tobacco control in south africa in the 1990s: a mix of advocacy, academic research and policy

This article was downloaded by: [New York University]On: 06 December 2014, At: 01:28Publisher: RoutledgeInforma Ltd Registered in England and Wales Registered Number:1072954 Registered office: Mortimer House, 37-41 Mortimer Street,London W1T 3JH, UK

South African Journal ofEconomic HistoryPublication details, including instructions forauthors and subscription information:http://www.tandfonline.com/loi/rehd19

Tobacco control in SouthAfrica in the 1990s: A mixof advocacy, academicresearch and policyCorné Van Walbeek aa University of Cape TownPublished online: 04 Mar 2010.

To cite this article: Corné Van Walbeek (2004) Tobacco control in South Africain the 1990s: A mix of advocacy, academic research and policy, South AfricanJournal of Economic History, 19:1-2, 100-131, DOI: 10.1080/10113430409511172

To link to this article: http://dx.doi.org/10.1080/10113430409511172

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of allthe information (the “Content”) contained in the publications on ourplatform. However, Taylor & Francis, our agents, and our licensorsmake no representations or warranties whatsoever as to the accuracy,completeness, or suitability for any purpose of the Content. Any opinionsand views expressed in this publication are the opinions and views ofthe authors, and are not the views of or endorsed by Taylor & Francis.The accuracy of the Content should not be relied upon and should beindependently verified with primary sources of information. Taylor and

http://www.tandfonline.com/loi/rehd19

http://www.tandfonline.com/action/showCitFormats?doi=10.1080/10113430409511172

http://dx.doi.org/10.1080/10113430409511172

Francis shall not be liable for any losses, actions, claims, proceedings,demands, costs, expenses, damages, and other liabilities whatsoeveror howsoever caused arising directly or indirectly in connection with, inrelation to or arising out of the use of the Content.

This article may be used for research, teaching, and private studypurposes. Any substantial or systematic reproduction, redistribution,reselling, loan, sub-licensing, systematic supply, or distribution in anyform to anyone is expressly forbidden. Terms & Conditions of accessand use can be found at http://www.tandfonline.com/page/terms-and-conditions

Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

01:

28 0

6 D

ecem

ber

2014

http://www.tandfonline.com/page/terms-and-conditions

http://www.tandfonline.com/page/terms-and-conditions

100

Tobacco control in South Africa in the 1990sA mix of advocacy, academic research and policy

by

CORNé VAN WALBEEKUniversity of Cape Town

1. INTRODUCTION

South Africa has seen many dramatic changes in government policysince the democratic transition in 1994. In the area of public health,the ANC-led government has taken a particularly strong approachagainst tobacco. Whereas the apartheid government was generally in-different to, and even supportive of, the tobacco industry, the new gov-ernment made it clear from the start that they were going to imposestrong controls against tobacco. The new government's tobacco con-trol strategy allowed South Africa to catch up rapidly with, and evenovertake, the tobacco control policies that many developed countrieshad implemented in recent decades. Ever since the detrimental medicalimpact of tobacco was scientifically proven in the 1960s, many Euro-pean and North American countries had imposed tobacco control poli-cies, aimed at reducing tobacco use and smoking prevalence.

This article aims to investigate how this process unfolded. It will beshown that this was a long process, going back to at least the 1960s. Italso indicates how lobbyists on both sides have tried to influence gov-ernment policy. The government, the tobacco control lobby, and the

* I would like to thank Prof Murray Leibbrandt, Dr Yusuf Saloojee and twoanonymous referees of this journal for useful comments and insights on aprevious draft of this article. All errors and omissions are my responsibility.

South African Journal of Economic History

Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

01:

28 0

6 D

ecem

ber

2014

101

tobacco industry all used academic research to give weight to their ar-guments. This article illustrates how academic research, together withfocused lobbying, can have a major impact of government policy, pro-vided that the government has the necessary political will to effectchange.

2. THE RISE OF THE TOBACCO INDUSTRY IN SOUTHAFRICA

Tobacco was first cultivated in South Africa after the arrival of theDutch settlers in the seventieth century.1 By the beginning of the twen-tieth century the main production area had shifted from the WesternCape to the Transvaal, with small production pockets in the Easternand Western Cape, and in the Little Karoo. Currently most tobacco inSouth Africa is produced in the North-West and Limpopo Provinces.

Between 1937 and 1996 the Tobacco Board controlled the productionand marketing of leaf tobacco through a single channel marketing ar-rangement. The Tobacco Board formalised and gave legal power to aco-operative culture that had its genesis in 1909 when the first co-op-erative association was founded.2 By managing prices and productionvolumes, the Tobacco Board made South Africa more or less self-suf-ficient in the supply of leaf tobacco. Between 1930 and 1950 tobaccoproduction increased from nearly 10 million tons to 19 million tons.3

Tobacco leaf production peaked at more than 40 million tons in 1979.4

By 1996, when the Tobacco Board was disbanded as part of the gov-ernment's strategy to deregulate the agricultural sector, tobacco pro-duction had decreased to 27 million tons.

1. Anonymous, The golden leaf: A survey of the tobacco industry in South Africa.Pretoria, published for the Tobacco Control Board, 1970, p 11

2. J. Fourie, There's sunlight in my leaf report published by request of the TobaccoBoard to commemorate the 500th birthday of Nicotiana tabacum, 1992, p 32

3. Anonymous (1970), op cit, p 114. Tobacco Board, Annual report, various years

Volume 19: September 2004

Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

01:

28 0

6 D

ecem

ber

2014

102

The number of tobacco farmers had decreased from about 5000 in theearly 1970s, to less than 700 in 1995.5 During the same period the av-erage tobacco yield has increased from about 800 kg/hectare to about1800 kg/hectare. The more efficient tobacco growers have survivedwhile the less efficient growers have been forced out of the industry.According to the now-defunct Tobacco Board,6 the number of peopledirectly employed by the tobacco industry, of which the majority areemployed in the agricultural sector, decreased from about 76 000 in1985 to 35 000 in 1995.

The other important tobacco industry is cigarette manufacturing. InSouth Africa the story of cigarette manufacturing is essentially the sto-ry of the Rembrandt Group. In 1940 Dr Anton Rupert established theVoorbrand Tobacco Company in Johannesburg, the predecessor ofRembrandt. In 1948 the Rembrandt Group was incorporated, and in1956 it was listed on the Johannesburg Stock Exchange and veryquickly become the Afrikaners' symbol of economic, and specificallyindustrial, liberation.7

The rapid growth of Rembrandt can be attributed solely to the businessacumen of Dr Rupert. He had forged many strategic partnerships, andwas able to grow his company rapidly, not only in South Africa, butespecially abroad. Already in the early years Rembrandt had made anagreement with Rothmans in London to exchange technical knowl-edge.8 Through "industrial partnerships" with a large number of for-eign companies, he was able to expand his company rapidly aroundthe world. In each country where the group established a new enter-prise, half the shares were sold to the local community.9 Also, localcitizens were appointed to serve as chairmen and directors of boards.By 1961, Rembrandt was selling cigarettes in 120 countries.10 Accord-

5. Ibid6. Ibid7. Rembrandt Group Limited, Annual Report, 2000, p 148. W.P. Esterhuyse, Anton Rupert: Pleitbesorger vir hoop, Cape Lown, 1986, p 459. Fourie, op cit, p 6110. A. Rupert, Wie in wondere glo, Cape Lown, 1967, p 19


Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

01:

28 0

6 D

ecem

ber

2014

103

ing to Esterhuyse11 the secret of the company's success was its strongemphasis on quality ("Each cigarette a masterpiece") and innovation.For example, Rembrandt was the first company, worldwide, to pro-duce king-size cigarettes and menthol filter cigarettes.12 The companybecame the fourth largest international cigarette company (ie exclud-ing the state monopolies in countries like China and the formerUSSR).13 Given the unprecedented amount of advertising of cigarettesin the first half of the twentieth century, it was difficult for new com-panies to enter this market.14 According to Peter Drucker, author of In-novation and entrepreneur ship}5 "there has only been one major new-comer in the world's cigarette industry since the 1920s, the South Af-rican Rembrandt Group".

Dr Rupert held remarkably liberal views for a White Afrikaans-speak-er of that period. He introduced minimum wages in his company in1961 at much higher levels than the average wage at the time. Hisviews on racial issues, especially in the labour context, put him in oc-casional confrontation with the government, and especially PrimeMinister H.F. Verwoerd. However, despite the ideological differencesbetween Dr Rupert and the ruling party, the government was stronglysupportive of Rembrandt, especially during the 1970s and 1980s.

In due course Rembrandt diversified its interests away from tobacco,although tobacco remained the mainstay of the company. Currently thecompany, renamed Remgro in 2000, is classified as an investmentholding company, and derives its income from its investments in to-bacco products (about 48 per cent of the group's headline earnings in2001), wine and spirits (2 per cent), mining (26 per cent), industry (12

11. Esterhuyse, op cit, p 4512. Rembrandt Tobacco Manufacturing Corporation of South Africa Limited, Quality

above all: The story of Rembrandt, Stellenbosch, 197613. Ibid, p 414. J.E. Brooks, The mighty leaf: Tobacco through the centuries, Boston, 195215. Cited in Esterhuyse, op cit, p 45


Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

01:

28 0

6 D

ecem

ber

2014

104

per cent), financial services (7 per cent) and industries like medicalservices and telecommunications (5 per cent).16

By the late 1990s Rembrandt had 85 per cent of the market in ciga-rettes in South Africa, compared to 10 per cent in 1951, 30 per cent in1958 and 60 per cent in 1962.17 In the 1990s the only other competitorof any note was British American Tobacco, trading under the name ofUnited Tobacco Company (UTC). UTC had been in South Africasince the 1880s and, although it had dominated the market in the firsthalf of the twentieth century, it had been rapidly losing market share toRembrandt after 1948.18

More consolidation in the cigarette manufacturing industry took placein 1999 when Rembrandt sold Rothmans International to the UK-based British American Tobacco pic (BAT), the world's second larg-est and geographically most diversified cigarette producer. RothmansInternational had been listed on the London stock exchange in 1972and represented Rembrandt's non-South African tobacco interests. In1988 a separation of Rembrandt's local and overseas interests hadbeen effected with the founding of Compagnie Financiere RichemontAG (Richemont), a Swiss-based luxury goods group, which also hadsignificant tobacco interests. As a result of the 1999 merger, Remgroand Richemont now hold more than 30 per cent of BAT's shares.19

After the Rothmans/BAT merger, BAT South Africa has a 95 per centmarket share in South Africa. The only other cigarette company withany influence in South Africa is Japan Tobacco International, whosemain brand is Camel.20 In April 2004 the Marlboro brand, owned bythe US Altria Group, was launched in South Africa. There are early in-dications that Marlboro has negatively affected the market share of

16. Remgro Limited, Annual Report, 200117. Fourie, op cit, p 6118. Ibid, p 5619. Remgro Limited, op cit, p 720. A. Castillo, South Africa: Position paper, internal R.J. Reynolds document, 1994


Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

01:

28 0

6 D

ecem

ber

2014

105

Camel, more than any other brand, but significant trends would requiremore time to develop.

3. OPPOSITION TO TOBACCO BY THE MEDICALCOMMUNITY IN SOUTH AFRICA

As was the case internationally, the medical community drove the op-position against tobacco in South Africa. The first South African stud-ies that linked smoking to lung cancer were published in 1963.21 TheSouth African Medical Journal (SAMJ) became the main vehicle forpublishing tobacco-related research and opinions in South Africa. In1963 the editor of the SAMJ argued that

the educational campaign should be the main weapon inthe fight against cigarette smoking, but some restrictivelegislation will also be necessary. There should be no hes-itation about banning smoking in public places and onpublic transport. Here the discomfort and disease of thenon-smoker must be considered before the convenience ofthe smoker. The law about providing cigarettes to chil-dren must be more strictly enforced and automatic vend-ing machines must be banned. Cigarette advertisingshould at first be restricted in quantity and content with aview to its eventual complete limitation. It might also beadvisable to insist that each cigarette packet should carrya notice to the effect that the contents are potentially dan-gerous to health. The Minister of Health may also attemptfurther restrictions of smoking by increasing the taxationon cigarettes.... The matter is important and urgent.22

21. A.G. Oettlé, "Cigarette smoking as the major cause of lung cancer". South AfricanMedical Journal, 37, 1963 - published in three parts

22. Cited in Y. Saloojee, "Tobacco control: A case study", in D. Yach & S. Harrison,"Proceedings of the All Africa Conference on Tobacco or Health, 14-17November 1993", Harare, unpublished SAMRC report, 1994, p 162


Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

01:

28 0

6 D

ecem

ber

2014

106

The editorial stance of the SAMJ has been consistently, and some-times aggressively, anti-tobacco.23 South African tobacco-related stud-ies, published mainly, but not exclusively in the SAMJ, often focusedon smoking prevalence. The SAMJ also published some studies on therelationship between smoking and the risk of contracting specific to-bacco-related diseases (ie the etiology of tobacco), but these fall be-yond the scope of this article. Since the early 1980s, economic aspectsregarding tobacco use started receiving more attention, even in themedical literature.

4. THE MEDICAL COMMUNITY ENTERS THE ECONOMICDEBATE

Derek Yach24 was the first to attempt to quantify the economic costsof tobacco use in South Africa. Costs were limited to lost earningsfrom tobacco-related premature death and illness, and the direct costof hospitalisation. The financial gain from tobacco was defined as thesum of tobacco excise revenue, the salaries of people employed in thetobacco sector, and the value of cigarettes sold. On the basis of his re-sults, Yach contended that "when one compares the monetary and non-monetary costs that result from smoking, it becomes readily apparentthat the 'benefits' are dwarfed by the total social and economic costsof the industry".25 Interestingly, although he recommended that the taxon cigarettes be increased, he proposed that the government should"find alternative sources of excise revenue and decrease the reliance

23. See, for example, H.C. Seftel, "The smoking epidemic - what you can do". SouthAfrican Medical Journal, 59, 1981, pp 617-18; Medical Association of SouthAfrica, "News from the secretariat: Statement on smoking", South AfricanMedical Journal, 67, 1985, p 519; and A.J. Brink, "The smoking trail - researchshows a path of death and destruction", South African Medical Journal, 73, 1988,p 385

24. D. Yach, "Economic aspects of smoking in South Africa". South African MedicalJournal, 62, 1982, pp 167-70

25. Ibid, p 169


Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

01:

28 0

6 D

ecem

ber

2014

107

on tobacco/cigarette revenue".26 Presumably he thought that an in-crease in the excise tax on tobacco would result in a decrease in gov-ernment revenue. The reality, clearly demonstrated in South Africaand other countries in the following 20 years, is that an increase in theexcise tax increases government revenue.

In 1988 the South African Medical Research Council published thefirst of two comprehensive reports on tobacco in South Africa, the sec-ond report being published in 1992.27 The 1988 and 1992 reports sum-marised most of the existing medical and epidemiological researchthat had been published to date,28 and provided updated estimates ofthe economic costs and benefits of tobacco, based on Yach's earlierstudy. Given the methodological framework and the available data, thereports clearly indicated that the costs of tobacco outweigh the bene-fits.

The 1992 report appealed for strong action by the government. It dis-tinguished between what it called "popular" and "unpopular" interven-tions. "Popular" interventions were described as low-key educationprogrammes, posters, leaflets, and other "cosmetic" activities aimed atreducing smoking, but that are not effective and thus not opposed bythe industry. On the other hand, "unpopular" interventions entail the

26. Ibid27. South African Medical Research Council, Smoking and Health in South Africa:

The need for action, Pretoria, 1988 and South African Medical Research Council,Smoking in South Africa: Health and economic impact, Pretoria, SAMRC, 1992

28. Amongst other things, the 1988 report noted that, in 1984, about 100 000potential life years of people aged 35-64 were lost due to premature tobacco-related mortality, primarily as a result of ischaemic heart disease, followed bylung cancer and chronic obstructive lung disease. The report also summarised anumber of prevalence studies, of which the most consistent results are thefollowing: (1) smoking prevalence increases as people's incomes increases, butpast a certain income level (which varies from one racial group to another)smoking prevalence decreases slightly, (2) people with post-matric training havelower smoking prevalence, and (3) there is wide support for tobacco controlinterventions, from a diverse range of social groups.


Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

01:

28 0

6 D

ecem

ber

2014

108

passing of tobacco control legislation and require the political will totake on a powerful industry. The SAMRC urged the government toban tobacco advertising29 and to increase the excise tax on cigarettes,arguing that this would reduce cigarette consumption, and raise gov-ernment revenue at the same time.

Other than banning tobacco advertising and increasing the tobacco ex-cise tax, the 1988 and 1992 reports made a number of recommenda-tions on how to decrease tobacco consumption, specifically throughlegislation. These recommendations are shown in Table 1. If one com-pares the situation in 1988 to that in 2003, it is obvious that the coun-try has made some huge strides in its tobacco control strategy. As willbe pointed out in Section 6, tobacco consumption has decreased dra-matically in the past decade. The medical benefits take longer to mani-fest themselves, but in due course one can expect a decrease in tobac-co-related disease and death.30

29. According to the SAMRC's 1992 report, the rationale for an advertising ban isbased on the fact that "the battle between these manufacturers of ill-health andhealth educators is an unequal one. The tobacco industry commands massiveresources and has access to sophisticated techniques of persuasion in maintainingand promoting smoking as a desirable activity. They are effective because they tierisk behaviours such as smoking to dominant cultural themes and images.Consequently, it is inaccurate to talk of freedom of choice when considering suchbehaviours. Individual choice is shaped and limited by environmental factors andcommercial interests, which profit from unhealthy lifestyles. Because free choicedoes not operate in these situations, legislation is required to curb the activities ofthe manufacturers of illness"

30. J.M. Nkuchia, G.A. Giovino & A.M. Malarcher, "Epidemiological surveillanceof tobacco use in South Africa". CHASA Journal of Comprehensive Health, 7(2),1996, pp 74


Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

01:

28 0

6 D

ecem

ber

2014

c3n

in

n

O

o

TABLE 1

Policy and legislative recommendations in the 1988 and 1992 SAMRC reports

Proposal

Control of advertisingand sales promotion

Situation in 1988No government-decreed ban; voluntaryagreement that direct cigaretteadvertisements are not televised

Requirements for health Packets carry a small health warningwarnings and statement "Smoking is a health risk"; tar and nicotineof tar and nicotine contents are stated on packet by voluntarycontents agreement

Limits on tar and No legislation; in 1978 ranges were asnicotine contents follows: 12-39 mg tar/cigarette and 0,5-2,4

mg nicotine/cigarette

Restrictions on sales Sales to minors are not prohibited

Taxation and pricepolicy

Cigarettes were subject to excise tax, butreal tax level had been eroded by nearly 70per cent between 1970 and 1988

Situation in 2003Complete ban of all tobacco advertising andsponsorship (Tobacco Products Control AmendmentAct of 1999)

Eight rotating health warnings, covering 20 per centof front and 30 per cent of back of packet; tar andnicotine contents are stated on packet (TobaccoProducts Control Act of 1993). Pictorial healthwarnings are proposed in Amendment Bill of 2003

Restricted by law (TPCAA of 1999); currentmaximum tar = 15 mg/cigarette and maximumnicotine =1,5 mg/cigarette, to reduce to 12mg/cigarette and 1,2 mg/cigarette, respectively, byJune 2006

Sales to children under 16 is prohibited (TPCAA of1999). Age limit to increase to 18 years according toAmendment Bill of 2003

Rapid increases in excise tax since 1994; excise taxesare increased annually to maintain a 50 per cent taxincidence. In 2004 the tax incidence was increased to52 per centD

ownl

oade

d by

[N

ew Y

ork

Uni

vers

ity]

at 0

1:28

06

Dec

embe

r 20

14

IIs

1

tir

ProposalEconomic incentives to Nonesubstitute other cropsfor tobacco

Restrictions onsmoking in publicplaces

Restrictions onsmoking in theworkplace

Mandating healtheducation

Establishing a nationalorganisation for policydevelopment and co-ordination

Situationin1988

No national legislation; provincial andmunicipal by- laws prohibit smoking incertain public places

No dedicated national legislation; theMachinery and Occupational Safety Act(1983) may be applicable in certaininstances

No comprehensive national programme, butgovernment does not want young people tosmoke; educational material is madeavailable to schools

No government agency; voluntary anti-tobacco groups exist

Situation in 2003None

Prohibited in terms of the TPCAA of 1999;hospitality industry may set aside a maximum of 25per cent of their floor space to smokers, provided it isseparated from the main area of the establishment

Prohibited in terms of the TPCAA of 1999

Comprehensive health education is part of Life Skillscurriculum

Department of Health has strong tobacco controlfocus; NGO sector (especially National CouncilAgainst Smoking) actively works with the DoH

Source: South African Medical Research Council, Smoking and Health in South Africa: The need for action, Pretoria, 1988;,pp 104-16

Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

01:

28 0

6 D

ecem

ber

2014

I l l

5. TOBACCO CONTROL POLICY IN PRACTICE IN THE LATE1980s AND EARLY 1990s31

Despite the medical evidence that tobacco was hazardous to people'shealth, and despite sporadic pleas from medical associations to imposeeffective tobacco control measures, the South African government formany years did practically nothing to curb tobacco use. The only na-tional tobacco control measures introduced before 1990 were a volun-tary agreement not to advertise tobacco directly on television (1975),the introduction of a weak and very small health warning (1987), andthe banning of smoking on domestic flights (1989). The turning pointcame in 1991 when the official opposition in Parliament accused thethen newly-appointed Minister of Health, Rina Venter, of "protectingthe vested interests of the powerful tobacco industry, and not the peo-ple of the country".32 The background to this accusation was theSAMRC's 1988 report that highlighted the high cost of tobacco interms of life years lost and the implied financial cost. In response tothe opposition's attack, the minister promised to look into the possibil-ity of introducing tobacco control legislation.

The minister was urged on in her tobacco control efforts by the Tobac-co Action Group (TAG), an anti-tobacco alliance consisting of theHeart Foundation of Southern Africa, the Cancer Association of SouthAfrica and the National Council Against Smoking. According toYusuf Saloojee, who was the executive director of the National Coun-cil Against Smoking, the role of TAG was to mobilise extra-parlia-mentary support for the proposed tobacco control bill.33 The TobaccoProducts Control Bill was published for comment in March 1992 andprovided for the control of smoking in public places, the printing of

31. For a detailed history of tobacco control in South Africa, see M. Malan & R.Leaver, "Political change in South Africa: New tobacco control and public healthpolicies", in J. de Beyer & L. Waverley Brigden, Tobacco control policy:Strategies, successes & setbacks, Washington and Ottawa, 2003

32. Saloojee, op cit, p 16333. Ibid


Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

01:

28 0

6 D

ecem

ber

2014

112

prominent rotating health warnings and nicotine and tar content on to-bacco packaging and advertising, and the prohibition of sales to mi-nors. By international tobacco control standards, even at that time, thelegislation was mild. For example, the bill did not contemplate a banon tobacco advertising or promotions, nor did it intend to prohibitsmoking in the workplace.

In 1992 a survey was conducted for the Minister of Health, which in-dicated that about two thirds of people acknowledged the harmful ef-fects of active and passive smoking, and that most people supportedthe tobacco control measures that were being proposed.34 The bill alsoreceived the full support of Nelson Mandela.

The tobacco bill was strongly opposed by the industry. They made anumber of presentations to the Minister of Health and, together withsome strong support in the cabinet, were able to water down the draftlegislation.35 Despite industry opposition, the Tobacco Products Con-trol Act was approved by Parliament in June 1993, and became effec-tive in May 1995.

Even though the legislation was comparatively weak, Yusuf Saloojee,who was actively involved in lobbying for strong tobacco control leg-islation, was very complimentary towards Rina Venter, the Minister ofHealth, on the way she handled the legislative process.36 She certainlyhad a difficult job. On the one hand, medical research and the tobaccocontrol lobby groups persuaded her that tough legislation was re-quired. On the other hand, she had to persuade her colleagues to takeaction against an industry with which the ruling party had historicallyhad very close ties. In an attempt to be democratic and "encompass-ing" in the legislative process, she listened to the appeals of the tobac-

34. G. Martin, K. Steyn & D. Yach, "Beliefs about smoking and health and attitudestoward tobacco control measures", unpublished South African Medical ResearchCouncil (Essential Health Research Group) report, c1992

35. Saloojee, op cit, pp 164-6536. Ibid, p 166


Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

01:

28 0

6 D

ecem

ber

2014

113

co industry; and there are strong indications that they were able to wa-ter down the proposed legislation. In retrospect, this "encompassing"approach might have been a mistake, although the political impera-tives and the historical relationships may have forced her to take thisapproach. In contrast, her successor, Nkosazana Zuma, did not engagemuch with the tobacco industry when the Tobacco Products ControlAmendment Bill was debated in 1998, and the result was a much morecomprehensive and rigid piece of legislation. Nevertheless, the Tobac-co Products Control Act of 1993 represented the "first major dent inwhat has until now been a solid wall of vested interest".37

6. PROFESSIONAL ECONOMISTS ENTER THE DEBATE

6.1 Reekie and Wang, 1992

Reekie and Wang38 criticised the cost-benefit analysis of theSAMRC's 1988 report on the grounds that smokers have already dis-counted the hazards associated with smoking. They argued that a per-son's decision to smoke or not to smoke is influenced by the expectedutility obtained from smoking and the probability of smoking-inducedillness or death. On the basis of the perceived risk and benefits, peoplewill then decide to smoke or not. Using a state dependent approach tostandard decision making (ie where the outcome or consequence of anaction is uncertain and dependent on the state of nature), they showedthat, given people's different preferences, it is possible that smokingconfers benefits on (some) consumers.

In a related article in Business Day, Reekie explained that all costs as-sociated with smoking are internalised by smokers.39 He argued that

37. Ibid38. W.D. Reekie & L.F.S. Wang, "The benefits and costs of cigarette smoking - A

state dependent approach". Journal for Studies in Economics and Econometrics,16(1), 1992, pp 1-12

39. D. Reekie, "Smokers pay their way, and are not a burden on society", BusinessDay, 12 November 1992


Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

01:

28 0

6 D

ecem

ber

2014

114

"public policy on smoking - or anything else - is necessary only ifthere are external costs which cannot be internalised", implying thatgovernment intervention in the form of tobacco control policy was notnecessary or desirable. He dismissed the notion that smoking was ad-dictive, claiming that the large number of people that had successfullygiven up smoking suggests that smoking is no more than a bad habit.

In response, Saloojee and Yach argued that Reekie's claims that smok-ing is a "free choice" and that smoking is not addictive were untrueand at odds with scientific evidence.40 According to them, Reekie did"not take cognisance of the need to smoke to allay withdrawal symp-toms, of the desire of most smokers to stop, and of their failed at-tempts to quit". They dismissed Reekie's analysis on the basis that itwas based on false assumptions.

Abedian and Dorrington also strongly contested Reekie and Wang's"facade of technical, empirical and scientific sophistication", arguingthat the latter's theoretical approach and empirical research methodol-ogy were flawed.41 Abedian and Dorrington proposed that consumersare not as rational and capable of processing information as the stan-dard theory would lead one to believe. They argued that this, togetherwith the addictive nature of smoking, rendered Reekie and Wang's"contrived" results unacceptable.

6.2 Reekie 1994

In a subsequent paper Reekie42 set out to show that the cost benefitanalysis performed by the SAMRC43 was flawed, and that the benefits

40. Y. Saloojee & D. Yach, "Public interest in curtailing smokers' rights undeniable",Business Day, 23 November 1992

41. I. Abedian & R.E. Dorrington, "An evaluation of a recent attempt to assess thesocial benefits of cigarette smoking". Journal for Studies in Economics andEconometrics, 18(1), 1994, pp 59-72

42. W.D. Reekie, "Consumers' surplus and the demand for cigarettes", Managerialand Decision Economics, 15, 1994, pp 223-34

43. SAMRC (1988, 1992), op cit


Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

01:

28 0

6 D

ecem

ber

2014

115

obtained from smoking, in fact, exceeded the costs. First, he arguedthat the SAMRC understated the monetary benefits received fromsmoking. If total expenditure on tobacco is a cost (as the SAMRC as-sumes), then the equivalent amount, transferred to the factors of pro-duction, must be a benefit to society. He argued that, "the 'balancesheet', as presented, fails to meet the principles of double-entry book-keeping and fails... to define social costs and benefits. It is tautologicaland provides scope for (false) policy inferences only because it is in-complete".44 He argued that, should one apply this methodology toother goods and services, none would ever provide a net benefit to so-ciety.

According to Reekie the SAMRC assumed that smoking provides con-sumers with zero satisfaction.45 While the SAMRC grudgingly admitsthat smoking provides some satisfaction to smokers, it does not at-tempt to measure its magnitude, and thus this benefit did not enter theSAMRC's cost benefit analysis.46 This is a clear shortcoming in theSAMRC's analysis. Using the well-known concept of consumer sur-plus, Reekie's primary aim was to quantify the benefit that smokersderived from smoking, by estimating the consumer surplus. Reekie es-timated the consumers' surplus at nearly R2 billion. Since the calculat-ed consumers' surplus was greater than the SAMRC's estimate of thehealth care cost of smoking, Reekie concluded that "on the most pessi-mistic of assumptions... the data suggested that consumers still receivesubstantial net benefits from smoking".47

6.3 Van Walbeek 1996

Using Reekie's approach of measuring the consumers' surplus, VanWalbeek tried to determine how much additional revenue the govern-

44. Reekie (1994), op cit, p 22445. Ibid46. SAMRC (1988), op cit, p 8747. Reekie (1994), op cit, p 231


Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

01:

28 0

6 D

ecem

ber

2014

116

merit would be able to generate, should it set the tax rate at a revenue-maximising level.48 On the assumption that the supply of tobaccoproducts was perfectly elastic, he estimated demand equations for to-bacco, and used these to simulate how an increase in the excise taxwould effect tobacco consumption, prices, and government revenue.His main conclusions were as follows:49

" 1. The government can raise the excise rate to at least 110 per cent ofthe "producer price" of tobacco if it wishes to maximise its exciserevenue....50

"2. The government can expect to double (at least) its revenues fromtobacco if it increases the excise rate to these levels.

"3. The analysis suggests that raising the tobacco excise rate to theproposed levels could lead to a reduction in consumption of be-tween 41 and 46 per cent.

"4. The real retail price of cigarettes should rise by between 44 and122 per cent from their 1989 levels, if the government were tomaximise excise revenues from tobacco."

In June 1994 the newly-elected Government of National Unity an-nounced that it would increase the excise tax on cigarettes to 50 percent of the retail price, to be phased in over a number of years. TheMinister of Finance mentioned that the primary aim of the excise taxincrease was to reduce smoking, rather than raising revenue.Since the policy prescriptions of Van Walbeek's study closely corre-spond to the actual policy imposed by the government, one can retro-spectively evaluate his predictions against actual experience. The re-

48. C.P. van Walbeek, "Excise taxes on tobacco: How much scope does thegovernment have?", South African Journal of Economics, 64(1), 1996, pp 20-42

49. Ibid, p 3550. The "producer price" is defined as the "tax-free" or "pre-tax" price. Taking into

account a VAT rate of 14 per cent, this implies that the excise tax would comprise46 per cent of the retail price of cigarettes.


Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

01:

28 0

6 D

ecem

ber

2014

53n

in

n

TABLE 2

Predicted and actual changes in some tobacco measures, based on Van Walbeek

Variable Prediction byVan Walbeek

Excise tax as Revenuepercentage of maximising exciseretail price tax rate = 46 per

cent of retail price

Real govern- At least 100 perment revenue cent increasefrom tobacco

Consumption Decrease ofbetween 41 and 46per cent

Real retailprice

Increase bybetween 44 and122 percent

Actual outcome1989-2002

2002: excise taxequalled 33 per cent ofretail price; total taxincidence (ie VATincluded) = 45,2 percent of retail price

133 per cent increase

32 per cent decrease inaggregate consumption;49 per cent decrease inper capita consumption

130 per cent increase

Comments

Despite claims by the Minister of Finance that its self-imposedtarget of 50 per cent tax incidence has been achieved, this isnot so, for technical reasons. See ETCSA, 2003, p. 53 fordetails on this.

Increases in the population and real disposable incomeincrease tobacco consumption, which raised governmentrevenue by more than the predicted amounts.

Adult (15+) per capita consumption decreased by 49 per cent.Other factors (especially legislation of 1993 and 1999) arelikely to have had an additional depressing impact on tobaccoconsumption, over and above the price impact.

Retail price increased not only because of excise tax increases,but because the industry increased the retail price by more thanthe increase in the real excise tax since the mid-1990s.

Sources: C.P. van Walbeek, "Excise taxes on tobacco: how much scope does the government have?", South African Journal ofEconomics, 64(1), 1996, pp 20-42 and Republic of South Africa, Budget Review, Pretoria, various years

O

o

Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

01:

28 0

6 D

ecem

ber

2014

118

suits are shown in Table 2. Van Walbeek did not attach a time dimen-sion to his predictions, but a comparison of the actual changes in thevariables between 1989, the year for which the study was done, and2002 indicates that the predictions were generally quite accurate.

6.4 Economics of Tobacco Control in South Africa Project, 1996-1998

In 1996 the Economics of Tobacco Control in South Africa (ETCSA)Project was established at the Applied Fiscal Research Centre of theUniversity of Cape Town. Whereas earlier economic studies with a to-bacco control agenda were generally limited in focus, the aim of theETCSA Project was to perform a comprehensive investigation into theeconomic impact of tobacco control policies. The focus of the Projectwas on "finding the level of taxation that will best meet government'scompeting objectives, and not jeopardise the economy in any way".51

In line with international experience, the Project found that, of all to-bacco control instruments available, rapid increases in the real price ofcigarettes would be most effective in reducing cigarette consumption.The price elasticity of demand for cigarettes was estimated at about-0,6, which is typical for a developing country.

Other significant findings of the ETCSA Project were the following:

• An increase in the excise tax increases government revenue. Eventhough an increase in cigarette taxes reduces cigarette consumption,the decrease in consumption is much smaller than the increase inthe tax per cigarette, with the result that government revenue in-creases.

• By allowing the real excise tax to decrease by about 70 per cent be-tween the mid-1970s and the early 1990s, real government revenue

51. Economics of Tobacco Control in South Africa (ETCSA) Project, The Economicsof Tobacco Control in South Africa, report submitted to the International TobaccoInitiative. Cape Town, Applied Fiscal Research Centre, School of Economics,University of Cape Town, 1998


Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

01:

28 0

6 D

ecem

ber

2014

119

from cigarette taxation decreased, despite a rapid increase in ciga-rette sales over this period. Using demand analysis, the opportunitycost (in the form of foregone government revenue) of not raisingthe level of excise tax in line with inflation was measured andfound to be substantial.

• It was found that advertising expenditure has a small but positiveimpact on cigarette consumption. On the basis of this result, it wasargued that an advertising ban would reduce cigarette consumption.

• Using a social accounting matrix, the researchers investigated thelikely employment impact of a reduction in the demand for ciga-rettes. It was found that the decrease in employment in tobacco-re-lated sectors would be more than compensated for by increased em-ployment in other sectors, because consumers would switch theirexpenditure on cigarettes to other goods and services.

• In comparison with some Mediterranean and Eastern Europeancountries, South Africa does not have a significant cigarette smug-gling problem. The researchers speculated that, based on a courtcase between two major cigarette manufacturers, Rembrandt andPhilip Morris (which has subsequently changed its name to theAltria Group), there are indications that the industry may be in-volved in cigarette smuggling in South Africa.

The policy implications of the ETCSA Project's research were impor-tant. Together with inputs from the medical community and other anti-tobacco lobby groups, this research was used to urge the South AfricanMinistry of Health to impose more comprehensive tobacco controllegislation than the Tobacco Products Control Act of 1993.

One of the ETCSA Project's most publicised findings was that "a 1 percent increase in the growth in advertising expenditures will increasegrowth in demand for cigarettes by between 0,18 and 0,24 per cent".52

52. Ibid, p 77


Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

01:

28 0

6 D

ecem

ber

2014

120

At first glance, this finding provides the empirical justification for anadvertising ban: if advertising expenditure increases tobacco consump-tion, then a ban on tobacco advertising would, presumably, reduce to-bacco consumption, ceteris paribus. The typical industry argument isthat advertising has little, if any, impact on aggregate tobacco con-sumption, but that it has a marked impact on market shares.53 The in-ternational literature on the determinants of demand for tobacco is spliton this issue. Studies that have an apparent bias towards the industrytend to find insignificant relationships between advertising and con-sumption, while studies with an apparent bias towards tobacco controlinterventions tend to find a positive relationship.54

The ETCSA Project's finding on the positive relationship between ad-vertising expenditure and consumption was severely criticised byLeach55 and High.56 They pointed out that the consumption57 and ad-vertising expenditure58 data were incorrect, that the relationship isonly marginally significant, and only in certain sub-periods of the peri-od being investigated. In the public hearings in October 1998, Leachand High submitted their reservations about the ETCSA Project's re-

53. H. High, Does advertising increase smoking? Economics, free speech andadvertising bans, London, 1999, pp 18-22

54. See surveys in H. Saffer & F. Chaloupka, "The effect of tobacco advertising banson tobacco consumption." Journal of Health Economics, 19, 2000, pp 1117-37and High, op cit

55. D. Leach, "Submission to Parliamentary Commission on the Tobacco ProductsControl Amendment Bill", unpublished, 1998

56. High, op cit, pp 65-6957. Leach, op cit and High, op cit point out that the ETCSA Project's data on

cigarette consumption is much higher, and shows a much more rapid growth ratein the early 1990s, than is actually the case.

58. According to the ETCSA Project, "prior to 1990, reliable data [on advertisingexpenditure] was only available for 1980, 1986 and 1989. However, tobaccoadvertising for these years was a consistent five per cent of total advertisingexpenditure. Therefore, for the remaining observations, total advertising spendwas multiplied by 0,05". Leach pointed out, quite correctly, that for 17 of the 25observations on which the analysis was based, the values for cigarette advertisingwere "basically made up".


Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

01:

28 0

6 D

ecem

ber

2014

121

suits to the Portfolio Committee on Health. Their submissions did notreceive much media attention, possibly because of the technical natureof their reservations. Despite their criticism, the legislation waspassed, but certainly not because of the rigour of the ETCSA Project'seconometric work on this issue.

Despite these criticisms, the overall impact of the ETCSA Project wasprofound. It provided the anti-tobacco lobby, and the Minister ofHealth in particular, with the empirical research to show that the eco-nomic impact of the proposed tobacco control legislation was not asdetrimental on the economy as the tobacco and related industries want-ed people to believe.

6.5 Economics of Tobacco Control in South Africa Project, Phase II,2000-2003

After the first phase of the ETCSA Project was completed in 1998, asecond phase was initiated in 2000, and completed in 2003.59 Whereasthe first phase of the Project had much policy impact locally, the sec-ond phase has had a much stronger international focus. The aim of thesecond phase was to extend the economic analysis and policy implica-tions of the original study, based on the possibility that South Africacould act as a role model in tobacco control for other developing coun-tries.

One of the main focus points in second phase of the Project was toconsider the impact of excise tax increases on the regressivity of theexcise tax. It was shown that, although the cigarette excise tax is re-gressive, the regressivity of the tax had decreased significantly duringthe 1990s. This is explained by the fact that the real price of cigarettes

59. Economics of Tobacco Control in South Africa Project (ETCSA) (Phase II), TheEconomics of Tobacco Control in South Africa, report submitted to Research forInternational Tobacco Control and the World Bank, Cape Town, Applied FiscalResearch Centre, School of Economics, University of Cape Town, 2003


Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

01:

28 0

6 D

ecem

ber

2014

122

increased sharply during the 1990s, and that poor people tend to bemore sensitive to price changes than the rich. Another important find-ing of the second phase of the ETCSA Project concerned the pricingstrategy of the tobacco industry. Given its high degree of concentra-tion, the tobacco industry was able to increase the real price of ciga-rettes by much more than the increase in the excise tax. The apparentrationale of this strategy was to increase industry revenue, rather thansales volumes. In simulation analysis it was shown that this strategyhas materially increased the industry's profitability, despite the factthat the quantity of cigarettes sold in South Africa decreased by abouta third during the 1990s.

7. TOBACCO CONTROL POLICY IN PRACTICE AFTER THEDEMOCRATIC TRANSITION IN 1994

In contrast to the outgoing National Party (NP) government, in 1994the newly-elected ANC government had no historical ties with the to-bacco industry. The long and close relationship between the NP andthe tobacco industry had resulted in a hesitant and cautious tobaccocontrol policy. Coining into power with a clean slate, the ANC was notbound by the informal agreements of the past. Even before assumingpower, the ANC made clear its position on tobacco control. In May1992, on World No-Tobacco Day, Nelson Mandela committed theANC to a strong tobacco control policy and publicly supported the To-bacco Products Control Bill that was being debated at the time. In No-vember 1993, at the All Africa Conference on Tobacco or Health,Nkosazana Zuma, the future Minister of Health, confirmed the ANC'santi-tobacco stance. She argued that a comprehensive tobacco controlprogramme would be an important component of the new govern-ment's commitment to improve the health of the population.60

60. D. Yach & S. Harrison, "Proceedings of the All Africa Conference on Tobacco orHealth, 14-17 November 1993, Harare, Zimbabwe", unpublished SAMRC report,1994, p 9


Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

01:

28 0

6 D

ecem

ber

2014

123

It did not take long for the new government to start implementing itstobacco control policy. In June 1994, less than two months after theANC took power, the Minister of Finance announced that

The increase in the excise duty on tobacco products is aspecial case [compared to other excisable products]. Ar-guments from the health community indicated a prefer-ence for an increase in the excise duty to fifty per cent ofthe retail price, which is the order of impost in many oth-er countries. After consultations with all interested groupsand taking into account industry-specific limitations andmarket conditions, Government has opted for a phasedapproach, which is reflected in the announced increase.Future budgets will have to deal with the remainder of the

This announcement was a clear victory for the health community, eventhough they were disappointed with the size of the tax increase in thatyear and the following two years. However, in 1997 the new Ministerof Finance, Trevor Manuel, announced a 52 per cent increase in theexcise tax, which would, he claimed, raise the expected total tax inci-dence to 50 per cent of the average retail price.62 Between 1997 and2003 the annual increases in the excise tax on cigarettes have beenquite predictable, with the government adjusting the excise tax in linewith the average cigarette price increase, so as to maintain the 50 percent tax incidence. In the 2004 budget speech the tax incidence was in-creased to 52 per cent of the retail price, to be maintained for at leastthe following three years.

As one could expect, the "excessive" excise tax increases were strong-ly opposed by the tobacco industry. They argued that they were a le-gitimate industry and that the excise tax increases unfairly discriminat-ed against them. Also, they claimed that the rapid excise tax increases

61. Republic of South Africa, Budget Review, Pretoria, 1994, p 5.762. Republic of South Africa, Budget Review, Pretoria, 1997, p 7.16


Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

01:

28 0

6 D

ecem

ber

2014

124

would exacerbate the smuggling problem, which, in their opinion, wasalready out of control, and that this would negatively affect govern-ment revenue.63

However, a long-term view of cigarette taxation indicates that the rap-id tax increases in the mid- and late 1990s was simply a reversal of a20-year trend of decreasing excise taxes. Between 1970 and the early1990s the real level of excise tax had fallen by 70 per cent. This hap-pened because the government allowed inflation to erode the real val-ue of the excise tax. Despite the rapid increases of the 1990s, the levelof real excise tax on cigarettes in 2002 was still lower than in the1960s and early 1970s.

The next major tobacco control move came in January 1998 when theMinister of Health announced that a Tobacco Products ControlAmendment Bill would be tabled in Parliament that year. The ration-ale, she argued, was to "protect children and to prevent them from be-ing bombarded with pro-smoking messages".64 The two most impor-tant, and controversial, provisions of the bill were the banning of alltobacco advertising and sponsorship, and the prohibition on smokingin all public and workplaces.

The cabinet approved the bill six months after the minister had an-nounced her plans. Shortly after the bill was published for public com-ment a few weeks later, the tobacco industry applied for an urgent in-terdict to stall the legislation, citing "a lack of consultation" in thebill's drafting process, and wanting access to the documents on whichthe legislation was based.65 Throughout the legislative process, the in-

63. Rembrandt Group, "An open letter to Dr. Nkosazana Zuma", advertisement inSunday Times, 20 October 1996

64. R. Leaver, "Tobacco control policy in South Africa through the eyes of the printmedia", in Economics of Tobacco Control in South Africa Project (Phase II), TheEconomics of Tobacco Control in South Africa, report submitted to Research forInternational Tobacco Control and the World Bank, Cape Town, Applied FiscalResearch Centre, School of Economics, University of Cape Town, 2003, p 25

65. Malan & Leaver, op cit, p 148


Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

01:

28 0

6 D

ecem

ber

2014

125

dustry complained that the Minister of Health would not consult withthem. In what the minister described as a "victory in the parliamentaryand legislative process", the industry's application was dismissed afew days later.66 The National Council of Provinces approved the billin October 1998. Public hearings into the proposed legislation wereheld later that month. The hearings before the Portfolio Committee onHealth attracted much media attention and were the climax of manyweeks of intense lobbying by both supporters of and opponents to theproposed legislation.

Opponents to the proposed legislation were against the banning of to-bacco advertising and sponsorship on the grounds that: (1) advertisingfor products that have achieved the "mature" stage of the product lifecycle (like cigarettes) is aimed primarily at increasing and maintaininga brand's market share, rather than increasing the overall size of themarket,67 (2) there is no empirical evidence to support the hypothesisthat advertising increases tobacco consumption, (3) the ban on adver-tising is an infringement of the right to free speech, (4) the ban on ad-vertising would have a very detrimental impact on the advertisingbusiness, and the ban on sponsorships would jeopardise many sportsbodies, and (5) should the government decide to ban tobacco advertis-ing on the grounds that it is potentially harmful, before long the adver-tising of other products, like alcohol, cars and unhealthy foods, will beprohibited as well (the so-called slippery slope argument).68

Proponents of the legislation, on the other hand, argued that tobaccoadvertising falsely presents smoking as a pleasurable and glamorousactivity. They argued that advertising distorts the consequences ofsmoking and, rather than informing, prohibits people, and especially

66. Leaver, op cit, p 2867. High, op cit, p 2068. For a detailed account of the arguments made before the Portfolio Committee, see

C.P. van Walbeek, "Arguments for and against advertising bans and clean indoorair policies: The South African experience", unpublished paper, 2001


Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

01:

28 0

6 D

ecem

ber

2014

126

children, from making an informed choice about it.69. It was arguedthat tobacco advertising, rather than being an expression of freespeech, in fact creates its own censorship; many magazines and news-papers are so dependent on tobacco advertising that they would notpublish anything that would anger the industry, out of fear of losingtheir tobacco accounts. Furthermore, they argued that, despite the to-bacco industry's claims that they do not target children in their mar-keting endeavours, secret documents reveal the opposite.70

Arguments against the prohibition of smoking in public places andworkplaces were primarily based on financial and economic grounds.The hospitality industry claimed that, should this bill become law, theywould lose a large proportion of their customers. Proponents of the"clean indoor air" legislation argued that the rights of non-smokers hadto be protected. They argued that medical science had shown beyondreasonable doubt that environmental tobacco smoke (ETS) is morethan just a nuisance, but is in fact harmful to non-smokers. Whereaspreviously smokers had the right to pollute the air with ETS, this legis-lation would grant non-smokers the right to smoke-free air.71

Despite strong opposition - about three quarters of the submissions atthe public hearings were against the proposed legislation - the Minis-ter of Health was undeterred. After the Portfolio Committee on Healthapproved the bill, it was presented to Parliament's House of Represen-tatives, where it was approved with 213 votes in favour and 106

69. A. Seidel Marks, Submission in support of the Tobacco Product ControlAmendment Bill, submission to the Parliamentary Committee on the TobaccoProducts Control Amendment Bill, 1998

70. In 1997, a US court ordered that millions of tobacco documents be made availablefor public scrutiny. Numerous anti-tobacco organisations have trawled throughthese documents and have compiled summary reports that indicate that thetobacco industry has, over a long period of time, been guilty of dishonest andunethical conduct. A particularly thorough report, "Trust us, we're the tobaccoindustry", can be accessed at:

http://www.ash.org.uk/html/conduct/html/trustus.html71. Van Walbeek (2001), op cit


Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

01:

28 0

6 D

ecem

ber

2014

127

against.72 The ANC and the African Christian Democratic Party votedin favour of the legislation. The final hurdle was for PresidentMandela to sign the bill into law. However, the bill was subject to anunexpected delay when President Mandela requested that some defini-tions in the bill be clarified because they were too vague, and might besubject to legal challenge. This was done and after both houses of Par-liament had approved the amendments, the President signed the To-bacco Products Control Amendment Act73 in April 1999.74

If one compares the legislative process followed by the AmendmentAct of 1999 to the original Tobacco Products Control Act of 1993, anumber of differences are clearly evident. First, the 1999 legislationwas much more comprehensive than the 1993 legislation. The industrywas easily able to exploit loopholes in the 1993 legislation. For exam-ple, advertisements that promoted tobacco-sponsored events were notrequired to carry health warnings, on the grounds that this was "indi-rect" advertising. Predictably, after 1993 there was a rapid increase inthe advertising of sponsored events. Having learnt from her predeces-sor's experience of the 1993 legislation, the Minister of Health took amuch harder line in 1998. Despite industry pleas for "appropriate" and"reasonable" legislation (which could, presumably, be easily circum-vented) the minister made the restrictions on smoking in public placesand on tobacco advertising as watertight as possible.

The second difference was the way in which the two ministers allowedthe industry to influence the legislative process. In 1993, the Ministerof Health, Rina Venter, allowed the industry to make representationsto her, and they successfully slowed down the process, and watereddown the legislation. In 1998 the new Minister, Nkosazana Zuma,largely ignored the industry's pleas to consult with her, arguing thatthey have no constructive role to play. At one point she remarked:"What is consultation? We did consult them, but consultation does not

72. Leaver, op cit, p 3073. Act 12 of 199974. Leaver, op cit, p 30


Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

01:

28 0

6 D

ecem

ber

2014

128

mean consulting until they agree. They will never accept tobacco re-strictions no matter how long we speak to them".75

The third difference was the support that the two Ministers of Healthreceived from their colleagues in the cabinet. Rina Venter had the oddsstacked against her. Many colleagues, including the President, F.W. deKlerk, were chain smokers and only grudgingly supported her tobaccocontrol efforts. The Minister of Agriculture openly supported tobaccofarmers in their fight against the proposed legislation;76 and by "buy-ing favour" with the governing party, the industry was able to place a"moral indebtedness" on the government.77 On the other hand,Nkosazana Zuma was operating in a much more favourable environ-ment: President Mandela openly supported tobacco control measures;the ANC's focus on primary health care implied a strong focus on to-bacco control; and the ANC did not have any long-term relationshipwith the tobacco industry that could possibly influence its policy deci-sions.

After a delay of more than a year, the regulations enforcing the Actwere published in September 2000. The Act came into effect in Janu-ary 2001. As a result of continuous pressure, the new Minister ofHealth, Mantombazana Tshabalala-Msimang, allowed the hospitalityindustry to set aside a maximum of 25 per cent of their floor space tosmokers, provided the smoking area was enclosed and separately ven-tilated. According to Saloojee78 this was a reasonable compromise to"guarantee non-smokers the right to clean air, while taking account ofthe need of smokers".

The degree of compliance to the new legislation has been relativelyhigh. All visible tobacco advertising and sponsorship has disappeared,although there has been an apparent increase in "below the line" ciga-

75. Ibid, p 32, citing The Star, 7 August 199876. Ibid, pp 14-1577. Malan & Leaver, op cit, p 12378. Y. Saloojee, personal communication, 2004


Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

01:

28 0

6 D

ecem

ber

2014

129

rette advertising. According to a survey performed in October 2002 inthree of the nine provinces, more than 90 per cent of respondents indi-cated that their workplaces have some form of smoking restrictions; 30per cent of hospitality establishments are completely smoke-free, 30per cent have a separate smoking section, and 40 per cent (mainlysmall and rural establishments) do not comply with the legislation.79

In October 2003 the Ministry of Health announced further amend-ments to the legislation. Other than closing some loopholes in theAmendment Act of 1999 (eg by disallowing "below the line" advertis-ing in the form of cigarette parties, and product stacking at the point ofsale) and increasing penalties, the amendment bill aims to:(1) introduce pictorial health warnings, as in Canada and Brazil,(2) ban "misleading" descriptors like "mild", "light" and "low tar",(3) ban the sale of "Duty-Free" and "Tax-Free" tobacco products, and(4) ban smoking in certain outdoor public places and within five me-tres of doorways and entrances.80

8. CONCLUSION

The debate about appropriate tobacco control policy in South Africahas been particularly acrimonious during the 1990s, given the magni-tude of the interests at stake. The tobacco industry has tried to empha-sise its economic importance, especially as a source of revenue for thegovernment and as a source of employment for many people. On theother hand, proponents of stronger tobacco control policy have arguedthat public health should take preference over economic interests, andthat the economic contribution of the tobacco industry is more illu-sionary than real. While the tobacco industry was able to rapidly in-

79. F. Steyn, M. Engelbrecht, C. Ngwena & D. van Rensburg, Assessing the levels ofcompliance of public places with current tobacco legislation, report to theNational Department of Health by the Centre for Health Systems Research andDevelopment, University of the Free State, Bloemfontein, 2003

80. Republic of South Africa, Government Gazette, 17 October 2003


Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

01:

28 0

6 D

ecem

ber

2014

130

crease its sales quantities in the period before 1990, the tide turned inthe early 1990s. At first the turning of the tide was hesitant, but after1994 it gained momentum when the newly-elected democratic govern-ment announced that it would increase the excise tax and imposestronger tobacco control legislation.

So what has been the impact of these legislative changes? Empirically,it is difficult to measure their impact on tobacco consumption. Giventhe paucity of appropriate tobacco data in South Africa and the factthat it is practically impossible to separate the effects of the variouslegislative interventions, one cannot determine accurately by howmuch cigarette consumption has decreased as a result of these legisla-tive interventions, ceteris paribus. Even the international literaturesuggests that the direct impact of legislative interventions is muchsmaller than changes in the tax on and price of cigarettes (eg Bardsleyand Olekalns, 1999 and Chaloupka and Warner, 1999).81

In this situation, why are tobacco control advocates so partial to legis-lative interventions? Primarily because it creates a social environmentwhere tobacco use is no longer regarded as normal and acceptable.There is nothing glamorous about smoking in the cold outside or be-tween the pigeon droppings. When smoking is denormalised, it createsa platform where economic disincentives to smoke (ie price increases)are more effective in encouraging people to quit smoking and in pre-venting youngsters from starting to smoke. Also, through "clean in-door air" policies property rights are transferred from smokers to non-smokers. Whereas previously the right to pollute the air with environ-mental tobacco smoke rested, by social convention, with smokers, thetobacco control legislation has granted non-smokers the right tosmoke-free air. Even though the legislation allows for fines to be im-posed on offenders of the "clean indoor air" legislation, this is not the

81. P. Bardsley & N. Olekalns, "Cigarette and tobacco consumption: Have anti-smoking policies made a difference?", The Economic Record, 75(230), 1999, 225-40 and F. J. Chaloupka & K.E. Warner, The economics of smoking, WorkingPaper 7047, National Bureau of Economic Research, Cambridge (Mass), 1999


Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

01:

28 0

6 D

ecem

ber

2014

131

primary point. The point is that non-smokers can now demand smoke-free air, and have a law to back up their demands.


Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

01:

28 0

6 D

ecem

ber

2014

tobacco control in south africa in the 1990s: a mix of advocacy, academic research and policy

Documents