A Voluntary Levy Instead of a Cap: Reading the Trade

An industry-funded levy trades a hard limit for a payment. Whether that is a good trade depends on who counts the nights.

Last updated: October 29, 2026

A cap and a levy are two different instruments pretending to be the same argument. One limits a quantity. The other sets a price and lets the quantity float. That difference decides who carries the risk, who does the counting, and what happens when someone declines to take part.

Key Takeaways

• A cap is a quantity; a levy is a price. They solve the same political problem in opposite ways.
• Who pays determines who changes behaviour. A platform payment and a per-night host payment are different instruments wearing one name.
• The base is the argument. Charged on gross or on net, on nights or on revenue, the same rate produces different outcomes.
• No rate has been settled, so any figure you've seen is a proposal rather than a term.
• Read the fallback. A voluntary scheme that triggers a cap when participation is low isn't voluntary for long.

What a levy replaces

The proposal on the table in Sydney is easy to describe and hard to price. Platforms would contribute a share of the revenue from listings in the city to community housing bodies. If they don't take part, the council has said it would consider lowering the annual cap on nights.

Read that as a mechanism rather than a policy and two things stand out. The payment is voluntary in name and conditional in practice, because the alternative is the instrument hosts already know. And the levy doesn't replace the cap so much as offer a way to avoid tightening it.

That framing matters for anyone trying to work out what it costs them. A cap acts on the edge of the year: you sell until the quantity is gone, then you stop. A levy acts on every night you sell, all year, whether the year is busy or quiet. One is a wall. The other is a toll on the road.

Walls produce certainty and blunt outcomes. Tolls produce revenue and a permanent argument about the rate. Neither is obviously better. The choice between them is usually made on politics, not on economics.

There's a third possibility that gets less attention. Both instruments can operate at once, with the levy buying headroom under a cap that still exists. That's the version most hosts would end up living with if the proposal progressed, and it's the version that's hardest to model from a headline.

Who pays, who counts

Two questions decide whether a levy changes anything for a host, and neither has been answered.

The first is who pays. A payment made by a platform out of its own margin is a cost to the platform, and platforms are good at passing costs through. A payment charged per booking is a cost to the guest. A payment charged per night to the host is a cost to the host, visible on a statement, and it changes behaviour at the margin.

Those three designs produce the same headline and completely different decisions. Under a per-night charge, a low-season night becomes less attractive, because it carries the same levy as a peak night. Under a platform-level contribution, the host's nightly decision doesn't move at all. The levy becomes a fixed cost of selling through that channel.

The second question is who counts. A levy needs a base, and a base needs a record. If the count comes from platform reporting, the platform defines the population. If it comes from a register, the register does. If it comes from self-reporting, you do, and the audit question arrives later.

Counting also decides whether the levy and the cap measure the same thing. A cap counts nights. A levy counts revenue. A host with cheap nights and a full calendar and a host with expensive nights and an empty one can look identical on one measure and nothing alike on the other.

We keep the sourced version of proposals like this one, with the outlet and the date attached, because the difference between a floated idea and a settled term is the whole story. localsbnb.com is where that version sits, so you can check what was actually said before you reprice anything.

The ratio nobody has fixed

As published by The Sydney Morning Herald, the share of revenue the levy would take hasn't been set. That's not a detail to skip past. It's the entire number.

Everything a host needs to plan sits behind it. Is the charge small enough to absorb in a nightly rate? Does it change which nights are worth selling? Does it make a long stay more or less attractive than several short ones? None of that can be answered while the rate is open, and any article that answers it is inventing a figure.

What can be read now is the shape. A revenue-based charge scales with price, so it bites hardest where nightly rates are high. A per-night charge scales with volume, so it bites hardest in properties selling many cheap nights. Those are different businesses, and a design that suits one will annoy the other.

There's a subtler question underneath the ratio: what counts as revenue. Gross booking value, net of platform commission, accommodation only or including cleaning fees. Each definition moves the effective charge without changing the headline rate, and each one is a negotiating position rather than a technicality.

The honest summary is that the trade can't be priced yet. What can be done is deciding in advance which version you'd accept and which you'd argue about, so that when a rate appears you already know what it means.

That's a different exercise from budgeting. Budgeting assumes a number and works out the consequence. This is the step before: working out which structure you'd want the number to sit inside, so that the first published version isn't also the first version you've thought about.

What to watch before 2027

Five things tell you whether a levy is becoming real. None of them is the headline rate.

Whether it's platform-level or host-level. That single choice decides whether you see the charge on a statement or never see it at all.

Whether non-participation triggers the cap. A voluntary scheme with a credible fallback is a mandatory scheme with better marketing, and the trigger is the part to read.

Where the money goes and who reports on it. A payment into a housing fund is a different proposition from a payment into general revenue, and the difference is usually stated once and then forgotten.

Whether the count is public. If nobody publishes how many nights and how much revenue the levy covers, nobody can tell whether it's working. The rate then becomes permanent by default.

And whether it survives a change of government. Instruments that depend on goodwill are the first to be revisited, and a levy is precisely that kind of instrument.

FAQ

Is the levy happening?

Not as a settled rule. It's a proposal, and the rate behind it hasn't been published. Anything firmer than that is a forecast, and forecasts about local government are worth what you paid for them.

Would a levy cost me more than a cap?

That depends entirely on the rate and the base, and neither exists yet. A cap costs you the nights you can't sell. A levy costs you a share of the nights you do. Without a rate, there's no comparison to make, only a preference to state.

What should I do in the meantime?

Nothing dramatic. Keep the night count accurate, keep the revenue record clean, and know which of the two instruments your market actually operates under today. If a rate is published, the first question to ask is what it's charged on.

A cap answers the question "how much can I sell". A levy answers "how much does selling cost". They're both answers to a housing argument, and only one of them is currently a rule. If you want the sourced account of where this proposal stands, localsbnb.com keeps it with the outlet and the date, so you can see what's been said and what's still open.

---

This article is general guidance for hosts and isn't legal, tax or insurance advice. Whether a levy applies, at what rate and on what base, is a decision for the authorities involved and may change before it takes effect. Confirm the current position with the competent authority before you change a rate or a calendar because of it.