Which income number are you looking at
Statistics Canada publishes several different figures that all get called "income" in casual conversation, and they are not interchangeable. Total income is the broadest: employment earnings, self-employment income, investment income, private pension income, retirement income sources such as CPP and OAS, and government transfers, all added together before anything is subtracted. Employment income is a narrower slice — wages, salaries and self-employment earnings only — leaving out investment income, transfers and retirement income entirely. Market income is total income minus government transfers, isolating what a person earned from work, investment and private pensions specifically, without what the state paid them.
"Government transfers" is itself a specific category, not a vague catch-all: it covers payments like the Canada Child Benefit, Employment Insurance, social assistance, and the CPP, OAS and GIS payments covered elsewhere on this site. A person's total income includes all of these; their market income excludes all of them; their employment income excludes them and also excludes investment income and private pension income on top of that.
A single Canadian can have a total income figure, an employment income figure, and a market income figure that are three different numbers describing the same person in the same year, and casual reporting on "average income" or "median income" frequently does not specify which one is meant. A retiree living mostly on CPP, OAS and investment income, with no employment earnings at all, shows the gap clearly: their employment income sits near zero while their total income can be substantial, and their market income falls somewhere between the two depending on how much of it came from private investments and pensions rather than public benefits.
Self-employment income adds its own wrinkle inside the employment-income category, because it is reported net of business expenses and can swing sharply from year to year in a way salaried employment income rarely does — a self-employed worker's income in any single year reflects that year's business results, not a stable annual wage, which is part of why a single year's snapshot of self-employment earnings is a noisier signal than a salaried income figure for the same dollar amount.
Individual, household, or economic family
The second axis of confusion is whose income is being measured. Individual income is one person's income, full stop. Household income adds together the income of everyone living at the same address, whether or not they are related to each other. Economic family income adds together the income of people related by blood, marriage, common-law partnership or adoption who live in the same dwelling — a narrower grouping than household in a home with unrelated roommates, and often the same grouping as household in a home containing only a couple, or a couple and their children.
A two-earner couple's household income is roughly the sum of both partners' individual incomes, which is an entirely different number from either partner's individual income considered alone. Comparing a household figure against an individual figure produces a comparison that looks far more dramatic than it should: a household with two earners is not, in any meaningful sense, twice as well off as a single person earning the same amount as one of those two earners, because the household also has to support twice as many potential dependants and expenses on that combined figure.
Published statistics do not always announce which of these three units they are using in the headline number, and the difference matters most exactly where the stakes are highest — comparisons meant to say something about ordinary living standards, where conflating an economic family figure with an individual one can make typical income look considerably higher than any one person in that family actually receives.
Before tax and after tax
Every income measure above can be quoted before or after income tax. Before-tax (gross) income is what shows up on a T4 slip or on a tax return's income lines; after-tax income subtracts federal and provincial income tax, leaving the amount actually available to spend or save. Because provincial tax rates and brackets differ from province to province, two people with identical before-tax income living in different provinces can end up with different after-tax income, even though their before-tax figures are, by definition, identical.
Comparisons that quote a national average or median without specifying before- or after-tax, and without specifying which of the total, employment or market income concepts is being used, are technically incomplete even when the number itself is accurate. Three axes — which income concept, before or after tax, and individual versus household versus economic family — combine into a large number of distinct valid figures for the same underlying population, and a single headline "average income in Canada" could plausibly refer to any of them.
Tax deductions and tax credits complicate the before/after-tax line further, because they change how much tax is owed on a given amount of gross income without changing the gross income figure itself. An RRSP contribution, for instance, lowers taxable income for the year it is claimed, which lowers tax owed and raises after-tax income for that year relative to an otherwise-identical filer who did not contribute — two people with the same before-tax income can post different after-tax income for reasons that have nothing to do with how much they were paid.
Why the average exceeds the median
Income in Canada is right-skewed: most people earn a moderate income, and a comparatively small number of people earn a very large one, and that small group pulls the average upward without moving the median much at all. The median is the income at which half of people earn more and half earn less; the average (mean) is the total income divided by the number of people, a calculation that weights every dollar earned identically regardless of who earned it. A small number of very high earners can add enormous absolute dollars to the total without adding a single additional person to either side of the median split.
This is why the average is consistently reported as higher than the median in Canadian income data, sometimes considerably so, and why the median is generally treated as the more representative figure for describing a "typical" income — the average can be dragged upward by a thin slice of very high earners in a way the median, by construction, cannot be. The gap between the two numbers is itself informative: a wide mean-median gap signals a more unevenly distributed income range than a narrow one, even before looking at any other measure of inequality.
The same skew shapes how a percentile is read, a subject covered in more detail in a separate guide on income percentiles: because high earners pull the top of the distribution out into a long tail, the difference in income between, say, the 90th and 99th percentile is typically far larger than the difference between the 50th and 60th, even though each gap covers the same number of percentile points.
The age gradient
Income does not move in a straight line across a working life. It typically rises through the twenties and thirties as workers gain experience, seniority and qualifications, peaks somewhere in the mid-career years, and then declines toward and through the traditional retirement age as employment earnings give way to part-time work, pension income and reduced hours. Comparing income across two people of different ages without accounting for this shape can attribute a difference to something else entirely — location, education, gender — when age alone explains a meaningful part of the gap.
The decline after peak earning years deserves its own caution, because it does not mean one single thing. Some of it reflects a genuinely lower total income in retirement; some of it reflects a shift in the mix of income sources rather than a fall in living standards, since a retiree's total income figure can include pension income and government benefits that do not show up as employment earnings at all, alongside assets like a paid-off home that do not appear in any income figure whatsoever.
This is also why comparing income meaningfully requires comparing within an age band rather than across the whole adult population at once — a 30-year-old and a 60-year-old sit at different points on the same life-cycle curve, not on two different curves, and a raw national average blends every point on that curve into one number that describes no single age well.
The education gradient
Income also varies with educational attainment, on average, across most working-age bands: workers with a bachelor's degree or a graduate degree tend to report higher median and average employment income than workers whose highest credential is a high school diploma or a college or trade certificate. This pattern holds broadly across age groups during the working years, though the size of the gap between education levels is not constant — it narrows or widens across a career rather than staying fixed from a worker's first job to their last.
The gradient is measured from employment income specifically, which matters because employment income and total income are not the same measure, as covered above: the education gradient describes a gap in what people earn from work, not necessarily a gap in every income source that makes up total income, and applying an employment-income gradient to a total-income base — which is exactly what this app does, described next — carries that distinction forward into the estimate.
How this app estimates income
This app's income comparisons start from a real total-income base — actual Statistics Canada figures by province, or nationally, and by age group — and then apply a national education pay-gradient, derived separately from employment income data, to split that base out by education level. The result is a hybrid estimate: the age and province numbers are real totals, and the education split layered on top of them is modelled from a different income concept than the one it is being applied to.
One adjustment in that model is disclosed here directly because it affects the oldest age group specifically. For the 65-and-older band, the education gradient is clamped to 1.0 — this app shows the same income figure across every education level for that age group, rather than applying the modelled gradient at all. The reason is that the raw employment-income gradient for workers 65 and older is not a clean measure of education's effect on earnings: at that age, employment income increasingly reflects who is still choosing or able to work at all, a selection effect rather than a stable premium tied to credential, since most people in that age band are not primarily earning employment income in the first place. Applying the working-age gradient to that band would overstate the effect education has on a retiree's total income, so the app removes the gradient for that one age group rather than publish a figure built on a selection artifact.
Why the measure problem matters
None of the distinctions above are unique to this app — Statistics Canada documents each of them, and any careful reader of the underlying tables can find the total-versus-market-versus-employment split, the individual-versus-household-versus-economic-family split, and the before-versus-after-tax split clearly labelled. What varies is whether a given comparison tool or news summary carries those labels forward, or collapses them into one unlabelled "average income" figure for simplicity. A reader comparing their own income against a headline number has no way to know which of the many valid figures that number represents unless the source says so.
This app is explicit, on the pages that follow, about which measure it is using at each point: individual total income from the census, split by province, age, gender and education, using the hybrid method and the clamp described above. That specificity is the whole reason a measure-by-measure guide exists at all — the number itself is only useful once the reader knows exactly what it is a number of.
Net worth is a related but separate measure, covered in its own guide, and the two are not interchangeable proxies for each other: a high-income household in an early career stage can carry a modest net worth, and a retired household with a low income can carry a substantial net worth built up over decades, so income data alone does not describe a household's overall financial position.
What this guide does not cover
This guide covers how income is measured and estimated in general terms. It does not walk through every province's figures individually — that comparison, including why a gross income comparison between provinces can mislead without accounting for tax and housing costs, is covered separately — and it does not explain how a specific dollar figure is converted into a percentile ranking, which is also covered separately given how much of that calculation depends on the shape of the income distribution rather than the income concepts described here.