Plain Packaging For Cigarettes

Shortly, the Australian government plans to introduce a bill requiring all tobacco products to be sold in plain or generic packaging.

The draft bill gives the government the power to (among many other things):

• Prohibit the use of any trade mark, logo, brand, business or company name, or other identifying mark on packaging or on tobacco products.
• Prohibit the use of any design of packaging or any design of a tobacco product.
• Provide that information of a specified kind is not to be included on packaging.

So, essentially the government is giving itself the power to dictate every aspect of cigarette, cigar and loose tobacco packaging: the shape, colour, markings and words.

Reading paragraphs 55 - 57 of former senator Steve Fielding’s 2009 bill gives some clues as to what the government is likely to mandate: the packaging may show only the brand name, contact details of the manufacturer and the number of cigarettes or weight of tobacco. It goes on to specify the exact number of cigarettes, size and shape of the pack and even the font face and size of the writing.

This must be the first time a government has legislated the removal of the right of manufacturers to use branding display to differentiate legal products.

To me, there is something absurdly hypocritical about this approach. You can’t say a class of products is legal to manufacture or import, allow competing, profit driven enterprises to do it, then not allow them to differentiate their products.

If tobacco is such a terrible public health hazard, then ban it. If you’re not prepared to do that AND you let corporations competitively make and sell tobacco products, then logically, you can’t remove a principal component of that competition: the development of a brand.

Most people agree with the current pre and proscriptions on the sale and advertising of tobacco products. There is nothing wrong with requiring the manufacturers to tell the truth about their products on the packaging: namely that smoking causes many diseases. There are plenty of other products whose advertising is restricted, either in the mass media or at point of sale, usually on the grounds of prohibiting advertisement to minors. Pornography is one example.

But the wholesale destruction of branding of a legal class of products is unprecedented and in my opinion, contrary to the commercial rights of a legal enterprise in an open society.

There have been several studies demonstrating a positive relationship between the display of brands’ designs and logos on tobacco packaging and the appeal of the product. A paper reviewing the evidence by the Cancer Council of Victoria summarises the results of these studies:

Virtually all the findings of these five studies converge on the following conclusions. Plain and generic packaging of tobacco products (all other things being equal), through its impact on image formation and retention, recall and recognition, knowledge, and consumer attitudes and perceived utilities, would likely depress the incidence of smoking uptake by non-smoking teens, and increase the incidence of smoking cessation by teen and adult smokers. This impact would vary across the population. The extent of change in incidence is impossible to assess except through field experiments conducted over time.

OK, so the badging effect of the brand on the packet makes it more likely young people will take up smoking and less likely existing smokers will quit, but quantifying these differences is difficult without actually bringing in the packaging change.

Coupled with the evidence of the negative health effects of smoking, plain packaging will almost certainly prevent illness and deaths and probably lead to decreased demands on the public health budget.

There have been claims from the tobacco lobby that plain packaging will lower prices and hence increase smoking. Philip Morris commissioned a behavioural econometric study purporting to demonstrate this.

The study makes standard, equilibrium economic modeling assumptions regarding market competition between a relatively small number of suppliers, then uses elasticity of demand and price arguments to conclude that prices will decrease by 5 – 20% and consumption rise by 2.5 – 16%.

So the epidemiologists and the economic consultants have produced opposing conclusions. Who is correct?

The epidemiologists.

The Philip Morris study neglects some important assumptions:

• If price reductions occur, governments will counter them by increasing taxes.
• People can still differentiate brands in a different way: by using cigarette cases. It used to be fashionable and will become so again.
• The existence of an equilibrium. Government intervention and non-independence of the suppliers will probably prevent the market from even exhibiting stationary state behaviour, let alone an equilibrium.

In my opinion, this study is a good example of the abuse of the most noble of all sciences: mathematics.

Having considerable experience with both economic modeling and consultants, I have seen this happen a lot. Constrain the environment by making the right assumptions, solve some equilibrium economic equations, derive the answers the client wants, add a disclaimer and send the invoice.

Changing the assumptions of the study to allow significant government interference in the market via taxes or other price regulation, or changing some of the assumptions about consumer utility and preferences will lead to very different model predictions.

Given these counter arguments, I give the Philip Morris study negligible credence.

So, the decision is a no brainer, right? Clear, evidence based policy in favour of plain packaging.

Well, to me, the above argues in favour of banning smoking tobacco altogether. That’s not to say there are no arguments against banning smoking, but public health is not one of them.

The main argument against banning smoking is liberty. Adults should have the right to harm themselves if the associated harm to others is not material. Currently, you can drink yourself to sleep every night if you wish, blow $1000 per week on the pokies if you can afford it or if you like racing cars, pay some leather clad ladies to thrash your arse to jelly with a riding crop.

People who don’t smoke should reasonably expect that smokers will not fill their environment with more than the background pollution from other sources, such as traffic fumes or industrial emissions. You can’t smoke in confined spaces with children, in office buildings or in certain areas of restaurants. Most people agree this is fair enough.

But if you are legally allowed to drink a case of beer or bottle of whisky every night, why shouldn’t you be able to smoke alone in the park, or on your own balcony? Is there really something so terrible about an old widower smoking a pipe in his study?

The health problems associated with smoking cause a drain on the public health budget. Analyses suggest the economic cost is billions of dollars per year. Consequently, it’s reasonable that smokers should bear this extra burden.

It is reasonable for governments to impose large taxes on tobacco products, which function as a proxy for a targeted increase in the Medicare levy. It is also reasonable to allow health and life insurance companies to charge significantly higher premiums for smokers, as long as their behaviour in assessing claims is rigorously and independently monitored. By this, I mean that insurers should be required to have both regular and occasional smoker categories, so that they cannot deny a claim to someone who ticked “non-smoker”, then had a cigarette when drunk one night and the insurance assessor found the picture on Facebook.

It is these alternative measures, allowing society to collect compensation for the costs of smokers’ self harm, combined with rules to protect others, governing where people can’t smoke which convince me that other people’s rights not to be unduly burdened by the harmful effects of smoking can be sufficiently protected and thus the liberty of smoking should not be banned.

The crux of the argument is that the harmful effects of smoking are mostly statistical and so therefore must be counter measures, such as insurance premia and sales taxes. By this, I mean that smoking does not necessarily cause any given disease. It increases the probability of contracting many illnesses, but there are many other genetic and environmental factors.

It is important to understand what is meant by the phrase “behaviour X causes disease Y”.

We could say “cutting someone’s head off causes death”. In this case, we mean 100% of the time. I have not tested this theory myself, however numerous experiments performed in mediaeval Europe and post revolutionary France argue in favour of the hypothesis.

We could also validly say “smoking causes cancer”, although this does not mean all smokers develop cancer. It means that the proportion of smokers who develop cancer is higher than in the general population, after controlling for other lifestyle factors which are also correlated with cancer and whose observation may be biased in smokers, such as lower rates of exercise and higher alcohol consumption.

So, smoking does not necessarily lead to any particular disease in an individual, but does almost certainly lead to higher incidences of that disease in the population.

This is why I think it shouldn’t be banned. You can’t prove that smoking will later definitely cause a particular disease in any particular individual (except perhaps emphysema). So, society is better off banning smoking in certain places and making smokers pay en masse for the overall costs of the smoking as a group, rather than going down the draconian, nanny state option of banning it altogether … and if it’s to continue being legally sold, by corporations, I don’t see that it’s reasonable to outlaw product differentiation, notwithstanding the decrease in smoking rates which it appears would ensue.

If you’re going to allow tobacco companies to manufacture and market a harmful product, is there anything extra you can do to ameliorate its public health effects?

It’s generally believed that smoking pure tobacco is less harmful to health than tailor made cigarettes, due to the many chemicals added to the latter by the manufacturers. This may well be true, however medical studies imply that pure tobacco smoke is itself carcinogenic, in addition to still causing emphysema.

OK, so any tobacco smoke is potentially harmful, but apparently the plethora of additives makes it worse, not just because some of them are carcinogens, but additionally because some of them have been implicated in increasing the addictiveness of nicotine.

Combined with the liberty argument, this tends to argue in favour of banning a large set of tobacco additives while allowing the sale of pure tobacco, or tobacco laced with natural flavours such as the cherry, vanilla, port wine etc. favoured by some pipe smokers.

I suggest this might be a better solution: allow smoking in certain environments, allow differential health and life insurance premia and prescribe a list of acceptable classes of “natural” tobacco additives such as the above, rather than proscribe a set of unacceptable ones and watch the cigarette companies try to get around your list.