Before you call a housing number a market signal, ask three questions

How to Read a Housing Number

A housing story can say that prices rose, rents fell, listings increased, or affordability improved. Before treating the number as a description of “the market,” ask three questions: what exactly was measured, where and when was it measured, and what does it leave out?

That short pause matters because Canadian housing data comes from different systems. An asking price is not a completed sale. A new-home absorption measure is not the same as a resale transaction. An average is not a typical household’s experience, and a national figure may hide large regional differences.

What the number actually measures

Start by naming the measure in plain language. Is it a listing, a completed transaction, a rent observation, a housing start, a completion, an unabsorbed unit, or a modelled affordability ratio?

CMHC’s Housing Market Information Portal offers data at several geographic levels, from national results to neighbourhood-level views. Its housing-market tables cover measures including starts, completions, units under construction, absorbed and unabsorbed units, and prices. Those measures answer different questions. A rise in starts can speak to construction activity; it does not prove that more completed homes are available today.

CMHC’s housing affordability and supply research also explains that house prices are recorded when a sale takes place and therefore reflect market transactions rather than the values of every property. Its affordability work uses a defined ratio that combines house prices, income, mortgage-rate assumptions, and homeowner expenses. That can be useful for comparing a modelled affordability trend, but it is not the same as a household’s actual monthly payment.

Statistics Canada’s Canadian Housing Statistics Program data-quality guide adds another qualification. Its data guide says that estimates can vary in availability and comparability because of differences in data sources, regional coverage, and processing. Some sale-price estimates are limited to market transactions and resident buyers. The definition is part of the result.

The regional qualification

Next, write down the geography. “Canada” may be appropriate for a national trend, but it is often too broad for a decision about a particular city, suburb, rental market, or neighbourhood. CMHC’s portal allows multiple levels of geography, while Statistics Canada warns that some property and assessment comparisons are affected by differences between provinces and territories.

The time period matters too. A monthly result, a year-to-date result, and a multi-year trend can point in different directions. A forecast is not a completed observation. A result from a small set of transactions may move sharply without representing every home in the area.

This does not make housing data useless. It makes the data label necessary. A reader should be able to tell whether a statement describes asking behaviour, completed sales, construction, inventory, rent, or a modelled ratio.

What the number cannot show

A market statistic cannot tell one household whether it should buy, sell, rent, refinance, or wait. It may not show the property condition, financing terms, concessions, household income, or the distribution hidden behind an average. It also cannot turn a national result into a neighbourhood forecast without additional evidence.

A workable reading habit is to keep the number and its definition together. Before sharing a housing claim, write:

• Measure:
• Geography:
• Reference period:
• Source and release date:
• What it does not measure:

If one of those lines is missing, the conclusion may be moving faster than the evidence.

Your next research step

Choose one housing number you have seen recently and find its original source. Re-label it with the five lines above. Then look for one second measure that answers a different question, such as transactions beside listings or completions beside starts. The aim is not to build a complete market forecast. It is to avoid making one number carry more meaning than its definition allows.

A second measure is most useful when it tests the first measure rather than simply repeating it. For example, pair a price observation with the number of completed sales, or pair construction starts with completions and unabsorbed inventory. The combination may still leave questions, but it makes the gap visible.

The extra measure does not need to be complicated. It only needs to prevent a definition from disappearing. A reader who knows what was counted can decide whether the statistic is relevant before using it to describe a market or a household decision.