Rent-to-Income Is a Blunt Instrument. Capacity-to-Pay Is a Scalpel
The 30% rule began as a subsidized-rent formula. A rent to income ratio alternative built on Canadian data, and where the ratio still wins.
Alfred BEditorial Reviews
A household bringing home $4,000 a month applies for a $1,400 apartment. Thirty-five percent, so the ratio says no. That household paid $1,350 somewhere else for four years without a late month. A second applicant lands at 27% and services $1,900 of consumer debt behind the paycheque. The ratio says yes.
Any rent to income ratio alternative starts from what a household's cash actually does rather than what fraction of gross income the rent represents. The 30% threshold was built between the 1940s and 1981 as a formula for setting rents in subsidized housing. It was never fitted against payment outcomes, and it is now used to decide them.
Where did the 30% rule actually come from?
From rent-setting in American public housing, in three steps. The US Department of Housing and Urban Development's own policy history puts the maximum affordable rent for federally subsidized housing at 20% of income in the 1940s, rising to 25% in 1969 and to 30% in 1981. Each step decided what a tenant would be charged.
HUD's Office of Policy Development and Research, in an article published 22 September 2014, attributes the 25% cap to a 1969 amendment sponsored by Senator Edward Brooke. Nothing in that lineage is a risk finding.
Canada imported the number rather than deriving it. CMHC's 2016 research report Defining the Affordability of Housing in Canada, prepared by Prism Economics and Analysis, states that the benchmark "emerged in Canada in the 1980s" after HUD adopted it as an eligibility guideline, and carries the line that settles the question: "Neither the current 30% benchmark, nor the earlier 25% benchmark, were based on studies of household budgets or actual housing costs in different regions."
What do the Canadian numbers actually show?
That a third of Canadian renters live above the threshold and almost none lose their housing over it. CMHC defines the affordability standard as housing costing less than 30% of before-tax household income. Statistics Canada reported on 21 September 2022, from 2021 Census data measured against 2020 incomes, that 33.2% of renter households spent 30% or more on shelter.
The Canadian Housing Survey puts the renter figure at 33.0% for 2022, and 34.0% in market rental housing, released 10 September 2024. CMHC research published in 2023, drawing on the same survey for 2021 and 2022, estimated that about 1% of renters were evicted in that period, roughly 49,000 households.
One in three above the threshold. Roughly one in a hundred evicted. Whatever the 30% line measures, it is not the probability that rent stops arriving.
How does the ratio fail in both directions?
By over-rejecting households whose money is well-behaved and under-rejecting households whose money is already committed. A ratio built on rent and gross income sees neither the rest of the balance sheet nor the rhythm of the deposits, so it declines paying applicants and clears non-paying ones with equal confidence.
The second failure is the quieter one. Statistics Canada reported on 12 June 2026 that household credit market debt reached 179.6% of disposable income in the first quarter of 2026, with 14.75% going to debt service. An applicant at 27% rent-to-income carrying a car loan, a line of credit and a card balance commits far more of the paycheque than one at 35% carrying nothing.
Mortgage lending never accepted a single ratio. CMHC's mortgage loan insurance criteria use two, gross debt service capped at 39% and total debt service at 44%. The second exists because the first cannot see other obligations.
What does a rent to income ratio alternative measure?
Residual income plus observed behaviour. Capacity to pay is a measurement of whether a household's actual income, obligations and cash buffer leave enough after rent to cover everything else, computed from observed flows rather than a fraction of gross income.
Michael Stone made the residual income case in Housing Policy Debate in 2006: a household has an affordability problem "if it cannot meet its nonhousing needs at some basic level of adequacy after paying for housing." The threshold therefore varies with household size, type and income, and CMHC's 2016 report lands in the same place.
The behavioural half is better documented. Balance behaviour, payment rhythm, negative-balance frequency and the fixed-versus-variable income mix are catalogued with their evidence in the cash-flow signals hiding in plain sight. A household paying a similar amount on time for three years is showing what the ratio guesses at.
What each measurement uses, what it costs to run, and what it can and cannot see.
| Rent-to-income ratio | Capacity measurement | |
|---|---|---|
| Inputs | Stated gross income, asking rent | Observed deposits, obligations paid, balance behaviour, household size |
| Origin of the threshold | US subsidized-housing rent formula, 1940s to 1981 | Fitted to the portfolio's own outcomes |
| Cost to run | Minutes, no infrastructure | Data connection, categorization, a written policy, monitoring |
| Legible to the applicant | Fully | Partly, and only if explained |
| Sees other debt obligations | No | Yes |
| Sees income variability | No | Yes |
| Sees payment history at a similar rent | No | Yes |
| Same answer from two reviewers | Yes | Only with a versioned policy |
How does income variability change the answer?
It moves the question from a yearly number to a monthly one. A ratio divides an annual income figure by a monthly payment, which works when the paycheque is the same every month and stops working when it is not.
The JPMorganChase Institute reported on 30 September 2025 that for hourly workers, take-home pay changes in seven of every ten months even while they stay in the same job, that the typical monthly change is about 9%, and that one month in four brings a change of 21% or more. Hourly wage rates change roughly once a decade. The instability is in the hours.
Canada has counts rather than distributions. Statistics Canada reported on 4 March 2024 that 871,000 Canadians did gig work in their main job in the fourth quarter of 2022. For them the annual figure at the top of the ratio averages months that did not resemble each other.
Who does the ratio exclude who would have paid?
Mostly people who moved recently, and people whose income does not arrive in a recognizable shape. Statistics Canada's Census in Brief on recent and existing renters, published 4 October 2023 from 2021 Census data, found 43.2% of recent renter households spending 30% or more of income on shelter, against 30.5% of existing ones.
That gap is market rent against sitting rent. Statistics Canada reported on 27 August 2025 that renters of five or more years paid about 19% less per month than renters of under a year in 2021. Applied at the moment of moving, the 30% test is partly a test of how long someone has already lived where they live.
The second excluded group is anyone whose earnings are real but illegible to a ratio: the self-employed with lumpy invoicing, the newcomer with two months of Canadian pay history and a decade of rent paid elsewhere. Twelve months of account activity shows what one annual number cannot, and applicants with no bureau file are covered in underwriting the credit invisible. The OECD's Affordable Housing Database puts the cut-off plainly: "the choice of the threshold, for instance, whether 30% of gross income is 'acceptable' and 40% is a 'burden', is arbitrary."
Does the ratio still make sense for a small landlord?
Yes, and pretending otherwise would be dishonest. The ratio is free, takes a minute, gives the same answer from any two people who run it, and can be explained to an applicant in a sentence. A capacity measurement has none of those four properties.
Most Canadian rental housing sits with owners who cannot run one. Statistics Canada reported on 7 July 2026, from 2022 data across six provinces, that small-scale individual investors held the largest share of assessed value among investor-owned residential properties in five of the six. Someone with four units has no data connection, no categorizer, and no book of outcomes to fit a threshold against.
The measurement wins where volume pays for the infrastructure. My view is that the threshold belongs at triage rather than at the decision, and that at real scale, deciding tenancies on a number invented to set subsidized rents is choosing to be wrong in both directions on purpose. Portable, pre-verified applicant files are one route there, discussed in the reusable applicant profile.
What we couldn't verify
Three things, and the first matters more than the article's whole argument. No published study establishes how well a rent-to-income ratio predicts non-payment at any threshold. The rule's documented history is about rent-setting and affordability measurement, and no source found here tests it as a screening cutoff.
The claim that the 25% rule descends from a nineteenth-century maxim of "a week's wages for a month's rent" could not be traced to a primary source. HUD's published history starts at 20% in the 1940s.
Canada publishes no national series on rental arrears or non-payment. CMHC's 2023 eviction research says so directly, noting that housing matters sit with the provinces and territories. The base rate any threshold would be calibrated against does not exist in published form.
Common questions
What is the 30% rent-to-income rule based on?
Rent-setting in US subsidized housing. HUD's policy history records maximum affordable rent at 20% of income in the 1940s, 25% after a 1969 amendment sponsored by Senator Edward Brooke, and 30% from 1981. CMHC's 2016 report states neither benchmark came from studies of household budgets.
Is there a rent to income ratio alternative that works better?
Residual income measurement plus observed cash-flow behaviour. Residual income asks whether enough remains after rent to cover non-shelter necessities, which varies by household size and income. Cash-flow behaviour adds obligations actually paid, balance patterns and income variability. Both need account-level data.
How many Canadian renters spend more than 30% of income on rent?
About a third. Statistics Canada reported from the 2021 Census, measured against 2020 incomes, that 33.2% of renter households spent 30% or more on shelter. The Canadian Housing Survey put it at 33.0% for 2022 and 34.0% in market rental housing.
Does exceeding 30% mean a tenant will not pay?
No published evidence says it does. Roughly a third of Canadian renter households sit above the threshold, while CMHC research using 2021 and 2022 survey data estimated about 1% of renters were evicted, around 49,000 households. Two orders of magnitude separate the rates.
Why do landlords with a few units still use the ratio?
Because it is free, fast, consistent between reviewers and explainable to an applicant in a sentence. Statistics Canada reported in July 2026 that small-scale individual investors held the largest share of investor-owned residential value in five of six provinces studied.
Carousel builds the intake and verification layer that turns a rental application into observed data rather than stated numbers. See how verification fits your flow
