HMRC put the latest
tobacco tax gap estimates out on Tuesday. As with the previous few years, they defy both belief and mathematics. Only 14% of tobacco is illicit, they claim, and illicit hand-rolling tobacco is at an all time low!
Simple maths shows that HMRC’s
figures cannot possibly be accurate. Converting kilograms of rolling
tobacco into sticks, the equivalent of 36.6 billion cigarettes were sold
legally in the UK in 2022. If this declined at the same rate as smoking
prevalence (-17 per cent), it would have been 30.4 billion by 2024. In
fact, there were only 22.9 billion cigarettes sold legally in 2024, a
shortfall of 7.5 billion.
If we imagine, for the sake of
argument, that there was no illicit tobacco at all in 2022, this would
mean that 25 per cent of the total market was illicit by 2024/25 (7.5
billion is 25 per cent of 30.4 billion). If you believe, as HMRC does,
that 13 per cent of the market was illicit in 2022/23, it means that 33
per cent of the market was illicit in 2024/25. If HMRC underestimated
the size of the black market in 2022/23, the real figure is higher
still, but the absolute minimum it could be in a wildly optimistic and
frankly impossible scenario is 25 per cent.
The calculation above is based on all other things being equal. HMRC always adjust the figures, but they must have really adjusted them in the last few years. They need to explain what they have done and why. Since 2020, the tobacco tax gap has been little more than a guess.
The only piece of solid data is the
legal sales figure. We have survey data for the number of smokers and
the number of cigarettes they consume, but the ONS survey HMRC relies on
had its status as an official accredited statistic withdrawn
in 2024 due to “ongoing challenges with the response rates, levels, and
weighting approach”. Moreover, smoking prevalence and cigarette
consumption figures always need to be adjusted upwards to account for
shy smokers and under-reporting. This requires a certain amount of
guesswork. Estimates of how much duty-free tobacco is brought into the
UK are also based on surveys and may be inaccurate.
A further problem is that during the
pandemic, the ONS stopped asking people how much they smoked. HMRC
therefore “imputed” the figures for 2020/21, 2021/22 and 2022/23. The
question was asked again in 2023/24 and so HMRC describes its estimate
for that year as “actual”, but it has gone back to guessing for 2024/25.
Blaming “modifications to question sequencing and funding limitations”
it says that “certain consumption data for 2024 to 2025 are incomplete”.
It has therefore “projected consumption data” for 2024/25 “based on
established historical trends.” But what we are seeing today is not in
line with historical trends. That is the whole point!
One point that I didn't make in the article is that the ONS's estimates of how many cigarettes smokers consume has gone up, not down, since 2022. Since 2019, the daily cigarette consumption averages are:
2019: 10.6
2020: 9.3
2021: 9.8
2022: 10.5
2023: 11.3
2024: 11.2
The figures for 2020 and 2021 are guesses. The survey question wasn't asked in those years so the ONS used "modelled predicted estimates". They seem suspiciously low. Since the question was asked again in 2022, the daily consumption estimate has risen from 10.5 cigarettes to 11.2 cigarettes.
That might not seem much but it represents a 7% increase in tobacco consumption or two and half billion extra cigarettes! That's another 2.5 billion cigarettes unaccounted for. More, in fact, because HMRC adjusts these figures upwards to account for under-reporting.
HMRC seems to have ignored the ONS figure completely in its latest estimates, preferring "projections" instead. If nothing else, this shows how sensitive these estimates are to small changes and small errors.
The current system has become farcical. Just use empty pack surveys for God's sake!
The best way to gauge the size of the
black market in tobacco is to gather a large and representative sample
of disused tobacco packs and see whether they were sold legally. Tobacco
companies have been doing this for years. KMPG, acting on behalf of
Philip Morris International (PMI), produce an annual empty pack survey
which recently found that 33 per cent of cigarette packs in the UK were
contraband or counterfeit. Japan Tobacco International (JTI) use a
slightly different methodology in which they interview smokers in their
homes and offer to exchange their empty or current pack of cigarettes
for a JTI brand. Using this system, they estimate that 33% of cigarettes
and 50 per cent of hand-rolling tobacco were illicit in 2024/25 (their
figures for 2025/26 suggest that this has risen to 42 per cent and 60
per cent respectively).
Many people are wary of claims made
by tobacco companies, but the Irish government has been using a close
cousin of the JTI survey since 2009 for its official estimates. With a
similarly high level of tobacco duty, the official Irish figures are
similar to the industry’s figures for the UK. Ireland’s 2025 survey found that 28 per cent of cigarettes and 37 per cent of hand-rolling tobacco was illicit.
