A gambling duty is a tax on what the operator keeps, not on what the player spends. That sounds like a technicality for accountants. It is actually the entire mechanism behind the gambling tax impact on casinos, because a duty charged on gaming yield is deducted before rent, wages and electricity are paid, and a venue only ever closes on what is left after all four.
The popular answer to “why do tax rises shut casinos?” is that casinos cannot afford the tax. Right conclusion, wrong reason. Rank Group’s most recent quarter makes the point neatly: group revenue grew five per cent to £216.1m, with online gaming up six per cent and land-based up five per cent, and the company still closed three casinos. Demand was not the problem. Company-level profitability was not the problem. The condition almost nobody checks is the one that matters: how much margin an individual site has left over its fixed costs, and how many percentage points of duty it takes to erase it.
Three closed venues and a group that grew
Rank closed three venues in the last quarter, in Swansea, Reading Central and Sunderland, and has warned that a further increase in Machine Games Duty would lead to more closures. Rank operates the Grosvenor Casino and Mecca bingo brands, so the warning covers a sizeable share of Britain’s regulated land-based estate rather than one isolated site.
Chief Executive Richard Harris framed the chain of consequences directly: any increase to Machine Games Duty “will increase costs, reduce venue viability, lead to venue closures and job losses, reduce tax receipts and halt investment.” He added that HM Treasury recognised these economic arguments last year, and that with the right conditions the company believes its casinos and bingo clubs can contribute to the Government’s growth agenda. He also noted that speculation about the upcoming Budget continues to weigh on investor sentiment, which is its own cost: capital allocation decisions get deferred while the rate is unknown.
Read the two facts together. Revenue rising, venues closing. That is the signature of a cost problem concentrated at the site level, not a trading problem across the business.
Machine Games Duty, defined
Machine Games Duty is a UK excise duty charged on the net takings of gaming machines, meaning the money players put in minus the winnings paid out. The operator of the premises is liable, not the player and not the machine supplier. HMRC sets the rates in bands according to the cost to play and the prize available, so the applicable rate depends on the machine category.
Two features of that definition do most of the work. First, the base is net takings rather than turnover, so the duty scales with the margin the machines actually generate. Second, it is a cost of occupancy as much as a cost of gaming: a venue pays it on the machines sitting on its floor, whatever else is happening in the business that quarter.
How MGD sits alongside the other UK gambling duties
A British casino does not pay one gambling tax. It pays Machine Games Duty on its gaming machines and Gaming Duty on the gross gaming yield from table games, the latter charged through a banded structure in which higher yield attracts higher marginal rates. Remote operators pay Remote Gaming Duty on their own gaming profits instead. Current rates and band thresholds for each are published by HMRC and change at Budgets, which is precisely why pre-Budget speculation is treated as a live operational risk rather than a news story.
The practical consequence is that a land-based casino’s tax exposure is split across machine income and table income, and a change to either rate lands on the same fixed cost base.
Duty comes off the top, not the bottom
Gambling duty is not a tax on profit. It is deducted from gross gaming yield before rent, salaries, utilities, compliance staffing and licensing are paid, which means a rate increase does not reduce profit proportionally. It reduces the pool of money available to cover costs that do not move.
Corporation tax takes a share of whatever is left at the end. Duty takes its share first, and takes it whether or not anything is left. For a venue running at a thin positive contribution, those are completely different animals.
An illustrative worked example
The figures below are hypothetical and chosen for arithmetic clarity, not drawn from Rank or any other operator. Take a venue with £4m of annual machine net takings and £150,000 of annual operating profit after all costs, including its current duty bill. Now apply duty increases expressed in percentage points of those net takings.
| Increase in duty rate | Extra annual duty cost | Resulting operating profit |
|---|---|---|
| +1 percentage point | £40,000 | £110,000 |
| +3 percentage points | £120,000 | £30,000 |
| +5 percentage points | £200,000 | Loss of £50,000 |
Nothing about the venue has changed. Footfall is identical, the games are identical, the staff rota is identical. A five point move on machine yield has turned a modestly profitable site into a loss-maker, and the only levers left are cutting hours, cutting machines, cutting headcount or cutting the lease. The first three reduce the revenue the duty is charged on, which is why operators describe the response as a spiral rather than an adjustment.
Scale that logic across an estate and you get Rank’s arithmetic. A group with growing revenue can still hold a tail of sites whose contribution sits within a few percentage points of zero. Raise the rate and the tail is no longer marginal; it is terminal.
Why a physical floor is harder to defend than a website
Land-based venues carry per-site fixed costs that online operators simply do not have: a lease or freehold, utilities on a large building, security and floor staff on fixed rotas, cash handling, premises licensing and local authority compliance for every single location. Those costs are almost entirely insensitive to revenue in the short term.
An online gambling operator’s cost base behaves differently. Its largest variable costs, marketing and payment processing, flex with activity, and its fixed costs sit in technology and licensing that are shared across the whole customer base rather than duplicated per location. When duty rises, a remote operator can throttle acquisition spend and protect margin. A casino with a twenty year lease on a city centre building cannot throttle rent.
That asymmetry explains the sequencing of closures under any duty increase. Physical venues with high fixed costs and thin contribution fail first, regardless of whether the tax change was aimed at them. Casino taxation designed to capture value from a growing market lands hardest on the part of the market least able to absorb it.
The Budget debate, and the same argument elsewhere
The immediate context is the UK Budget, where a gambling duty increase has been under active discussion and where Rank says investor sentiment is already being affected by the uncertainty. Harris’s reference to HM Treasury recognising the economic arguments last year points to the real question in front of officials: whether a higher rate raises more revenue, or raises less because the base shrinks as venues close.
That is an empirical question, not an ideological one, and it does not have a universal answer. It depends on how many sites in a market sit near breakeven, how concentrated the estate is, and how quickly closures feed through to employment and local spending. Operators have an obvious interest in forecasting the worst case. Treasuries have an obvious interest in forecasting the best. Neither forecast settles it.
The structure of the argument is not peculiarly British. Any jurisdiction with a mature land-based sector and a growing remote sector faces the same design problem: a single rate applied across both captures very different economics, and tax rules written for machine halls do not map cleanly onto a market where most growth happens on phones.
The trade-off, stated honestly
Closures carry real costs. Jobs disappear, a building goes dark on a high street that may already have vacancies, and the local economic contribution and tax receipts from that site go to zero rather than falling slightly. Duty receipts also fund public spending, and a government weighing a rate increase is weighing those receipts against those closures. Both sides of that are genuine, and the honest position is that the net effect depends on the distribution of venue-level margins, which outsiders cannot see.
What operators can see, and should be modelling, is which of their own sites sit inside that danger band. The useful question is not whether a duty rise is affordable at group level. It is how many percentage points of additional duty each individual site can absorb before its contribution turns negative, and what happens to the estate at each step. Rank has effectively published that answer for three of its venues. The rest of the sector has the same sum to do, and until the Budget lands, nobody knows which number to run it with.