Anti-Tobacco Activism at the World Bank


Richard Tren* and Hugh High**



A review of:





*RJT Environmental Economics and Consultant to SA Government, South Africa

**Principal Lecturer Economics, University of Cape Town, South Africa




We have written this review with the intention of encouraging an economic debate on the highly politicised subject of tobacco control. We would like to thank three anonymous reviewers for their comments.





Preface: Up in Smoke by Deepak Lal

Executive Summary


1.       Global Trends in Tobacco Use

2.       The Health Consequences of Smoking

3.       Do Smokers Know Their Risks and Bear Their Costs?

4.       Measures to Reduce Demand for Tobacco

5.       Measures to Reduce the Supply of Tobacco

6.       The Costs and Consequences of Tobacco Control

7.       An Agenda for Action



Preface: UP IN SMOKE

Deepak Lal, Coleman Professor of Development Economics, UCLA


The World Bank (WB) has joined the World Health Organisation (WHO) in a world war on tobacco. As its initial broadside, the WB recently published an astonishing report. The following paper exposes the many follies and economic errors in what is not a decent economic study, but a document for crusaders. Whatever one's personal attitude to the poisonous weed, most past studies of the economics of smoking found that it has net social benefits (see e.g. K. Viscusi: Rational Risk Policy, OUP, 1998). Thus, for the US it was estimated that in 1993, the social costs and benefits per packet of cigarettes smoked were as follows. Social Costs: Medical care $0.55; Sick leave $0.01; Group life insurance $0.14; Fires $0.02; Second hand smoke $0.25; Local taxes on earnings $0.40. Total costs to society were therefore $1.37. Social benefits: Nursing home savings $0.23; Pensions and social security payments saved $1.19; Excise taxes paid $0.53. The total social benefits were $1.95, yielding a net social benefit of $0.58 per pack of cigarettes. The benefit could be even higher as this estimate takes at face value the highly disputed scientific basis for the dangers of secondary smoke.


The World Bank report does an intellectual somersault in denying this standard economic methodology based on consumer sovereignty. In post-modernist garb, it characterises a divided self. It claims a smoker has a physical addiction which prevents rational decisions, so that the mind (I presume) imposes life-threatening smoking on the body. This means that, within this divided self, there is an externality that the mind imposes on the body. Thus the usual consumer surplus benefits that individuals derive from the utility of smoking are in fact, benefits that accrue to the mind at the expense of the body, and hence can, from the viewpoint of the body, be completely discounted and be looked upon as costs! For mainstream economists who have (rightly) always taken a more robust view of the mind-body problem, this assumption can only be an example of higher nonsense.


Moreover, as the WB report recognises, the poor – in both developed and developing countries – are preponderantly smokers. So the consumer surplus losses from taxes and regulations of tobacco accrue in disproportion to the poor. For a majority of the world's poor the ingestion of tobacco provides one of their few pleasures. To deny them this enjoyment, in lives which for the foreseeable future remain 'nasty, brutish and short', in the name of Western political correctness, is questionable.


In estimating the benefits from giving up tobacco, the Report uses a variant of the quality of life adjusted years (QALYs) that would be gained. Ignoring the serious conceptual flaws in QALYs (and DALYs, disability adjusted life years as used in the Report), they still require some estimate of standard life expectancy. The Report takes Japan as a standard, which has the highest life expectancy in the world, and defines premature deaths from smoking in middle age to include deaths up to 69 years. It is ripe to tell a landless labourer in rural India that he is dying prematurely at the age of 69 because of his addiction to 'bidis'!


I shall not detail the numerous other illogicalities and inconsistencies in this report as the authors do this admirably. More important are two other issues. The first concerns the reasons for the success of this form of political correctness. Resistance to smoking has a long history in the Christian West. But the crucial element which gave a recent impetus to the campaign by anti-tobacco pressure groups was the issue of passive smoking.


In most liberal Western societies ordinary people realised that if one was damaging one's body wilfully, it was one’s own and not the State's business. But once the scare about passive smoking began in the 1980s, these anti-tobacco pressure groups had the ammunition to inflame popular passions.


Passive smoking studies are littered with poor methodology, science and statistics. The most influential report, which promoted passive smoking fears, came from the US Environmental Protection Agency in 1992. But this report was discredited by a US Court last year for being inherently biased.


Similarly, the WHO had its fingers burnt when its review of passive smoking studies also claimed that there was an increased risk of lung cancer. When it was pointed out that the results were not statistically significant by the UK Sunday Telegraph, they were accused by the WHO of being pro-tobacco, as if this rebutted the scientific counter-claim. But these WHO and EPA studies are still cited as though they were quality research and have indirectly encouraged numerous places/countries to enact laws on smoking in public, including cities like New Delhi. This concern with passive smoking must seem a cruel joke to anyone who has to breathe the polluted air of New Delhi or Xian.


This false scare has led to this international war on tobacco, which raises the second issue. As is becoming common, attempts to legislate Western 'habits of the heart' worldwide, are now made through a treaty like the proposed WHO Framework Convention on Tobacco Control, which it is hoped, health ministry officials from Third World countries will readily sign. But apart from the welfare costs involved – particularly to the poorest – there are also serious fiscal costs. The WB report is illogical, claiming that a proposed annual 10 per cent real increase in taxes on tobacco for 10 years will save millions of lives and not hurt tax revenues. Either demand is elastic and so people will stop smoking at the expense of the exchequer or it is inelastic whence revenues will not fall nor will the numbers smoking. Finance ministries need to keep a careful eye and be part of any WHO negotiations to prevent any agreements which are against the interests of their poor and of their fisc.





Executive Summary


Curbing the Epidemic: Governments and the Economics of Tobacco Control (World Bank, 1999), hereafter referred to as the Report, attempts to provide an economic justification for increased intervention in the tobacco market. The Report relies upon several assumptions, evidence and studies (most of which are unpublished) in order to justify this argument. However, not only are the arguments flawed, but they, and the resulting recommendations, are muddled and incoherent.


The basis for intervention relies on four main arguments. First, smokers, especially the young and poorly-educated, are not aware of the risks associated with smoking. Second, smoking imposes greater costs on those in lower socio-economic groups than those in higher socio-economic groups, and that the former are not able to make rational decisions about whether they should smoke. Third, smoking imposes external costs on others and lastly, that far from imposing economic costs and endangering jobs, tobacco control will actually result in increased employment opportunities in many countries.


There are several problems with these economic arguments. Among others, the Report fails to recognise that it is entirely normal for individuals to perceive risks to themselves differently to their perception of risks to others. It is also entirely rational for the young and for those in poor countries, where there are generally high risks of disease, to significantly discount the future. That an increasing proportion of smokers are poor and probably poorly informed about the harms from tobacco is not sufficient justification for intervention, as smoking is only one of many cultural, dietetic educational and lifestyle differences between the socio-economic classes within and between countries.


The Report asserts that tobacco is addictive, hence smokers are not able to make rational choices about whether or not to continue smoking. From this, the authors argue that intervention is justified, but the argument is considerably weakened by their statement that tobacco consumption is greatly influenced by price rises. This would suggest that far from being a ‘special good’ tobacco is a normal, albeit inelastic, good and consumers are able to make rational choices about consumption. Again, the Report does not provide sufficient motivation for intervention (or, indeed, supporting evidence for most of its claims).


The Report argues that smoking imposes external costs on others and diverts scarce healthcare resources away from other needs and illnesses. The Report however goes on to state that the size of these costs is not known and admits that some studies have found that smokers do not incur higher healthcare costs than non-smokers. Neither does the Report take adequate account of the considerable tax contributions made by smokers in the form of direct taxation. Despite acknowledging that neither the size nor even the existence of external costs is known, the Report boldly recommends policies, such as tobacco tax increases, set at levels which “maximise revenues”.


For economists to admit that the size of the externality is not known and that there are difficulties in determining the optimal level of taxation, and then to propose taxes be set so as to maximise revenue, displays a remarkable interpretation (if not perversion) of the nature of external costs.


The Report argues that, far from imposing costs on the economy through losses in tax revenue and higher unemployment, increased tobacco controls will actually create new employment opportunities in many countries. The authors argue that smokers, or ex-smokers, will spend the money previously spent on tobacco on other goods, which will stimulate the economy and create more jobs. None of the studies upon which this argument is based are properly cited and many (perhaps all) appear to be unpublished. Furthermore, the authors do not provide any information as to the underlying assumptions of these studies.


The Report argues that tobacco control is worth paying for and estimates the cost of intervention based on the number of disability adjusted life years (DALYs). DALYs, which measure the discounted value of a life burdened with disease, are based on Japanese life expectancy of 80 years, which is wholly unrealistic for developing countries, and indeed for many developed countries.


Much of the data and the findings that underpin the Report’s findings are not published, which not only makes it difficult to evaluate the paper, but demonstrates the unscientific manner in which the authors have approached the study. An example of this is the fact that the projection of tobacco related deaths after the year 2000 is based on a single personal communication.


In addition to the authors’ failure to provide any coherent economic justification for significant intervention in the tobacco market, the Report is riddled with inaccuracies and illogical statements. The authors attempt to argue, for example, that although tobacco taxes are in themselves regressive, the fact that poorer consumers will decrease consumption faster than wealthier consumers means that overall, tobacco taxes are progressive. Furthermore, the Report proposes intervention strategies that do not take into account the actual external cost (if there is any) of smoking and it underplays the expected economic impacts of intervention.


With this publication the World Bank has missed an opportunity for rational economic analysis and is close to becoming a fully fledged anti-tobacco activist organisation (see Thomas Sowell, The Vision of the Anointed, Harper Collins, 1995, for similar ideologically driven government projects). The Report presents a case, much like an anti-tobacco prosecution lawyer might, rather than trying to find out what is the ‘case’, of the economics of tobacco farming, manufacture and use. It uses every trick in the book to make its argument, abandoning good economics, and the Bank’s reputation, in its wake.


It provides the reader with no source material from which to judge its claims. It ignores published data in favour of ‘new studies’ which the authors themselves are undertaking. It relies on its own authors for much of its ‘review’ and on personal communications for key projections. It mixes facts with its authors’ own prejudices and cleverly uses non sequiturs to trap the unwary. Furthermore, it reinterprets many economic concepts for its own ends (such as the law of demand, the impact on the poor of flat taxes, major parts of welfare economics and even concepts of psychological economics – the mind/body phenomenon referred to by Professor Lal in his preface). In short, if there is any valuable and reliable information in the World Bank tobacco control report, it is masked by the apparent desire of its authors to obfuscate and delude the reader into thinking there will only be gain from the end of the use of tobacco.


One can only speculate on why the Report was published, since its biases discount it from being a useful economic study. But perhaps it will serve a useful political purpose by furthering calls for greater international control of tobacco. Perhaps the World Bank and the Report’s authors may benefit from the earmarked tobacco taxes the Report advocates. Therefore the incentive to overplay the impacts and externalities associated with tobacco consumption is significant. Probably of greatest concern is that the Report’s recommendations will lead to many measures being forced on developing countries that will not be appropriate to their economic wellbeing. Developing country finance ministries (to whom this report is evidently directed) should reject its findings because they are based on numerous flaws and problems mentioned above and detailed in the following critique.




Curbing the Epidemic: Governments and the Economics of Tobacco Control, published in May 1999 by the World Bank, consists of seven chapters. These are:


1.       Global Trends in Tobacco Use

2.       The Health Consequences of Smoking

3.       Do Smokers Know Their Risks and Bear Their Costs?

4.       Measures to Reduce the Demand for Tobacco

5.       Measures to Reduce the Supply of Tobacco

6.       The Costs and Consequences of Tobacco Control

7.       An Agenda for Action.


This review will discuss the issues raised in each of these chapters in turn, plus a short conclusion.


A visitor to the World Bank web site ( will find numerous documents and press releases dealing with tobacco. The majority of these address the negative health effects of tobacco and seem to concentrate on the alleged unsustainability of the tobacco industry. There are no fewer than six separate publications that promote the Report, five of which are press releases with sensationalist figures of the health impact of tobacco consumption and on the need for tobacco control. This wide coverage and promotion of the Report demonstrates the high regard in which the Bank holds the Report and the conclusions that it reaches. 


It is interesting to note that, among the numerous anti-tobacco press releases and documents, is an abstract of World Bank teaching notes entitled ‘Proposal for a Joint Venture between a Multinational Company and a Local Tobacco Factory’. Although the authors of this review have not read the teaching notes, the abstract makes no mention of the negative health impacts of tobacco consumption or the supposed unsustainability of the tobacco industry, rather it highlights the main decisions that should be made in cases such as these. This could be a case of the left hand not knowing what the right hand is doing, or perhaps genuine confusion as to the role the Bank is supposed to be playing – promoting economic development or protecting individuals’ health.


One might reasonably ask why a Bank, which has economic development as its avowed mission, has chosen to publish a report on tobacco and methods of controlling tobacco growing and consumption. While one can agree with the authors that there is a relationship between economic development and tobacco consumption, the same can be said of a number of variables which indirectly affect economic growth and development, such as religious practices and familial structures. Perhaps the World Bank is set to expand its remit, or perhaps this foray into lifestyle choices is an aberration.


A possible answer as to why the World Bank would spend scarce resources on funding this study and its authors with flights (and all other expenses) to Beijing, Cape Town, Lausanne, and Washington, is the observation by Milton Friedman that: “Nothing is as long-lived as a government program”. Tobacco control is in vogue and the World Bank is following the trend.


Nevertheless, it is not too much to suggest that this entire Report is a superb example of appeasement to pressure groups by the Bank and successful rent-seeking by the Bank’s authors and affiliates, which the Bank has supported. In recent years the Bank has come under attack, from a mixture of grass roots activists to multinational pressure groups, for financing ‘imperialistic’ projects in the third world. It is therefore keen to find a mission which will serve to justify its continuing existence and expansion, and gain support of the increasingly powerful global consumerist and protectionist stakeholders (see Jeremy Rabkin and James Sheehan, ‘Global Greens: Global Governance’, IEA, 1999, for a discussion of this phenomenon). It is difficult to believe that John Maynard Keynes and his contemporaries, envisioned that their creation – the World Bank – would come to be particularly concerned with the personal decisions of consumers in developing countries, as opposed to their general economic development.


A second major observation about the Report concerns its objectivity and independence. It has become good general practise among government agencies to submit reports to independent review and criticism prior to their publication, particularly if the findings are likely to prove controversial and/or have significant policy implications. One might reasonably expect that this Report was subjected to rigorous independent assessment, as indicated by the Acknowledgements to a number of individuals, and the note that prior to the Report’s publication, there was an ‘External Experts (sic) Review’ (see pp. 99-101). Indeed from reading the February draft, substantial changes were made prior to the May publication. Yet a careful reading of the Report casts doubt on its ‘independence’. For instance, of the 23 participants at the external experts’ review, six were listed as writers of the Report, while 11 were listed as having participated in the planning. This leaves only six truly independent assessors at the experts’ meeting. These six may be independent, but their expertise is in doubt, as none have been cited as authors of any works listed in the Report’s bibliography.


On the other hand, many of those present at the economists’ technical meeting, and earlier reviewers of outlines and the contents of papers are listed as authors of papers on which the Report is based (see pp 99-101). One cannot but wonder as to the objectivity of the individuals who, in compiling this Report, found it desirable to rely on their own works and those of colleagues who were instrumental in the design of the Report itself.


These preliminary observations noted, we now turn to the Report itself.


The project team consists of many analysts led by Prabhat Jha. The Report originated at the 10th World Conference on Tobacco in Beijing, China, where it was considered that insufficient economic analysis had been applied to tobacco control. At the same time, economists at the University of Cape Town began a project on the economics of tobacco control in Southern Africa. The project included a conference in Cape Town in February 1998, the proceedings of which are entitled ‘The Economics of Tobacco Control: Towards an Optimal Policy Mix’, Abedian et al, University of Cape Town, 1998.


The World Bank Report, which draws on the work of many of the analysts who met in Cape Town, attempts to tackle some of the weighty and difficult issues surrounding the use of tobacco, health, economics, risk and development. The Report is essentially an economic report, however it relies on epidemiological and scientific studies into the effects of smoking. It is not surprising that the economist authors do not question studies outside their expertise, but as we have noted above, it is troubling to note that some of the works on which the Report relies are unpublished and thus do not allow easy independent enquiry.


The standpoint of the researchers is obvious from the title of the Report, ‘Curbing the Epidemic.’ In choosing the word epidemic, which is usually reserved for disease and not for a lawful everyday activity, prompts one to question their approach. Individuals undertake hundreds of activities every day that carry risks and dangers; smoking is just one of them. It is unheard of to hear people discussing the driving epidemic, the drinking epidemic, or the eating epidemic.


The first sentence of the Report sets its tone. The affirmative statement is made that “500 million people now alive will eventually be killed by tobacco use, if current smoking patterns are maintained”. Such a statement cannot but alert the reader to what seems to be a problem of enormous proportions.


Yet, the statement is terribly flawed and indicative of the entire Report. It is an unsubstantiated statement. Absolutely no footnotes or citations are provided for such a sweeping allegation. And, it will not save the statement to say that various papers on which the Report relied, by chapter, are given in the Bibliographic Note. It is common practice among scholars that statements require specific support, and not a generality that ‘you will find support for my allegations in one of several of the following works’. The reader is then left to read all such works, and ferret out the right reference for himself –or he must accept the statement is true as a matter of faith.


Perhaps one can forgive this lack of substantiation in a Foreword. But the first sentence of the Summary, which appears at the beginning of the Report, states that: “smoking already kills one in 10 adults worldwide”. There is no citation for this statement. It is only in the Bibliographic Note, which appears 100 pages later, that the interested reader is told, “the discussion on consumption and epidemiology draws on [one or some of the five papers listed.”]


The reason we call attention to this rather inferior form of citation and scholarship is that the Report is replete with it. More importantly, the Report, either through sloppy scholastic method, or by design (as one suspects), is not specifically referenced so as to enable the reader to verify the validity of several sweeping statements; statements which, if true, will have immense policy significance. The writers of the Report appear to have decided their conclusions in advance. Statements are made which are clearly designed to support a particular point of view.


Lest it be thought these observations are petty quibbles, it is important to note that the absence of specific citations is important in a number of those areas which are addressed by the Report’s authors. To note but one and obvious example: the epidemiological statements are often eye-catching and dramatic. Those noted above clearly are. Yet, there is much controversy regarding, for example, the methodology employed in estimating the incidence and death rate from smoking-related illnesses (see the much disputed paper ‘Lies, Damned Lies and 400,000 Smoking Related Deaths’ by Robert Levy and Rosalind Marimont, Regulation, 21.4.1998; for a more scholarly treatment of estimation techniques see Ashford, ‘Problems in Assessing Smoking –Related Mortality’, Journal of Smoking-related Diseases 1992, 3,3, pp263-274). If a statement in the Report relies on a flawed technique or estimate, then without knowing the source which gave rise to the particular statement, the reader is hardly in position to affirm, or deny, the veracity of the Report’s statement.


1. Global Trends in Tobacco Use


The Report highlights the fact that global patterns of smoking have changed, from predominantly men in high-income countries, to women in high-income countries and men in low-income countries. The Report states that the “…upward trend may have slowed in these (low and middle-income) countries a little since the early 1990s”. (p. 13.) Unfortunately, however, the authors fail to include any data since 1992. It would seem unlikely that this data is not available and one therefore has to question whether it has been omitted simply because it may undermine their findings and conclusions.


One of the alleged causes of increased consumption in low and middle-income countries is cited as the relaxation of trade barriers, with greater competition, greater promotion and advertising. The support for this allegation is “an econometric model developed for this report” (p.14). Yet there is no citation of the model, and no way for the reader to independently check the statement. At most, the reader might discover on page 101 of the Report, that “the discussion in the Report on trade liberalisation draws on an article by Chaloupka” (who was one of the main authors of the Report itself), and his co-author, Laizuthai in a 1996 paper. Or the statement might be attributable to a paper merely cited as “Taylor and others, background paper”, which is not cited at all in the Bibliography. It is impossible to know anything about the alleged econometric model, since we don’t know where, if at all, it was published. The authors might have learned much from reading the alleged article and/model, but the reader of this Report has no way of sharing in this knowledge and is, then, reduced to merely accepting the Report’s allegation, or remaining in the dark. As it is a truism in scientific circles that statements which cannot be falsified are statements without merit, one can but assume this statement is without merit; it is certainly without support.


The second point to make concerns the benefits that the relaxation of trade barriers has brought to both high and low-to-middle-income countries. The period during which cigarette consumption increased in the Asian countries that the Report cites (Japan, South Korea, Thailand and Taiwan) coincided with a period of strong economic growth in those countries (with the possible exception of Japan). During this period, the consumption of all goods would have increased, including washing machines, televisions and alcohol.


Trade liberalisation would have brought many economic benefits, as well as economic costs (such as increases in the amount of waste). The Report however fails to discuss the benefits to consumers of having a wider range of cigarettes, at lower prices and also the introduction of low-tar and other ‘healthier’ cigarettes. Indeed, at least one scholarly economics paper on the effects of trade liberalisation comes to precisely that conclusion; that lowering tariff barriers permitted locals to consume more, and presumably healthier, foreign low tar/low nicotine cigarettes (see Hsieh, C-R, Hu, T-W & Lin, C-F J. ‘The Demand for Cigarettes in Taiwan: Domestic versus Imported Cigarettes’. Contemporary Economic Policy, Vol. 17, No. 2, April 1999, pp 223 to 234). In asserting that “increased trade liberalisation contributed significantly to increases in cigarette consumption, particularly in low and middle-income countries” (p.15), they merely state the economically obvious: that, as long as the marginal propensity to import is positive, when prices of any foreign goods are lowered, domestic consumers will tend to consume rather more of them. This is true of cigarettes, but it is also true of tea towels, CD’s and foreign magazines. To make this statement is to belabour the obvious. One can only speculate as to the reason the authors did so. In addition, no mention is made of the economic benefits that arise when smokers have more money in their pocket (as a direct result of cheaper cigarettes) and are able to spend that money on other goods and services, perhaps even health care and education.


The Report highlights the fact that smoking has changed from being a habit of the wealthy, towards being a habit of the poor and less educated. The trend emerged in the 1950s and 60s when the health consequences of smoking became publicised. It is not surprising that the poorer and less educated members of society should smoke more than the wealthy and more educated. The same is true of the more harmful eating and general living habits of the poor.


Furthermore, it must be recognised that the poor in low and middle-income countries face risks and uncertainties that are completely different to those living in high-income countries. For example, the prevalence of diseases such as Tuberculosis (TB), malaria and cholera in low and middle-income countries is far higher than in high-income countries. This leads to people in low-income countries having higher discount rates than people do in high-income countries, and this will play an important part in their rational decision making of whether or not to take up or continue smoking (see a discussion on discounting in Pearce, D. Markandya, A. Barbier, E. 1989 Blueprint for a Green Economy, Earthscan, London  pp 132 – 152).


This chapter concludes with a ‘boxed’ statement/estimate of the number of persons, to age 20, who begin smoking daily. Quite apart from the quality of the estimates, one would think that ‘parallelism’ in a chapter ostensibly concerned with global patterns of tobacco use, would have also produced an estimate of the number of persons, worldwide, who quit smoking, daily. Its absence is not only regrettable, and well indicates the absence of parallel treatment of the issue of tobacco consumption, but also indicates the authors’ willingness to ‘play to the crowd.’ This is regrettable.


2. The Health Consequences of Smoking


The Report explains that tobacco contains nicotine, which is an addictive substance that can cause addiction quickly thereby reinforcing the smoking habit. There is, of course, no citation given, nor is any mention made of the rather extensive literature on the very nature of addiction, nor is mention made of the fact that pharmacologists, generally, do not speak of ‘addiction’ but of ‘degrees of addiction’ and are among the first to note that the very term ‘addiction’ is none-too-useful or descriptive (for an interesting series of papers see David Warburton ed. Addiction Controversies, Harwood Publishers 1990). The point of mentioning this is twofold. First, psychopharmacologists have generally distinguished between drugs (such as heroin) that have physiological impacts requiring increasing doses to maintain the same effect for the taker, and drugs, which simply may be habit forming (such as caffeine and nicotine). Second, that other food substances (such as caffeine or chocolate) can commonly be thought of as ‘addictive’, that in and of itself should not be a reason to recommend regulation.


The Report goes on to estimate the disease burden of tobacco use and states that tobacco is responsible for one in 10 adult deaths, and by 2030 the figure is expected to be one in six. It is not possible to query these figures as no detail is given as to how they are calculated or the underlying assumptions that were used. Furthermore, the authors, in the same paragraph, assert that “500 million people alive today will eventually be killed by tobacco, half of them in productive middle age, losing 20 to 25 years of life” (p. 22). This dramatic statement rests on the Report’s very peculiar definition of ‘middle age’, about which we comment below. Essentially, the authors neglect to inform the reader that this ‘middle age’ is decidedly greater than the life expectancy of most of the world’s population. Moreover, the authors fail to note that many of the estimates of ‘tobacco-related deaths’ rest on death statistics which attribute death to tobacco if the deceased ever smoked and are often filled with recollection inaccuracies (see Ashford op. cit.).


Later in the chapter in discussing the health impacts of smoking, the authors refer to ‘low tar’ and ‘low nicotine’ cigarettes, stating that smokers believe these cigarettes are safer than ordinary cigarettes. The Report goes on to state that “… the difference in the risk of premature death for smokers of low-tar or low-nicotine brands compared with smokers of ordinary cigarettes is far less than the difference in risk between non-smokers and smokers” (p. 24). It is not clear what point the authors are trying to make here. Low-tar and low-nicotine cigarettes allow smokers to enjoy their habit at a lower risk of disease. To state that it would be safer not to smoke at all is fatuous. It should be clear to most people that the risks of flying in an old Dakota aircraft are higher than the risks of flying in a modern Boeing, but one’s risks of dying in an air accident are lowest when staying at home and not flying at all. The introduction of the modern Boeing aircraft has reduced the risks of air travel and increased welfare; the argument for low-tar and low-nicotine cigarettes is surely the same. If the authors have some evidence that low-tar, low-nicotine cigarettes are as potentially harmful as others, they should cite it. That they have not is merely part of the continuing problem with this Report.


In discussing smoking and the health disadvantage of the poor, the Report states on page 25 that tobacco is responsible for more than half of the difference in adult male mortality between the highest and lowest socio-economic status in four countries (Canada, Poland, United Kingdom and the United States). There are many lifestyle differences between the highest and lowest socio-economic groups, of which smoking is just one. In Scotland for example, unhealthy eating by those of lower socio-economic status leads to higher rates of heart disease than among those in higher socio-economic groups. This has not, however, led to calls for the government to attempt to ban deep fat frying, or to impose a cholesterol tax on black pudding. Likewise, vehicular deaths are greater for high income earners than for low income ones – for the simple reason that low-income earners are less likely to own a car. Furthermore, the number of deaths from vehicular accidents, per thousand population, is greater in high income countries than low – for like reasons (See John Adams, Risk, UCL Press, 1995). It is difficult to understand why (a) this is not obvious; and (b) what is the source for the authors’ statements, since it is not given and thus not verifiable (and thus we don’t know whether the statement rests on raw data, or is adjusted for other differences in income classes).


The Report also states that “among white women in the United States, smoking alone has been found to be responsible for 63 percent of the difference in birth weight between babies born to college-educated women and babies born to those who received a high school education or less”. It appears from this statement that the authors are confusing the impact of smoking on birth weight and the impact of education on birth weight. In the absence of any citation, we cannot further judge the statement, but since it raises more questions than it answers, the reader might well wonder why such an obtuse and narrow statement about college educated US women is here at all, given the number of statements they might have made about the more general issue. One is left with the impression the authors are groping for something to say.


Figure 2.2 in the Report details deaths in middle-aged males in Poland and the key assumes that middle age is between the ages of 35 and 69. While this may be an appropriate age span for middle age in a few developed countries, it is certainly not a suitable age span in any developing country. The United Nations Development Programme estimates that 65 years is the longest a baby born in a non-industrialised Southern Hemisphere country could expect to live. Without access to the unpublished report, from which these figures are derived, it is not possible to judge how this age span has been used elsewhere in the Report.


It is interesting to note that, in the subsection ‘The Risks from others’ smoke’, the authors first make the eye-catching statement that “smokers affect not only their own health but the health of those around them” (p. 26). This is followed by a discussion, without citation, of the weight at birth of babies born to mothers who smoke, and to those who don’t smoke. This is followed by an equally dramatic statement that “cigarette smoking accounts for much of the health disadvantage of babies born to poorer women”(p. 26). Not only is there no citation for this assertion, but the “health disadvantage” of being born to a poor woman is not discussed.


This impression is reinforced by the glaring absence of any statements about ETS, or ‘second hand smoke’. This would have been an obvious place to discuss this issue if it were perceived as a danger. Presumably, in view of the evidence that has been mounted against this once-widely denounced ‘danger’, the authors determined to ignore the issue. We can, then, presume they do not find ETS, or ‘second hand smoke’, a danger to others (see Gio Gori and John Luik, Passive Smoke, Frazer Institute, 1999).


3. Do Smokers Know Their Risks and Bear Their Costs?


Chapter 3 of the Report addresses three main issues, firstly whether smokers are aware of the risks of smoking; secondly it deals with issues of young people, addiction and the capacity to make sound decisions; lastly it deals with costs imposed on others. These are each dealt with in turn.


Awareness of risks


The Report states that people’s knowledge of the health risks of smoking, particularly in developing countries, is limited (again no supporting citations are given). But assuming there is evidence to support this hypothesis, which seems plausible, there is a strong argument for making the risks of smoking known to both smokers and non-smokers so they can make informed decisions as to whether to smoke. However, adequate information of the dangers of tobacco does not equate with rejecting smoking.


The fact that smokers perceive the risks to themselves of smoking as lower than the risks to smokers of smoking in general is not surprising. It is human nature to perceive risks individually and also to assume that ‘it could never happen to me’ (see Warburton op. cit.). The fact that, once made aware of the risks of smoking, individuals choose to interpret the information differently, is not sufficient justification for controlling tobacco. If it were, the door would be open for government to regulate the production and sale of all manner of goods, from butter and sugar to nylon clothing. Furthermore, there is evidence that smokers and non-smokers significantly overestimate the risks to themselves from smoking (see Kip Viscusi, Smoking, New York: Oxford University Press, 1992).


Youth, addiction and the capacity to make sound decisions


The Report explains that teenage smokers do not understand the risks of smoking, underestimate the risks of becoming addicted to nicotine and do not make informed choices. As before, there is a strong motivation for making the risks of smoking known to all smokers and potential smokers. However, it should not be surprising that the young interpret risks differently and have difficulties in imagining old or even middle age. There is some evidence, which appeals to intuition, that an individual’s discount rate depends on his or her life expectancy (see Pearce, op. cit. 132 – 152). Thus, those who anticipate many years of life ahead, as do the young, often disregard present risks. This is a general phenomenon and hardly confined to consumption of tobacco. More, this is certainly not sufficient justification for controlling individual’s tobacco consumption (see Warburton op. cit.).


The Report quite rightly points out that young people are attracted to many different types of risky behaviour “… such as fast driving or alcoholic binge-drinking” (p. 32), to which one could add far more innocent yet risky activities, such as skate boarding and rugby. The Report goes on to say that much risky behaviour, such as driving under the influence of alcohol, is controlled by law. This, however, seems to be an outlandish comparison to make. It is disproportionate to compare the risks a drunk driver imposes on his passengers, other drivers and pedestrians with the risks to which a smoker exposes others. Furthermore, in most countries it is illegal to sell tobacco to minors, so inasmuch as a law can control underage smoking, curbs are already in place.


The Report states that smoking is different to other risky behaviours, as over a lifetime it is much more dangerous than other risky activities. Extrapolations from data on high-income countries are used to estimate the number of 15-year-olds in low and middle-income countries that will die from smoking and other causes (such as road accidents and violence.) There are several difficulties in extrapolating data from high-income countries and applying it to low-income countries, not least the fact that such extrapolations are unlikely to pick up other diseases prevalent in low-income countries, such as malaria, TB and cholera, and will therefore give a biased outcome (see Ashford op. cit.).


Costs imposed on others


The Report explains that external costs to smoking do exist and that if smokers took these costs into account, they would smoke less. This concept of an optimal level of smoking is not a new one and is accepted. The authors assert, without the inconvenience of providing the reader with a citation and/or evidence, that treating smoking-related diseases consumes between 6 per cent and 15 per cent of total healthcare costs in high-income countries. The consumption of healthcare services by smokers means that these resources cannot be used in other healthcare areas. The Report however states that smokers generally die earlier in life than non-smokers and therefore tend to consume the same overall amount of healthcare resources. Quite correctly, the authors note that “the assessment of these costs is complex, and therefore it is not yet possible to conclude anything about how they may influence smokers’ consumption choices” (p. 33). This quite correct and judicious observation does not, however, prevent the authors from then asserting that as taxpayers, non-smokers contribute to the healthcare costs of smokers and this is unfair.


Furthermore, the question of fairness prompts one to question whether taxpayers who do not eat fatty foods should be exempt from contributing to the healthcare costs of those that need medical treatment related to an unhealthy diet. By the same logic, it would be unfair for cautious drivers to contribute to the healthcare costs that could be incurred by reckless drivers. Smoking is an activity that carries with it benefits and costs, just as eating fatty foods does, and it is not clear why it should be treated differently to any other good that carries benefits and costs.


Overall, this chapter is characterised by a large measure of ‘indirection’. For example, the authors assert that “in high income countries, public expenditure on health accounts for about 65 per cent of all health expenditures, or about six per cent of GDP” (p. 33). Quite apart from the fact this is an unsupported assertion, it is followed by the observation that “if smokers have higher net lifetime healthcare costs, then non-smokers will subsidise the healthcare costs of smokers” (p. 34). This is, of course, true. But it is equally true if we reverse the statement to read “if non-smokers have higher net lifetime healthcare costs, then smokers will subsidise the healthcare costs of non-smokers”. Presumably the only reason it is included is merely journalistic – not scholarly. This is especially the case in view of the authors’ acknowledgement on the previous page that “assessment of these costs is complex”.


The authors only briefly mention the tax contribution that smokers make through tobacco duties and fail to analyse the extent to which smokers place, or do not place costs on others. Given that the rates of taxes on tobacco vary from below 40 per cent in low-income countries and over 70 per cent in high-income countries of the total price per pack, the contribution from smokers is likely to be significant. In failing to include these tax contributions, save for stating that “… smokers may not be imposing costs on others” (p. 34), the bias of the authors is clearly illustrated (see Viscusi, op. cit. for statistical analysis). Bias is again clearly indicated by the last full paragraph of this section by a leap of logic in which the authors, having earlier acknowledged the difficulty of measuring net costs, boldly assert “in sum, smokers clearly impose direct costs, such as health damage, on non-smokers”. Such an assertion most conveniently fails to include a statement regarding ‘net’ costs.


It is accepted economics that should externalities to smoking exist, policies could be introduced to internalise those external costs and attempt to achieve an optimal level of smoking. The Report tells us that the size of the externalities is unknown and there are numerous reasons to suggest that smokers do not actually impose costs on others at all. Despite not knowing what the external costs might be, nor even whether they exist, the authors propose numerous anti-smoking measures designed to repay these costs to society. This has no economic rationale.


The following section of this chapter is entitled ‘Appropriate responses for governments’. In the first paragraph of this section, we are informed that since smokers might not know the full extent of the risks they undertake, or to bear the full consequences of their choices, “governments may therefore be justified in intervening to adjust the incentives to consumers so they smoke less” (p. 34). This is a refreshing admission that the authors’ real goal is to provide not more information, but to give governments worldwide a rationale for dictating to consumers.


4. Measures to Reduce Demand for Tobacco


The Report proposes three main ways of reducing demand for tobacco, namely raising cigarette taxation, non-price measures to reduce demand, such as consumer information, bans on advertising and smoking restrictions and lastly, nicotine replacement therapy (NRT).


Tobacco Taxation


The Report states that increases in taxation will ensure that the demand for cigarettes decreases, but need not decrease revenues. The Report also makes the point that the revenues from tobacco taxes can be used for “… anti-smoking activities or other specific activities”(p. 38). It is unfortunate for society that the massive revenues at stake can lead to rent-seeking by anti-tobacco campaigners, including the authors of this Report (see Mitchell and Simmons, Beyond Politics, Boulder: Westview, 1995, for a discussion of rent-seeking). A wider example of this rent-seeking could be pharmaceutical companies that produce anti-smoking devices such as nicotine patches overstating the dangers of smoking or the size of the smoking ‘problem’ in order to increase the government subsidy on the sale of their products.


One of the major justifications for intervention in the tobacco market is that tobacco consumers are addicted to the substance and therefore are unable to make rational choices about the risks involved. If this were so, demand for tobacco would be highly inelastic and the price could rise to exorbitant levels, and smokers would continue to consume the product because of their addiction. The Report claims the contrary in saying “… demand for tobacco, while inelastic, is nevertheless strongly affected by its price” (p. 39). If this is the case, it cannot still be claimed that smokers are unable to make rational decisions about how much tobacco to consume. The fact that smokers would react to a price increase in tobacco, as they would in any other normal good indicates that tobacco is not a special good, but should be treated just like every other good that individuals choose either to consume or not to consume.


In asserting that an increase in the price of cigarettes may well lead to a decrease in consumption, the authors refer to the Canadian case in which tax increases between 1982 and 1992 led to a “substantial decrease in consumption” (without, of course, a citation). This assertion is accompanied by a graph which takes up half the page on which the assertion is made (p. 39/40). However, the graph also shows that, while consumption was allegedly falling as a result of the tax increase, smuggling was rising. Eventually, the smuggling problem contributed to the Canadian government’s decision to remove the extra tax. Put differently, while the authors would have the reader believe that an increase in taxes is an effective ‘tool in the war against smoking’, they ignore the secondary effects of higher taxes. And, of course, no mention is made at all as to the likelihood that smokers may switch to cheaper, higher-tar, higher-nicotine brands. The authors also assert that tax increases have led to decreased consumption in the UK (again and typically, an assertion without citation), yet ignore the tremendous increase in smuggling which is alleged to have occurred in the UK (‘One in five cigarettes are bootlegged, says Gallaher’, The Guardian 10th Sept. 1999; ‘Death and taxes’, Sunday Telegraph 31st October 1999). Some contraband tobacco could have come from eastern European countries which are important producers of high-tar, high-nicotine cigarettes (Craemer Raoul and George Yarrow, ‘Effects of National Regulation in the Single Market: The Case of Tobacco’, Oxford Regulatory Policy Institute Research Paper, April 1997; ‘Romania approves new excise levels for cigarettes’, Reuters News Service 28th October 1999).


If it is the case that individuals perceive the risk of contracting a tobacco-related disease incorrectly, the correct approach would be to recognise that it is human nature to perceive risks differently, and then to inform the smoker or potential smoker of the dangers of tobacco. Taxing a smoker because of his risk perception is not only a crude instrument to use, but it is unfair.


The Report shies away from proposing an optimal level of taxation (the level that equates the marginal social cost of the last cigarette consumed with its marginal social benefits) but then goes on to propose various methods of determining an appropriate tax rate. First it suggests that the tax should be set so as to achieve a specific reduction in cigarette consumption and meet a public health target. Alternatively, the Report suggests setting the tax level so as to maximise revenues. The first point to make is that the authors of the Report clearly have not understood that smoking is a lawful activity that brings benefits to those who enjoy it. To tax it as they have suggested would in no way reach the optimal level of consumption and shows a clear lack of understanding of the nature of externalities. The second point refers to the discussion earlier on rent seeking. If governments set tax rates to maximise revenues for anti-tobacco activities, the incentive for any organisation that has an interest in reduced smoking rates to promote anti-smoking scare stories and to bias the debate, is enormous.


The Report then suggests that countries should harmonise tax rates so that they are in line with countries with “comprehensive and effective tobacco control policies” (p. 45). Again, the authors have shown a complete lack of understanding of the external costs. For a low-income country to raise tax rates on tobacco in line with a high-income country, where the levels of economic development are different, the impacts on the economies will be different, the types of smoking externalities are different, would be economic madness. From the statements made in relation to tax rates, it is clear that the authors of the Report do not aim to achieve an optimal level of smoking, rather they are striving to hamper a lawful activity, a measure that will cause significant economic costs.


Non-price measures to reduce demand


The Report proposes a number of non-price measures to reduce demand, such as publicising findings on the health effects of smoking, introducing warning labels, counter advertising, educational programmes and banning cigarette advertising and promotion.


It would be difficult to argue that increasing the amount of information that a consumer has about a product is a bad thing. The more information that a consumer has, the better he or she is able to exercise rational choices. The proposals, however, severely affect commercial freedom of speech, which as a fundamental freedom should not be limited.


The authors mention, without citing, reports that purportedly show that aggregate consumption of tobacco is affected by advertising. The authors seem to be trying to persuade the reader that while no relationship between advertising and consumption has been found or measured, even despite the authors’ own efforts, a relationship does actually exist. This allows the authors to bypass a discussion of the evidence, that advertising increases smoking, and move straight to discussing advertising bans, using as support their own, unpublished, background papers.


However, the literature shows that the alleged aggregate consumption/advertising relationship is non-existent, (see Hugh High, Does Advertising Increase Smoking, IEA, 1998, for a review of the studies). The Report’s authors seem to wish to totally ignore this with their assertion that “cigarette advertising and promotion almost certainly does affect consumption although the data are not straightforward”. This statement is most surprising when, in the next paragraph the authors state that “[o]n the face of it, empirical studies of the relationship between advertising and sales have tended to conclude either that advertising has no positive effect on consumption, or that it shows only a very modest positive effect” (p. 49). Still, the authors are reluctant to rely on these studies and assert that, since advertising expenditures are high, any increased consumption resulting from increased advertising will be small and marginal and thus not ‘caught’ by the studies; moreover, the effects might not be felt for long periods. They refer to studies which do show a positive effect of advertising on consumption, but do not cite them although there is implicit reference to the widely discredited studies by Laugensen and Meads. Finally, the authors conclude that, as it is so difficult to interpret the effects of increased advertising, it is easier to look at post-ban consumption instead.


Having dismissed the empirical evidence of studies showing no positive effect the authors, again resort to ‘indirection’. This is well illustrated by the statement (p. 50) that modelling, within the authors’ possession, and which is based on the European Union’s advertising ban, suggest that advertising bans like that of the EU would be effective in reducing consumption. The authors provide an entire page which summarises the EU ban, but this in no way allows the putative model to be tested or evaluated. The reader would have been vastly better served by some detail about the underlying model, or citations, or something.


In yet a further effort to ‘demonstrate by indirect evidence’, that advertising bans or restrictions would curb consumption, the authors then assert that (again uncited) studies demonstrate that children recognise tobacco adverts. So what? There is also some evidence that men remember adverts for brassieres, but it is difficult to know what this may suggest about consumption. To ‘bootstrap’ recollection into consumption is hardly the basis for sound public policy decisions.


The authors conclude the section on the ‘impact of non-price measures on global demand for tobacco’ with the assertion that something between 5 & 23 million “lives could be saved” (p.53). This sort of statement is indicative of the entire work. It is based on a model which is very generally described, but not in sufficient detail to permit replication, or even to understand its underlying assumptions. The model is unavailable to the reader and uncited, yet the authors suggest various public policy proposals based on the model’s results. This does not serve the credibility of the authors, or of the World Bank, well.


Nicotine replacement therapy and other cessation interventions


The Report promotes the use of nicotine replacement therapy (NRT) as an effective way of helping smokers reduce tobacco consumption. According to the Report, the use of NRT along with brief anti-smoking advice increases the percentage of smokers abstaining for six months or more by 6 per cent, compared with brief advice alone. If NRT is coupled with intensive advice, a further 8 per cent will abstain for six months or more. While NRT is widely available in high-income countries, it is far less widely available in middle and low-income countries, and then only in the major urban centres.


The Report pushes for the wider distribution of NRT, reducing the regulation on the products, increasing the range of outlets through which they can be sold, increasing the hours during which they can be sold and reducing the restrictions on packaging. Alternatively, the Report proposed subsidising the price of NRT for limited periods to smokers on low incomes.


NRT requires a significant one-off payment to begin the course, which may be out of reach of many in middle and low-income countries, and the governments in those countries may not be in a position to offer subsidies. The Report admits that no attempt has been made to test the cost effectiveness of NRT, which should be done before it is recommended, especially to middle and low-income countries.


5. Measures to Reduce the Supply of Tobacco


The limited effectiveness of supply-side interventions


The Report highlights the incentives to supply tobacco. For example, in Zimbabwe it is approximately 6.5 times more profitable that the next alternative crop. The Report also highlights the impracticalities and undesirability from an economic and political standpoint of introducing a tobacco prohibition.


Of the 8 048 000 metric tonnes of tobacco produced in 1997, 6 272 800, or 78 per cent, was produced by middle and low-income countries. For many of these countries, particularly Zimbabwe, Malawi, India and Brazil, the tobacco industry forms a large and important part of the economy. It is also a major foreign exchange earner for several countries, notably Zimbabwe, Malawi and Brazil.


Any attempt to interfere directly with the crop decisions of these countries would be met with strong political opposition and accusations of imperialism. The Report is therefore careful not to propose any policies that might directly interfere with a country’s decisions on what crops to grow. It goes on to say that crop substitution attempts in the past have not succeeded because of the significant incentives that exist to grow the crop.


The Report argues that numerous high-income countries and China use price supports and subsidies in order to support tobacco growers. The Report further asserts that price supports, while raising the price of tobacco, “have but a small effect on consumption”. This is a curious statement to make since, earlier, the Report asserted that taxes, by raising the price of tobacco products, would lead to a decrease in consumption. Indeed, the discussion of these alleged effects in Canada and the UK was extensive. Later, in their discussion of smuggling, the authors assert that control of smuggling “helps the effective implementation of price increases that reduce demand”(p. 63). It is as if the authors forgot what they wrote both earlier and later. The reader is left confused as to whether the authors think the Fundamental Law of Demand operates or not. It is widely acknowledged that subsidies in the agricultural sector lead to misallocation and inefficient use of resources, but reductions in tobacco subsidies should happen in line with reductions in subsidies of all agricultural goods. Sudden removal of subsidies to one, politically incorrect, crop, could cause significant hardship in some of the poorest countries of the world.


The Report discusses the impact of trade liberalisation on tobacco consumption and argues that, although free trade brings with it economic benefits, the fact that tobacco is more harmful to health than most goods should prompt trade restrictions. The Report then correctly highlights the disadvantages of restricting free trade, such as retaliatory action. The Report, again, does not highlight the benefits that smokers and the wider economy enjoy through trade liberalisation. As pointed out above, these benefits include a wider choice of cigarettes at lower prices and the introduction of new types of cigarettes such as low-tar and low-nicotine cigarettes that carry reduced health risks. In addition, with smokers spending less on cigarettes as a direct result of free trade, there is more money available to be spent on other goods and services.


Firm action on smuggling


Tobacco smuggling occurs on a wide scale and in almost all markets. According to the Report, 30 per cent of internationally exported cigarettes are lost to smuggling. Smuggling has occurred on wide scale in developed and developing countries alike, notably between Canada and the United States, within Europe and in China. Smuggling occurs when the expected high-risk returns of illegal trade are greater than the returns of selling tobacco legally. The larger the product price differential between two countries, the stronger the incentive to smuggle.


The Report goes on to state that the determinants of smuggling are not based on price differentials alone, but also on the degree of corruption in a country. That there should be a relationship between the amount of smuggling and the degree of corruption should not be surprising, however it is important to remember that smuggling takes place in some of the least corrupt high-income countries of the world, such as Canada. According to the Report, the success of smuggling depends on cigarettes passing through a large number of owners in a short time, so that the cigarettes are difficult to trace.


The Report states that cigarette manufacturers will benefit from tobacco smuggling, as the average price of cigarettes, taxed and untaxed, will fall and therefore overall sales will increase. It is difficult to disagree with the statement that when the average price of a good falls, average consumption should rise; however, it should be recalled that is precisely what the authors did in their discussion of the demand for cigarettes and taxation in the previous chapter, as we have noted. Following this assertion is the trite statement on p. 65, that “the presence of smuggled cigarettes in a market that has hitherto been closed to imported brands will help to increase the demand for those brands, and hence increase their market share”. Self-evidently, introducing a new good to a market will increase the good’s market share.


Smuggling will also pressurise governments to keep tax rates low, according to the Report. Recommendations are made that could reduce the incidence of smuggling, such as placing prominent stamps with the country of origin on cigarette packs, licensing all parties involved in the tobacco trade, placing serial numbers on cigarette packs and improved record keeping by the cigarette manufacturers. These, plus the other measures proposed in the Report are likely to be very costly and will add further to the price of cigarettes, but may indeed reduce smuggling.


Successful prosecutions for cigarette smuggling depend on effective policing and an efficient judicial system, which are often lacking in middle and low-income countries. Smuggling is likely to occur when one country raises its taxes significantly. When this happens, not only is smoking reduced to levels that may be economically inefficient, but also criminal activity is stimulated to the detriment of the country as a whole. While the policies proposed in the Report may go some way to improving prosecution rates for smugglers, it does not remove the underlying incentive to smuggle (see Mark Thornton, The Economics of Prohibition, Utah University Press, 1993, for a discussion). It is the incentives to smuggle, and particularly the price incentives, which are relevant. As the price of a good rises in a country, so does the incentive to smuggle that good, which is one reason the Report advocates tax harmonisation. However, the possibilities of at least one nation not agreeing to tax harmonisation (and employing lower rates) is considerable, because the gain for their treasury, in the form of tax ‘arbitrage’ advantages from purchases by smugglers in their lower-tax country, are significant. But even if harmonisation is achieved, the incentive for officials in each country is to accept bribes from smugglers which are lower than the duty, rather than collect taxes for the nation state. This is especially important for nations which are major importers of tobacco products. Clearly, if a nation imports tobacco products, and its taxes are identical with those of another nation, the only way which consumers can avoid taxation is via bribery of the tax collectors. The greater the tax rate on a given good the greater the incentive to smuggle and evade that tax (see Thornton, op. cit., and Craemer and Yarrow, op. cit., for a discussion).


6. The Costs and Consequences of Tobacco Control


Tobacco control and employment effects


The authors ask the question whether tobacco control will result in job losses and conclude that for the net and full importers of tobacco, more jobs are likely to be created than lost if tobacco is controlled. For the agrarian economies that are dependent on tobacco, the opposite is likely to be true, with net national job losses.


The Report reaches this conclusion by arguing that money that would have been spent on tobacco would be spent on other goods and services and therefore jobs need not be lost, indeed they could be created. With a tobacco ban, the amount of money spent on tobacco consumption will undoubtedly be spent on other goods and services but since smoking is a singular pleasure ex-smokers may spend their money on other politically incorrect activities (see Thornton op. cit., for a discussion of the US alcohol prohibition). Without any detail on the underlying assumptions, it is impossible to comment further on this point except to say one thing. To achieve a reduction in smoking will require massive tax increases, and much of the money (that could be used to buy other goods) will be accrued to governments via tax, who do not have a superb record of job creation.


The Report claims that independent research of the UK market shows that if smokers spent their money elsewhere, 100 000 full time jobs would be created. The Report does not detail the assumptions behind any of the models used in their analysis and therefore it is not possible to fully assess their findings, but equally, neither is it possible to give much credence to such statements.


The Report claims that the net change in employment as a percentage of the economy for Zimbabwe, if all domestic tobacco consumption and production were eliminated, would be minus 12.4 per cent. The Report does not model any other net exporters that are middle or low-income countries such as Malawi, Turkey or Brazil, however, it would seem plausible that the effects on these economies would be equally severe.


Our own research (to be published in early 2000) into the economic impact of the tobacco industry in South Africa, shows that an increase in excise taxes would lead to a significant fall in employment (we expect the Reader to treat our assertions sceptically since our report is not yet published). South Africa is termed a balanced tobacco economy, in that it consumes about as much as it produces, yet even so, the impacts on the economy are likely to be significant and negative.


If tobacco control measures are introduced slowly, the impact on the economy is likely to be less severe, however there will be an impact, and the low-income countries are likely to suffer the most. It appears that the authors are content to allow low-income countries, such as Zimbabwe, to pay for the health policies of high-income countries.


In the discussion of the effects on national economies of eradication of a national tobacco industry, the authors suggest that the impact would be small, since that money spent on tobacco would be spent on other goods. This may be true, but provides no insights or direction for public policy as regards tobacco. The same statement could be made, with equal usefulness, that if a nation absolutely prohibited the sale of – say – tomato soup, or rye bread, or nail polish, the money which was previously spent on these goods would be spent on something else. But, having made such a trite observation, the authors then, in what can only be described as a most cavalier treatment of the effects on workers who may become unemployed if sales of the good are prohibited, assert that “[o]f course, industry transitions can be difficult and may create social and political problems in the short term. But economies go through many such transitions, and this one would not be exceptional”. One might add that those who are made better off, and profit from tobacco abolition, such as the authors of the Report, manufacturers of nicotine replacement devices, and a wide variety of rent-seekers in governmental, inter-governmental and non-governmental organisations, can afford to be cavalier about the ‘mere transition effects’.


Increased tobacco taxes and government revenues


The Report claims that increased tobacco taxes will not erode tax revenues, rather in the short and medium term, tax revenues will increase. The Report quotes a model developed for the study that shows “…modest increases in cigarette excise taxes of 10 per cent world-wide would increase tobacco tax revenues by about 7 per cent overall, with effects varying by country”. As before, no further detail is given and therefore it is difficult to criticise or even comment on the model.


It should be pointed out that the researchers could have an incentive to promote higher excise taxes and to downplay any negative impacts on tobacco revenues, as they could be the beneficiaries of earmarked tax revenue. This rent-seeking behaviour is not uncommon and when taken into account, shows that their research is no more independent than industry-funded research.


Tobacco taxes and smuggling


While the Report claims that tobacco smuggling tends to increase when a country increases its excise taxes, it underplays the effects of smuggling. For example, Canada and South Africa introduced significant increases in excise duties, and legal tobacco sales fell. However, the Report does not resolve whether smuggled cigarettes simply displaced demand for the legal product.


The Report recommends that, not only should policing activities be improved so as to reduce smuggling, but tax rates should be harmonised between countries so as to remove the incentives to smuggle. Firstly, many developing countries, such as South Africa simply do not have the resources to devote to smuggling activities as they are already stretched in keeping law and order in city centres. To recommend “… a crack down on crime” shows little appreciation of the realities facing low-income countries.


Secondly, the proposals for the harmonisation of tax rates are problematic for several reasons. First, this type of tax harmonisation is usually agreed upon by developing countries after pressure by international bodies, such as the UN. Usually, this has followed pressure by non governmental organisations (NGOs) and usually does not take into consideration the needs or requirements of the particular country. Frequently, the middle and low-income countries are under-represented among the parties that reach harmonisation agreements, which is not only unfair, it is undemocratic (see Rabkin and Sheehan, op. cit.).


Second, that a harmonisation of tax rates is recommended, again shows a lack of understanding of the nature of externalities and the methods that can be used to internalise them. Supposing intervention in the tobacco market is required on the grounds that smokers do not account for the full costs of their habit, the external costs are likely to vary widely from country to country. Harmonised tax rates will in no way reach the optimal level of consumption that is referred to in the Report.


Will poor consumers bear the heaviest financial burden?


The Report correctly claims that although tobacco taxes are regressive (in that poorer consumers spend more of their income on tobacco) the fact that their consumption will fall faster than a rich consumer’s consumption means that tax increases would, in actual fact, be progressive. The Report relies on two studies from the United States and the United Kingdom, which have found that tobacco tax increases are progressive, even if the tax in itself is regressive.


The logic of these statements is absurd, and is reminiscent of that of the Red Queen in Alice in Wonderland. One should be reminded that tobacco is a legal product, which is consumed because people perceive benefits in consuming it and feel that their utility is increased when it is consumed. Forcing poor consumers to consume less of the product by penalising them financially is in no way progressive. If it were the case that tax increases on goods such as tobacco were progressive, one has to ask why governments have not applied tax increases on other goods. If governments raise petrol taxes, no one would argue that this is actually being kinder to poorer consumers because it means that they travel less and expose themselves less to the risks of accidents and pollution.


Will tobacco control impose costs on individuals?


The Report argues that tobacco is not a typical consumer good because of addiction and information problems. The avoidance of withdrawal is listed as one of the benefits of smoking and the Report goes on to state that “… if tobacco control measures reduce individual smokers’ consumption, those smokers will face significant withdrawal costs” (p. 75).


First, it is important to point out that there are ‘information problems’ associated with the consumption of any good, be it shoes or margarine. No consumer will have full information about the risks associated with each good or of all the alternative possible brands. While it is important that consumers make as informed a choice as possible, it is unlikely to be the case that any consumer will have full information on any good.


Secondly, as pointed out above, the fact that the authors claim that tobacco consumption is strongly influenced by price undermines their claim that it is not a typical consumer good. If tobacco consumption rises when prices fall and falls when prices rise, this indicates that it is a normal good and one has to question whether the special treatment that it receives is justified. As we have noted, there seems to be confusion among the authors as to just how the Fundamental Law of Demand operates.


The authors go on to justify the fact that tobacco control measures will impose relatively and possible absolutely bigger costs on poor individuals than richer individuals by comparing it to any other health initiative, such as child immunisation. Child immunisation or family planning programmes are often more costly to poor households as they “may have to walk longer distances to clinics than rich families, and may lose income in the process” (p. 76). The crucial point that the authors miss is that child immunisation programmes and the like are in place to prevent diseases such as measles and mumps. There are no economic benefits to the victim or the economy associated with any of these diseases, and the victim of such a disease does not contract it by engaging in a lawful activity that he or she feels provides a welfare improvement.


The Report states that “welfare losses are likely to be minimised if control interventions are implemented as a package” (p. 76). It is not clear what this statement means. If it implies that if tobacco taxes, advertising bans and restrictions on where smokers can smoke are all implemented together, as a package, then it would seem unlikely to minimise welfare losses in any way.


Is tobacco control worth paying for?


The Report investigates the cost-effectiveness of the various tobacco control measures and uses a model described in a background paper to the Report. The Report concludes that, “depending on the assumptions made about the administrative costs of raising and monitoring higher tobacco taxes, the cost of implementing a tax increase of 10 per cent could be less than $5 per DALY (and is unlikely to be more than $17 per DALY) in low and middle-income countries”. The model estimates the cost effectiveness of non-price tobacco control measures to be between $68 and $272 per DALY for low/middle-income countries. Publicly provided NRT is estimated to cost between $276 and $297 per DALY for low/middle-income countries. Without any more information on the assumptions made, or any detail of the model, it is difficult to criticise the specific figures that are calculated.


What is clear though, is that the use of disability adjusted life years (DALYs) as a measure of the effectiveness of tobacco policies in low and middle-income countries is naive. DALYs try to assess the discounted value of a life burdened with disease, added to the years of ‘life lost’ to smoking for each individual, so that a smoker’s life is ‘worth’ less than a non-smoker’s. Not only does a smoker not live as long, but also the quality of his remaining years is degraded by the disability from his habit. This ignores any other diseases that the non-smoker may have and also ignores any disease or disability that the smoker may have. The age to which the dying person could have expected to live (if he had not smoked) is based on the life expectancy at birth equal to that of the world’s longest-living population, in Japan. The life expectancy that is used is 80 years, however the life expectancy of many middle and low-income countries is far below this.


The model clearly does not account for the numerous diseases that are found in low-income countries, such as malaria, TB, cholera and sexually transmitted diseases, which would dramatically change life expectancy (see Mooney and Bate (eds.), Environmental Health, Butterworth Heinemann, 1999). In many Sub-Saharan countries, life expectancy is only 50 years and therefore to apply their model to these countries is absurd and could result in their true ‘worth’ measured in DALYs being a negative number.


The Report admits that only rudimentary estimates of the costs of implementing a comprehensive tobacco control programme exist. The authors estimate that low and middle-income countries should be able to afford a comprehensive tobacco control programme. More research should be done, however, to assess the impact of such measures before they are recommended.


7. An Agenda for Action


The Report claims that only two causes of death are large and growing worldwide: HIV and tobacco. There are in fact a number of diseases that are growing and that are responsible for numerous fatalities, such as malaria (see Mooney and Bate op. cit.). It appears as if the authors have omitted common communicable diseases in order to make the impact of tobacco related deaths seem more severe. The authors also estimate that number of tobacco deaths will rise dramatically over the next 50 years, unless current smokers quit. The Report finds the cumulative number of deaths between 1950 and 2050 will reach 520 million if no action is taken. Different figures are given for different intervention scenarios, however it is very alarming to note that all the estimates of tobacco deaths after the year 2000 are based on a personal communication (with Peto). To base these important projections on a personal communication is unprofessional and unscientific and has to undermine the entire validity of their recommendations.


The Report recommends strong action to reduce the number of smokers, such as raising tax rates, publishing and disseminating research results into the effects of tobacco and widening access to NRT. These issues have all been dealt with above and there is no need here to further address problems associated with them. The Report also goes on to recommend that international organisations such as the UN should ensure that tobacco control is given “due prominence.” This would involve “facilitating international agreements on smuggling control, discussions on tax harmonisation to reduce incentives for smuggling, and bans on advertising and promotion involving the global communications media”.


These proposals will clearly increase the influence of activist NGOs. They are likely to impair the sovereignty of nation states and will increase costs significantly. The harmonisation of tobacco control activities will not, as has been mentioned above, take into account the externalities of specific countries and will therefore not achieve the optimal level of tobacco consumption that the authors strive to achieve.




The Report attempts to provide an economic justification for the continuation and strengthening of tobacco controls. This review has highlighted some of the inconsistencies and problems with the Report’s proposals. Among others the authors have relied upon unpublished data and papers, which apart from being unprofessional, does not allow the reviewer to fully analyse the findings.


The Report favours the introduction of tax increases as the most effective way of stopping myopic smokers (the young, poor and less educated) from smoking. It even states that tobacco tax increases will be progressive and therefore beneficial to the poor. In order to justify intervention, the authors have argued that because of the addictive nature of tobacco, it is not a normal good and consumers are not able to make rational decisions about how much to consume. This argument however is undermined by the authors who argue that, as with any normal good, price increases will cause a reduction in demand. In many areas, the research and arguments do not stand up to criticism on economic grounds.


The impartiality of the authors has to be questioned as they have relied heavily on their own unpublished research in trying to support contentious issues, particularly regarding their proposals to ban tobacco advertising.


The recommendations to remove political barriers and push through tobacco control measures via international bodies, such as the UN is very worrying as many measures are likely to be forced on countries that will have limited representation, and probably will not be appropriate in any case.