She rang us annoyed, which is fair enough.

Her investment townhouse — one property, held twelve years, alongside the family home — was empty on Census night. Not because she was holding out for a better rent. The bathroom had failed, the repair took longer than quoted, and the place was between tenants for about six weeks.

So she will be counted as one of the million.

More than a million Australian homes are expected to come back from this Census marked "unoccupied" — and within days that number will be everywhere, usually rendered as a million empty homes while people sleep in cars.

Her question was blunt: is that number about me? And is it going to cost me?

The first answer is no. The second is where it gets serious.

What the word actually means

"Unoccupied" on Census night means one thing only: nobody slept there that night.

That is the entire test. If you were on holiday, away for work, in hospital, or staying at your sister's, your own home — the one you live in, the one with your furniture in it — is recorded as unoccupied. Not vacant. Unoccupied.

Historically about one in ten Australian dwellings lands in that category, which is where the million comes from. Inside it sit homes between tenants, homes mid-renovation, deceased estates being sorted out by grieving families, holiday homes, and dwellings so newly completed nobody has moved in yet.

None of that is a home deliberately withheld from the market. It's a photograph of where the country slept on one particular Tuesday.

And the cohort who genuinely can leave good, rentable housing empty for years is small and specific: owners with many properties and no debt against them. For everybody else the arithmetic makes it impossible. Rates, land tax, insurance, body corporate fees and a mortgage don't pause because the house is empty. Plenty of leveraged properties run at negative cash flow while a tenant is paying rent. Switch the rent off and you're not making a strategic decision, you're bleeding.

There's a geographic problem too. A holiday house on the coast does not house a nurse in Sydney. A farmhouse four hours from any employment does not solve a rental shortage in Melbourne. Empty and useful are different words, and the Census counts neither — it counts slept in last night.

Why she's right to be worried anyway

Here's the part that turns a statistical quibble into a real financial question.

Levies priced off this idea already exist. Victoria taxes vacant residential land at 1% of the property's value — on a million-dollar property, $10,000 a year. And the mechanism that catches you is not an inspector; it is self-reporting. Owners must declare it themselves, and a great many have no idea the obligation exists.

Read that again with an investor's eyes. The most likely way an ordinary owner gets hurt by a vacancy levy is not being targeted. It's not knowing they were supposed to tell someone.

And the pressure is building. In South Australia an advocacy group has proposed taxing empty homes at 0.5% of value — around $5,055 a year on the Adelaide median — escalating to a 20% marginal rate for owners of multiple empty properties, with compulsory acquisition for anyone refusing to pay. It would apply to any house not slept in overnight, including unbooked gaps in a short-stay listing.

To be straight about it: that proposal is not government policy, and the state government has said it has no plans to adopt it. But the Victorian levy is real, it operates now, and every one of these schemes is argued for using a number that counts a family on holiday as an empty home.

If you accept the premise, you eventually accept the tax. Which is why the definition matters.

The bit that should change how you look at every other number

We spend our working lives on a version of this problem, and it shows up in two different disguises.

The first is resolution — the number is real, but it averages things that behave nothing alike. A suburb median is built from streets that don't move together, which is why a renovation can add nothing to a valuation. We routinely measure a 20–30% spread in effective yield between the best and worst streets inside a single suburb — same postcode, same median, same council — using achieved rents rather than advertised ones, real vacancy duration rather than a quoted rate, and actual days on market.

The second is definition, and it's the more dangerous one, because it survives any amount of accuracy. The Census figure isn't wrong. It's counted properly, published honestly, and it measures exactly what it says. It just doesn't measure the thing its name suggests to a reader in a hurry.

That's the trap: a number can be entirely correct and still be the wrong number for the decision you're making. An advertised rent is a real number and it is not what the property achieves. A quoted vacancy rate is a real number and it is not how long your place will sit empty — as when a forecast said $2 and the rent went up $50. "Unoccupied" is a real number and it is not "vacant".

Her townhouse is the perfect example. It was empty six weeks because a bathroom failed. In the national data that is indistinguishable from a mansion nobody has entered since 2019. Only property-level information tells those apart — and it's the only kind that answers an owner's actual question.

So what should she do

Three small things.

Check whether a vacancy obligation applies where the property is, and whether it's self-reported — the same quiet risk as rental rules that changed without most landlords noticing.

Keep a paper trail for genuine vacancies — the repair quote, the invoice, the dates it was advertised, the management agreement. Legitimate reasons are only useful if you can evidence them.

Stop underwriting off headline numbers. If a six-week gap between tenants can put her in a statistic used to argue for a tax, a suburb-level yield figure is nowhere near good enough to decide what to buy next.

The reason none of this is an argument against owning property

It's very nearly the opposite.

A market like this rewards work precisely because it is priced off these headline figures — medians, advertised rents, quoted vacancy rates, national counts — averages and snapshots dressed as facts about individual assets. In an efficient market, homework earns nothing. In this one, the gap between the published number and the real one is where the entire return lives.

She wasn't hoarding a home. She was fixing a bathroom. But she has just learned the more valuable version of that lesson: the numbers everyone argues about are not the numbers that decide what happens to your property.

The mistake was never owning an investment property. It's owning one while taking the country's biggest statistics at face value — and never once checking what they actually count.

This article is general information only and does not take your personal circumstances into account. It is not financial, tax or legal advice.