Personal income tax concentration by income group

Canada's personal income tax is paid by a narrow slice of filers. In 2023 the top 1% earned 10.1% of all income and paid 22.1% of all federal and provincial income tax — 2.19 times their share of the income. The bottom half of filers paid 5.0% of that same tax — 3.6% counting federal tax alone — and 28.5% of filers owed none at all. Measured per dollar earned, the burden on Canada's top 1% is heavier than the burden on America's.

"Paying your fair share" is one of the most-used phrases in Canadian tax politics and one of the least-measured. The underlying question — how much of the income tax bill each part of the distribution actually carries — has a precise answer, published annually by Statistics Canada and the Canada Revenue Agency from tax filings themselves. This page reports it. It does not settle what a fair share ought to be; that is a political judgment. It establishes what the current shares are, so the argument can start from the same facts.

The headline measure used throughout is the burden ratio: a group's share of the income tax paid, divided by its share of the income earned. A ratio of 1.0 means a group pays exactly in proportion to what it earns. Above 1.0 means it pays more than proportionally; below 1.0 means less. The ratio is used in preference to the raw share because it is the one number that survives the definitional differences between tax bases and between countries — raw shares do not.

How the burden is distributed

Statistics Canada publishes the share of income and the share of federal plus provincial income tax held by each percentile group, from the top 0.01% down to the top half of all filers. Every group above the 90th percentile pays more of the tax than it earns of the income; the bottom 90% pays less. The top 10% of filers earn just over a third of the income and pay more than half of the income tax, and the top half of filers pay 95.0% of it.

Splitting the same data into non-overlapping bands shows where the progressivity actually sits — and where it stops. The burden ratio climbs steadily up the distribution and peaks in the band between the 99.9th and 99.99th percentiles at 2.39×. Above that, at the very top 0.01%, it falls back to 2.31×. At the other end, the 10–50% band — the broad middle, four filers in every ten — pays 0.83× its income share, and the bottom half pays 0.28×. These seven bands account for the entire distribution: their income shares sum to 100% and so do their tax shares.

Who pays the same as whom

"The rich don't pay their fair share" is a claim about proportion, and proportion is measurable. It is worth being exact about what a measurement can and cannot settle. It can establish who currently pays what. It cannot establish what anyone ought to pay — that is a judgment about values, and no table decides it. What follows is only the first half: the part that is a matter of record.

Set the groups against each other directly and three comparisons fall out of the 2023 data.

The obvious reply is that they earn more, so of course they pay more. It deserves a direct answer rather than a deflection, and the answer is that the earnings gap runs the other way. The top 5% earn roughly a third as much income as the bottom 90% — 23.2% against 65.8% — and pay very nearly the same amount of income tax. The top 0.1% earn a fifth of what the bottom half earns and pay 1.7 times as much. Higher income explains part of a higher tax bill; it does not explain a bill that inverts the income ranking.

The cleanest way to see it is to put the two distributions side by side. Both columns below total 100%. The left column is who earns the income; the right column is who pays the tax. Every band occupies a different amount of space in the two columns, and the direction of that change is the progressivity of the system.

None of this makes the bottom half free riders, and the page should not be read that way. The bottom 50% of filers pay little income tax because they earn little income — 17.9% of the national total shared among 15.5 million people — and because Parliament has deliberately built a schedule of basic personal amounts, credits, and refundable benefits designed to lift low earners off the tax rolls. A bottom half paying 5.0% of the income tax is the system working as intended, not evidence of avoidance. Those same filers also pay GST and HST on what they buy, fuel and excise taxes, property tax through their rent, and CPP and EI premiums on their earnings — none of which is counted here.

What the numbers settle is narrower and firmer than the argument they are usually enlisted in. On personal income tax, the burden is already carried disproportionately by the top of the distribution, by a margin that has widened over forty years, and by more per dollar earned than in the United States. Whether that margin ought to be wider or narrower is a question about values, and reasonable people land in different places on it. It is not a question about facts, and it is not well served by being argued as though the facts were in dispute.

The bottom of the distribution

The CRA's own bracket-level statistics reach the parts of the distribution that Statistics Canada's high-income table does not. Reading the 2024 tax year from the bottom up: the lowest quarter of filers by income pay almost nothing in federal income tax, and the lower half pays a low single-digit share.

Source: Canada Revenue Agency, T1 Final Statistics, 2024 tax year, Final Table 2. 31,564,150 filers; $2,023.6B total income; $223.1B net federal tax.

Alongside this, 8,994,320 returns — 28.5% of all filers — were non-taxable, meaning nil tax payable after credits and deductions. That figure is frequently quoted loosely, so it is worth stating precisely: it means zero income tax, not zero tax. Those filers still pay GST/HST on what they buy, fuel and excise taxes, property tax through their rent, and CPP and EI premiums on employment earnings. It is also a count of filers, not of Canadians — a large number of people file a return specifically in order to receive the GST credit, the Canada Child Benefit, and provincial credits, which is why the filing rate is high among people with no tax liability at all.

The mechanism producing this distribution is the rate schedule itself, visible directly in the effective rate — net federal tax as a percentage of total income — at each bracket.

Forty years of rising concentration

The concentration is not new, but it has intensified. In 1982 the top 1% earned 7.1% of income and paid 13.4% of income tax — a burden ratio of 1.89×. By 2023 the income share had risen to 10.1% and the tax share to 22.1%, a ratio of 2.19×. The top 1% income share actually peaked in 2007 at 12.1% and has fallen since, while its tax share has held roughly flat — which is precisely why the ratio now sits at the top of its 40-year range. The five highest readings in the 42-year series all fall between 2016 and 2023; the single highest is 2020 at 2.22×, a pandemic-year distortion, and 2023 is second at 2.19×. Over the same four decades the bottom half of filers has never paid more than 6.5% of the income tax or less than 3.9%, ending 2023 at 5.0% — a flat line beneath a rising one.

Canada against the United States

The comparison most often reached for is the United States, where the top 1% pays 38.4% of federal individual income tax — a far larger headline share than Canada's 22.1%. That comparison is misleading on its own, because the American top 1% also earns twice as large a share of the income: 20.6% against Canada's 10.1%. Set the tax share against the income share and the ranking reverses. Per dollar earned, Canada's top 1% carries a heavier income tax burden than America's — 2.19× against 1.86×.

A wider G7 panel is not available. The United Kingdom publishes a top-1% share of income tax of 28.2% for 2024–25, but HMRC counts income taxpayers only — people with no income tax liability are excluded from the denominator entirely — which mechanically inflates every share and makes the UK ratio non-comparable with the Canadian and American figures. It is recorded in the companion workbook for reference and is deliberately not charted. Germany, France, Italy, and Japan publish no equivalent top-percentile series at all.

Findings

Sources & methodology