors, which means this income turns up elsewhere. (Even Jack Mintz, who doesn’t seem to have seen a business tax cut he doesn’t like, seem to wholeheartedly argue or concede that much of the ETI related to CIT changes was the result of tax shifting);

    4) related to that, the finance ETI estimates involve marginal tax rates; what I discuss above is the average tax rate. I note that the focus is as much on closing loopholes/ preferences (which provide opportunities for tax avoidance and shifting) as much as changing marginal rates;

    5) Piketty and Saez, among others, have dealt directly with this issue. There’s not a lot of evidence that changes in these top tax rates have substantial real economic effects as they have shown. It is unlikely that the top 1% would reduce their productive economic activities by 60 or 70% in relation to higher taxes, but that’s what those elasticities imply. Piketty and Saez instead suggest lower taxes on the top 1% make “rent seeking” exploitation of the 99% by these top 1% more lucrative. This brings us back to Warren Buffet, who explained (in the quote above) it’s the rich class that’s waging war, and winning.

    Here’s a nice accessible summary of the issue by them (Thanks to David M for sending this)

    (There’s also another kinder gentler factor that could explain the lack of correlation of changes in top marginal tax rates with economic growth: the opportunity benefits (or costs) of additional tax revenues being put to (or taken away from) productive use through public spending, investment or redistribution to others).

  • @Toby: I did not suggest a poll tax, which I agree would be regressive.

    If you were going to increase the average tax rate paid without increasing the top marginal rate (the paper linked by the voxeu article suggests a total elasticity of 0.5, which is a long way from the 0 assumed in a static revenue estimate), how would you do it?

  • @rcp

    To increase the top average tax rates, I’d start by reversing some of the cuts made since then, including restoring tax rates on stock options, capital gains and investment income so they are taxed at the same rate as employment income (for cap gains, adjusted by inflation). I didn’t suggest doing it w/o increasing top marginal rates, but much of the impact of these cuts were at the provincial level, such as Alberta’s shift to a flat tax. I mention all these in the blog post.

    Also to be clarify, the revenue figures I cite above are of how much would be raised if the average effective tax rate on those income levels had been collected, however you get there. They are not forecasts or predictions of what would be raised from specific policy proposals, such as raising rates or broadening bases. That’s obviously a different discussion which of course involves elasticities and behavioural factors.

Leave a Reply

Your email address will not be published. Required fields are marked *