- Email: [email protected]
Taxes on tobacco, alcohol and sugar sweetened beverages: Linkages and lessons learned
Social Science & Medicine 136-137 (2015) 175e179
Contents lists available at ScienceDirect
Social Science & Medicine journal homepage: www.elsevier.com/locate/socscimed
Short report
Taxes on tobacco, alcohol and sugar sweetened beverages: Linkages and lessons learned Evan Blecher a, b, * a b
Economic and Health Policy Research, American Cancer Society, USA Southern African Labour and Development Research Unit, University of Cape Town, South Africa
a r t i c l e i n f o
a b s t r a c t
Article history: Available online 14 May 2015
Increased consumption of sugar-sweetened beverages (SSBs) has been linked to increases in obesity in both high-income and low- and middle-income countries. Tobacco and alcohol taxes have proven to be effective tools to reduce tobacco and alcohol use. Many public health advocates propose using similar taxes to reduce consumption of SSBs. South Africa is a middle-income country that is considered a leader in the area of tobacco tax policy. A case study of tobacco and alcohol taxes is used to better understand optimal tax structures for SSBs. The case study tracks aggregate data over time on taxes, prices, consumption, tax revenues, and marketing expenditures at the brand level. Tobacco and alcohol taxes are shown to be effective in reducing the demand for tobacco. Additionally, taxes on the dose of alcohol rather than the volume of the beverage may incentivize producers to reduce the volume of alcohol in beverages through the supply side. While specific taxes based on the volume of beverages are likely to reduce the demand for SSBs, policy makers should also consider taxes on alcohol and SSBs that tax the dose of the alcohol and calories in order to create supply-side incentives for producers to lower alcohol and calorie levels in existing products or promote products with lower levels of alcohol and calories. © 2015 Elsevier Ltd. All rights reserved.
Keywords: Alcohol Tobacco Sugar-sweetened beverages Tax South Africa
The use of taxation as a public health intervention to reduce the consumption of products that have negative externalities is not a new concept. Adam Smith is famously quoted in the Wealth of Nations in 1776 as supporting the taxation of tobacco, alcohol, and sugar since none were considered necessary for life (Smith, 1776). While the taxation of tobacco for fiscal and health purposes is now universally accepted, the use of taxes on alcohol to influence health is still in its infancy even though it is commonly used as a fiscal tool (Nugent et al., 2006). The taxation of sugar-sweetened beverages (SSBs) is rare, and where used, it is typically used for health rather than fiscal purposes. This should not be confused with excise taxes on carbonated beverages which are typically used for fiscal purposes, however, these taxes usually do not distinguish between those sweetened with nutritional or non-nutritional sweeteners. This paper argues that while the experience of tobacco is very useful in the public policy space, the structures of taxes may need to be different when applied to alcohol and SSBs. While the experience of tobacco tends to favor specific taxes per cigarette, this paper will show that specific taxes on alcohol and SSBs may be more
* 250 Williams St NW, Atlanta, GA, 30303, USA. E-mail address: [email protected]. http://dx.doi.org/10.1016/j.socscimed.2015.05.022 0277-9536/© 2015 Elsevier Ltd. All rights reserved.
effective when taxing the volume of alcohol or sugar rather than the volume of the beverage. The case study of South Africa is used throughout. Ethical approval was not required since the study used only secondary data and no human subjects were involved. 1. Tobacco Higher taxes that result in higher prices result in lower consumption of cigarettes and smoking prevalence (IARC, 2011). A prominent example of this policy instrument is South Africa where between 1990 and 2012, real (i.e. inflation adjusted) excise taxes rose by 552% leading to a 217% increase in real prices (Fig. 1). The resultant decline in consumption amounted to 41% and 66% in aggregate and per capita terms, respectively. This coincided with a decline in adult smoking prevalence from 33% to 20% between 1993 and 2012 (Blecher, 2011). Given the growth in population during this time there are presently 4,5 million fewer smokers than there would have been had smoking prevalence in 1993 still prevailed. At the same time, annual revenue collections increased in real terms by 284%. South Africa shows a clear fiscal benefit from increasing tobacco taxes through increased revenues, and health benefits through lower tobacco use. This is a demand-side policy in that it intends to
176
E. Blecher / Social Science & Medicine 136-137 (2015) 175e179
Fig. 1. Cigarette excise taxes, price and consumption in South Africa. Source: National Treasury and StatsSA.
reduce the population and individual level demand for tobacco by increasing the price of the product. Tobacco is most commonly consumed as cigarettes, and while there is a clear link between smoking and certain diseases including cancers and cardiovascular disease, there is no significant link between the strength of the addictive component (i.e. the nicotine) and the health consequences thereof. Reducing the nicotine content of the cigarette will not result in reduced health consequences of cigarettes (Surgeon General, 2004). Health consequences are almost exclusively linked to the quantity of cigarettes smoked and the number of years that a smoker smokes (Surgeon General, 2010). As a result, there is consensus that taxation is most effective as a blunt instrument which makes all cigarettes more expensive and less affordable (WHO, 2010), acknowledging that no one cigarette is healthier than another. South Africa implements a uniform specific tax on cigarettes with the same tax applied to every cigarette. This tax is also appropriate from a fiscal policy perspective since it is easier to administer and provides the most predictable and stable revenue streams compared to tiered or ad valorem systems. The World Health Organization (WHO) recommends that countries rely more on specific taxes as the most optimal tax structure (WHO, 2010) and the WHO Framework Convention on Tobacco Control recommends that countries implement a uniform specific tax, or mixed tax systems with a minimum specific tax floor (WHO, 2014). These systems reduce the incentive for smokers to avoid tax increases by switching to cheaper brands but also ensure that all prices rise with similar magnitudes when taxes increase. This focus on tobacco tax structure rather than just increasing tobacco taxes is prominent in recent literature (Blecher et al., 2013) (Liber et al., 2014) (Chaloupka et al., 2014). 2. Alcohol There is a clear link between the alcohol consumption and alcohol-related diseases and injuries. However, unlike tobacco there is also a link between the dose of alcohol and harm. While the relationship is not linear, a beverage with higher alcohol volume is likely to cause more harm than a beverage with lower alcohol volume. Furthermore, “optimal” doses of alcohol may also be able to provide a protective effect from some diseases (Briasoulis et al.,
2012) (Corrao et al., 2004). A tax system like that recommended for tobacco would make all products less affordable and would lower overall volumes of alcoholic beverages while not necessarily lowering the volume of alcohol consumed (consumers may respond by drinking stronger alcohol). However, a system whereby the tax is applied to the volume of alcohol rather than the volume of beverage could create incentives to reduce harm. South Africa implements such a system on beer and spirits, where the excise tax is applied on the volume of pure alcohol rather than on the volume of the beverage (wine has a flat tax). In the 2013/14 fiscal year, South Africa applied an excise tax of R63,81 per liter of pure alcohol. (Beer comprises 60% of the volume of alcohol consumed in South Africa (Van Walbeek and Blecher, 2014) so we will focus this discussion on beer). Thus, on a standard 330 ml can of 5,5% alcohol beer the excise tax is R1,16, while a lighter alcohol beer of 4,0% alcohol would attract an excise tax of R0,84, or 27% less tax (Van Walbeek and Blecher, 2014). This system has been employed since the 1998/99 fiscal year, before which a flat tax per liter of beer rather than alcohol was applied. Such a system would reduce demand if producers pass on the difference in tax to consumers. Lower alcohol products would be sold at cheaper prices than higher alcohol products, and thus consumption would shift from higher to lower alcohol products. This might occur in a competitive market where producers compete on price. However, the South African beer market is highly concentrated with the largest producer having a 77% market share and the three largest a combined 95% in 2012 (Euromonitor, 2014). A more provocative response is likely to occur on the supply side. Producers could choose not to pass the tax differences onto consumers and rather absorb the higher taxes on products with higher alcohol content, but instead choose to save the taxes on those with lower alcohol content. In such circumstances they could shift advertising and promotion from the higher-tax to lower-tax products. All else remaining constant these would now be more profitable to sell. Alternatively, producers could choose to reduce the alcohol content of existing products and reduce costs through lowering the tax yield or could bring new lower alcohol products to the market. All of these possibilities are likely to reduce the total volume of alcohol consumed and would result in a lowering of the average alcohol volume per liter of beverage.
E. Blecher / Social Science & Medicine 136-137 (2015) 175e179
Since the current tax system was implemented the beer excise tax has increased in real terms by 35%, and spirits and wine have increased by 96% and 90%, respectively. As the excise tax rate increases, the tax difference between beers of varying alcohol content rises. Fig. 2 shows this relationship, with the excise tax per liter of absolute alcohol on the left axis, and the difference in tax between a 4,0% and 5,5% beer on the right axis. The difference in tax between the two beers increases from R0,10 per can to R0,32 (in 2013/14 prices) between 1998/99 and 2013/14. Thus, as the excise tax rate increases, so does the incentive for producers to move to lower alcohol beer through the supply side. The use of 4,0% and 5,5% alcohol volume for the example in Fig. 2 is not accidental. In fact, it refers to two popular beers in South Africa, namely, Castle Lite (4,0%) and Carling Black Label (5,5%). Since the implementation of the dose tax system, there has been a significant shift in advertising (as predicted) from higher alcohol to lower alcohol beers. Fig. 3 shows the weighted average alcohol volume per Rand spent on advertising for major South African beers on the left axis. (The advertising spend for each brand is weighted by the alcohol volume of that brand. The figure shows a downward trend on the left axis meaning that beer advertising has been moving towards lower alcohol beers, coinciding with the increased incentive towards producing lower alcohol beer. South African beer producers have increasingly favored advertising spending on beers with lower alcohol and thus responding to the supply side incentives. As an example, Castle Lite advertising share grew from 7% in 1997 to 26% in 2013. The demand sided incentives do not seem apparent since Castle Lite is priced as a premium beer. The changes in advertising patterns have coincided with sales. Credible market share data is not available, however, we are able to calculate the liters of absolute alcohol consumed for beer per adult in each year using excise revenue data (right axis). The downward trend coincides with the trend in advertising, and shows that South Africans are consuming less alcohol from beer, declining from 4,4 L of alcohol from beer per adult in 1999 to 3,9 L in 2013. A low of 3,7 L occurred in 2010, one year after the lowest weighted average advertising spent.
177
3. Sugar sweetened beverages There has been an increase in recent decades in the consumption of SSBs. This has been reported in both high-income (e.g. the United States) (Duffey et al., 2007) and low- and middle-income countries (e.g. Mexico) (Barquera et al., 2008). The increase in consumption of SSBs has been linked to increases in obesity and diseases including diabetes and heart disease. Malik, Schulze and Hu (Malik et al., 2006) conducted a systematic review of the link between SSBs and weight gain. Studies with long follow-up periods showed a positive association between increased intake of SSBs and weight gain and obesity in adults and children. In a follow-up, Malik et al. (Malik et al., 2010) found that individuals in the highest quintile of consumption of SSBs had a 26% increased risk of developing type two diabetes, and a 20% increased risk of developing metabolic syndrome, than those in the lowest quintile. From a public health perspective the market for non-alcoholic beverages is dichotomous. There are beverages that are sweetened with caloric or nutritional sweeteners (most commonly sucrose sugar, high fructose corn syrup or fruit juice concentrates e we call this SSBs for simplicity) and there are those that have no sweeteners or non-caloric or non-nutritional sweeteners. The distinction between non-sweetened beverages (e.g. water) and those sweetened with non-caloric or non-nutritional sweeteners is controversial. While some non-nutritional sweeteners have been found to be carcinogenic in rats there is little evidence that they are carcinogenic in humans (Weihrauch et al., 2004) (Gallus et al., 2007) (Bosetti et al., 2009). Furthermore, studies indicate that substituting non-sweetened beverages or non-caloric or non-nutritional sweetened for SSBs reduce overweight and obesity at similar magnitudes (Tate et al., 2012). Some argue, however, that substituting to non-nutritional sweetened beverage may not reduce total calorie intake or reduce the risk of obesity (Pereira and 2014). Substantial evidence exists to show that decreasing consumption of SSBs and/or substituting to non-caloric beverages reduces obesity (Brownell et al., 2009). Taxes on SSBs have been proposed as a means of reducing their consumption (Brownell et al., 2009) but few countries have done so. Those that do, follow tobacco control's
Fig. 2. Beer excise taxes in South Africa. Source: National Treasury and StatsSA.
178
E. Blecher / Social Science & Medicine 136-137 (2015) 175e179
Fig. 3. Advertising spend on major South African beer brands and volume of alcohol in beer. Source: South African Advertising Research Foundation and National Treasury.
blunt instrument approach with a uniform specific tax implemented on SSBs and no tax on the unsweetened or non-caloric/ non-nutritional sweetened beverages. The intention is that by making some products more expensive than others that there will be substitution, at the population level, to “healthier” products. This assumes that changes in demand will occur because taxes are passed on to consumers, and that the taxes are significant enough to induce behavioral changes. This model is clearly operating on the demand side. Soft drinks are price inelastic and the price elasticity of demand is estimated to be 0,8, meaning that for every 10% increase in price, consumption is expected to decline by 8% (Andreyeva et al., 2010). An even greater price effect is expected with SSBs since some consumers will switch to other beverages rather than quitting altogether (Brownell et al., 2009). Most prominently, Mexico recently (1 January 2014) introduced a tax of one Peso per liter (approximately US$0,08 per liter or 10% of the average price) on all SSBs (including carbonated and noncarbonated), thus excluding diet drinks (Instituto Nacional de Salud Pública, 2014). The aim is to reduce the demand for SSBs and to increase the demand for other beverages resulting in a decline in calories from beverages. Early evaluation conducted shows a 10% decline in purchases of taxed beverages in the first quarter of 2014 compared to the same quarter of the previous year, and a 7% increase in untaxed beverages (including a 13% increase in plain water purchases) (Instituto Nacional de Salud Pública, 2014) Untaxed beverages include diet sodas, sparkling and still plain water, 100% juices, flavored water with non-caloric sweeteners and milk without added sugar. The early evidence thus suggests that the tax is working as planned. 4. Discussion Increasing rates of obesity, particularly among vulnerable population groups, including the young and poor, are a cause for concern in the public health community. This is contributing to an increase in non-communicable diseases and placing significant burdens on health systems and economies. As policy makers look towards policies to arrest or reverse the increase in obesity they will inevitably look towards fiscal policy. Taxation has been widely used as a fiscal policy tool to influence health behavior, particularly on tobacco and alcohol. While the power of taxation as an
intervention has been well established there are some critical differences between tobacco, alcohol and SSBs. Particularly, there is a different relationship between the dose of the most prominent ingredients and the harm that is causes. The example of South Africa is important. It shows that smartly designed tax structures, in addition to higher tax rates are able to bring about important public health benefits. Using a uniform specific tax, cigarette consumption has declined significantly. Using a dose tax on alcohol, it has applied a harm reduction strategy to beer, reducing the per capita volume of alcohol consumed by incentivizing producers to produce and market lower alcohol beers. In fact, the effectiveness of the dose taxes on beer may have been even larger had the tax increases on beer been equal to or greater than those on spirits and wine. Since spirits and wine taxes increased by significantly more than beer taxes there may have been a substitution from spirits and wine to beer, even while per capita alcohol consumption from beer was declining. However, just like alcohol there is an opportunity to influence the supply side in addition to or instead of attempting to reduce the demand for SSBs. By taxing the dose of caloric sweeteners (i.e. the quantity of sugar) rather than the volume of beverage we are able to incentivize producers to produce and market lower or zero calorie beverages. Lowering the calorie content of beverages could have extraordinary public health effects since it would influence the entire population who consume them. With rates of obesity rising rapidly in many countries, attention has been drawn to the role that SSBs play. Many public health advocates have proposed using taxes as a means to reduce their use and/or encourage substitution to healthier products. The uptake of these policies has been slow and there is an apparent lack of political appetite for it. Governments likely fear the potential economic consequences of tax policies. However, if translating the success of beer taxes in South Africa to SSBs becomes apparent governments might become more receptive. Why are governments hesitant to tax SSBs? Firstly, while tobacco use has a clear externality, the externality of SSBs at low levels of consumption may be insignificant. Secondly, the prevalence rates of SSBs are significantly higher than smoking in many countries. Nelson (Nelson, 2003) shows that tobacco advertising bans were more likely to be imposed in countries where tobacco consumption was already falling. Thus, even though advertising
E. Blecher / Social Science & Medicine 136-137 (2015) 175e179
bans significantly reduce consumption (Saffer and Chaloupka, 2000) (Blecher, 2008) they are less likely in countries with high or increasing consumption. Thus it would be more difficult to implement policies to reduce SSBs use in countries with high or increasing prevalence. Brownell et al. (Brownell et al., 2009) proposed the use of a uniform specific tax on SSBs or alternatively by applying the tax only to beverages that exceed a threshold of added caloric sweetener or kilocalories per ounce. The threshold approach is argued to promote reductions in calories and encouraging manufacturers to reformulate their products. This is likely to be even more effective when applying the tax per calorie rather than a fixed threshold which would likely move many producers to produce just under the threshold rather than a linear incentive to reduce calories. They argue against the application of the tax based on the unit of added sugar on the grounds that it would be difficult to administer. However, as this paper has shown, South Africa has successfully applied this model on beer for over a decade, overcoming these administrative challenges. While the structure of taxes is important, it is also equally important under both approaches that tax rates are also increased over time. While increases in taxes will reduce the overall demand for tobacco, alcohol or SSBs, increases in tax rates under a dose tax will also create further incentives for producers to lower doses. Acknowledgments I would like to thank Jeffrey Drope and Binh Nguyen for comments and suggestions on earlier drafts, and thanks to the WHO country office in South Africa for financial assistance for an earlier project that led to this paper. References Andreyeva, T., Long, M.W., Brownell, K.D., 2010 Feb. The impact of food Prices on consumption: a systematic review of research on the price elasticity of demand for food. Am. J. Public Health 100 (2), 216e222. Barquera, S., Hernandez-Barrera, L., Tolentino, M.L., Espinosa, J., Ng, S.W., Rivera, J.A., et al., 2008 Dec 1. Energy intake from beverages is increasing among Mexican adolescents and adults. J. Nutr. 138 (12), 2454e2461. Blecher, E., 2008. The impact of tobacco advertising bans on consumption in developing countries. J. Health Econ. 27, 930e942. Blecher, E., 2011. The Economics of Tobacco Control in Low- and Middle-income Countries. University of Cape Town, Cape Town. Blecher, E., Ross, H., Stoklosa, M., 2013. Lessons learned from cigarette tax harmonisation in the European Union. Tob. Control 23 (e1), e12ee14. Bosetti, C., Gallus, S., Talamini, R., Montella, M., Franceschi, S., Negri, E., et al., 2009 Aug 1. Artificial sweeteners and the risk of gastric, pancreatic, and Endometrial cancers in Italy. Cancer Epidemiol. Biomarkers Prev. 18 (8), 2235e2238. Briasoulis, A., Agarwal, V., Messerli, F.H., 2012 Nov 1. Alcohol consumption and the
179
risk of hypertension in men and women: a systematic review and meta-analysis. J. Clin. Hypertens. 14 (11), 792e798. Brownell, K.D., Farley, T., Willett, W.C., Popkin, B.M., Chaloupka, F.J., Thompson, J.W., et al., 2009 Oct 15. The public health and economic benefits of taxing sugarsweetened beverages. N. Engl. J. Med. 361 (16), 1599e1605. Chaloupka, F.J., Kostova, D., Shang, C., 2014 Jan 1. Cigarette excise tax structure and cigarette prices: evidence from the global adult tobacco survey and the U.S. National adult tobacco survey. Nicotine Tob. Res. 16 (Suppl. 1), S3eS9. Corrao, G., Bagnardi, V., Zambon, A., La Vecchia, C., 2004 May. A meta-analysis of alcohol consumption and the risk of 15 diseases. Prev. Med. 38 (5), 613e619. Duffey, K.J., Popkin, B.M., 2007 Nov 1. Shifts in patterns and consumption of beverages between 1965 and 2002. Obesity 15 (11), 2739e2747. Euromonitor, 2014. Beer in South Africa. Euromonitor International. Gallus, S., Scotti, L., Negri, E., Talamini, R., Franceschi, S., Montella, M., et al., 2007 Jan 1. Artificial sweeteners and cancer risk in a network of caseecontrol studies. Ann. Oncol. 18 (1), 40e44. IARC, 2011. Handbooks of cancer prevention, tobacco control. In: Effectiveness of Tax and Price Policies for Tobacco Control, vol. 14. International Agency for Research on Cancer, Lyon. Instituto Nacional de Salud Pública, 2014. Resultados preliminares sobre los efectos xico. Instituto Nacional de del impuesto de un peso a bebidas azucaradas en Me Salud Pública [cited 2014 Sep 10]. Available from: http://www.insp.mx/epppo/ blog/preliminares-bebidas-azucaradas.html. Liber, A.C., Ross, H., Ratanachena, S., Dorotheo, E.U., Foong, K., 2014. Cigarette price level and variation in five Southeast Asian countries. Tob. Control 0, 1e5. Malik, V.S., Schulze, M.B., Hu, F.B., 2006 Aug 1. Intake of sugar-sweetened beverages and weight gain: a systematic review. Am. J. Clin. Nutr. 84 (2), 274e288. s, J.-P., Willett, W.C., Hu, F.B., 2010 Nov 1. Malik, V.S., Popkin, B.M., Bray, G.A., Despre Sugar-sweetened beverages and risk of metabolic syndrome and type 2 diabetes a meta-analysis. Diabetes Care 33 (11), 2477e2483. Nelson, J., 2003. Cigarette demand, structural change, and advertising bans: international evidence, 1970-1995. Contrib. Econ. Anal. Policy 2 (1). Article 10. Nugent, R., Knaul, F., 2006. Fiscal policies for health promotion and disease prevention. In: Jamison, D.T., Breman, J.G., Measham, A.R., Alleyne, G., Claeson, M., Evans, D.B., et al. (Eds.), Disease Control Priorities in Developing Countries, second ed. World Bank, Washington. Pereira, M.A., 2014 Nov 1. Sugar-sweetened and Artificially-sweetened beverages in relation to obesity risk. Adv. Nutr. Int. Rev. J. 5 (6), 797e808. Saffer, H., Chaloupka, F., 2000. The effect of tobacco advertising bans on tobacco consumption. J. Health Econ. 19, 1117e1137. Smith, A., 1776. An Inquiry into the Nature and Causes of the Wealth of Nations. The Modern Library, New York. Surgeon General, 2004. The Health Consequences of Smoking: A Report of the Surgeon General. Centers for Disease Control and Prevention, Atlanta. Surgeon General, 2010. How Tobacco Smoke Causes Disease: The Biology and Behavioral Basis for Smoking-attributable Disease: A Report of the Surgeon General. Centers for Disease Control and Prevention, Atlanta. Tate, D.F., Turner-McGrievy, G., Lyons, E., Stevens, J., Erickson, K., Polzien, K., et al., 2012 Mar 1. Replacing caloric beverages with water or diet beverages for weight loss in adults: main results of the choose healthy options consciously Everyday (CHOICE) randomized clinical trial. Am. J. Clin. Nutr. 95 (3), 555e563. Van Walbeek, C., Blecher, E., 2014. The Economics of Alcohol Use, Misuse and Policy in South Africa. University of Cape Town, Cape Town. Weihrauch, M.R., Diehl, V., 2004 Oct 1. Artificial sweetenersddo they bear a carcinogenic risk? Ann. Oncol. 15 (10), 1460e1465. WHO, 2014. Guidelines for Implementation of Article 6 of the WHO FCTC: Price and Tax Measures to Reduce the Demand for Tobacco. WHO, 2010. Technical Manual on Tobacco Tax Administration. World Health Organization, Geneva.
Contents lists available at ScienceDirect
Social Science & Medicine journal homepage: www.elsevier.com/locate/socscimed
Short report
Taxes on tobacco, alcohol and sugar sweetened beverages: Linkages and lessons learned Evan Blecher a, b, * a b
Economic and Health Policy Research, American Cancer Society, USA Southern African Labour and Development Research Unit, University of Cape Town, South Africa
a r t i c l e i n f o
a b s t r a c t
Article history: Available online 14 May 2015
Increased consumption of sugar-sweetened beverages (SSBs) has been linked to increases in obesity in both high-income and low- and middle-income countries. Tobacco and alcohol taxes have proven to be effective tools to reduce tobacco and alcohol use. Many public health advocates propose using similar taxes to reduce consumption of SSBs. South Africa is a middle-income country that is considered a leader in the area of tobacco tax policy. A case study of tobacco and alcohol taxes is used to better understand optimal tax structures for SSBs. The case study tracks aggregate data over time on taxes, prices, consumption, tax revenues, and marketing expenditures at the brand level. Tobacco and alcohol taxes are shown to be effective in reducing the demand for tobacco. Additionally, taxes on the dose of alcohol rather than the volume of the beverage may incentivize producers to reduce the volume of alcohol in beverages through the supply side. While specific taxes based on the volume of beverages are likely to reduce the demand for SSBs, policy makers should also consider taxes on alcohol and SSBs that tax the dose of the alcohol and calories in order to create supply-side incentives for producers to lower alcohol and calorie levels in existing products or promote products with lower levels of alcohol and calories. © 2015 Elsevier Ltd. All rights reserved.
Keywords: Alcohol Tobacco Sugar-sweetened beverages Tax South Africa
The use of taxation as a public health intervention to reduce the consumption of products that have negative externalities is not a new concept. Adam Smith is famously quoted in the Wealth of Nations in 1776 as supporting the taxation of tobacco, alcohol, and sugar since none were considered necessary for life (Smith, 1776). While the taxation of tobacco for fiscal and health purposes is now universally accepted, the use of taxes on alcohol to influence health is still in its infancy even though it is commonly used as a fiscal tool (Nugent et al., 2006). The taxation of sugar-sweetened beverages (SSBs) is rare, and where used, it is typically used for health rather than fiscal purposes. This should not be confused with excise taxes on carbonated beverages which are typically used for fiscal purposes, however, these taxes usually do not distinguish between those sweetened with nutritional or non-nutritional sweeteners. This paper argues that while the experience of tobacco is very useful in the public policy space, the structures of taxes may need to be different when applied to alcohol and SSBs. While the experience of tobacco tends to favor specific taxes per cigarette, this paper will show that specific taxes on alcohol and SSBs may be more
* 250 Williams St NW, Atlanta, GA, 30303, USA. E-mail address: [email protected]. http://dx.doi.org/10.1016/j.socscimed.2015.05.022 0277-9536/© 2015 Elsevier Ltd. All rights reserved.
effective when taxing the volume of alcohol or sugar rather than the volume of the beverage. The case study of South Africa is used throughout. Ethical approval was not required since the study used only secondary data and no human subjects were involved. 1. Tobacco Higher taxes that result in higher prices result in lower consumption of cigarettes and smoking prevalence (IARC, 2011). A prominent example of this policy instrument is South Africa where between 1990 and 2012, real (i.e. inflation adjusted) excise taxes rose by 552% leading to a 217% increase in real prices (Fig. 1). The resultant decline in consumption amounted to 41% and 66% in aggregate and per capita terms, respectively. This coincided with a decline in adult smoking prevalence from 33% to 20% between 1993 and 2012 (Blecher, 2011). Given the growth in population during this time there are presently 4,5 million fewer smokers than there would have been had smoking prevalence in 1993 still prevailed. At the same time, annual revenue collections increased in real terms by 284%. South Africa shows a clear fiscal benefit from increasing tobacco taxes through increased revenues, and health benefits through lower tobacco use. This is a demand-side policy in that it intends to
176
E. Blecher / Social Science & Medicine 136-137 (2015) 175e179
Fig. 1. Cigarette excise taxes, price and consumption in South Africa. Source: National Treasury and StatsSA.
reduce the population and individual level demand for tobacco by increasing the price of the product. Tobacco is most commonly consumed as cigarettes, and while there is a clear link between smoking and certain diseases including cancers and cardiovascular disease, there is no significant link between the strength of the addictive component (i.e. the nicotine) and the health consequences thereof. Reducing the nicotine content of the cigarette will not result in reduced health consequences of cigarettes (Surgeon General, 2004). Health consequences are almost exclusively linked to the quantity of cigarettes smoked and the number of years that a smoker smokes (Surgeon General, 2010). As a result, there is consensus that taxation is most effective as a blunt instrument which makes all cigarettes more expensive and less affordable (WHO, 2010), acknowledging that no one cigarette is healthier than another. South Africa implements a uniform specific tax on cigarettes with the same tax applied to every cigarette. This tax is also appropriate from a fiscal policy perspective since it is easier to administer and provides the most predictable and stable revenue streams compared to tiered or ad valorem systems. The World Health Organization (WHO) recommends that countries rely more on specific taxes as the most optimal tax structure (WHO, 2010) and the WHO Framework Convention on Tobacco Control recommends that countries implement a uniform specific tax, or mixed tax systems with a minimum specific tax floor (WHO, 2014). These systems reduce the incentive for smokers to avoid tax increases by switching to cheaper brands but also ensure that all prices rise with similar magnitudes when taxes increase. This focus on tobacco tax structure rather than just increasing tobacco taxes is prominent in recent literature (Blecher et al., 2013) (Liber et al., 2014) (Chaloupka et al., 2014). 2. Alcohol There is a clear link between the alcohol consumption and alcohol-related diseases and injuries. However, unlike tobacco there is also a link between the dose of alcohol and harm. While the relationship is not linear, a beverage with higher alcohol volume is likely to cause more harm than a beverage with lower alcohol volume. Furthermore, “optimal” doses of alcohol may also be able to provide a protective effect from some diseases (Briasoulis et al.,
2012) (Corrao et al., 2004). A tax system like that recommended for tobacco would make all products less affordable and would lower overall volumes of alcoholic beverages while not necessarily lowering the volume of alcohol consumed (consumers may respond by drinking stronger alcohol). However, a system whereby the tax is applied to the volume of alcohol rather than the volume of beverage could create incentives to reduce harm. South Africa implements such a system on beer and spirits, where the excise tax is applied on the volume of pure alcohol rather than on the volume of the beverage (wine has a flat tax). In the 2013/14 fiscal year, South Africa applied an excise tax of R63,81 per liter of pure alcohol. (Beer comprises 60% of the volume of alcohol consumed in South Africa (Van Walbeek and Blecher, 2014) so we will focus this discussion on beer). Thus, on a standard 330 ml can of 5,5% alcohol beer the excise tax is R1,16, while a lighter alcohol beer of 4,0% alcohol would attract an excise tax of R0,84, or 27% less tax (Van Walbeek and Blecher, 2014). This system has been employed since the 1998/99 fiscal year, before which a flat tax per liter of beer rather than alcohol was applied. Such a system would reduce demand if producers pass on the difference in tax to consumers. Lower alcohol products would be sold at cheaper prices than higher alcohol products, and thus consumption would shift from higher to lower alcohol products. This might occur in a competitive market where producers compete on price. However, the South African beer market is highly concentrated with the largest producer having a 77% market share and the three largest a combined 95% in 2012 (Euromonitor, 2014). A more provocative response is likely to occur on the supply side. Producers could choose not to pass the tax differences onto consumers and rather absorb the higher taxes on products with higher alcohol content, but instead choose to save the taxes on those with lower alcohol content. In such circumstances they could shift advertising and promotion from the higher-tax to lower-tax products. All else remaining constant these would now be more profitable to sell. Alternatively, producers could choose to reduce the alcohol content of existing products and reduce costs through lowering the tax yield or could bring new lower alcohol products to the market. All of these possibilities are likely to reduce the total volume of alcohol consumed and would result in a lowering of the average alcohol volume per liter of beverage.
E. Blecher / Social Science & Medicine 136-137 (2015) 175e179
Since the current tax system was implemented the beer excise tax has increased in real terms by 35%, and spirits and wine have increased by 96% and 90%, respectively. As the excise tax rate increases, the tax difference between beers of varying alcohol content rises. Fig. 2 shows this relationship, with the excise tax per liter of absolute alcohol on the left axis, and the difference in tax between a 4,0% and 5,5% beer on the right axis. The difference in tax between the two beers increases from R0,10 per can to R0,32 (in 2013/14 prices) between 1998/99 and 2013/14. Thus, as the excise tax rate increases, so does the incentive for producers to move to lower alcohol beer through the supply side. The use of 4,0% and 5,5% alcohol volume for the example in Fig. 2 is not accidental. In fact, it refers to two popular beers in South Africa, namely, Castle Lite (4,0%) and Carling Black Label (5,5%). Since the implementation of the dose tax system, there has been a significant shift in advertising (as predicted) from higher alcohol to lower alcohol beers. Fig. 3 shows the weighted average alcohol volume per Rand spent on advertising for major South African beers on the left axis. (The advertising spend for each brand is weighted by the alcohol volume of that brand. The figure shows a downward trend on the left axis meaning that beer advertising has been moving towards lower alcohol beers, coinciding with the increased incentive towards producing lower alcohol beer. South African beer producers have increasingly favored advertising spending on beers with lower alcohol and thus responding to the supply side incentives. As an example, Castle Lite advertising share grew from 7% in 1997 to 26% in 2013. The demand sided incentives do not seem apparent since Castle Lite is priced as a premium beer. The changes in advertising patterns have coincided with sales. Credible market share data is not available, however, we are able to calculate the liters of absolute alcohol consumed for beer per adult in each year using excise revenue data (right axis). The downward trend coincides with the trend in advertising, and shows that South Africans are consuming less alcohol from beer, declining from 4,4 L of alcohol from beer per adult in 1999 to 3,9 L in 2013. A low of 3,7 L occurred in 2010, one year after the lowest weighted average advertising spent.
177
3. Sugar sweetened beverages There has been an increase in recent decades in the consumption of SSBs. This has been reported in both high-income (e.g. the United States) (Duffey et al., 2007) and low- and middle-income countries (e.g. Mexico) (Barquera et al., 2008). The increase in consumption of SSBs has been linked to increases in obesity and diseases including diabetes and heart disease. Malik, Schulze and Hu (Malik et al., 2006) conducted a systematic review of the link between SSBs and weight gain. Studies with long follow-up periods showed a positive association between increased intake of SSBs and weight gain and obesity in adults and children. In a follow-up, Malik et al. (Malik et al., 2010) found that individuals in the highest quintile of consumption of SSBs had a 26% increased risk of developing type two diabetes, and a 20% increased risk of developing metabolic syndrome, than those in the lowest quintile. From a public health perspective the market for non-alcoholic beverages is dichotomous. There are beverages that are sweetened with caloric or nutritional sweeteners (most commonly sucrose sugar, high fructose corn syrup or fruit juice concentrates e we call this SSBs for simplicity) and there are those that have no sweeteners or non-caloric or non-nutritional sweeteners. The distinction between non-sweetened beverages (e.g. water) and those sweetened with non-caloric or non-nutritional sweeteners is controversial. While some non-nutritional sweeteners have been found to be carcinogenic in rats there is little evidence that they are carcinogenic in humans (Weihrauch et al., 2004) (Gallus et al., 2007) (Bosetti et al., 2009). Furthermore, studies indicate that substituting non-sweetened beverages or non-caloric or non-nutritional sweetened for SSBs reduce overweight and obesity at similar magnitudes (Tate et al., 2012). Some argue, however, that substituting to non-nutritional sweetened beverage may not reduce total calorie intake or reduce the risk of obesity (Pereira and 2014). Substantial evidence exists to show that decreasing consumption of SSBs and/or substituting to non-caloric beverages reduces obesity (Brownell et al., 2009). Taxes on SSBs have been proposed as a means of reducing their consumption (Brownell et al., 2009) but few countries have done so. Those that do, follow tobacco control's
Fig. 2. Beer excise taxes in South Africa. Source: National Treasury and StatsSA.
178
E. Blecher / Social Science & Medicine 136-137 (2015) 175e179
Fig. 3. Advertising spend on major South African beer brands and volume of alcohol in beer. Source: South African Advertising Research Foundation and National Treasury.
blunt instrument approach with a uniform specific tax implemented on SSBs and no tax on the unsweetened or non-caloric/ non-nutritional sweetened beverages. The intention is that by making some products more expensive than others that there will be substitution, at the population level, to “healthier” products. This assumes that changes in demand will occur because taxes are passed on to consumers, and that the taxes are significant enough to induce behavioral changes. This model is clearly operating on the demand side. Soft drinks are price inelastic and the price elasticity of demand is estimated to be 0,8, meaning that for every 10% increase in price, consumption is expected to decline by 8% (Andreyeva et al., 2010). An even greater price effect is expected with SSBs since some consumers will switch to other beverages rather than quitting altogether (Brownell et al., 2009). Most prominently, Mexico recently (1 January 2014) introduced a tax of one Peso per liter (approximately US$0,08 per liter or 10% of the average price) on all SSBs (including carbonated and noncarbonated), thus excluding diet drinks (Instituto Nacional de Salud Pública, 2014). The aim is to reduce the demand for SSBs and to increase the demand for other beverages resulting in a decline in calories from beverages. Early evaluation conducted shows a 10% decline in purchases of taxed beverages in the first quarter of 2014 compared to the same quarter of the previous year, and a 7% increase in untaxed beverages (including a 13% increase in plain water purchases) (Instituto Nacional de Salud Pública, 2014) Untaxed beverages include diet sodas, sparkling and still plain water, 100% juices, flavored water with non-caloric sweeteners and milk without added sugar. The early evidence thus suggests that the tax is working as planned. 4. Discussion Increasing rates of obesity, particularly among vulnerable population groups, including the young and poor, are a cause for concern in the public health community. This is contributing to an increase in non-communicable diseases and placing significant burdens on health systems and economies. As policy makers look towards policies to arrest or reverse the increase in obesity they will inevitably look towards fiscal policy. Taxation has been widely used as a fiscal policy tool to influence health behavior, particularly on tobacco and alcohol. While the power of taxation as an
intervention has been well established there are some critical differences between tobacco, alcohol and SSBs. Particularly, there is a different relationship between the dose of the most prominent ingredients and the harm that is causes. The example of South Africa is important. It shows that smartly designed tax structures, in addition to higher tax rates are able to bring about important public health benefits. Using a uniform specific tax, cigarette consumption has declined significantly. Using a dose tax on alcohol, it has applied a harm reduction strategy to beer, reducing the per capita volume of alcohol consumed by incentivizing producers to produce and market lower alcohol beers. In fact, the effectiveness of the dose taxes on beer may have been even larger had the tax increases on beer been equal to or greater than those on spirits and wine. Since spirits and wine taxes increased by significantly more than beer taxes there may have been a substitution from spirits and wine to beer, even while per capita alcohol consumption from beer was declining. However, just like alcohol there is an opportunity to influence the supply side in addition to or instead of attempting to reduce the demand for SSBs. By taxing the dose of caloric sweeteners (i.e. the quantity of sugar) rather than the volume of beverage we are able to incentivize producers to produce and market lower or zero calorie beverages. Lowering the calorie content of beverages could have extraordinary public health effects since it would influence the entire population who consume them. With rates of obesity rising rapidly in many countries, attention has been drawn to the role that SSBs play. Many public health advocates have proposed using taxes as a means to reduce their use and/or encourage substitution to healthier products. The uptake of these policies has been slow and there is an apparent lack of political appetite for it. Governments likely fear the potential economic consequences of tax policies. However, if translating the success of beer taxes in South Africa to SSBs becomes apparent governments might become more receptive. Why are governments hesitant to tax SSBs? Firstly, while tobacco use has a clear externality, the externality of SSBs at low levels of consumption may be insignificant. Secondly, the prevalence rates of SSBs are significantly higher than smoking in many countries. Nelson (Nelson, 2003) shows that tobacco advertising bans were more likely to be imposed in countries where tobacco consumption was already falling. Thus, even though advertising
E. Blecher / Social Science & Medicine 136-137 (2015) 175e179
bans significantly reduce consumption (Saffer and Chaloupka, 2000) (Blecher, 2008) they are less likely in countries with high or increasing consumption. Thus it would be more difficult to implement policies to reduce SSBs use in countries with high or increasing prevalence. Brownell et al. (Brownell et al., 2009) proposed the use of a uniform specific tax on SSBs or alternatively by applying the tax only to beverages that exceed a threshold of added caloric sweetener or kilocalories per ounce. The threshold approach is argued to promote reductions in calories and encouraging manufacturers to reformulate their products. This is likely to be even more effective when applying the tax per calorie rather than a fixed threshold which would likely move many producers to produce just under the threshold rather than a linear incentive to reduce calories. They argue against the application of the tax based on the unit of added sugar on the grounds that it would be difficult to administer. However, as this paper has shown, South Africa has successfully applied this model on beer for over a decade, overcoming these administrative challenges. While the structure of taxes is important, it is also equally important under both approaches that tax rates are also increased over time. While increases in taxes will reduce the overall demand for tobacco, alcohol or SSBs, increases in tax rates under a dose tax will also create further incentives for producers to lower doses. Acknowledgments I would like to thank Jeffrey Drope and Binh Nguyen for comments and suggestions on earlier drafts, and thanks to the WHO country office in South Africa for financial assistance for an earlier project that led to this paper. References Andreyeva, T., Long, M.W., Brownell, K.D., 2010 Feb. The impact of food Prices on consumption: a systematic review of research on the price elasticity of demand for food. Am. J. Public Health 100 (2), 216e222. Barquera, S., Hernandez-Barrera, L., Tolentino, M.L., Espinosa, J., Ng, S.W., Rivera, J.A., et al., 2008 Dec 1. Energy intake from beverages is increasing among Mexican adolescents and adults. J. Nutr. 138 (12), 2454e2461. Blecher, E., 2008. The impact of tobacco advertising bans on consumption in developing countries. J. Health Econ. 27, 930e942. Blecher, E., 2011. The Economics of Tobacco Control in Low- and Middle-income Countries. University of Cape Town, Cape Town. Blecher, E., Ross, H., Stoklosa, M., 2013. Lessons learned from cigarette tax harmonisation in the European Union. Tob. Control 23 (e1), e12ee14. Bosetti, C., Gallus, S., Talamini, R., Montella, M., Franceschi, S., Negri, E., et al., 2009 Aug 1. Artificial sweeteners and the risk of gastric, pancreatic, and Endometrial cancers in Italy. Cancer Epidemiol. Biomarkers Prev. 18 (8), 2235e2238. Briasoulis, A., Agarwal, V., Messerli, F.H., 2012 Nov 1. Alcohol consumption and the
179
risk of hypertension in men and women: a systematic review and meta-analysis. J. Clin. Hypertens. 14 (11), 792e798. Brownell, K.D., Farley, T., Willett, W.C., Popkin, B.M., Chaloupka, F.J., Thompson, J.W., et al., 2009 Oct 15. The public health and economic benefits of taxing sugarsweetened beverages. N. Engl. J. Med. 361 (16), 1599e1605. Chaloupka, F.J., Kostova, D., Shang, C., 2014 Jan 1. Cigarette excise tax structure and cigarette prices: evidence from the global adult tobacco survey and the U.S. National adult tobacco survey. Nicotine Tob. Res. 16 (Suppl. 1), S3eS9. Corrao, G., Bagnardi, V., Zambon, A., La Vecchia, C., 2004 May. A meta-analysis of alcohol consumption and the risk of 15 diseases. Prev. Med. 38 (5), 613e619. Duffey, K.J., Popkin, B.M., 2007 Nov 1. Shifts in patterns and consumption of beverages between 1965 and 2002. Obesity 15 (11), 2739e2747. Euromonitor, 2014. Beer in South Africa. Euromonitor International. Gallus, S., Scotti, L., Negri, E., Talamini, R., Franceschi, S., Montella, M., et al., 2007 Jan 1. Artificial sweeteners and cancer risk in a network of caseecontrol studies. Ann. Oncol. 18 (1), 40e44. IARC, 2011. Handbooks of cancer prevention, tobacco control. In: Effectiveness of Tax and Price Policies for Tobacco Control, vol. 14. International Agency for Research on Cancer, Lyon. Instituto Nacional de Salud Pública, 2014. Resultados preliminares sobre los efectos xico. Instituto Nacional de del impuesto de un peso a bebidas azucaradas en Me Salud Pública [cited 2014 Sep 10]. Available from: http://www.insp.mx/epppo/ blog/preliminares-bebidas-azucaradas.html. Liber, A.C., Ross, H., Ratanachena, S., Dorotheo, E.U., Foong, K., 2014. Cigarette price level and variation in five Southeast Asian countries. Tob. Control 0, 1e5. Malik, V.S., Schulze, M.B., Hu, F.B., 2006 Aug 1. Intake of sugar-sweetened beverages and weight gain: a systematic review. Am. J. Clin. Nutr. 84 (2), 274e288. s, J.-P., Willett, W.C., Hu, F.B., 2010 Nov 1. Malik, V.S., Popkin, B.M., Bray, G.A., Despre Sugar-sweetened beverages and risk of metabolic syndrome and type 2 diabetes a meta-analysis. Diabetes Care 33 (11), 2477e2483. Nelson, J., 2003. Cigarette demand, structural change, and advertising bans: international evidence, 1970-1995. Contrib. Econ. Anal. Policy 2 (1). Article 10. Nugent, R., Knaul, F., 2006. Fiscal policies for health promotion and disease prevention. In: Jamison, D.T., Breman, J.G., Measham, A.R., Alleyne, G., Claeson, M., Evans, D.B., et al. (Eds.), Disease Control Priorities in Developing Countries, second ed. World Bank, Washington. Pereira, M.A., 2014 Nov 1. Sugar-sweetened and Artificially-sweetened beverages in relation to obesity risk. Adv. Nutr. Int. Rev. J. 5 (6), 797e808. Saffer, H., Chaloupka, F., 2000. The effect of tobacco advertising bans on tobacco consumption. J. Health Econ. 19, 1117e1137. Smith, A., 1776. An Inquiry into the Nature and Causes of the Wealth of Nations. The Modern Library, New York. Surgeon General, 2004. The Health Consequences of Smoking: A Report of the Surgeon General. Centers for Disease Control and Prevention, Atlanta. Surgeon General, 2010. How Tobacco Smoke Causes Disease: The Biology and Behavioral Basis for Smoking-attributable Disease: A Report of the Surgeon General. Centers for Disease Control and Prevention, Atlanta. Tate, D.F., Turner-McGrievy, G., Lyons, E., Stevens, J., Erickson, K., Polzien, K., et al., 2012 Mar 1. Replacing caloric beverages with water or diet beverages for weight loss in adults: main results of the choose healthy options consciously Everyday (CHOICE) randomized clinical trial. Am. J. Clin. Nutr. 95 (3), 555e563. Van Walbeek, C., Blecher, E., 2014. The Economics of Alcohol Use, Misuse and Policy in South Africa. University of Cape Town, Cape Town. Weihrauch, M.R., Diehl, V., 2004 Oct 1. Artificial sweetenersddo they bear a carcinogenic risk? Ann. Oncol. 15 (10), 1460e1465. WHO, 2014. Guidelines for Implementation of Article 6 of the WHO FCTC: Price and Tax Measures to Reduce the Demand for Tobacco. WHO, 2010. Technical Manual on Tobacco Tax Administration. World Health Organization, Geneva.