Australia’s tobacco problem is disappearing from the till. Is it disappearing from people’s lives?

The Parliamentary Budget Office’s nicotine analysis raises an uncomfortable question: how much of the decline in legal tobacco sales represents quitting, and how much represents shopping elsewhere?

A government can count every taxed cigarette and still misunderstand the market in cigarettes, because falling legal sales can mean several different things: people might be quitting, smoking less, switching products or buying from someone who does not collect tax. Each possibility has different implications for public health, government revenue and the effectiveness of the law.

The Parliamentary Budget Office’s updated nicotine pricing paper, published on 3 September 2026, deserves attention because it puts legal and illegal supply inside the same analytical frame. Its baseline assumes approximately 75% of tobacco volume is illicit, a modelling assumption rather than a definitive measurement, but one that makes the alternative market central to the calculation. PBO, Nicotine prices

My reading is that Australia needs a more demanding test of tobacco policy: whether it reduces harm across the market people actually use. A shrinking taxable market cannot answer that question by itself.

The PBO’s illustrative 2026–27 assumptions put annual spending on factory-made cigarettes at $10,892 through legal suppliers, compared with $3,277 through illicit suppliers, for consumption of roughly 13 cigarettes a day. These are modelled averages rather than universal retail prices, but the difference works out to $7,615 a year, or about $146 a week. PBO price assumptions

For a household weighing rent, groceries and electricity, that is a substantial financial incentive. Recognising it does not excuse illegal selling or purchasing; it helps explain why a policy can produce a different response from the one its designers want. Someone confronted with a higher cigarette price might stop smoking, but they might also find cheaper cigarettes.

Imagine a smoker who changes suppliers while consuming exactly the same amount. Tax receipts and legal sales fall, even though their smoking does not. Now imagine someone who quits completely: tax receipts and legal sales fall in that case too, which means the two people can look remarkably similar in the revenue figures despite having very different health outcomes.

That distinction is essential to assessing whether the policy is working.

The Australian Bureau of Statistics has independently identified a substantial measurement problem. Its experimental analysis estimates that total nicotine consumption increased by almost 40% between 2017 and 2025, while the population grew by 14%, and reports that household spending on legally purchased cigarettes and tobacco has almost halved since 2020. The estimates use wastewater evidence and additional assumptions to examine consumption outside the recorded legal market. ABS, Household consumption of illicit tobacco and nicotine products

Those findings warrant investigation, but they do not establish that every measure of declining smoking is false. The Australian Institute of Health and Welfare reports that daily smoking among people aged 14 and over fell from 8.3% in 2022–23 to 5.6% in 2025, while current vaping declined from 7% to 6% and daily vaping remained broadly stable. AIHW smoking overview, AIHW vaping findings

These measures ask different questions: surveys estimate how many people report particular behaviours, wastewater analysis helps estimate aggregate consumption, and legal sales measure transactions within a particular channel. Their populations, periods and methods also differ, so a rise in nicotine consumption does not automatically establish a rise in the number of smokers or an equivalent rise in smoking-related disease.

Rather than choosing whichever statistic supports our preferred position, we should be trying to reconcile them by examining how much remaining users consume, which products they use, where they buy them and how much uncertainty comes from the measurement methods themselves. Until those questions are better answered, confident claims of either comprehensive success or comprehensive failure run ahead of the evidence.

The PBO assumes nicotine participation stays around 17% for costing purposes, explicitly stating that this is not a prevalence forecast. Its analysis estimates direct fiscal effects rather than broader health and social outcomes and assumes no illicit-price response while acknowledging substantial uncertainty. PBO methodology and limitations

Those limits matter when politicians and commentators draw conclusions from the paper. A calculation of tax receipts cannot establish the best public-health policy by itself, just as a fixed participation assumption cannot demonstrate that future price changes will leave the number of users unchanged. An assumption is an input to an argument; treating it as a finding reverses the direction of the reasoning.

The fiscal and health questions therefore need to be examined together. A serious assessment should ask whether a proposed change would bring existing purchases into the legal market, increase consumption, affect uptake or quitting, and alter who benefits financially or bears additional harm.

There is a strong general case for tobacco taxation. The World Health Organization’s tobacco-taxation knowledge hub summarises evidence that a 10% price increase reduces consumption by approximately 4% to 8%, and that evidence should remain part of the discussion. WHO FCTC tobacco-taxation evidence

Applying it, however, requires attention to the price consumers actually face. Increasing the price at a legal retailer may have a different effect when an accessible alternative supplier offers a much cheaper product, which makes the relevant policy question quite specific: what would the next change in Australian excise achieve under current market conditions?

Past effectiveness does not remove the need to evaluate the next decision, while the existence of illicit trade does not prove that taxation caused the entire problem. Availability, enforcement, distribution and consumer preferences also deserve investigation.

The debate should turn on whether a particular package of measures produces better outcomes, rather than on loyalty or hostility towards tobacco taxes.

It is tempting to move straight from a large price gap to a promise that lower excise would restore the legal market and government revenue, but the arithmetic requires more discipline. With a hypothetical 50% reduction in excise per cigarette, taxed sales would have to double merely to preserve cigarette excise revenue, before allowing for other effects.

Some additional legal sales might represent purchases moving out of the illicit market, while others could represent additional consumption. Those possibilities have different health implications even if they produce the same revenue.

Competitive responses also matter because an illegal seller could respond to cheaper legal cigarettes by cutting their own price, legal suppliers could retain some tax relief through higher margins, and buyers could value convenience or established purchasing relationships alongside price. These are scenarios to test rather than outcomes to assume away.

Anyone advocating a tax cut should show how their proposal performs under less favourable assumptions, and anyone advocating further increases should face the same requirement. The standard of evidence should survive a change in political preference.

I would start with three requirements, beginning with a coherent public account of the whole market. Survey results, legal sales, wastewater estimates and illicit-market research should be brought together in a way that explains their differences and uncertainty, allowing the public to see what each measure establishes.

Second, policy packages should be assessed against several outcomes simultaneously, including smoking prevalence, consumption, youth uptake, illicit availability, household costs and net revenue. When a measure improves one indicator while worsening another, the public deserves an honest account of that trade-off.

Third, taxation and enforcement should be connected with practical support for quitting. If the objective is fewer people smoking, the route out of smoking must be part of the policy design, rather than leaving price pressure to carry the entire burden.

Enforcement needs an outcome test too: activity figures can establish that authorities are working, but the harder question is whether illegal supply becomes less accessible and less commercially attractive. None of this requires pretending there is an easy tax rate that solves every problem; it requires making the objectives clear enough to judge whether the policy is working.

The budget debate needs that same clarity. If people quit and tobacco revenue falls, the loss is compatible with public-health success, and governments should be prepared to accept it. If purchases simply migrate outside the tax system, falling receipts tell a different story that demands a different response.

Australia should judge tobacco policy by what happens to people’s health and behaviour across the entire market, because counting what passes through the legal till is only the beginning.