Affordable Housing: What the Word Measures, and How a Stock Shrinks Without Demolition

The word is everywhere and rarely defined, which produces conversations where two people use one term for two different things. This piece carries no figures: we reached no primary source on the volumes involved, and the mechanism is fully understandable without them. On entry prices into ownership, a separate subject, see our read of the Quebec City condo market.

The word names a ratio, not a price

This is the original confusion, and everything else follows from it. People hear affordable as an adjective attached to the unit, like spacious or renovated: a property would be affordable the way it is electrically heated. That is not what the word measures.

Affordability is a ratio between two quantities: what the housing costs, and what the household occupying it has. Past a certain share of income absorbed by housing, it stops being counted as affordable. The threshold chosen varies between organisations and programs, but the structure does not: it is a fraction, never an amount.

Two consequences follow immediately, and they surprise people. Two units at the same rent are not equally affordable: they are affordable for one household and not another. And a single unit can change category without its rent moving by a dollar, if the reference income it is measured against has shifted. The unit did nothing; the denominator moved.

How a stock shrinks with no unit disappearing

When a number of affordable units is reported lost, the image that comes to mind is buildings demolished or converted. That happens, and it is not the main mechanism.

The main exit is by threshold crossing. A unit whose cost rises faster than the income it is measured against eventually passes the chosen fraction. It leaves the category. The building still stands, the unit is still occupied, nobody moved, and the count has nonetheless dropped by one.

This mechanism has a property worth seeing: it is silent and continuous. A demolition is noticed, contested, documented. A threshold crossing is not noticed, because it produces no visible event on the street. That is why a stock can contract for years with no construction site involved, and why new building fails to compensate: construction adds units while the threshold removes them.

Affordable and mid-range are not the same word

The two travel together and are not interchangeable. Affordable describes a ratio, as above. Mid-range describes a product category, sitting between social housing and market housing, defined by the rules of the program that funds it rather than by an income threshold.

From that follows a situation that looks contradictory and is not: a unit can be perfectly mid-range under the program that made it possible, and not be affordable in the ratio sense for the households applying. The two statements describe different objects. The person talking about the program and the person talking about the rent are both right, and they are not talking about the same thing.

Why a unit delivered as affordable can be judged too expensive

This is the second mechanism, and it comes from a gap between two dates. A housing project is designed long before anyone lives in it: financing, approvals, construction. The target rent is set on conditions prevailing at design.

Between approval and delivery, three quantities evolve at their own pace and not necessarily in the same direction: construction cost, financing cost, and the income of the households targeted. The rent actually reached at delivery incorporates the first two; the judgement passed on that rent is made against the third.

A project can therefore fully respect the rules under which it was approved and deliver a rent the public judges too high, with no rule broken and no commitment abandoned. That is not an execution failure, it is a benchmark that moved while the building went up. Naming it that way does not make it acceptable; it lets the argument land on the right thing, namely how the benchmark is indexed rather than whether the parties acted in good faith.

What we are not writing here, and why

No count of units lost or gained. No income share. No reference rent. We reached no primary source we can cite for those quantities, and a figure taken from an intermediary without being traced to its source is not a measurement, it is a quotation.

We say so rather than fill the gap, because on this subject the number is precisely what gets quoted most and verified least. The mechanism itself needs none: it is true from the definition alone. When a primary source becomes available, it will quantify a mechanic already understood, which is the right order.

The three checks that concern you

The first is a calculation rather than a reading: what share of your net income, charges included, would this unit absorb. That is the only affordability measure that concerns you, and it cannot be derived from any published statistic, because the denominator is yours.

The second is about pace rather than level: how much has that cost risen in recent years, compared with what you have. A unit affordable today whose cost rises faster than your income will stop being affordable, and the date that happens can be calculated.

The third applies only to a unit from a program: which rule sets its rent, and for how long. A time-limited advantage is not planned like a permanent one, and that duration sits in the program documents, not in the listing.