Background: Increasing the price of tobacco products is arguably the most effective method of curbing the prevalence and consumption of tobacco products. Price increases would reduce the global burden of disease brought about by tobacco consumption.
Objectives: To compare cigarette price data from more than 80 countries using varying methods, examine trends in prices and affordability during the 1990s, and explore various policy implications pertaining to tobacco prices.
Design: March 2001 cigarette price data from the Economist Intelligence Unit are used to compare cigarette prices across countries. To facilitate comparison and to assess affordability, prices are presented in US dollars, purchasing power parity (PPP) units using the Big Mac index as an indicator of PPP and in terms of minutes of labour required to purchase a pack of cigarettes. Annual real percentage changes in cigarette prices between 1990 and 2000 and annual changes in the minutes of labour required to buy cigarettes between 1991 and 2000 are also calculated to examine trends.
Results: Cigarette prices tend to be higher in wealthier countries and in countries that have strong tobacco control programmes. On the other hand, minutes of labour required to purchase cigarettes vary vastly between countries. Trends between 1990 and 2000 in real prices and minutes of labour indicate, with some exceptions, that cigarettes have become more expensive in most developed countries but more affordable in many developing countries. However, in the UK, despite recent increases in price, cigarettes are still more affordable than they were in the 1960s.
Conclusions: The results suggest that there is ample room to increase tobacco prices through taxation. In too many countries, cigarette prices have failed to keep up with increases in the general price level of goods and services, rendering them more affordable in 2000 than they were at the beginning of the decade. Opportunities to increase government revenue and improve health through reduced consumption brought about by higher prices have been overlooked in many countries.
- health gains
- CPI, consumer price index
- EIU, Economist Intelligence Unit
- EU, European Union
- FCTC, Framework Convention on Tobacco Control
- GCC, Gulf Cooperation Council
- LCU, local currency unit
- PPP, purchasing power parity
- UBS, Union Bank of Swizerland
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
- CPI, consumer price index
- EIU, Economist Intelligence Unit
- EU, European Union
- FCTC, Framework Convention on Tobacco Control
- GCC, Gulf Cooperation Council
- LCU, local currency unit
- PPP, purchasing power parity
- UBS, Union Bank of Swizerland
In a 1999 report, the World Bank examined the effectiveness of an array of interventions and concluded that both price (taxes) and non-price measures (advertising bans, information campaigns, smoking restrictions, etc) can reduce the demand for cigarettes.1 This article will briefly review the effect that price increases have on smoking behaviour and compare cigarette price data from more than 80 countries using different methods. As well, trends in prices and affordability during the 1990s will be examined* Finally, this article will explore various policy implications that pertain to tobacco prices.
PRICE INCREASES AND SMOKING BEHAVIOUR
Increasing the price of tobacco products is arguably the most effective method of curbing the prevalence and consumption of tobacco products. Individuals who do not use tobacco may refrain from starting, and thus avoid addiction. It can also induce current users to consume less tobacco or persuade them to quit, as well as prevent ex-users from starting again†. Price increases would therefore reduce the global burden of disease brought about by tobacco consumption.
In a 1999 report, the World Bank concludes that on average, a price rise of 10% would be expected to reduce demand for tobacco products by about 4% in high income countries and by about 8% in low and middle income countries.1 Using a model of cohort smokers alive in 1995, it is estimated that tax increases that would raise the real price of cigarettes by 10% worldwide would cause about 42 million of these smokers to quit and prevent a minimum of 10 million tobacco related deaths.6 These conclusions have tremendous implications for public health.
Furthermore, increases in tobacco prices affects more the behaviour of the young and the poor who tend to be more responsive to price changes than older and wealthier individuals for several reasons. Firstly, because of the addictive nature of tobacco, long term users are less able to curb consumption and therefore adjust less rapidly to changes in tobacco prices, compared to younger individuals who may not yet be addicted to nicotine.7 Secondly, youth smoking is said to be determined more by peer behaviour than adult smoking—that is, an increase in price will first reduce the number of young smokers; then, through less peer smoking, it will again reduce the number of young smokers, hence multiplying the effect of price changes.7 Thirdly, youths and the poor spend a larger share of their relatively smaller disposable income on tobacco than wealthy adults do.8 Therefore, these individuals tend to be more responsive to increases in the price of tobacco products. There is supporting evidence. For example, in the UK, it was found that the price responsiveness was inversely related to social class.9 In the USA, less educated individuals were found to be more responsive to price than educated ones,4,10 and smokers from lower income and minority groups were more likely to quit in response to price increases.5 Similarly, in South Africa, young adults (16–24 years old) and low income individuals appeared to be more responsive to increases in prices.11
The tobacco industry realises the implications that higher taxes would have on their sales volume. Secret industry documents obtained in US litigation, from Philip Morris and British American Tobacco, express well the industry's concerns:
“Of all the concerns, there is one—taxation—that alarms us the most. While marketing restrictions and public and passive smoking do depress volume, in our experience taxation depresses it much more severely. Our concern for taxation is, therefore, central to our thinking about smoking and health.”12
“Increases in taxation, which reduce consumption, may mean the destruction of the vitality of the tobacco industry.”13
It is therefore not surprising that the tobacco industry vehemently opposes increases in tobacco taxes. The tobacco industry usually contends that increasing tobacco taxes will inevitably lead to illegal contraband of tobacco products, notably cigarettes. Discrepancies in tobacco prices between countries, it is argued, create an incentive to smuggle.14 In 1994, this argument persuaded the Canadian government to lower dramatically its tobacco taxes. Sweden in 1998 and more recently the Ukraine, too, lowered their taxes or excise duties in the hope that it would diminish the magnitude of the contraband market. Not surprisingly, government's revenue from tobacco dipped after the tax cut in Sweden, Canada, and the Ukraine.14,15 It is also important to note that there does not appear to be any documented cases of reduced tax revenues when tobacco taxes were increased.14 Moreover, the World Bank stresses that the determinants of smuggling are much more than price alone. Using indicators of corruption levels based on the Transparency International's Index, the World Bank observed that the level of tobacco contraband tends to increase with the degree of corruption in a country.1
For many years, many have believed that the tobacco industry was not only trying to influence public policy by publicising the myth of the negative impact that tobacco taxes could bring, but also that it was involved directly or indirectly in the contraband of cigarettes. Recent developments from Canada and the UK seem to indicate that these fears were not unfounded. In 1997 two British American Tobacco managers pleaded guilty to charges related to tobacco smuggling between Canada and the USA.16 In early 2000, the Guardian and the International Consortium of Investigative Journalists published revelations about British American Tobacco's involvement with smuggled tobacco products17 which prompted the UK Department of Trade and Industry to launch an investigation in October 2000. In August 2001, the European Commission filed a complaint on behalf of the European Community and several member states against two US cigarette manufacturers‡.
SOURCES AND METHODS
The Economist Intelligence Unit (EIU)§ conducts every six months a worldwide cost of living survey, collecting prices in about 130 cities covering nearly 90 countries per survey. The price data for over 160 items are collected in the first week of March and September and include three tobacco products: (1) cigarettes, Marlboro or nearest equivalent international brand¶ (pack of 20); (2) cigarettes, local brand (packs of 20); and (3) pipe tobacco, MacBaren type (50 g).
March 2001 cigarette price data are presented in local currency units (LCU) and in US dollars in order to allow for some comparison between countries. The US dollar figures were calculated from the EIU cigarette price expressed in local currencies and the exchange rate at the time of the survey (exchange rate data provided by the EIU). For countries where prices were sampled in more than one city, averages of all the city prices were calculated**.
Price comparisons in terms of US dollars are often flawed. The fundamental problem in comparing tobacco prices in US dollars stems from the fact that there are numerous determinants of exchange rates, they are influenced not only by inflation differentials, but also by interest rate differentials, current account deficit, political stability, etc. A more appropriate measure of comparison would therefore be based on the theory of purchasing power parity (PPP). The PPP conversion factor (the number of units of a country's currency required to buy the same amount of goods and services in the domestic market as $1 would buy in the USA) exists, but price data become available much sooner than PPP conversion factor estimates. It is therefore not feasible to construct recent cigarette affordability indices using available PPP conversion factors. There is also considerable debate over the choice of an appropriate “basket” for making purchasing power comparisons.18
The Economist circumvented these problems by using the price of a Big Mac (McDonalds hamburger) as an indicator of PPP. Using the Big Mac is an attractive option because its composition is uniform in most countries. However, Big Mac local prices may be distorted by trade barriers on beef, sales tax or significant differences in the cost of inputs such as rents. More recently, it was pointed out that bovine spongiform encephalopathy may soon begin to distort Big Mac prices in Europe††. Despite these flaws, several academic studies have concluded that the Big Mac index is not only a good indicator of PPP, it is also an unexpectedly accurate predictor of exchange rates in the long run.18–20
Using Big Mac prices to assess the affordability of cigarettes was proposed by Michelle Scollo21 in 1996 and provides a light hearted comparison tool for recent cigarette price estimates by weighting cigarette prices by the Big Mac PPP index. Simply put, the Big Mac PPP index is the exchange rate that would mean hamburgers cost the same in the USA as abroad. Big Mac prices published in The Economist were used to construct an affordability index. March 2001 cigarette prices were in turn weighted by the implied Big Mac PPP.
The World Health Organization also proposed to assess tobacco affordability by examining how many minutes of labour are required to purchase a pack of cigarettes.22 The Union Bank of Switzerland (UBS) Swiss Economic Research conducts a survey of international prices and wages in more than 50 cities every three years. Data from the most recent survey, conducted in the second quarter of 2000,23 were utilised with the September 2000 EIU tobacco price data to calculate how many minutes of work‡‡ are required to purchase one pack of Marlboro and one pack of local brand cigarettes.
Annual real percentage changes in price between 1990 and 2000 were calculated by taking the percentage difference in local currency prices while taking into account or “discounting” for inflation. Inflation is the proportionate rate of change in the general price level within each country. The calculation was facilitated by creating an inflation index (CPI) based upon the consumer price index estimates provided by the EIU. Unless otherwise noted, the data shown represent prices in the first week of September of 1990 and 2000. One cannot understate the importance to correct nominal changes in price for changes in inflation. For example, the price of one pack of Marlboro in Turkey jumped from 2500 Lira in 1990 to 875 000 Lira in 2000. However, during the same period, general prices increased by about 28 000%. Even in countries with low inflation, it is instrumental to correct for inflation. For example, in an economy in which inflation averages just 5% annually, prices would more than double in less than 15 years.
It is important to note that an increase in the real price of cigarettes does not necessarily mean that cigarettes are less affordable (or more costly) since increases in income levels are not taken into account. Generally, income is positively related to cigarette consumption.24 That is, everything else remaining the same, an increase in income will lead to an increase in tobacco consumption. In other words, even if the real price increased, increased income may offset, partially or even fully, the increase in real price. To control for changes in income levels, annual changes between 1991 and 2000 in the minutes of labour required to buy cigarettes were calculated based on the UBS 1991 and 2000 surveys.
Table 1 presents March 2001 cigarette prices in LCUs and in US dollars for 87 countries, provinces, and territories. As expected, cigarette prices are higher in wealthier countries and in countries that have strong tobacco control programmes such as the UK, Norway, Hong Kong Special Administrative Region of China, New Zealand, and Australia.
Figures 1 and 2 presents Marlboro and local brand price data in terms of their domestic affordability calculated from prices in local currencies weighted by the Big Mac implied PPP conversion factors. As expected, countries in which tobacco taxes were used as a pubic health instruments, such as Hong Kong Special Administrative Region of China, the UK, Sweden, New Zealand, and Ireland, appear towards the top of the index while countries like Switzerland and Japan where the household incomes are very high and tobacco control policies poor show up towards the bottom of the scale.
Table 2 presents cigarette affordability in terms of minutes of labour required to buy one pack of cigarettes in 56 cities. In order to also allow us to assess the affordability of cigarettes within countries, these results are presented alongside minutes of labour required to purchase one Big Mac, 1 kg of bread, and 1 kg of rice calculated by UBS. Minutes of labour required to purchase cigarettes vary vastly between countries, from about 10 minutes in Japan and Switzerland to close or more than 100 minutes in Kenya and India. It is interesting to compare the affordability of cigarettes relative to that of bread, rice, and a Big Mac. In most developed countries, cigarettes are more expensive than bread, rice, and Big Macs while in many developing countries cigarettes are cheaper. In some developing countries such as India, Panama, and Kenya, cigarettes appear prohibitively expensive which may explain why prevalence and consumption of manufactured tobacco tends to be very low in many low income countries.
Figures 3 and 4 provide trends in Marlboro and local brand price data from September 1990 to September 2000. Although cigarettes prices are determined by many factors other than taxes—and it is important to understand these competing determinants—what is ultimately more important for public health is how affordable tobacco products are. Even though the changes in cigarette prices presented may only partially reflect changes in domestic tobacco tax policy that have occurred during the 1990s, these estimates provide vital insight into trends in the price of cigarettes.
The diverging trends in price changes between certain countries are startling. For example, in the European Union (EU), while the real price of both local brand and Marlboro cigarettes increased by more than 5% per year in France and the UK, it remained fairly stable or even decreased in Austria, Germany, and Denmark. With the exception of Poland, the real prices of cigarettes have decreased during the past decade among EU applicants. Real Marlboro prices have decreased in more countries than local brand prices (33 out of 68 v 19 out of 64). These results are not surprising considering the recent evidence uncovered by The Economist alleging that major transnational companies conspired to fix cigarette prices.25 The companies were not initially fixing prices high to raise margins; they were allegedly fixed to mislead governments into thinking that new foreign brands would not become very popular, and thus avoid strong and effective tobacco control measures.
Table 3 presents annual changes in the minutes of labour required to purchase one pack of Marlboro and one pack of local brand cigarettes between 1991 and 2000. Trends in minutes of labour required to purchase cigarettes show somewhat similar trends to those witnessed by changes in real prices§§. In 11 countries out of 42, cigarettes were more affordable in 2000 than they were at the beginning of the decade. The direction of the changes in prices were mostly similar but not identical while the magnitude of the change in prices showed no discernible pattern. Comparable data were available for 35 countries and the direction of the changes in Marlboro and local brand prices were the same in 29 and 28 countries, respectively.
These results also indicate, with some exceptions, that over the past 10 years, cigarettes have become more expensive in most developed countries, but relatively more affordable in many developing countries¶¶. It is also quite troubling to see that cigarette prices have decreased by more than 50% between 1990 and 2000 in the Islamic Republic of Iran, Egypt, and Viet Nam while they steadily increased in countries that have fairly strong tobacco control programmes such as Australia, Hong Kong Special Administrative Region of China, New Zealand, Singapore, and South Africa. In the UK, despite recent increases in price, cigarettes are still more affordable than they were in the 1960s.26
Ample room for increases in tobacco taxes
The trends outlined above suggest that there is ample room to increase tobacco taxes. In too many countries, cigarette prices have failed to keep up with increases in the general price level of goods and services, rendering them more affordable in 2000 than they were at the beginning of the decade. Opportunities to increase cigarette prices, increase government revenue, and improve health have been overlooked in many countries.
Recently, several countries have acknowledged the impact that increases in tobacco prices can have on the health of their population. In December 1998, in “Smoking Kills –A White Paper on Tobacco”, the UK Secretary of State for Health recommended that tobacco prices be increased to improve health outcomes.26 The UK Chancellor announced increases in tobacco taxes of, on average, at least 5% a year in real terms. Similarly, in September 1999 in France, the “Rapport Recours” on Health Policy and Tobacco Fiscal Policy recommended a 20% increase in the price of tobacco products to discourage smoking.27 Armenia, Canada, France, Ireland, South Africa, Thailand, and the UK are examples of this increasing commitment by national governments to use fiscal policy to advance public health.
Tax all tobacco products
The data presented above, combined with the price data available for pipe tobacco (not shown here), suggests that in many countries there are wide discrepancies in changes in cigarette prices across brands and across tobacco products. In order to maximise the policy objective, taxes should be implemented uniformly to all products so as not to encourage substitution.
Earmark revenues to fund tobacco control
The WHO recommends earmarking a portion of government revenues from tobacco to fund activities that will advance tobacco control and activities that will ease the effects of short-run transition of tobacco farmers, whose livelihoods may be affected by reduced consumption.28
Recently, several countries have decided to fund tobacco control activities or broader public health programmes through tobacco higher taxes. In the UK in November 1999, Gordon Brown, the Chancellor of the Exchequer, announced that the National Health Service was to benefit from increases in tobacco taxes.29 In December 1999, Ireland's Minister for Finance and Minister for Health and Children announced that the revenue equivalent to a new tax increase would fund health provisions. During the 52nd World Health Assembly, Saudi Arabia's health minister proposed to Gulf Cooperation Council members to standardise the levy imposed on tobacco products and that 5% be dedicated to tobacco control activities.30 To date Qatar has earmarked tobacco taxes to fund similar efforts. Other countries such as Australia, Egypt, the Islamic Republic of Iran, Thailand, and several US states such as California and Massachusetts earmark a portion of tobacco taxes to fund tobacco control programmes activities such as counter-advertising or broader public health activities.31,32
As indicated earlier, discrepancies in tobacco prices between countries can create an incentive to smuggle. However, neighbouring countries can minimise the incentive to smuggle by harmonising taxes on tobacco products. The EU adopted three directives in October 1992 (92/78/EEC, 92/79/EEC, and 92/80/EEC) that aimed at decreasing tobacco price disparities between EU member countries. The first directive defined the taxation structure of tobacco products while the latter two fixed a minimum tax level of at least 70% of the retail price. These three directives have been in force since 1 January 1993. A new directive was adopted in 1995 (95/59/CE) and directive 92/78/EEC was amended to harmonise further this process.
In 1996, the six nation Gulf Cooperation Council (GCC) agreed to increase gradually customs duty on tobacco and related products to 100% by 2000 from less than 30% in the 1980s. Saudi Arabia, Bahrain, Qatar and Oman have already increased tobacco duty to 100% while the United Arab Emirates increased it to 90% in 1999 and Kuwait to 70% in 1997.33 However, these achievements are at risk. The tobacco industry is pressuring the GCC members to lower their duties, stressing that their fiscal revenue from tobacco duties will soon begin to fall because of increased smuggling.17 In early 2000, Lithuania, Latvia, and Estonia announced plans to harmonise their respective tobacco fiscal policies as they are required to raise their rates in order to qualify for membership in the EU.34 The EU should continue its efforts to harmonise tobacco prices, and countries of economic groups such as the Commonwealth of Independent States (CIS), Latin American Economic System (SELA), Caribbean Community (Caricom), and South Asian Association for Regional Cooperation (SAARC) should follow suit.
Implement CPI adjustments and remove tobacco from the CPI
Because general prices and wages tend to rise over time, and because even small annual changes can significantly affect price levels over a decade (a 10% annual change will double prices in less than 8 years), it is recommended, at the very least, to adjust cigarette prices with increases in the CPI. Australia has adopted such a measure and now adjusts cigarette prices twice a year.35
In today's economies, the CPI is an indicator of great importance. Not only is it often used within the framework of monetary policy, but changes in CPI (inflation) are often used, for example, to index pension funds and adjust wage settlements. Because tobacco taxes generally lead to higher prices, increasing tobacco taxes can affect, albeit only marginally, a country's inflation. Given the importance of the CPI as the benchmark for inflation, raising taxes on tobacco products (which provide upward pressure on CPI) is in conflict with low inflationary policies set by central banks and may create a disincentive to raise tobacco taxes. Also, the impact of increasing tobacco taxes in an attempt to discourage tobacco consumption would be offset, to some extent, by adjustments in income tied to CPI movements.36 The aforementioned potential barrier to higher tobacco taxes should not be downplayed. The 2000 Ireland budget speech by Charlie McCreevy, Minister for Finance, illustrates the importance of the CPI in setting tax policies:
What this paper adds
Increasing the price of tobacco products is arguably the most effective method of curbing the prevalence and consumption of tobacco products. International comparisons of levels and trends of cigarette prices and affordability are scarce, notably in developing countries.
This paper examines trends in real cigarette prices and their affordability in more than 80 countries. It also compares prices across countries using different methods. The results suggest that there is ample room to increase tobacco taxes, which would reduce consumption, increase government revenue, and improve health.
“I propose accordingly to increase the excise duty on cigarettes from midnight by 50p per packet of 20 inclusive of VAT with corresponding increases in other tobacco products. This will raise £132 million in a full year and add 0.75% to the CPI.”37
Fortunately, there exists a solution to these problems: the calculation of two distinct CPIs with and without tobacco. The EU has already recommended its member countries to exclude tobacco products from their respective CPIs.38 Luxembourg (1 January 1991), France (1 January 1992), and Belgium (1 January 1994) have removed tobacco products from their respective CPI.39
Support the WHO Framework Convention on Tobacco Control
The member states of the WHO are currently negotiating a Framework Convention on Tobacco Control (FCTC). The FCTC has the potential to enhance various aspects of tobacco control and could include provisions for cooperation in research, programme, and policy development and protocols that aim to foster better price harmonisation and anti-smuggling measures.40,41 The FCTC may provide a framework similar to that of the EU where countries agree to harmonise tobacco prices, thus minimising the incentive to smuggle, so that the full benefits of increases in tobacco taxes—lower smoking rates and better health—are achieved.
Empirical work to date indicates that increasing the price of tobacco products will indeed reduce consumption while also increasing government revenue. As we are all well aware, reducing the consumption of tobacco will not only reduce the global burden of disease but also, among other things, increase the wellbeing of the individuals around us. Therefore, the policy implications inherent in pricing tobacco products take on a vital role. The evidence presented heretofore illustrates that there is indeed room to increase the prices of tobacco products. Specifically, in order to maximise the policy objective, taxes should be implemented uniformly to all products so as not to encourage substitution. Policy makers should encourage regional coordination to reduce the incentive to smuggle and remove tobacco prices from the CPI. These measures will ensure that tobacco consumption will be subject to the full effect of the price increase.
However, from the limited evidence provided here it is sufficient to say that more data need to be gathered and analysed. The dataset currently being worked on will enable us to understand better the relation between price and consumption and in turn the vital role of policy makers in increasing taxes and reducing the global burden of disease attributable to tobacco consumption.
We would like to thank Joy de Beyer, Karen Klimowski and two anonymous reviewers for their comments, and Helen Green for her editing assistance. We alone are responsible for the remaining errors in this paper.
↵* Affordability is defined as cost in terms of income.
↵† Grossman and colleagues2 conclude that one consistent result throughout most price elasticity studies is that about 50% of the change generated by price increases is caused by a reduction in consumption among remaining smokers. More recently, Harris,3 Chaloupka and Wechsler,4 and Farrelly and Bray5 found similar results.
↵§ The Economist Intelligence Unit, part of The Economist Group, is a business information provider with offices in London, New York, HongKong Special Administrative Region of China, Singapore, and Cambridge (USA).
↵¶ Where the Marlboro brand is not available, the EIU surveys the nearestequivalent international brand. For example, in Zimbabwe where theMarlboro brand is not available, the data refers to Benson and Hedgesand Dunhills.
↵** Cigarette prices can be significantly different within countries.
↵†† For more details on the advantages and disadvantages of using theBig Mac as an indicator of PPP see Pakko and colleagues20and The Economist Big MacCurrencies, 29 April 2000. URL: http://www.economist.com/markets/bigmac/displaystory.cfm?story_id=305167
↵‡‡ Average wages are based on actual wages in 12 occupations aftertaking into account working time, holidays, and vacations. The 12occupations are: primary school teachers, bus drivers, automobilemechanics, building labourers, skilled industrial workers, cooks,department managers, electrical or mechanical engineers, bank creditclerks, secretaries, saleswomen, and female industrial workers.
↵§§ Correlation coefficient ( r) between annual changes in real price andannual change in minutes of labour is 0.55 for Marlboro and 0.53 forlocal brand (n = 35).
↵¶¶ Correlation coefficients between real change in prices (Marlboro andlocal brand) and gross domestic product per capita at PPP in the year2000 are 0.45 and 0.29 (n = 67 and n = 63), respectively. The 2000per capita gross domestic product PPP data were obtained from the EIU(http://countrydata.bvdep.com/).