January, 2013:

http://cordis.europa.eu/result/rcn/54569_en.html

Executive summary:

Smoking is the largest single cause of death and disease in the European Union (EU). Effective and equitable control of tobacco in the EU by the use of fiscal policies is both significant for addressing the tobacco disease burden and highly complex. PPACTE aimed to develop evidence-based policy recommendations to improve market regulation of tobacco products, for more effective, equitable control of tobacco in Europe.

The PPACTE consortium undertook several studies to inform the design of tobacco tax policy in the EU. PPACTE convened an international working group to review and critically appraise the international literature on the effectiveness of tobacco tax and price policies for tobacco control. PPACTE built on existing evidence by:
– surveying over 18 000 citizens in 18 European countries on their attitudes towards and responses to tobacco tax and price policies;
analysing time series from 11 EU Member States to estimate the impact of price, income and tobacco control policies on the demand for tobacco products;
– developing dynamic simulation models to assess independently for each of 15 European countries the present impact of their tobacco tax policies and other tobacco control policies on smoking prevalence and smoking-attributable deaths and to make predictions for the future;
– examining industry influence on and responses to tobacco taxation in selected countries through interviewing key informants, analysing tobacco industry documents and analysing data on of the British cigarette market; and
– integrating research findings to distil policy recommendations.

Substantial evidence suggests that tobacco taxation improves public health through preventing initiation amongst never smokers, promoting cessation among current smokers and reducing consumption among continuing smokers. This is substantiated by robust econometric analysis of time series data from 11 European countries, which confirm convincingly that price is a major determinant of cigarette demand. A 10% increase in cigarette price reduces demand for cigarettes by 3-4%, while a 10% increase in income increases demand by 3-4%; thus, tobacco tax increases must increase price above inflation and income growth to achieve tobacco control aims. SimSmoke simulation modelling indicates that increasing taxes has immediate effects on smoking prevalence and smoking-attributable mortality, with the effects growing over time. The PPACTE survey indicated that the majority of smokers would attempt to quit in response to a substantial increase in price. Furthermore, survey responses demonstrated strong public support for tax increases, particularly where at least some of the tax is earmarked for smoking cessation support and prevention.

Future EU tobacco tax directives must address the following: the availability of cheap and ultra-low price cigarettes; the relatively low taxes on alternative products such as fine cut tobacco used for hand-rolled cigarettes; illicit trade or smuggling; and the pernicious interference and influence of tobacco companies on the development of tobacco tax policy. In this regard, the PPACTE consortium formulated 15 recommendations for tobacco fiscal policy and supportive legislation, which if implemented has the potential to reduce tobacco use, tobacco-attributable mortality and health inequalities in the EU. The PPACTE consortium has also delivered tools to support advocacy for policy implementation in EU MS including: 15 country-specific simulation models and usersmanuals enabling advocates to demonstrate the potential impact of stronger policy, documentation exposing industry arguments and strategies to undermine and influence tobacco control policy, and a handbook to support researchers in conducting econometric analysis of demand in their own country.

Project context and objectives:

Pricing policy and control of tobacco in Europe: The PPACTE project

Introduction

In EU, smoking continues to be the largest single cause of death and disease, accounting for over 650 000 premature deaths each year (1). Europe has only 15% of the world population but some 30% of the worldwide burden of tobacco-related diseases (2) and is estimated to cost the economy EUR98-130 billion, or 1.04-1.39% of the EU’S Gross domestic product (GDP) for 2000 (2).

Despite increasing awareness of the health consequences of smoking, a third of all citizens in the EUover the age of 15 currently smoke tobacco products (3). While smoking prevalence trends across the EUhave shown a decline in recent years, the rates remain alarmingly high and continue to rise among females in some Member States. In addition, the average age of initiation has dropped to 11 years of age (2).

In response to the health, economic and social costs of tobacco use, governments have implemented increasingly stringent tobacco control regulations during the past two decades. Action at Member State level has been reinforced and reinvigorated by the WHO FCTC, a widely embraced public health treaty that represents a legally binding agreement between Parties to implement evidence-based tobacco control measures (4). The FCTC calls on Parties to implement measures to reduce the demand for tobacco products, including price and tax measures, protection from exposure to tobacco smoke, product content regulations, packaging and labelling regulations, education and awareness-raising campaigns, smoking cessation support and bans on tobacco advertising, promotion and sponsorship. The EUand all but one Member State (the Czech Republic, at the time of writing) are Parties to the FCTC.

The degree of implementation of tobacco control policy varies among Member States. Joossens and Raw (5) quantified implementation of tobacco control policies at country level on the ‘tobacco control scale’ and ranked 30 countries by their total score on the 100-point scale. In 2005, only four countries scored 70 or more (Iceland, Ireland, Norway and the United Kingdom), two countries scored above 60 (Malta and Sweden), seven scored above 50 (Belgium, Cyprus, Finland, France, Italy, the Netherlands and Poland) and the rest scored 49 or below (6). In 2010, while the average score on the scale had increased by 5%, a similar pattern emerged. The new EUaccession states of central and eastern Europe continue to be strongly represented among the countries scoring lowest on this scale (7).

Taxation of tobacco products is an important means of tobacco control under EUcompetence. Previously, tobacco tax policies reflected the sole priority of creating a strong single market economy. Early tobacco tax directives issued by the EU were primarily concerned with harmonisation of tax structures and approximation of tax rates to prevent market distortions and create a functioning single market. More recently, with the acceptance of health protection of all EUcitizens as a mandate, the importance of tobacco pricing policies in controlling tobacco has been increasingly emphasised.

Currently, there are large price discrepancies among EU Member States, despite attempts to harmonise tobacco tax rates. As of March 2011, tobacco prices for 20 cigarettes in the most popular price category ranged from as high as EUR8.50 in Ireland and EUR6.90 in the United Kingdom to as low as EUR1.87 in Poland, EUR2.03 in Estonia and EUR2.14 in Lithuania. Total tax (inclusive of VAT) as a percentage of total tax-inclusive retail sales price ranged from 71.6% in Sweden to 88.7% in Bulgaria (8).

Increased integration within the EU and the large price discrepancies in tobacco products among Member States provide incentives for tax avoidance. Tobacco taxation for tobacco control is further complicated and undermined by the presence of an extensive eastern land border with Belarus, the Russian Federation and the Ukrainecountries with high prevalence of smoking, very low real prices of tobacco products and relatively weak tobacco control policies. This border complicates the policing of illicit trade and exaggerates ‘grey-market’ activity, with a particular influence in new EU Member States such as the Baltic States, Bulgaria and Romania. Tax avoidance and tax evasion have implications for government revenues and the effectiveness of tobacco control policies for public health.

Furthermore, tobacco control is one of the few public health issues that has an active opponentthe tobacco industrywhich is making calculated strategic attempts to undermine policy and thereby minimise public health gains, in the interest of protecting profits.

Effective and equitable control of tobacco in the EU by the use of fiscal policies is both significant for addressing the disease burden caused by tobacco use and highly complex, because of the diversity among Member States in their stage in the tobacco epidemic, their level of tobacco control and their tobacco market structure, as well as their economic, cultural and political environments.

Objective

The main aim of the PPACTE project is to make evidence-based policy recommendations to improve market regulation of tobacco products, for more effective, equitable control of tobacco in Europe. To achieve this aim, the PPACTE consortium undertook several studies with different methods to provide insight into the challenge of designing a tobacco tax policy within the complex policy environment of EU. It was structured into work packages as described below.

Methods
Integration and dissemination
PI: Dr Maria Leon-Roux, International Agency for Research on Cancer, France

A working group of international experts was convened to produce a handbook, ‘The effectiveness of tax and price policies for tobacco control’ (IARC Handbooks of Cancer Prevention, Volume 14), as a deliverable of PPACTE, referred to herein as ‘the handbook’. It consists of a review and critical appraisal of the current international literature on the effectiveness of tobacco tax and price policies for tobacco control. Topics covered include: tobacco industry pricing, price related marketing and lobbying strategies, the effectiveness of tobacco tax and price policies based on aggregate demand studies, the effectiveness of tax and price policies for youth, adults, and the poor based on individual or household-level survey data, tax avoidance and tax evasion, and the health and economic impact of tobacco taxation.

European survey on the economic aspects of smoking
PI: Dr Silvano Gallus, Istituto di RicercheFarmacologiche Mario Negri, Italy

An extensive survey was conducted with over 18000 respondents in 18 EU Member States to collect recent, comparable, individual-level survey data. The purpose of the survey was to estimate the prevalence, consumption and social acceptability of smoking, purchasing patterns and attitudes towards fiscal policy.

Data were collected during face-to-face interviews between January 2010 and July 2010 in 18 European countries by standardised methods. The 18 056 participants (8653 men and 9403 women) aged 15 years and older were representative of the national general population in terms of age, sex, rural and urban location and socioeconomic characteristics in each country. Surveys were conducted in Albania, Austria, Bulgaria, Croatia, the Czech Republic, England, Finland, France, Greece, Hungary, Ireland, Italy, Latvia, Poland, Portugal, Romania, Spain and Sweden. The survey included questions on demographic, socio-economic and anthropometric characteristics; smoking status, number of cigarettes smoked per day, age at starting and age at stopping, smoking dependence, channels of cigarette distribution and weekly expenditure on tobacco products; a ‘show your pack’ section focused on the latest pack of cigarettes purchased by smokers in order to validate direct questions on smuggling; information on perceptions and attitudes towards an increase in cigarette price including the intention to quit and smoking behaviours according to an increase in cigarette prices; and information on smokeless tobacco use.

Econometric analysis of demand for tobacco
PI: Professor Gunnar Rosenqvist, National Institute for Health and Welfare, Finland

Few studies have been conducted on the impact of tobacco price on tobacco product demand with data from EUMember States. The available published European studies used varying specifications for the included variables, inconsistent data sources and divergent empirical strategies and modelling approaches. The PPACTE project analysed demand for tobacco in similar data sets from different countries over longer periods (30-60 years), with a consistent estimation strategy and model-building approach. The aims of these analyses were to:

– estimate the price elasticity of demand for selected tobacco products in countries with suitable data sets;
– investigate whether cigarettes, pipe and hand-rolling tobacco and snus are substituted for each other;
– evaluate the impact of non-price tobacco control policies on consumption;
– evaluate the extent to which the estimates of price elasticity of demand for tobacco products differ in Europe; and assess the extent to which demand for tobacco products can be controlled by price measures.

Analyses were conducted in 11 Member States where suitable national aggregate time series datasets could be obtained for a sufficient duration. The countries were Austria, Finland, France, Germany, Ireland, Italy, the Netherlands, Portugal, Spain, Sweden and the United Kingdom.

The analyses addressed factors affecting per capita consumption of tobacco products (cigarettes and, where appropriate, pipe and hand-rolling tobacco or snus). Per capita consumption was the dependent variable, while the real price of tobacco products, real disposable income per capita and a tobacco control policy index quantifying implementation of tobacco control policies at country level were the explanatory variables. On the basis of the theory of demand and addiction, conventional, partial adjustment and rational addiction models were applied. In view of the non-stationarity of time series data, error correction models were also considered. Dynamic models were estimated by instrumental variables methods (2SLS) and the Engle-Granger two-step procedure. The estimated models were tested for autocorrelation in residuals, and recursive estimation was used to assess the estimated error correction models.

Simulation modelling of tobacco control policies
PI: Professor David Levy, Pacific Institute for Research and Evaluation, United States

SimSmoke, the dynamic simulation model of tobacco control policy, was adapted to assess independently for each of 15 European countries the present impact of their tobacco tax policies and other tobacco control policies on smoking prevalence and smoking-attributable deaths and to make predictions for the future. SimSmoke contains a population model, a smoking model, a smoking-attributable death model and policy modules. It considers the effect of cigarette taxation, smoke-free legislation, advertising bans, health warnings, media and educational campaigns, cessation treatment and restrictions on sales to and by young people both independently and as part of a comprehensive tobacco control strategy. The model begins with a baseline year, before major policy changes, for which a large-scale survey of smoking rates is available. To validate the model, predictions of smoking rates are compared with actual rates from surveys, from the baseline year through 2010 or the most recent date of a national survey. SimSmoke then predicts the effect on smoking prevalence rates and smoking-attributable deaths of stronger policies fully consistent with the WHO FCTC and MPOWER recommendations and compares the effect to the status quo, in which policies are held constant at 2010 levels. Country-specific models were developed for Albania, the Czech Republic, Finland, France, Germany, Great Britain, Ireland, Italy, the Netherlands, Poland, the Russian Federation, Spain, Sweden, Turkey and the Ukraine.

Industry and market response
PI: Professor Anna Gilmore, University of Bath, United Kingdom

The work carried out to examine Industry influence on and responses to tobacco taxation can be categorised into four main areas:
1) An examination of efforts by tobacco companies to influence tax policy via country case studies in Czech Republic, Bulgaria, Poland, and France. This work was primarily based on analysis of internal tobacco industry documents and key informant interviews.
2) An examination of transnational tobacco company strategies to develop and market smokeless tobacco products in the EU. This work was primarily based on analysis of internal tobacco industry documents, key informant interviews, and snus test purchases.
3) An examination of the impact of tobacco control policies on tobacco product and cigarette brand choice to determine the extent of substitution or down-trading, using the UK as a case-study. The work is based on secondary data analysis using UK survey data. 4) An examination of state of the art systems for supply chain tracking and tracing of tobacco products.

WP6 – Policy analysis and recommendations
PI: Ms Fiona Godfrey, International Union against tuberculosis and lung disease

This WP analysed the development of EU tobacco fiscal policy and the current EU policy context within which emerging policy recommendations from PPACTE would be situated. An expert policy panel was then convened to integrate the research findings and distil recommendations for EU fiscal policy.

Project results:

Effectiveness of tobacco taxation for public health
Econometric analyses of demand for tobacco in 11 European countries

There is substantial evidence from the vast body of international literature reviewed in the Handbook that tobacco taxation improves public health by preventing initiation of smoking among people who have never smoked, promoting cessation among current smokers and reducing consumption among continuing smokers. The effectiveness of tobacco taxes for tobacco control, as reported in the literature, is substantiated by econometric analysis of time series from 11 European countries, which confirm convincingly that that price is a major determinant of cigarette demand.

The results obtained with the preferred demand models indicate a negative relationship between cigarette consumption and cigarette price, higher prices being associated with lower consumption. This relationship was apparent in both the short and the long term. This negative relationship is illustrated in figures 1-4, which show trends in per capita cigarette consumption and the real price of cigarettes over the 30-60year period in selected countries. For cigarettes, short-run price elasticity estimates of demand obtained from our preferred models ranged from -0.30 to -0.40, suggesting that a 10% increase in the real price of cigarettes will reduce cigarette consumption by 3-4%. Outside this range are Ireland, with an elasticity estimate of -0.27, Germany, with an elasticity estimate of -0.79 (or -0.67 depending on the model chosen), and Austria, with a price elasticity estimate close to zero and statistically insignificant. In agreement with other studies, these estimates of price elasticity suggest that demand for tobacco is price-inelastic: the reduction in consumption is proportionately less than the increase in price. Therefore, increasing the price by raising taxes will discourage consumption and increase government revenues. These results lend weight to the considerable body of international evidence for using tobacco taxes as a public health measure to dissuade consumption of tobacco in the European context.

Similarly, a negative relationship was found between the price of pipe and hand-rolling tobacco and its consumption in Finland, with a price elasticity of demand of -0.43, and between price and consumption of snus in Sweden, with an elasticity estimate of -0.24. These own-price elasticity estimates lie between -0.2 and -0.4, similar to the estimates for cigarettes. Moreover, pipe tobacco was found to be a substitute for cigarettes in Finland, with a cross-price elasticity of 1.7, implying that a 10% increase in the price of cigarettes would result in a 17% increase in the consumption of pipe and hand-rolling tobacco.

These analyses show a positive relationship between real disposable income and cigarette consumption, suggesting that consumption of cigarettes increases as income increases. While the estimates of income elasticity in this study vary between 0.1 in Italy and 1.2 in Sweden, most fall within a range of 0.1-0.6, with the median and typical estimates between 0.3 and 0.4. In other words, a 10% increase in income increases consumption by 3-4%. In contrast, a negative relationship was found between income and the consumption of pipe and hand-rolling tobacco in Finland and the consumption of snus in Sweden. This might suggest that users of pipe and hand-rolling tobacco in Finland and users of snus in Sweden with higher income will switch from these products to cigarettes and that users of pipe and hand-rolling tobacco and snus tend to be poorer than cigarette smokers.

These findings suggest that the planning of tobacco price and tax policies must take into account the effect of real income growth on cigarette consumption. The real price of cigarettes must increase above the rate of income growth to prevent increases in the affordability of cigarettes and consequent increases in consumption. Furthermore, these findings suggest that taxation of pipe and hand-rolling tobacco and snus should be kept in line with that of cigarettes to discourage product substitution.

Predicting the impact of raising tobacco taxes on smoking prevalence and smoking-attributable deaths in the SimSmoke model of tobacco control policy.

While the econometric analyses described above are empirical evaluations of the impact of price, income and tobacco control policies on the consumption of tobacco products, the SimSmoke simulation model of tobacco control policy draws together data from various sources to distinguish the effect of tobacco control policies implemented by European countries from long-term trends, to evaluate the effects of tobacco control policies on smoking prevalence and related mortality, and to consider the potential impact of stronger policy alternatives in these countries.

In the SimSmoke model, when taxes change, an equation translates changes in the tax rate (as a percentage of price) into changes in price, assuming that tax increases are passed on to consumers and not under- or over-shifted. Changes in price are then translated into changes in smoking prevalence by an equation dependent on price elasticity estimates. In the models for Germany, Great Britain, Ireland, the Netherlands and Sweden, a price elasticity of -0.3 was applied to people aged 15-17 years, -0.2 to those aged 18-24 years, -0.15 to those aged 25–34 and -0.1 to those aged 35 and over. In the models for Finland, France, Italy, Poland, the Russian Federation, Spain, Turkey and the Ukraine, a price elasticity of -0.3 was applied to people aged 15-24, -0.2 to those aged 25-34 and -0.1 to those aged 35 and over.

Country-specific data on excise duty rates were used for 2010, as the baseline from which future tax changes were measured. Rates of excise (specific and ad valorem, exclusive of VAT) were obtained from the WHO MPOWER report (9) or the European Commission excise duty tables for 2010. For all years before 2010, a cigarette price index of actual prices deflated by the consumer price index was used.

The SimSmoke models consider the impact of increasing tobacco taxes (specific and ad valorem, exclusive of VAT) to 70% of the price. Table 2.1 shows the absolute change in smoking-attributable deaths (deaths averted) as a result of increasing tobacco taxes to 70% of tobacco price, relative to the status quo scenario in which all policies are held constant at 2010 levels. The estimate for a particular year represents deaths in that year alone, whereas the cumulative estimate is for the years 2011-2040. Tables 2.2 and 2.3 show the percentage change in smoking prevalence (measured relative to the status quo in a particular year) if the tax were increased to 70% of the price, with all other policies held constant at 2010 levels, for males and females. The variation in the percentage effects on prevalence is due mainly to the difference between the 2010-specific tax rate and the 70% tax rate.

In general, this model showed how the prevalence of smoking among young people declines more than among adults as a result of tax increases. This is the main reason that taxes continue to reduce adult smoking rates over time. The projected number of deaths reflects the effectiveness of a tax policy in reducing smoking. The effects of tobacco taxes on the number of deaths are delayed not only because the effects of cessation on death rates are relatively slow but also because the greatest effects are on the prevalence among young people. Changes in the prevalence among the young do not avert deaths for at least 20 years.

The predictions shown above are based on the assumption that all other policies are held constant at 2010 levels. The model predictions suggest that, in each country, substantial reductions in smoking prevalence and lives lost due to smoking can be realised by increasing tobacco taxes alone. Moreover, when tobacco tax increases form part of a comprehensive tobacco control strategy in which stronger laws on smoke-free air and restrictions on sales to and by young people are implemented, strict tobacco advertising and marketing bans are promulgated, strong tobacco health-warning labels are required, high-publicity media campaigns are coordinated with other policies, strong comprehensive smoking cessation treatment services are provided and all policies are enforced, considerable reductions in smoking prevalence and smoking-attributable mortality can be achieved. Because of the natural progression of tobacco-related illnesses, early reductions in smoking prevalence have a relatively small short-term impact on the number of smoking-attributable deaths but a much larger long-term impact.

In some of the countries considered, such as the Czech Republic and Poland, tobacco product prices are lower than in other countries and in relation to the standard of living. In these countries, excise duties, as a proportion of the tax-inclusive retail-selling price, are relatively high. In other countries, such as Ireland and the United Kingdom, tobacco product prices are relatively high and the excise duties as a proportion of the tax-inclusive retail-selling price are somewhat lower. The divergence amongstEU Member States in the rates of tobacco excise as a proportion of price and the absolute value of excise duties highlights the need for a specified monetary minimum specific tax floor to achieve appreciable increases in the tax-inclusive retail selling price and corresponding improvements in health outcomes.

European survey on economic aspects of smoking: perceptions and attitudes to increasing tobacco taxes

The PPACTE European survey examined current smokers’ declared responses to a hypothetical 20% increase in the price of a pack of cigarettes. Overall, 14.2% of current smokers said they would quit smoking, 30.6% would consume fewer cigarettes, 21.5% would engage in compensatory behaviour (13.7% would switch to cheaper brands and 3.8% to hand-rolling tobacco, 3.5% would switch at least part of their smoking consumption to illegal or smuggled cigarettes, 0.5% would switch to or also use smokeless tobacco including snuff, snus or chewing tobacco, and 33.6% would not change their smoking habit). Participants in Romania and Spain indicated that they would be most responsive to hypothetical price increases, with 80.3% and 75.9% of current smokers reporting that they would change their smoking habit in some way in response to the price increase, respectively, while participants in Finland and France would be least responsive, with 45.8% and 49.3% of respondents claiming they would not change their smoking behaviour in response to a 20% price increase, respectively. The proportion of current smokers who would switch to smokeless tobacco was lower in Portugal (0.0%) and Hungary (0.3%) and was higher in Bulgaria (11.2%) and Latvia (18.6%), while the proportion of current smokers who would switch to smokeless tobacco was higher in Poland (1.4%) and Sweden (2.0%).

Current smokers were asked if they intended to quit smoking within the next 6 months. Overall, 36.0% of current smokers said they intended to quit smoking (34.1% of men and 38.3% of women). Among male current smokers, the proportions of men who said they intended to quit smoking were lowest in Hungary (7.7%) and the Czech Republic (10.9%) and highest in Spain (56.7%) and Romania (61.0%). Among female current smokers, the proportions of women who said they intended to quit smoking were lowest in Hungary (10.0%) and Austria (14.7%) and highest in Poland (54.3%) and Spain (57.6%).

Respondents were then asked what percentage increase in the current cigarette price would encourage them to quit completely. Overall, 20.5% of current smokers reported that they would quit in response to a 20% increase in price, 19.1% reported they would quit in response to a 21-40% increase, 18.2% would quit with an increase of 41-60%, 6.3% would quit with an increase of 61-80%, and 35.7% said that it would take a price increase of more than 80% for them to quit. The proportions of current smokers who reported that they would quit smoking completely in response to an increment of 20% or less were lower in Hungary (6.8%) and Latvia (6.9%) and higher in Ireland (30.0%) and Romania (31.3%). The proportions of current smokers who reported that they would quit smoking completely in response to an increment of five or more times the current price were lower in Hungary (0.5%) and Portugal (4.0%) and higher in Albania (16.3%) and Latvia (16.5%).

Current, ex- and ‘never’ smokers were asked about their attitudes to an increase in the price of a pack of cigarettes and were asked to assume that the revenue from the increase would be allocated to support smoking cessation, such as free access to anti-smoking centres and free smoking cessation products. Overall, 78.7% of non-smokers and 49.2% of current smokers were in favour of a 5% increase in prices if the revenues were used for tobacco control. Moreover, 73.6% of non-smokers and 39.6% of current smokers were in favour of a 20% increase in price.

Respondents were asked how useful they perceived various tobacco control measures to be in reducing tobacco use. Overall, 55.0% of respondents (61.8% of non-smokers and 36.9% of current smokers) perceived tobacco taxation to be a useful means for reducing tobacco use. There were substantial differences among countries in perceptions of the effectiveness of pricing policies to control tobacco: only 19% of respondents in Hungary perceived price to be a useful tobacco control measure, while over 65% of respondents in Albania, Finland and Italy considered tobacco price and tax increases as useful tobacco control measures.

Summary

The vast body of international literature provides substantial evidence that tobacco taxation improves public health by preventing initiation of smoking, promoting cessation among current smokers and reducing consumption among continuing smokers. The effectiveness of tobacco taxes for tobacco control is further substantiated by econometric analyses covering periods of 30-60 years for 11 European countries. The SimSmoke simulation model suggests that increasing taxes has immediate effects on smoking prevalence and smoking-attributable mortality, with the effects growing over time. Moreover, the European survey indicates that most smokers would attempt to quit in response to a 60% increase in price and that approximately two thirds of non-smokers and over one third of smokers perceive price to be an effective measure for limiting tobacco use. Not only are tobacco tax increases effective in reducing tobacco use and its associated burden of disease, but they are also acceptable to European citizens, with support from smokers and non-smokers alike.

The extent to which public health benefits from tobacco taxation is influenced not only by the structures and rates of excise taxes but also by the extent to which these structures and rates are harmonised across regions, thereby decreasing the incentives for illicit trade. In addition, the effectiveness of tobacco taxation for public health depends on the extent to which the tobacco industry passes on tax increases to the consumer, rather than under-shifting the tax by decreasing their profit margins, and the success of tobacco companies’ efforts to lobby governments for weaker tax policies. Tobacco tax structures and rates, illicit tobacco trade and tobacco industry influence on tobacco tax policy are discussed in turn in the following sections.

Tobacco tax structures and rates

The structure and rates of excise duties influence the prices of different products, government tax revenues, the quality and variety of products on the market, the administrative burden, the profits and competitive positions of tobacco producers and the distribution of income (1). Excise duties can be specific or ad valorem. A specific excise is levied as a fixed monetary amount of tax per quantity, volume or weight of tobacco, while an ad valorem excise is levied as a percentage of some measure of product value (currently the weighted average price of tobacco). In a purely specific excise structure, the same fixed monetary amount is applied to all tobacco products of a given quantity, volume or weight, irrespective of their pre-tax price. This structure tends to discourage consumption of tobacco products, irrespective of their price. A specific excise structure has the advantage of being easy to administer and narrows the gap between low- and high-priced tobacco brands. The fixed monetary amount does not, however, automatically keep pace with inflation and must be adjusted regularly. A purely ad valorem structure tends to lead to lower prices, with a wider gap between low- and high-price brands. While a purely ad valorem structure keeps pace with inflation, it is more complex to administer. To combine the best elements of both the specific and the ad valorem structure, the two can be combined into a mixed structure that can give preference to the ad valorem or the specific element, depending on the objectives.

The most complex structure is a mixed specific and ad valorem excise with a minimum tax floor. Minimum excise duties are effectively specific excise duties and represent a fixed monetary amount per quantity, volume or weight that applies if the ad valorem excise falls below a minimum floor. In this system, lower-priced products are taxed at the specific minimum rate, while higher-priced products are taxed at the ad valorem rate.

Structures and rates of taxes as set out in excise directives

Member States of the EUmust apply a mixed structure with a minimum tax floor. Effective 1 January 2011 to 31 December 2013, the specific component applicable to cigarettes may not be less than 5% or more than 76.5% of the total tax burden, including the specific excise, and the ad valorem excise and VAT levied on the weighted average retail selling price. From 1 January 2014, the specific component must fall within the range 7.5-76.5% of the total tax burden. Directive 2010/12/EU applies a substantially lower overall minimum excise rate to other tobacco products including fine-cut tobacco for hand-rolled cigarettes, than the rate applied to manufactured cigarettes.

Adjustment of the overall minimum tax

As incomes and costs of living vary among Member States, the minimum monetary tax is currently set at only EUR64 per 1000 cigarettes. This minimum tax must be raised to EUR90 per 1000 cigarettes over the 5 years following the 2010 directive. Even at this low level, some Member States, under pressure from tobacco companies, have negotiated for derogation for several years, to allow their taxes to remain even lower. This has the effect of causing governments to lose valuable tax revenue and increasing cigarette use and the prevalence of disease related to smoking. This also aggravates problems of cross-border shopping for neighbour countries with higher taxes.

To account for differences in income and cost of living and ensure that tobacco remains unaffordable, the minimum monetary tax could be adjusted by the comparative price level or purchasing power parity, which are available for all EUMember States. The basic minimum tax could be set for the States with the lowest income and adjusted upwards for each other State. For example, the minimum tax could be set at EUR 125 per 1000 cigarettes for one low-price country and adjusted by comparative price level for all other countries. In countries with higher prices, the minimum would be set as above and adjusted upwards annually above inflation and income changes. The basic minimum tax would then be adjusted annually in line with inflation and income levels and the relative cost of living by comparative price level or purchasing power parity.

The index considered for adjusting the overall monetary minimum tax should be: available for most countries (both Member States and candidate countries), provided by an institutional authority (Eurostat), stable, updated annually and published by a single source.

Substitution of hand-rolling tobacco for cigarettes

The effectiveness of tobacco taxation can only be fully realised to the extent that tax increases raise the price of the cheapest cigarettes or other tobacco product. Otherwise, there is an opportunity and incentive for the most price responsive smokers to trade down to cheaper brands or products. Directive 2010/12/EU applies a much lower rate of excise on fine cut tobacco than that applied to manufactured cigarettes. Planned excise rate increases on fine-cut tobacco between 2010 and 2020 do not go far enough to reduce the incentive for consumers to substitute preferentially taxed and thus cheaper fine-cut tobacco for cigarettes.

To levy comparable rates of excise on fine-cut tobacco for rolling cigarettes and on manufactured cigarettes, the weight of tobacco used to prepare hand-rolled cigarettes and manufactured cigarettes must be established. The current conversion rate suggests that 1 kg of fine-cut tobacco is equivalent to 1000 cigarettes (1 g = 1 cigarette); however, this is considered to be inaccurate, particularly with the increasing use of dry ice-expanded tobacco manufacturing processes, which reduces the mass of tobacco in a manufactured cigarette. According to the International Standards Organisation norm on measuring tar and nicotine in hand-rolled cigarettes (ISO 15592–3), there are 0.4–0.75 g of tobacco per hand-rolled cigarette. This suggests that 1333-2500 hand-rolled cigarettes can be made from 1 kg of fine-cut tobacco, rather than 1000.

An estimate of the average weight in grams of a hand-rolled cigarette was obtained from data in the PPACTE European survey using data on daily consumption of and weekly expenditure on manufactured and hand-rolled cigarettes. Respondents were asked to show or recall the weight in grams of the last pack of hand-rolling tobacco they had purchased and how much they paid for it. The weight in grams per cigarette was calculated on the basis of data for 185 smokers of hand-rolled cigarettes:
Grams per cigarette = weekly expenditure / cost of latest pack x grams per pack / days per week / cigarettes per day

This data estimates that one hand-rolled cigarette uses approximately 0.7-0.8 g of fine cut tobacco implying that that 1 kg of fine-cut tobacco yields 1250.0-1482.6 cigarettes, a narrower and lower range than that indicated by the ISO standards. A reasonable conversion rate is required to achieve full alignment of taxes on manufactured cigarettes and fine cut tobacco for hand-rolled cigarettes.

Down-trading from more expensive to cheaper brands

When faced with tobacco tax increases, price-sensitive consumers may continue smoking manufactured cigarettes but ‘trade down’ to a cheaper brand rather than switching to a cheaper product (e.g. hand-rolled cigarettes). Industry pricing can undermine the effectiveness of tobacco tax increases by creating price differentials between brands, so that consumers can trade down to cheaper brands when taxes increase.

Detailed analysis of data from the British market suggests that a multifaceted strategy is being used to keep prices low on the ultra-low price segment of the cigarette market. Between 2006 and 2009, the gap between the most and least expensive brands widened, with a broader range of prices available within each segment; the weighted average price of ultra-low price brands did not increase in real terms, while the average prices of the other brand segments did. An examination of trends in the prices of the best-selling brands during the same period showed that the price of ultra-low-price brands increased by less than 1%, with real price (i.e. inflation-adjusted) decreases in some cases. Meanwhile, the price of premium brands increased by 3-5% and the price of mid-price and economy brands by 5-6%. Furthermore, examination of patterns of price changes (net of tax) over the same 3-year period suggests that, while tax increases are generally being over-shifted, tax increases are under-shifted on the ultra-low-price brands.

Taken together, these findings suggest that the industry is cross-subsidising cheaper brands with profits from more expensive brands. As a result of this pricing strategy and the growth in the number of ultra-low-price brands, consumers have greater opportunities to down-trade from more expensive to cheaper cigarettesand the market share of ultra-low-price brands has increased in response. These findings, combined with an analysis of British survey data showing who is smoking these cheap cigarettes, industry documents on the role of cheap cigarettes (10), industry’s willingness to under-shift taxes, plus interviewee responses in the country case studies, suggest that the availability of cheap cigarettes undermines tobacco tax policy, allowing price-sensitive smoker – particularly the young and poor to continue to initiate and maintain their smoking habits. While this analysis is based on British data, Euromonitor data suggest that the market share of cheap cigarettes is growing in other countries, including Austria, the Czech Republic, Denmark, Hungary, Lithuania, Portugal, Slovakia and Sweden (11).

Under-shifting of tobacco taxes can be discouraged by substantially increasing tobacco excise taxes with a predominantly specific excise. Furthermore, an excise structure that is predominantly specific or has a high minimum floor and a low ad valorem component helps to reduce the price differential between the highest- and lowest-priced brands.

While over-shifting of excise increases is not a public health issue, it reflects a missed opportunity for governments to increase excise rates and thus increase revenues. It is clear that the transnational tobacco companies prefer small, gradual tax increases (12), and evidence is beginning to appear that gradual tax increases are more easily absorbed by consumers and therefore facilitate tax over-shifting, leading to greater industry profit margins. Large excise increases are likely to benefit public health to a greater extent than incremental increases, the difference accruing as government revenue rather than industry profit. Further empirical work is needed to explore this issue.

Summary

The main issues to be addressed in future EUdirectives are: the availability of cheap and ultra-cheap cigarettes, including ‘dumped’ cheap cigarettes; the relatively very low taxes on alternative products, such as fine-cut tobacco for roll-your-own cigarettes; illicit trade and smuggling; and the pernicious interference and influence of tobacco companies on the development of tobacco tax policy. PPACTE evidence suggests that selling cigarettes below cost and low price-based marketing, including selling below the tax level, should be addressed. It is no easy matter to address all these problems or loopholes, but specific changes to the tobacco tax structure and rates and other supportive legislation could go a long way in this regard.

Illicit tobacco trade

Tobacco tax avoidance and evasion undermine the effectiveness of tobacco taxation by providing access to cheaper tobacco products, and they weaken the impact of other tobacco control policies and increase health disparities, while reducing government revenues (13).

Extent of illicit tobacco trade in Europe

Measuring illicit tobacco trade is methodologically challenging, for many reasons. First, it is an illegal activity, and illegal traders are unlikely to document their activities. Secondly, data on illicit trade are difficult to collect, as law enforcement agencies often do not publish information about all their activities, in the interests of security and confidentiality. Thirdly, all the available methods of measuring illicit trade are limited, and the data sources used may bias the estimates. As a result, transparent and public data on illicit trade in Europe.

The limited empirical evidence measuring the extent of illicit trade indicates that tax evasion is much more widespread than tax avoidance, that cigarette tax evasion is more prevalent in countries that have lower cigarette prices and lower cigarette taxes and that the size of the illicit market is inversely related to a country’s income.

On the basis of an analysis of data collected by the professional services company KPMG, the European Commission estimated that, in 2004, total market penetration of the illicit cigarette trade represented approximately 8-9% of cigarette sales within the EU(which had 25 Member States at the time) (14). It also noted that the illicit market share in the new EUMember States (Estonia, Hungary, Lithuania, Poland and Slovakia) was far higher than the previous average. This report is limited because it is based on cigarette seizures in the EUand on studies provided by the tobacco trade and governments; however, the overall figure of 8-9% appears to be a reasonable estimate, as it falls between the higher estimates from the United Kingdom and eastern and central European countries and the lower estimates from southern European countries like Italy and Spain.

KPMG continued its research on illicit trade as part of its obligations under the 2004 agreement between Philip Morris International and EU. According to the KPMG report, total cigarette consumption in the EUin 2009 was 685 billion units, and contraband trade accounted for 8.9% of total consumption (15). The content of the KPMG report was made public only in August 2011, after a formal request based on EUlegislation regarding public access to documents (Regulation No. 1049/2001 of May 2001).

Case study findings from Bulgaria and Poland suggest that the tobacco industry might exaggerate the extent of illicit trade as part of their argument against increases in excise duty. For example, Phillip Morris reported in 2010 that the prevalence of illicit trade in Bulgaria represented 34% of total market sales (16), while independent data from the PPACTE European survey suggest a 14.5% prevalence in that year.

European survey on the economic aspects of smoking: purchasing patterns and latest pack

To obtain updated, comparable estimates of the extent of tax avoidance in 18 strategically selected EUMember States, respondents to the European survey on the economic aspects of smoking were asked about their purchasing patterns and to show their latest purchased pack of cigarettes or hand-rolling tobacco. Among current smokers, an average of 88.1% bought cigarettes from legal tobacco shops (including vending machines), 4.9% from other countries or duty-free shops and 3.6% from smuggled sources. On average, 3.4% smoked cigarettes offered by their peers and 0.1% bought cigarettes over the Internet.The proportion of current smokers who reported that they smoked cigarettes from other countries or duty-free shops was higher in Austria (12.3%), Finland (13.2%) and France (13.2%). More current smokers reported smoking smuggled cigarettes in eastern European countries, particularly in Bulgaria (12.2%) and Latvia (25.9%). Overall, 8.4% of current smokers had bought smuggled cigarettes in the past 30 days, representing at least 1% of their total cigarette purchases.

Overall, 73.9% of current smokers agreed to show the interviewer their latest purchased pack of cigarettes or hand-rolling tobacco. Figure 4.2 shows the percentage distribution of current smokers by the type of their latest purchased tobacco product and by country. Overall, 81.6% of current smokers had bought a pack of 20 cigarettes, 4.3% a pack of 10 cigarettes, 10.9% hand-rolling tobacco and 3.3% another type of tobacco product. The highest proportion of smokers showing hand-rolling tobacco was observed in England (31.8% overall, 38.2% of men and 24.9% of women), followed by France (17.0% overall) and Finland (14.0% overall).

Overall, 93.1% of current smokers showed a tobacco product with a health warning in the local language and 1.1% a pack with no health warning. The prevalence of current smokers showing a pack with a health warning in a foreign language was lowest in Portugal (0%) and Greece (0.3%) and highest in Austria (12.2%) and Latvia (26.1%). The prevalence of current smokers showing a pack with no health warning was lowest in England, France, Portugal and Sweden (0%) and higher in Latvia (5.2%) and Croatia (8.6%).

Overall, 89.5% of current smokers showed a product with a tax stamp in the local language and 4.5% in a foreign language, 1.7% showed a pack with a tax stamp removed or destroyed and 4.4% had either a duty-free pack or a pack with no tax stamp. Current smokers showed a tobacco product with a foreign stamp most often in Latvia (26.3%), followed by France (11.0%) and Austria (9.6%). The largest proportion of smokers showing a duty-free pack or one with no tax stamp was in England (15.2%), followed by Bulgaria (8.3%) and France (7.7%).

Summary

The tobacco industry are being presented as partners in the fight against illicit trade, while the same industry uses data on illicit trade that are not publicly available and have not been subject public scrutiny, to attack tobacco control legislation developed and supported by health officials. Transparent and public data on illicit tobacco trade are missing in most European countries. Preliminary data from a survey, undertaken in 18 countries as part of the PPACTE project in 2010, indicate that the illicit cigarette trade is highest in Latvia, Romania, Bulgaria and Poland, countries with low prices, but close to Russia and the Ukraine, important suppliers of illicit cigarettes in Europe. Hence, in Europe, supply side factors (such as the supply from manufacturers in Russia and the Ukraine) appear to play a key role in determining levels of illicit tobacco trade. This finding contradicts industry arguments on this topic and highlights the need for data on smuggling to be made public. A global tracking and tracing system should be combined with better regulation of the legal tobacco trade. Major tobacco companies are now collaborating in promoting Phillip Morris International in-house marking system on cigarette packs. This agreement suggests the industry fears the uptake of alternative systems. This and the willingness of these companies to collaborate on this issue raises concerns that such a system would not be in the public interests.

Industry influence on tobacco taxation

EUand all Member States with the exception of the Czech Republic are parties to the WHO FCTC and, under article 5.3, are bound to prohibit the influence of the tobacco industry on the formulation of public health policy. The WHO FCTC states that ‘in setting and implementing their public health policies with respect to tobacco control, parties shall act to protect these policies from commercial and other vested interests of the tobacco industry in accordance with national law’.

The PPACTE case studies and research clearly show that governments continue to engage with the tobacco industry when formulating tobacco taxation policy. Tobacco companies lobby Member State governments constantly and persuade them to keep tobacco taxes low, arguing incorrectly that if taxes increase tobacco revenue will decrease and smuggling will occur. It is clear from documents on countries acceding to the EUthat transnational tobacco companies were greatly concerned to prevent any significant increase in excise duties on accession of the country to the EUand to ensure that any such increases would be gradual. The companies worked collectively to prevent and postpone increases in excise and lobbied successfully for derogation of the excise level. As a result of the derogations, and with EUaccession also leading to higher incomes, cigarettes actually became slightly more affordable in some accession countries. To influence policy, industry targeted key government officials at both national and EUlevel.

Despite the importance of the influence of tobacco companies on tax policy, most of the empirical studies identified in the literature review for the Handbook were conducted in the United States and a few other high- or middle-income countries and most address the influence of transnational tobacco companies on tax levels, rather than structures. PPACTE addressed this research gap in order to add to understanding of tobacco industry influence on taxation policy and industry pricing strategy in EU.

Understanding tobacco industry pricing strategy: the cigarette market in Great Britain

Industry pricing strategy was reviewed by examining the British cigarette market. Four price segments were identified in the academic and trade literature: premium, mid-price, economy and ultra-low price, the last emerging since 2006. Brands were categorised into price segments on the basis of recommended retail price data from PriceChecker (1999-2005) and actual sales prices from Nielsen (a global leader in market research, measurement and information) (2006-2009). Trends (by volume) in the market share by price segment between 2001 and 2009 were observed.

Around half the market was held by economy segment brands: The share held by this segment grew until 2007-2008, when it fell slightly because of gains in the ultra-low-price segment. As the tobacco companies acquired supermarket brands and launched their own ultra-low-price brands after 2006, the share of these brands increased substantially. In contrast, the market share of premium and mid-priced brands fell sharply from 2001 onwards. While the three segments (premium, economy and ultra-low priced) have clearly separate price ranges, the price of mid-price brands now overlaps entirely with the price of lower-end premium brands. Furthermore, the range of prices available within each segment appears to have widened, and the gap between the most and the least expensive brands has doubled.

When trends in price increases by price segment were examined for 63 individual brands, each with a market share of 0.2% or greater (as of November 2006) between 2006 and 2009, it became apparent that the prices of brands in the ultra-low-price segment had increased much less than those of brands in other segments. The price of every brand in the ultra-low-price group increased by less than a 1%, and the real price fell in some cases. Meanwhile, the price of premium brands increased by 3-5% between 2006 and 2009 and that of mid-price or economy brands by 5-6%.

By examining the patterns of price changes in real terms (deflated against the consumer price index, with 2010 = 100), net of tax between November-May when taxes increase and May-November when taxes do not change, it was possible to establish whether tobacco companies over-shifted, under-shifted or simply passed on tax increases to consumers between 2006 and 2009. Overall, the results suggest that tax increases were over-shifted, with greater price increases in November-May when tobacco taxes rise than in May-November. The extent to which tax increases were passed on to smokers differed by brand segment, the price increases for premium and mid-price segments being much higher during November-May than May-November. This suggests that, for these brands, cigarette manufacturers were using the increases in tobacco duty to disguise additional price increases, thus over-shifting taxes. The prices for the economy segment were also being over-shifted, but the larger price increase was timed so that it did not coincide with the tax increase. For the ultra-low-price segment, the prices net of tax actually fell between November and May when the tax burden rose, with a very slight increase in net prices between May and November. Overall, taxes on ultra-low-price brands were under-shifted.

These findings suggest that the tobacco industry is using a sophisticated pricing strategy in Great Britain to cross-subsidise cheaper brands with profits from more expensive brands. This study shows real price increases and over-shifting of taxes in the premium, mid-price and economy brand segments, while the real price of ultra-low-price brands did not increase and taxes on this segment were under-shifted. Consistent with this strategy, the price increases were greatest on mid-price and economy brands and lowest (and often negative in real terms) on individual ultra-low-price brands. Furthermore, the price increases appeared to be timed to accentuate differences in prices between brands in different segments at the time when duties were increased; however, overall prices increased and taxes were over-shifted to consumers, a pattern that contributes to rising industry profits.

As a result of this pricing strategy and the increase in the number of ultra-low-price brands, consumers have more opportunities to down-trade from more expensive to cheaper cigarettes. The market share by volume of ultra-low-price brands has increased in response. The availability of cheap brands undermines tobacco tax policy because it ensures that price-sensitive smokers, particularly the young and the poor, continue to initiate and maintain their smoking habits.

Tobacco industry influence on tobacco excise policy in four European countries: case studies in Bulgaria, the Czech Republic, France and Poland

The efforts of transnational tobacco companies to influence tobacco control policy (specifically tobacco excise policy) were examined in four EUMember States by detailed analyses of tobacco industry documents, in which a socio-historical approach was used. Analysis of documents was triangulated and updated with the results of interviews and secondary data analysis. Countries examined include: Bulgaria (1988-2011), the Czech Republic (1989-2011), France (1990-2011) and Poland (1990-2011).

The countries selected have several attributes that make them interesting for analysis. At the time of market entry, each country was of particular interest to the tobacco industry for its strategic importance and the opportunity it presented for market expansion. Bulgaria was seen as a lucrative market for several reasons. First, the existing production infrastructure, low average salaries and high production provided opportunities for improving profit margins. Secondly, Bulgaria was seen as the gateway to accessing the then closed markets of Turkey and former socialist countries. Thirdly, joint-venture initiatives with Bulgartabac provided opportunities for expansion of the existing brand portfolio. The Czech Republic was of interest because of its strategic location in central Europe, bordering former socialist countries that the transnational tobacco companies hoped to access. France has seen two major tax increases in the past twenty years, and also has close relationships between the industry and state, partly due to the state’s former monopoly on tobacco products. The size of the Polish market and the potential for even further growth underpinned the transnational tobacco companies’ interest in penetrating the Polish market.

The lack of effective government intervention in tobacco control provided a favourable business environment in each of the case countries. Bulgaria’s tobacco control is weak, its ranking on the ‘tobacco control scale’ falling from 13 out of 30 European countries in 2007 to 24 out of 31 countries in 2010. The Czech Republic has one of the poorest tobacco control records in Europe, ranked fourth lowest in Europe for tobacco control policy implementation in 2010 (7), with senior political figures publicly supporting the tobacco industry. Although France ranked sixth on the 2010 tobacco control scale, rising one place from 2007, the Government’s policy on tobacco taxation and pricing is still considered friendly to the industry. Unlike most other central and eastern European countries, Poland had a fledging tobacco control movement in place in the late 1980s and early 1990s, providing some opposition to the transnational tobacco companies as they entered the Polish market. While Poland was a leader in implementing tobacco control policies in 1995, ranking 14 out of 30 European countries in 2007, it has been described as ‘losing momentum’, and its ranking fell to 19 out of 31 countries in 2010 (7).

In each of the case studies in the central and eastern European countries, the period considered covers the entry of the transnational tobacco companies into the country, privatisation of the tobacco industry (except in Bulgaria where privatisation is ongoing) and accession to EU. Transnational tobacco companies use industry privatisation as an opportunity to manipulate tobacco excise structures in their favour, lobby against cigarette excise rate increases and influence the broader regulatory environment (17). Much of this evidence is, however, based on studies from the former Soviet Union and applies to British American Tobacco in particular. Accession to the EUrequired elimination of any remaining barriers to European imports and implementation of EUexcise directives, as well as other tobacco control legislation; it also provided an opportunity for industry influence. In France, the study specifically addressed the two large tax increases in 1991-1993 and 2003-2004 and the industry’s response to them, as there is little evidence on the impact of large tax increases.