Friday, 11 July 2025

HMRC greatly underestimates the black market in tobacco

Since writing about HMRC's implausible claim that only 6.9% of cigarettes in the UK were 'non-duty paid' in 2022/23, they have produced new figures claiming that the figure for 2023/24 was 10.5% (oddly, the previous year's number has been upgraded to 9.1%). I have been trying to think of any scenario in which this could be true. It is not just that the estimate doesn't not match what I'm seeing on the street, but that it is mathematically impossible under any reasonable set of assumptions.
 
Put simply, HMRC estimates the amount of non-duty paid tobacco being sold by estimating the total consumption of tobacco (number of smokers x average annual cigarette consumption per smoker) and subtracting the amount of legal tobacco sold (via tax receipts). 

HMRC's tobacco bulletin shows that the number of cigarettes sold legally fell from 23.6 billion in 2021 to 13.2 billion in 2024, a decline of 44.4%

The ONS's Annual Population Survey says that the number of smokers in the UK fell from 6.6 million in 2021 to 6 million in 2023, a decline of 9%. ONS figures are not yet available for 2024, but the Smoking Toolkit Study reports that smoking prevalence in England fell from 14.7% in 2021 to 14.2% in 2024, a decline of less than 4% (accounting for population growth the decline in the number of smokers is even smaller). Quite clearly, there has not been a 44.4% decline in the number of smokers since 2021, nor anything close to it. 

Are smokers consuming fewer cigarettes? Apparently not. On the contrary, the ONS's Adult Smoking Habits in Great Britain survey finds that average daily cigarette consumption per smoker rose from 9.8 in 2021 to 11.3 in 2023. (The figure for 2021 is a modelled predicted estimate, but the 2023 is nonetheless higher than in any year since 2016.) An academic study published last year found that the decline in average cigarette consumption per smoker has plateaued since 2019. 

Have smokers switched to hand-rolling tobacco? Perhaps, but the tobacco bulletin also shows a decline in the legal sale of rolling tobacco, from 8.6 million kilograms in 2021 to 4.5 million kilograms in 2024, a fall of 47.6%. Overall, using the conventional estimate of how many cigarettes are made from a kilogram of loose tobacco, the number of cigarettes bought on the legal market in the UK fell by 45.5% between 2021 and 2024. 

HMRC also produces a tax gap estimate for hand-rolling tobacco. It is higher than the cigarette tax gap estimate at 22.9% but, implausibly, the figure for 2023/24 is the lowest estimate on record. Its estimate for the tobacco tax gap overall is 13.8%, fractionally higher than in 2022/23 but still one of the lowest on record. 

None of this stands up to basic arithmetic. We have a 45.5% decline in the quantity of duty-paid cigarettes sold (including hand-rolled cigarettes), but no decline in the number of cigarettes consumed per smoker and an overall decline in the number of smokers that is vastly smaller than 45.5%. 

It is a mathematical impossibility than only 10.5% of cigarettes consumed in the UK are non-duty paid. Even if there was no black market for tobacco in 2021 - an impossible proposition - it would now be much bigger than that now. 

For the sake of argument, let's assume that the next ONS smoking prevalence survey shows a 15% reduction in the number of smokers between 2021 and 2024 (although there is no reason to think that the decline is that large). Given what we know about daily cigarette consumption, this should translate into a 15% reduction in duty-paid tobacco sales if the non-duty paid market stays the same. This would mean legal sales of manufactured cigarettes fell from 23.6 billion in 2021 to 20.1 billion in 2024. In reality, only 13.6 billion cigarettes were sold in 2024, a shortfall of 6.5 billion cigarettes that can only have come from the non-duty paid sector. 6.5 billion equates to 32% of the total and that is with the profoundly unrealistic assumption that no non-duty paid cigarettes were sold in 2021. 

In other words, if there was no non-duty paid market in 2021, non-duty paid cigarettes would make up 32% of the market in 2024. If we assume, as HMRC does, that 8.8% of the market was non-duty paid in 2021, total cigarette sales (including non-duty paid) would have been 25.9 billion. If cigarette consumption then fell by 15%, the total market would be 22 billion in 2024, but we know that legal sales were 13.2 billion in 2024 so the non-duty paid market must therefore be 40% of the total.  A similar calculation can be done with hand-rolling tobacco and will produce a similar finding (since the decline in legal sales is almost identical). 

One can play around with these figures as much as one likes using different assumptions, but in no reasonable scenario do non-duty paid cigarettes make up 10.5% of the market, nor does non-duty paid tobacco overall make up 13.8%. It seems to me that HMRC's model is broken.



No comments: