England’s new tourist tax: what it raises, who pays, and what it costs

Mayors across England are to be given the power to charge a tourist tax on overnight stays, expected to be around 5%. It would apply to British and foreign visitors alike.¹²

The proposal has already generated competing claims about its impact. So we’ve built an independent, fully open model to answer three questions: how much will the tax raise, who will pay it, and what will it do to England’s hospitality industry? You can see exactly how the model works, change its assumptions, and test the results for yourself.³

Our central estimate is that a 5% tourist tax across England raises about £600m a year.⁴ We are reasonably confident in the overall revenue estimate: changing the behavioural assumptions makes surprisingly little difference to it. Trips by British tourists generate 63% of receipts in our central case, although the British/foreign split is less certain than the overall total.

The consequences for tourism are much harder to estimate. Our central assumptions imply about six million fewer nights in taxed accommodation and a fall of about £700m in gross visitor spending, of which about £400m is diverted abroad or no longer arrives. The remainder is (we expect) diverted to other spending in the UK. These numbers are much more speculative because they rest upon assumptions that are highly uncertain – how strongly visitors respond, whether they cancel or merely change their trips, and where their spending goes instead. They should therefore be treated as illustrative estimates, not precise forecasts. You can see directly in the model quite how sensitive it is to these assumptions, and substitute your own.

So whilst our model is informed by the considerable body of research into existing tourist taxes in Europe and the United States, we’ve tried to be explicit about every assumption we’ve made. That research shows diverse outcomes, some reflecting differences in the taxes, some reflecting differences in the tourist destinations, and some undoubtedly reflecting the difficulty and uncertainty inherent in this kind of work.

Combining our more cautious and more generous assumptions (reflected as the two illustrative presets in the model) gives receipts between £494m and £977m and lost spending between £46m and £1,550m. These are illustrative scenarios, not upper and lower limits.

Here’s the model – below that is a detailed explanation of why we think this is the correct approach, what the assumptions mean, and why we believe their starting points are appropriate.

These buttons jump to different sections of this report:

The proposed tax

The proposed tax is simple: a percentage of the accommodation bill per night. That’s different from many countries where the tax is a flat charge per night. The idea, of course, is to protect budget holidays.

The innovative element, for the UK at least, is that this is a tax entirely devolved to mayors and other local authorities, with the revenues retained locally.⁵ Whether to apply a tax and the rate of the tax is up to local mayors (or the new “foundation strategic authorities” in places without a mayor, such as the Cotswolds).⁶ There isn’t a cap⁷, and in theory the rate could be anything, but ministers expect it not to exceed 5%.

The tax would cover commercial visitor accommodation, with British and foreign tourists treated alike. Hotels, B&Bs, guesthouses and platform lets such as Airbnb are all in scope. Campsites, touring pitches and hostels can be exempted locally but are not exempt by default.⁸ Non-commercial accommodation would be outside the tax, with registered gypsy and traveller sites, temporary accommodation and charitable shelter given as examples. A threshold is also planned to keep occasional informal providers outside the tax.⁹

Calculating tourist tax revenues

Given that the tourist tax is a percentage of the accommodation bill, it should be simple to calculate the revenues. Just multiply the average rate by the total spent on accommodation by tourists staying in England.

The problem is that you can’t get the answer directly from any published statistics. There’s a domestic survey that asks people what they spent on accommodation and publishes a total, but doesn’t split by accommodation and region. There’s an overseas survey records the whole cost of the trip (without separating-out accommodation cost).¹⁰ The ONS Tourism Satellite Account then estimates spending by product using those surveys and other sources. Our estimate uses the surveys and the accounts and compares the result – but these aren’t independent data sources (and that has consequences we discuss later).

The first approach is what we use in our model – it builds up from the surveys:

• Each type of accommodation is given a share of trip spending (40% for a hotel stay, 45% for a rented cottage or flat, 30% for camping), set so that British visitors’ total accommodation spending reproduces the figure the Great Britain Tourism Survey publishes separately.¹¹ The model lets you change these percentages.
• Overseas visitors are different: nearly half their nights are with friends or family, and shopping is a far bigger part of what they spend, so their shares are scaled to the Tourism Satellite Account, which puts their accommodation at 19.6% of what they spend in the UK.¹²

The second approach is best viewed as a limited sense-check on the first:

• We simply take the Tourism Satellite Account’s own figure for tourism consumption of accommodation.
• Then scale it to England, to 2024 volumes and to today’s prices.
• Then allow for accommodation excluded from the tax, either nationally or by local choice. We deduct 8%, but no source measures the right percentage, so we show a range: at 4% the accounts give £14.4bn and at 12% they give £13.2bn.¹³

Here’s the result:¹⁴

How the liable accommodation bill is estimated	England, 2024 volumes, 2026 prices, incl VAT	Tax at 5% on a common basis
Built up from the visitor surveys (the model)	£14.5bn	£606m
Tourism Satellite Account, scaled to England	£13.2bn to £14.4bn	£548m to £598m

Our survey figure sits between 1% and 10% above the accounts, and above them on any reading of that 8%. That gives some reassurance that our accommodation shares are not wildly wrong, though not much more than that.¹⁵

There are then two further adjustments we have to make. One of those is easy, and the other one is very difficult.

The easy one is VAT. The table above assumes every provider charges it, so every bill is divided by 1.2 before the tax is applied. That isn’t true for many Airbnb hosts and small B&Bs, whose whole bill is taxed, so allowing for them pushes receipts up. Our model lets you vary the share of the accommodation bill that carries VAT, with our central estimate at 85%.¹⁶

The difficult one is how tourists respond to the tax, the “behavioural response”.

If you tax something, you generally get less of it – so if we tax UK tourism, we can expect less UK tourism. But how much less?

There’s plenty of research on existing and historic hotel and short-let taxes, but it’s frustrating and limited in many respects. A complicating factor is that there are two families of study. One asks how many rooms get booked when the room price goes up.¹⁷ The other asks how visits or spending change when the whole cost of a trip goes up.¹⁸ The numbers in the second family look much bigger, but they get applied to a much smaller price change, because a 5% tax on the room is only about 1.5% of the cost of a trip. Our analysis in our model is therefore careful to distinguish the two families.

Our model starts with what we think is a defensible central estimate for the behavioural response, and lets you adjust it. Some evidence suggests local and foreign tourists don’t behave the same way, so the sliders in our model start at different values and let you adjust them separately.¹⁹

It matters a lot less than you might think. The modelled revenue changes only slightly, even with dramatic changes to the behavioural response.²⁰ That’s because the reduction in tax caused by any reduction in visitor stays is much less than the amount that the tax raises. A tax that cuts demand by 2% still collects 98% of what it would have done.

We are, if you like, nowhere near the top of the Laffer curve. But the reduction in visitor stays is a very real effect with very real consequences, discussed below.²¹

The revenue figure changes much more when you change the accommodation share, because a larger share means a larger bill to tax.

How much does the tax raise prices for British tourists?

Our model’s central estimate is that, at a 5% rate, trips by British tourists generate £379m of receipts and trips by overseas tourists £224m.²² That’s 63% from trips by people who live here, which is a lot of domestic money for a tax being sold as a way to “raise money from the 42 million international visitors who came to the UK”.

Asked whether ministers had considered a version aimed only at international visitors, Number 10 refused to say. We expect they didn’t because a discriminatory tax would be problematic from a legal perspective, and burdensome to hotels from an administrative perspective.

Note: we say trips by British tourists are generating 63% of the revenue, not that 63% of the tax comes from British tourists. This is because hotels and other providers may have to cut their prices somewhat, i.e. absorbing some of the tax themselves. The “economic incidence” of the tax is, therefore, shared between tourists and providers. Our model starts with the assumption that three-quarters of the tax is passed on to visitors, but you can change this.

For an individual family the sums are modest, but not nothing. Take a room that would otherwise cost £100 a night, VAT included. If three quarters of a 5% tax is passed on to the customer, that’s £3.75 a night, or about £26 for a week.²³ Whether that’s trivial or objectionable is a political question rather than an economic one.

What’s the impact on hotel businesses?

The impact is going to vary widely between different businesses, but it’s helpful to set out some illustrative examples.

The chart below looks at the impact of the tourist tax (in terms of both reduced income and absorbing some of the cost of the tax itself). It puts this in the context of recent increases in the minimum wage, employer national insurance and business rates. We construct three notional examples: a small B&B, a 20-room hotel and a large hotel chain. We believe they are fairly representative, but that involves subjective judgement on the part of our industry contacts and research team.²⁴

The tourist tax would take about a quarter of starting profit in the two smaller examples and 11% in the chain:²⁵

We would, therefore, hope that the government looks carefully at the impact on the smallest businesses. They’ve been largely protected from other recent changes (thanks to small business rate relief and the Employment Allowance), but face a significant hit to their (limited) profits from the tourism tax.²⁶

It’s a challenging problem to solve. Any small-business relief needs careful design: a cliff-edge threshold could discourage growth and shift the burden onto otherwise similar competitors.²⁷ One answer would be to give each business²⁸ an allowance, with the first £1,000 (say) of the tourist tax it collects retained by the business and not paid across as tax. That avoids a cliff-edge and cushions the impact on small businesses. For some very small operators, it could more than offset the cost. Such an allowance could be expensive: an illustrative calculation puts the annual cost at around £100m if the tax applied throughout England, given the large number of second homes used for holiday lets.²⁹ So it would make sense to restrict the allowance to actual hotels and B&Bs, not second homes/holiday lets.

The question is how to do that. One approach would be to restrict the allowance to properties assessed for business rates rather than council tax. That’s an existing test rather than a new one: a holiday let only enters the rating list if it’s available 140 days and actually let 70 days a year. Most of these properties pay no rates at all because of small business rate relief, but they are on the list, and it’s a good proxy for a commercial operation rather than an occasionally-let second home. It would reduce the cost by somewhere between about a third and a half.³⁰

What’s the economic impact?

In our central scenario, a 5% tax raises £603m after visitors respond, and visitors book 5.8 million fewer nights in taxed accommodation, about 2% of the nights that would carry the tax. The spending that’s lost with those vanished nights is £689m (noting again that this is highly sensitive to the assumptions).

That’s a bigger number than the receipts, but it would be a category error to subtract one from the other and conclude that the tax loses money. The two numbers measure different things.

We think this is a better way to look at it.

Central scenario, 2026 prices	Tax collected	Visitor spending displaced
Total	£603m	£689m
From trips by British tourists	£379m	£402m
From trips by overseas tourists	£224m	£287m
Of the displaced spending, leaves the UK		£387m
Of the displaced spending, stays in the UK		£302m

£287m is our modelled reduction in spending by overseas visitors in England, and we assume all of it is lost to the UK.³¹

£402m is our modelled reduction in spending on trips by British tourists in England. That money is lost to the businesses serving those trips, but some of it will appear elsewhere in the UK, whether in other hospitality businesses or in a different sector entirely. Some of that money will leave the UK as people take a holiday abroad instead. We assume a quarter of this deterred spending goes abroad, which gives £101m.³² Add that to the overseas spending that no longer arrives and the total diverted abroad or no longer arriving is £387m. The remaining £302m is displaced to other spending within Britain.

Given the approximations and assumptions, it’s best to use round numbers here and conclude that the loss to English tourism is around £700 million, of which about £400 million is diverted abroad or no longer arrives. It would be a bad mistake to use the £700m figure as an overall economic cost. The £400m is also gross spending, not a measure of lost GDP or the overall economic cost.³³ And compared to the overall size of the UK economy, about £3.1 trillion, it’s a very small number.

These figures are highly dependent on how responsive visitors are to the price change. Here’s what happens when you use the different price elasticities found in the research:

What the visitor reacts to	Price elasticity used	Tax raised	Visitor spending lost	Of which leaves the UK
The room price (our default)	−0.5 foreign, −0.625 British	£603m	£689m	£387m
The room price	−0.13 foreign, −0.1625 British	£613m	£179m	£101m
The room price	−0.7 foreign, −0.875 British	£598m	£964m	£542m
The whole trip	by purpose of visit	£610m	£426m	£200m
The whole trip	Peng, −1.489 for the destination	£602m	£940m	£460m

Two things stand out from that table. The spending lost swings by a factor of five, from £179m to £964m. But (as we noted above) the tax raised barely moves at all, staying between £598m and £613m, because a tax that cuts demand by 2% still collects 98% of what it would have done. The argument about elasticities is an argument about the damage and not about the money.

Other published numbers

There are some other numbers being discussed in the public debate about the new tourist tax, and we thought it would be helpful to summarise what they are and how they compare with our figures.

UKHospitality/Oxford Economics

Easily the most substantive estimate is by Oxford Economics, commissioned by UKHospitality and published in May 2026.

The Oxford Economics study puts gross receipts from a 5% tax at £1,388m in 2026, rising to £1,595m in 2030, of which domestic visitors contribute £708m rising to £818m.³⁴ That is more than double our £603m.

Oxford Economics’ modelling is more sophisticated than ours. It forecasts tourism to 2030, estimates how overseas visitors respond to prices, and follows spending changes through supply chains. We deliberately do much less. But their report doesn’t disclose enough for outsiders to check all those calculations.³⁵ That’s problematic for any research, but particularly when it’s commissioned by an industry body with a direct interest in the outcome.

The differences between our results and Oxford Economics are substantial:

Measure	Our central case: 2024 volumes, 2026 prices	Oxford Economics: 2026 scenario	Oxford Economics: 2030 scenario
Levy receipts	£603m	£1,388m	£1,595m
Receipts from British tourists’ trips	£379m	£708m	£818m
Fewer accommodation nights	5.78m	10.87m	11.87m
Reduction in visitor spending	£689m	£1,473m	£1,787m
Reduction in GDP	Not estimated	£1,827m	£2,241m

There are some immediate points of difference:

• We assume the levy applies before VAT, following Scotland. Oxford Economics describe theirs as applying “on top of VAT”. If they mean the VAT-inclusive bill, that could account for around £100m of the difference.³⁶
• Oxford Economics’ figures imply that overseas visitors spend about 37% of their money on accommodation, against about 20% in the national accounts. That difference alone could mean hundreds of millions in tax receipts.³⁷ The report doesn’t explain this difference.
• There’s a specific problem with the receipts figure that we can’t figure out. The report’s own baseline puts accommodation spending at £26bn. Five per cent of £26bn is £1.3bn. Yet receipts are said to reach £1,595m by 2030, which needs almost £32bn of spending for the rate to apply to. On the face of it, the tax collected is more than 5% of the thing it’s collected on.³⁸
• Oxford Economics’ separately reported domestic and inbound spending reductions add to £1,872m, while their reported combined result is £1,787m. It’s unclear where the £85m difference comes from.³⁹

The biggest difference appears to be how much accommodation spending they expect the tax to reach. The published information doesn’t let us reconcile their estimate with ours.⁴⁰

It would be helpful if Oxford Economics could clarify these points and set out their methodology in more detail.

There are two further aspects of the Oxford Economics report, and the way UKHospitality has used it, which we don’t think can be justified.

First, the GDP calculation counts the damage from reduced tourism, but leaves out what happens when councils spend the tax receipts and British households spend their unused holiday budgets elsewhere. It therefore cannot tell us whether the tax makes the economy better or worse off.⁴¹ The figures UKHospitality is using in press releases are, therefore not justifiable – there will not be a “£2bn hit to the economy” unless the tax revenues, and diverted spending, all go up in smoke.

Second, UKHospitality’s publicity emphasises £688m of reduced Treasury receipts: a forecast fall in other taxes as tourism activity declines. It also reports £1.6bn of new tax on holidaymakers, but leaves out the combined result. Its own model predicts £907m more revenue across central and local government in 2030.⁴²

Other published figures

There are other figures being mentioned which we think are much less reliable than UKHospitality’s:

• The Taxpayers’ Alliance claims that a £10 per night tax could cost the economy over £13bn and deter 42 million domestic overnight visits. It gets there by applying opinion poll responses to visitor numbers and spending. As we said recently, asking people whether a tax would stop them booking a holiday is a poor guide to what they will actually do.⁴³ None of the studies cited here finds a tax-induced collapse on that scale.⁴⁴ We don’t think the polling supports the claimed economic loss.
• PwC carried out a study for the European Commission in 2017 which modelled a hypothetical 3% occupancy tax in every member state that didn’t have one. It projected 23,140 lost jobs in the UK, the largest absolute job-loss figure in the study. They obtain the figure by applying the percentage fall in producer revenue to employment.⁴⁵ Some degree of approximation is inevitable, and we are certainly guilty of it ourselves, but assuming employment falls in exact proportion to revenue takes it too far: businesses can adjust through hours, vacancies and margins as well as headcount. More seriously, money lost to the sector is not disappearing. Most of it will remain in the UK economy, but go elsewhere, either as redirected tourist spending or as new local authority spending, financed by the tax. Any calculation of the jobs impact needs to take account of new jobs created as well as old jobs destroyed. We therefore don’t think PwC’s number should be used.
• The Times, in an explainer for readers, said: “There are more than 130 million overnight stays in England each year, all of which would be liable for the charge when it is introduced”. This appears to be an error.⁴⁶ England had over 500 million overnight nights in 2024, not 130 million: 249.9m from overseas visitors and 250.6m from British residents.⁴⁷ Not all of these would be taxed: nearly 200 million of those nights are spent with friends and family, or in a second home. Our assumed scope leaves about 268 million taxable nights.

We’d very much welcome additional analysis from independent researchers.

Will we see tax competition between regions?

This is an unusual tax for the UK because it is genuinely being devolved. Neighbouring authorities are free choose different rates or decide against a tourist tax altogether.

So we won’t see one tax at all. We’d see a separate tourist tax for every strategic authority that chooses to charge one, each set by local leaders facing a different tourist economy and a different set of incentives.⁴⁸ Each tax may have different exclusions. In theory, at least, each tax could have a different rate. Some will charge 5%, but some may charge nothing at all, either because they think it will hurt them or because they’d rather their neighbours went first.

This creates the potential for tax competition, which some view as dangerous and others as positive. Early indications are that we will see coordination rather than competition, but that may change. Many localities already compete, particularly when trying to attract large conferences.

The tourist tax doesn’t apply if you visit Manchester. It applies if you sleep in Manchester. So somewhere just outside Manchester (Wilmslow or Alderley Edge, perhaps) might decide to grab a bit more tourist revenue and poach Manchester’s hotel nights. Bradford-on-Avon may see an opportunity to poach hotel stays from Bath.

One can identify many such examples. The Peak District is split down the middle: the Derbyshire side, with Bakewell and Buxton, sits inside a mayoral authority, and the Staffordshire Moorlands side doesn’t. And London is ringed by towns that are a short train ride from a mainline terminus and outside the mayor’s boundary: Watford, Slough, Windsor, Dartford.

However, with one exception, which we’ll mention below, we see the actual potential for competition here as limited. There was already a big price difference between central hotels and hotels on the fringe. The tourist tax will just slightly exacerbate competitive pressure that already exists.

There is some evidence from other countries of small competitive effects between different regions. Along the Italian coasts, strings of resorts sit next to each other and set their own rates, and there is evidence of “substitution effects” towards resorts with lower taxes, though only on one coast.⁴⁹ A similar effect has been found in Texas.⁵⁰ This has also been used to explain why the responsiveness of tourists towards Georgia’s tourist tax is higher than for Hawaii’s tourist tax, because travellers to Georgia can sleep in another state, and travellers to Hawaii usually can’t.

However, larger studies have found very few effects of this type. Californian cities set their own rates and hotels sit on either side of city lines, making for a neat natural experiment. There was, however, no significant effect found.⁵¹

Some of the difference in findings may be down to how visible taxes are.⁵² Some of the US research concerns booking systems that showed pretax prices, with the tax appearing only at checkout. A tax shown late in the booking process may have less influence on the initial choice.⁵³.

So, how much tax competition becomes a factor may depend on how visible the tax is. But, unlike in many countries, any UK tax would have to be visible when people book. Tax competition may, therefore, be more of a factor in the UK than elsewhere.

Evidence from published research

There’s a considerable body of published research on the impact of tourist taxes.

What the evidence doesn’t tell us

Tourist taxes are small, and their expected effects will therefore be small. Statistical studies are generally not good at picking up small tax-driven effects, particularly when they’re across complex visitor economies that are subject to significant seasonal and annual variation.⁵⁴

A small percentage change can mean a substantial sum across a £57bn visitor economy, while being hard to separate from the effects of weather, exchange rates and other changes in hotel trading.

For example, treating Manchester’s £1 charge as roughly 1% of a room bill, our central assumptions and the approach in our model imply a fall in nights of about 0.4%. An effect that small may be difficult to detect in monthly data.

A study finding no statistically significant effect therefore does not, by itself, confirm or disprove our estimate. We would need to compare the size of the predicted effect with the study’s confidence interval.

The studies, therefore, do not establish a single, reliable estimate for England – but we can use them to inform the model’s inputs and to check its predictions.

What the evidence tells us

As a broad principle, we would say that the studies we reviewed generally find modest effects on hotel trade, sometimes indistinguishable from zero.

That is true of Manchester’s £1 charge, the only English study we have located and a smaller charge than a 5% tax on a typical hotel room,⁵⁵ of Hawaii’s 5% room tax,⁵⁶ of a panel of Italian municipalities,⁵⁷ of the Balearic sustainable tourism tax,⁵⁸ and of eight American destinations.⁵⁹ Swenson, using establishment-level data across Californian cities with different rates (the best identification in the American literature), is blunt: “Hotel occupancy tax rates in California cities had no measurable impact on hotel revenues or employment.”⁶⁰

One of the most striking findings was in a study funded by the American Hotel & Lodging Educational Foundation. Hudson and colleagues concluded: “Hotels appear to have absorbed any tax increases with little impact to their businesses, but there was concern among stakeholders as to how the lodging tax was spent.”⁶¹

As one might expect, things are different for group bookings, such as conferences, which are both highly organised and highly price-competitive. Sharma, Perdue and Nicolau, looking at more than 100 American urban submarkets, found that the effect on revenue per available room “was non-significant for transient but significant and negative for groups”, suggesting group bookings are more responsive to tax differences than individual bookings.⁶² So if anything holds English mayors back from introducing tourist taxes, it could be the conference trade, and the rules won’t let them exempt it.

There’s also research on “incidence”, the question of who absorbs the cost of the tax: visitors or accommodation providers. The two cleanest incidence studies, on a Georgia room tax and on Airbnb tax enforcement across American cities, both find hoteliers and hosts absorbing roughly a quarter to a third of the tax.⁶³ Our model uses 75%, near the middle of that range. So the simple take, that the tax is paid by tourists, isn’t right. Some of it is paid by the accommodation sector.

The Scottish Government’s 2021 evidence review found no direct evidence for Scotland.⁶⁴ Evidence on accommodation taxes also points to differences between destinations.⁶⁵

There’s also a strand of theory arguing that taxing tourism is unusually efficient, because foreigners pay it. ⁶⁶ But England’s tourist tax applies to everyone, not just foreigners, so it doesn’t get that benefit.

So if studies find only very limited effects, why does our model predict an effect?

Our model predicts that a 5% tax cuts nights in taxed accommodation by 2.2%, and total visitor spending in England by 1.2%.⁶⁷ Some studies may struggle to detect effects of that size. We cannot assume that all would.⁶⁸

The paradox is that a fall of around 1.2% in £57bn of visitor spending is nearly £700m, which is not such a small number.

We think our model and its assumptions are justified by the research, and we set out why in the methodology below. But we could easily be wrong.

Check back here in 2032.

Caveats and limitations

• The accommodation share is the biggest assumption. The domestic survey records accommodation spending, but the published breakdown does not give us everything needed to allocate it across accommodation types and regions. The overseas data used here records trip spending. For British visitors we calibrate the split so that total English accommodation spending falls within a range based on the figures the Great Britain Tourism Survey publishes separately (£6,657m of direct accommodation spending, plus part of £4,994m of packages, both 2024 figures), and the model re-checks that on every run. For overseas visitors we anchor to the ONS Tourism Satellite Account. But no source gives the split by accommodation type directly.
• Several assumptions have no direct measurement behind them: the share of accommodation bills that carry VAT (85%), the share of lost accommodation demand counted as lost trip spending (three quarters), the share of deterred British spending that goes abroad (a quarter), the share of deterred overseas spending retained elsewhere in the UK (zero), and the share of the national accounts’ accommodation total that the tax doesn’t reach (8%, which is what sets the 1% to 10% gap between our two estimates of the base). We think our central choices are reasonable, but they are judgments, and you can change them in the model. The two presets combine our more cautious and more generous settings. They illustrate how far the answer moves; they aren’t confidence limits.
• Our estimates use 2024 visitor numbers and spending uprated to 2026 prices by CPI. We do not forecast the visitor numbers or prices for the first year of the tax.
• 63% of receipts are from trips by British tourists in the central case. That is a share of receipts associated with trips, not the share of the burden borne by those tourists. Changing only the overseas accommodation assumption to its upper setting gives an overseas majority of about 51%.
• Our headline splits lost spending into money that leaves the UK and money that doesn’t. That rests on assuming a quarter of deterred British spending goes abroad and that no deterred foreign visitor switches to Scotland or Wales. Set the British share to zero and £287m leaves rather than £387m. Set it to zero and also let 30% of deterred overseas visitors go elsewhere in the UK and it’s £201m. Assume half of deterred British spending goes abroad and it’s £488m. You can change both.
• Our response assumptions draw largely on evidence from other countries and periods. The English study of Manchester’s £1 charge found no significant effect, but that charge is smaller than 5% on a typical hotel room. Evidence from Edinburgh’s 5% levy should provide a closer UK comparison.
• We assume the tax is charged at the same rate everywhere and that the response to it is linear in the rate. Neither is likely to be exactly true, and things would be very different if, for example, a locality applies a 10% rate. And our overall figure will obviously be an overstatement if some localities don’t apply the tax or apply it at lower rates.
• The web app splits both British and overseas visitors across nine English regions. These are not the strategic authorities that will set the tax.
• Regional overseas estimates use smaller survey samples than the England total, so the regional results are less reliable. Allocating visitors to individual strategic authorities would require further work.
• Domestic trips that the survey cannot place in a region (national parks and unspecified destinations, 5.1 million nights) are left out. Including them adds about £8m, so it’s not terribly material.
• Northern Ireland is outside the scope of this model, which estimates an English tax.
• We use 2024, the latest full year with published results for both surveys. Both sets of estimates are still classed as “official statistics in development” and may be revised.⁶⁹ A fault in the domestic survey may have understated overnight trips, but we don’t yet know what correcting the historical figures would mean for spending in taxable accommodation.⁷⁰ Our choice of 2024 doesn’t mean that our estimate is conservative.⁷¹

Methodology

We’ve published all the code implementing the model and the web app on our GitHub.

The model is based on these key parameters:

Assumption	Central value	Status	Basis	What it moves
Tax rate	5%	Policy	Modelled rate; the government proposes local discretion with no national cap	Everything, roughly in proportion
Accommodation share of British visitors’ trip spending	40% hotels, 45% rentals, 30% camping	Calibrated	Chosen to fall within a range based on the GBTS published accommodation spending figures (£8.3bn to £10.4bn)	Receipts: −£48m at the bottom of the band, +£76m at the top
Accommodation share of overseas visitors’ spending	19.6% of UK spend	National accounts estimate	ONS Tourism Satellite Account, Table 1	Receipts: −£17m at the low end, +£121m if British visitors’ shares are used instead, or +£176m at the model’s upper setting. Also decides whether British or overseas trips generate the majority
Share of accommodation bills that carry VAT	85%	Judgment	Hotels are registered; many Airbnb hosts and small B&Bs sit under the £90,000 threshold. Nothing measures the split	Receipts: −£18m at 100%, +£29m at 60%
Share of the tax passed on to the customer	75%	Judgment informed by evidence	Incidence studies suggest roughly 70% to 86% (Georgia, Airbnb, Hiemstra and Ismail). PwC assumes 60%	Receipts barely; spending lost £551m to £790m
Room-price elasticity, overseas visitors	−0.5	Evidence range	Published hotel and short-let demand estimates, −0.13 to −0.7	Receipts £598m to £613m; spending lost £179m to £964m, varying both elasticities and keeping British visitors 25% more responsive
Room-price elasticity, British visitors	−0.625	Judgment on direction	Set 25% above overseas, after Biagi and colleagues found domestic demand fell and international did not	Spending lost −£80m if set equal to overseas
Share of lost accommodation demand counted as lost trip spending	75%	Judgment	Lost paid nights need not mean lost trips: visitors may shorten their stay or switch accommodation. There is no direct measurement of this share	Spending lost £230m at 25%, £918m at 100%; receipts unchanged
Share of deterred British spending that goes abroad	25%	Judgment	Nothing measures it. DCMS’s UK model shows domestic and foreign holidays are substitutes (substitute-price elasticities 0.84 to 2.83) but not what share goes abroad. Oxford Economics assume 3%, also a judgment	Leaves the UK: £287m at 0%, £299m at 3%, £488m at 50%
Share of deterred overseas visitors who go elsewhere in the UK	0%	Judgment	We assume no substitution to elsewhere in the UK. There is no direct measurement	Leaves the UK: £301m at 30%
Campsites and touring pitches	Taxed	Policy reading	In scope unless an authority exempts them (government response, 4.3)	Receipts −£27m if exempted everywhere
Hostels, halls of residence, boats	Not taxed	Scope assumption	We exclude this mixed survey category. Commercial stays in halls or boats would not automatically be exempt	Receipts +£28m if all taxed
Domestic trips with no recorded region	Excluded	Editorial	5.1m nights the survey can’t place in an authority	Receipts +£8m
Share of the national accounts’ accommodation total the tax doesn’t reach	8%	Judgment	Used only in the cross-check; nothing measures it	Whether the survey figure sits 1% or 10% above the accounts
Prices	2026, uprated from 2024 by CPI	Measured	ONS CPI 141.5 against 133.9, a factor of 1.0568	Our model’s £ figures by 5.7%; volumes unchanged
Visitor volumes	2024	Measured	Latest complete year of both surveys; not projected to 2028	First-year receipts depend on prices, visitor numbers and local adoption

The methodology then works like this.

1. The data

We assemble the critical data sources like this:

• Overseas visitors come from the ONS International Passenger Survey for 2024, in the regional release published by VisitBritain. British visitors come from the 2024 Great Britain Tourism Survey trip-level file, using its England-weighted columns, so that Scottish and Welsh residents visiting England are counted and English residents’ trips to Scotland are not.
• Both are reduced to a grid by origin, journey purpose and accommodation type, holding nights, trips and total trip spending. The Python model’s default grid splits domestic trips across the nine English regions and uses a single England total for overseas trips: 195 rows. The web app splits both across the nine regions. Its data file has 363 rows, including domestic trips with no recorded region, which are excluded from results by default. The nine English regions of the IPS reconcile to its England total to within 0.002% on nights and on spending, so no cell is counted twice.⁷²
• Recorded 2024 spending is £27,114m inbound and £26,806m domestic. Every figure from here on is multiplied by 1.0568, which is ONS CPI for 2026 so far (141.5) against 2024 (133.9), so those become £28,652m and £28,328m in 2026 prices. Nights and trips aren’t uprated, because uprating is about the value of money and not about how many people came. Totals: 249.9m inbound nights and 250.6m domestic, 500.5m in all.⁷³

2. What the tax reaches

Accommodation is grouped into five types.

• Three are liable: serviced accommodation (hotels, guest houses, B&Bs), commercial rentals (self-catering and platform lets) and camping or caravanning.
• Two are not: someone’s private home, whether a friend’s or a second home, because there is no commercial bill;
• The residual “other” group, covering hostels, university halls and boats. We exclude this mixed category as a modelling assumption. The government’s response permits local hostel exemptions, but does not automatically exclude commercial visitor stays in halls or on boats.

That leaves 267.7m liable nights out of 500.5m.

3. From trip spending to an accommodation bill

We need accommodation shares to turn the published trip spending into the breakdown required by the model. GBTS provides an accommodation total, but not the full breakdown we need by accommodation type and region. For each row:

bill = trip spending × accommodation share of that type × (0.649 if the visitor is from overseas)

The shares are 40% for serviced accommodation, 45% for commercial rentals, 30% for camping, 25% for the residual and 2% for a private home.

For British visitors they are set so that total English domestic accommodation spending reproduces the figure GBTS publishes separately: £6,657m of direct accommodation spending plus £4,994m of package spending, of which some part is accommodation. Taking a third to three quarters of the package gives a band of £8,305m to £10,402m, and the shares above produce £9,234m, inside it.

Overseas visitors get a lower share, because their spending pattern is different: surprisingly (at least to us) nearly half their nights are with friends or family and shopping is a far larger part of what they spend. The ONS UK Tourism Satellite Account 2023 puts their accommodation at £6,131m of £31,203m spent in the UK excluding air fares, or 19.6%. Scaled to England’s 85.8% share of inbound spending and to 2024, that is £5,813m, and 0.649 is the multiplier on the domestic shares that reproduces it (£5,813m in the model).

Across all accommodation types this gives a £15,368m accommodation bill, of which £14,540m is liable to be taxed.

The national-accounts anchor described above puts the liable bill between £13,157m and £14,353m (depending on that pesky 8% assumption).

4. The tax

We assume the tax is charged on the price before VAT and then itself attracts VAT.⁷⁴

Survey spending is what the visitor actually paid, so it is VAT-inclusive, and the bill is divided by 1.2 for the share of it that carries VAT, which we take to be 85%.⁷⁵

Providers who absorb part of the tax do so by cutting their own price, so with pass-through p and rate t the pre-tax price falls to (1 + p × t) ÷ (1 + t) of what it was: 0.9881 at the defaults.

The taxable base is £12,332m, and at 5% the tax raises £616.6m before anyone changes their plans.

5. How visitors respond

Three steps: work out the price rise a visitor sees, multiply by an elasticity to get the change in demand, then apply that change. The model has two complete specifications and never mixes them.

The hotel-bill reading (default). Not all of the tax reaches the visitor, because hotels and hosts absorb some of it. We assume 75% is passed on. The incidence studies suggest roughly 70% to 86%; PwC uses a lower assumption.⁷⁶ Salience is set to 1, meaning visitors see the whole of what is passed on. The price rise is therefore rate × 0.75 = 3.75%. The elasticity is a room-price elasticity, −0.5, between the four published estimates of −0.13, −0.44, −0.48 and −0.7⁷⁷, and demand is measured in nights. Each row’s elasticity is multiplied by a responsiveness factor for where the visitor lives: 1.0 for overseas visitors and 1.25 for British ones, because the study we found separating the two, Biagi, Brandano and Pulina, found domestic demand fell after a visitor tax while international demand did not. Weighted across liable nights this gives an average elasticity of −0.576 (−0.625 British, −0.5 overseas) and a fall in demand of 2.16% (2.34% British, 1.88% overseas). Because a lost night in a hotel is not always a lost trip, only three quarters of the lost accommodation demand is counted as lost trip spending.⁷⁸

The whole-trip reading (illustrative). The price rise is diluted by the accommodation share of what the visitor spends in the destination, so 1.50% for a British visitor and 0.99% for an overseas one across the rows that carry the tax, and we apply whole-trip demand elasticities as proxies for changes in visits: either the purpose-level set (business −0.35 and visiting friends and relatives −0.80 from Peng, other −0.61 and rural holiday −1.23 assigned using Blake and Cortes-Jiménez for DCMS, 2007, urban holiday −0.70 blending PwC with DCMS, the holiday figure blended by each region’s share of holiday nights spent camping) or Peng’s −1.489 for a destination as a whole. Every lost visit is a lost trip, so the three-quarters factor above does not apply.⁷⁹

6. The outputs

Tax raised = taxable base × rate × (1 + change in demand), row by row = £603.2m

Nights lost = liable nights × proportionate reduction in demand = 5.78m

Visitor spending lost = total trip spending on liable rows × proportionate reduction in demand × 0.75 = £688.8m, of which £286.6m is overseas visitors and £402.3m is British

Leaves the UK = overseas spending lost + British spending lost × 0.25 = £387.1m

Spent elsewhere in the UK = British spending lost × 0.75 = £301.7m

7. The presets

The “cautious” preset is an illustrative lower-revenue case: accommodation shares scaled by 0.92, inbound multiplier 0.60, all bills VAT-registered, pass-through 86%, room elasticities of −0.70 inbound and −1.05 British, every unit of lost accommodation demand translated into lost trip spending, half of deterred British spending going abroad and no inbound spending retained elsewhere in the UK. Camping is exempt. It gives £494m raised and £1,550m of gross spending lost.

The “generous” preset is an illustrative higher-revenue case: shares scaled by 1.1265, inbound multiplier 1.16, 60% of bills VAT-registered, pass-through 60%, both room elasticities −0.13, a quarter of lost accommodation demand translated into lost trip spending, no British spending diverted abroad and 30% of deterred inbound spending retained elsewhere in the UK. Hostels and halls are in scope. It gives £977m raised and £46m lost.

---

All the code implementing the model and our web app is available on our GitHub.

Many thanks to G for the original model and code, and B, V and O for their industry expertise that helped build it. Thanks to K and C for working on this month’s update to the model and on the draft of this report. Thanks to B for their review.

ONS International Territorial Level 1, January 2025, © Crown copyright and database right 2025, under the Open Government Licence v3.0. Contains OS data © Crown copyright and database right 2025.

And thanks to the authors of Apache ECharts and the other open source tools we rely upon so heavily.

Further reading

What would a land value tax actually do?

I just wrote in the Sunday Times about the political and economic case for a land value tax. I think the principled argument for…

What if Andy Burnham lowered the mansion tax threshold to £1.5m?

The November 2025 Budget introduced a “mansion tax” – formally, the “high value council tax surcharge” – an annual charge on homes worth £2m…

Has Britain run out of “other people” to tax?

British politics has spent a generation avoiding the same uncomfortable question: what if the government people want costs more than most voters are willing…

The UK has 90 different taxes. France has 348. Germany has 60. Why?

The UK has 90 different taxes – more than at any time since 1834. France, on the other hand, has 348. But Germany, which…

Why does Britain have more taxes than at any time since 1834?

Tax in the UK is the highest it’s been since 1945. That’s well understood – but another trend has been largely missed: the number…

90 UK taxes. On one chart.

The UK has 90 taxes. This chart shows them all, and this article explains why we have so many. Update 29 May 2026: the…