Capping London's tourist tax at 5pc caps the rate, not the bill

Five per cent sounds like a limit. It behaves like an index.

The mayor’s assurance this week — that the new charge on overnight stays in London will not exceed five per cent — was offered as the reassuring part of the package. It reassures about the rate. It says nothing about the money. Five per cent of a nightly room rate rises whenever room rates rise, and room rates in London rise fastest exactly when the city is busiest and rooms are scarcest. The ceiling holds the percentage still and lets the pound figure float. That is a real design choice with real consequences, and it deserves to be argued about rather than accepted as a concession.

The word “capped” is carrying too much

A cap on a percentage is a cap on a ratio, not on a bill. Nobody pays a ratio. The guest pays a number, and that number is determined by a price the tax authority does not set and cannot forecast. This is not a scandal — plenty of good taxes work this way, and a proportional charge has the considerable virtue of never needing to be re-legislated as prices drift. But it means the headline figure answers a question almost nobody is asking, while the question people will actually notice — what does a night now cost — is left entirely to the market.

The real decision is flat fee or percentage

The legislation unveiled this week lets local authorities charge either a flat fee per night or a percentage of the stay. These are not two routes to the same destination. A few pounds a night is a meaningful share of a hostel bed and a rounding error on a suite. A percentage does the reverse: it takes proportionally from everyone and lands hardest, in cash terms, on the most expensive rooms in the most expensive weeks.

So the instrument decides who pays. Handing that decision to individual authorities is not a technical detail devolved for convenience. It is devolving the distributional question itself, one borough at a time, with no requirement that the answers agree.

There is a second problem underneath. A flat fee needs one fact: how many nights. A percentage needs a definition — a percentage of what. Does the base include breakfast, cleaning charges, resort or service fees, the commission a booking platform takes, a non-refundable premium? Every proportional tax generates a boundary dispute about its base, and those disputes get settled either in statute or, expensively, one assessment at a time. If that definition is left to each authority, one accommodation market acquires as many taxable bases as it has councils.

A borough boundary is not an economic boundary

This is what makes London different from a levy in a self-contained resort town. A traveller choosing between two hotels twenty minutes apart on the same line is choosing within a single market. If one side of that line applies a proportional charge and the other a fixed one, the budget segment and the premium segment will quietly re-sort themselves across it. That is not speculation about behaviour; it is the arithmetic of two different instruments applied to the same bed. Meanwhile the operators and platforms that straddle the boundary carry the compliance cost of running two systems in one city.

It is not a tax on foreigners

The charge applies to domestic and international guests alike, and that is worth sitting with. Visitor levies are usually sold on the comfortable theory that the burden falls on people who do not vote here. This one falls on someone travelling in from elsewhere in the country for a training course, a funeral, a hospital appointment, a job interview. The underlying justification still works — visitors consume local services without contributing to local property taxes, and domestic visitors do that too. But if the tax is honest about who it reaches, it should stop being described as a tourist tax and be priced as what it is: a charge for occupying a night in the city.

The strongest case for leaving it alone

The case for local discretion is genuinely good. Westminster’s hotel stock is not the same business as an outer borough’s, and a council that knows its own market is better placed than a department to judge whether a fixed charge or a proportional one suits it. A cap constrains the worst outcome while leaving room to differ. And the proportional option is arguably the more durable instrument precisely because it tracks prices, sparing everyone the lumpy, politicised revisions that fixed statutory amounts always eventually require.

That last point is correct, and it is why the proportional charge is the better instrument. But it is an argument for mandating the percentage, not for permitting either. The discretion worth devolving is over the rate: anything from nothing up to the ceiling, chosen locally and defended locally. The discretion not worth devolving is over the instrument and the base, because those are the two variables that generate arbitrage across boundaries and duplicated compliance inside a single market.

A cap is the easiest thing to legislate and the easiest thing to announce. It is not the thing that will determine whether this charge works.