Smoking restrictions began being introduced when the weight of evidence consolidated about the harms of chronic exposure to secondhand smoke. This evidence was almost exclusively obtained from indoor domestic and occupational exposures, where non-smokers – including infants1 – spent hours on most days exposed to others’ tobacco smoke, sometimes for decades in small enclosed conditions. Notwithstanding slogans about “no safe level of exposure”, as with active smoking, the harms of exposure to secondhand tobacco smoke arise from chronic exposure, not occasional fleeting encounters with single plumes.

With almost all indoor public spaces now smoke-free in nations with comprehensive tobacco control policies, some are now emboldened to turn the attention onto outdoor spaces like parks and beaches. One such proposal comes from Lord Ara Darzi’s London Health Commission report.

The ethical justification for restricting where smoking can occur derives entirely from the Millsian precept of preventing harm to others.2 Evidence soon also mounted about important collateral benefits of banning smoking in workplaces: smokers reduce their daily consumption by about 21 percent,3 and more importantly many quit, welcoming the bans as a form of imposed self-discipline on a behaviour that 90 percent of smokers regret ever starting.4 Smoking bans fomented a rapid denormalisation of smoking:5 venues associated with relaxation, pleasure and conviviality like restaurants, bars and cinemas have no smoking, while smokers excuse themselves to go outside to footpaths in any weather or sit morosely in the fug of smoke in those desperate airport smoking rooms, wondering about how much they really enjoy smoking.

The proliferation of smoke-free areas certainly contributes to reducing both the frequency of smoking and the proportion of people who smoke. But so would forced incarceration, forfeiting smokers’ rights to healthcare or other draconian strategies too tame for the Ottoman Sultan Murad IV, who had smokers executed. The ethical test of any policy is plainly not just its efficiency in achieving outcomes.

Since I last contributed to this then-nascent debate in the British Medical Journal in 2008,6 the evidence base about the risks of outdoor smoking has grown. Our 2012 review7 found only studies of real world or simulated outdoor exposure in crowded cheek-by-jowl settings like bar patios, beer gardens and bus shelters. No studies looked at exposures in parks or beaches, almost certainly because researchers with any knowledge of airborne exposures would appreciate that such exposures would be so small, dissipated and transitory as to be of no concern.

The momentum to outdoor bans has incorporated three arguments that go well beyond direct health effects. First, large majorities of the population do not like being exposed to tobacco smoke. Outdoor bans premised on communities’ amenity preferences are not about public health but akin to ordinances about playing music in parks, bans on public nudity and littering. Outdoor smoking bans based on amenity should not be dressed up in the language of public health. Second, cigarette butts and packaging constitute a significant proportion of litter. Local governments wanting to abate this relentless source again should not appropriate public health arguments in justifying their decisions but be upfront about the litter problem.

Third, some have invoked the virtues of shielding children from the sight of smoking as worthy evidence in this debate.8 They may concede that smoking in wide open spaces like parks and beaches poses near homeopathic levels of risk to others, but point to indirect negative impact from the mere sight of smoking. This line of reasoning is pernicious and redolent of the worst excesses of totalitarian regimes’ penchants for repressing various liberties, communication and cultural expression not sanctioned by the state. North Koreans are routinely subjected to such fiats, but many would recoil at the advance of such reasoning elsewhere. If it is fine to tell smokers that they cannot smoke anywhere in public, why not extend the same reasoning to drinkers or people wolfing supersized orders in fast food outlets?

Coercing smokers to stop smoking in settings where their smoking poses negligible risk to others is openly paternalistic. Well-intentioned advocates for such policies argue, as paternalists always do, that such actions are for smokers’ own good, that many will be sooner or later grateful (which is often true). Paternalistic for-you-own-good laws about seats belts and motorcycle helmets involve trivial restrictions on liberty. Telling smokers they cannot smoke in public sight is a restriction of a different magnitude.

Last weekend I attended a twilight rock concert at Sydney’s Taronga Zoo. Announcements were repeatedly made that the open air event was non-smoking, but that smoking areas were available on the periphery of the crowd area. This arrangement struck me as totally civilised. Most, like me, don’t want to spend a few hours jammed up next to smokers. We all hope that smokers will quit, particularly those we love or care for. But when they don’t, we should not cross a very sacrosanct line and force them not to smoke if they are adults.

In Australia, daily smoking prevalence is now only 12.8 percent9 and highly likely to keep falling. This has been achieved without the unethical coercion of smokers. Those of us who have resolutely refused to cross that line, even knowing that it would likely bring public health benefit, have won the respect of a wide cross section of the community. Political support for dissuasive but not coercive policies like plain packaging and high tobacco tax rates has been bipartisan from the left and right of politics. That would almost certainly not have happened had we abandoned the ethical concerns that some are urging should occur.