Implications of a Sugar Tax in New Zealand: Incidence and Effectiveness

Gardiner, Alasdair Working Paper Implications of a Sugar Tax in New Zealand: Incidence and Effectiveness New Zealand Treasury Working Paper, No. 16/09 Provided in Cooperation with: The Treasury, New Zealand Government Suggested Citation: Gardiner, Alasdair (2016) : Implications of a Sugar Tax in New Zealand: Incidence and Effectiveness, New Zealand Treasury Working Paper, No. 16/09, ISBN 978-0-947519-48-3, New Zealand Government, The Treasury, Wellington This Version is available at: https://hdl.handle.net/10419/205703 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by/4.0/ Implications of a Sugar Tax in New Zealand: Incidence and Effectiveness Alasdair Gardiner New Zealand Treasury Working Paper 16/09 November 2016 DISCLAIMER The views, opinions, findings, and conclusions or recommendations expressed in this Working Paper are strictly those of the author(s). They do not necessarily reflect the views of the New Zealand Treasury or the New Zealand Government. The New Zealand Treasury and the New Zealand Government take no responsibility for any errors or omissions in, or for the correctness of, the information contained in these working papers. The paper is presented not as policy, but with a view to inform and stimulate wider debate. NZ TREASURY WORKING PAPER 16/09 Implications of a Sugar Tax in New Zealand: Incidence and Effectiveness MONTH / YEAR November 2016 AUTHOR Alasdair Gardiner Policy Advisor, Financial Markets Ministry of Business, Innovation and Employment PO Box 1473 Wellington 6140 New Zealand Email [email protected] Telephone +64 4 901 8547 This paper was written as a summer internship project while in the Tax Strategy team at Treasury. ISBN (O NLINE ) 978-0-947519-48-3 URL Treasury website at November 2016: http://www.treasury.govt.nz/publications/research-policy/wp/2016/16-09 ACKNOWLEDGEMENTS I am grateful to John Creedy for his advice throughout the drafting of this paper, Kristie Carter and John Marney for discussions in the preliminary stages, Matt Cowan and Suzy Morrissey for constructive suggestions. Christopher Ball provided valuable help with the data. Thanks must also go to the Policy and Strategy Team at Inland Revenue for inviting me to present these findings to them and providing helpful feedback. Access to data used in this paper was provided by Statistics New Zealand under conditions designed to give effect to the security and confidentiality provisions of the Statistics Act 1975. The graphs presented in this report are the work of the author and not Statistics New Zealand. The views, opinions, findings and conclusions are strictly those of the author and do not necessarily reflect the views of the New Zealand Treasury. NZ TREASURY New Zealand Treasury PO Box 3724 Wellington 6008 NEW ZEALAND Email Telephone Website [email protected] 64-4-472 2733 www.treasury.govt.nz WP 16/09 | Implications of a Sugar Tax in New Zealand: Incidence and Effectiveness i Abstract This paper has two aims. First, it surveys some of the literature on the likely effectiveness of sugar taxes as a policy instrument for reducing morbidity and mortality associated with obesity. There is a wide range of estimates among the literature of the price elasticity of demand for sugary products. A plurality of studies found that groups most at risk from obesity have greater price sensitivity. Studies also found there is a risk of consumers substituting unhealthy but non-taxed products for taxed products, negating any potential health improvements from a tax. The paper’s second aim is to build on the literature review by analysing the possible incidence of a sugar tax in New Zealand, based on New Zealand household expenditure data. The empirical analysis presented is consistent with international evidence that a sugar tax would be regressive at the general population level. JEL CLASSIFICATION H2; H3 KEYWORDS Sugar sweetened beverages; tax progressivity WP 16/09 | Implications of a Sugar Tax in New Zealand: Incidence and Effectiveness ii Executive Summary Excessive consumption of sugary products, especially sugar-sweetened beverages (SSBs), is a major contributor to New Zealand’s high rate of obesity, as well as causing a range of other health problems, such as type two diabetes and dental caries. Taxes on sugary products have been implemented by several countries to incentivise their substitution with healthier alternatives, in an attempt to reduce obesity rates. This paper explores the underlying logic for, and some of the challenges associated with, sugar taxes as a policy instrument for reducing obesity. Empirical analysis of New Zealand data is consistent with international evidence that a sugar tax is likely to be regressive. Low-income consumers spend a higher proportion of their income on the targeted food groups and so bear a relatively higher burden of the tax. However, this varies based on household type and ethnicity. The highest obesity rates in New Zealand occur amongst lower socio-economic groups, Māori, Pasifika and those with high consumption of SSBs. An argument sometimes advanced in favour of the tax is that the health gains are likely to be progressive, with the largest reductions in obesity rates occurring among those with high obesity rates or at most risk of becoming obese. Whether this is true depends on the price elasticity of demand of the targeted groups. The general population was found to have slightly inelastic demand for SSBs. Some studies suggested that the high-risk groups had a higher price elasticity of demand, but this was not conclusive. Substitution towards other unhealthy products rather than towards the intended healthy alternatives is another potential problem, which is accentuated with higher price elasticities. WP 16/09 | Implications of a Sugar Tax in New Zealand: Incidence and Effectiveness iii Table of Contents Abstract ............................................................................................................................. i Executive Summary ......................................................................................................... ii 1 Introduction............................................................................................................. 1 2 Background to sugar taxes .................................................................................... 2 2.1 Theoretical basis ............................................................................................................. 2 2.2 Sugar taxes in practice ................................................................................................... 2 2.3 Rationale for targeting SSBs ........................................................................................... 3 2.4 Indirect costs of obesity .................................................................................................. 3 2.5 Intervention logic ............................................................................................................. 4 3 Literature review ..................................................................................................... 5 3.1 Share of energy consumption from SSBs ....................................................................... 5 3.2 Summary of literature on SSB elasticities ....................................................................... 7 3.3 Review of effectiveness of sugar taxes ........................................................................ 10 4 Incidence of a sugar tax ....................................................................................... 12 4.1 Method of analysis ........................................................................................................ 12 4.2 Budget shares ............................................................................................................... 13 5 Conclusions .......................................................................................................... 18 References ..................................................................................................................... 19 List of Tables Table 1 – Summary of literature on share of total energy consumption from SSBs and other beverages ............................................................................................................................ 6 Table 2 – Summary of literature on own-price elasticities for SSBs .................................................... 8 Table 3 – Summary of literature on cross-price elasticities of SSBs with respect to related products ............................................................................................................................... 9 Table 4 – Summary of literature on effectiveness of sugar taxes ..................................................... 10 List of Figures Figure 1 – Intervention logic for a sugar tax (adapted from Nnoaham et al, 2009) ............................. 4 Figure 2 – Weighted average expenditure proportion for all households .......................................... 14 Figure 3 – Weighted average expenditure proportion for two or more adults without children ......... 14 Figure 4 – Weighted average expenditure proportion for two or more adults with children .............. 15 Figure 5 – Weighted average expenditure proportion for single adult households ........................... 15 WP 16/09 | Implications of a Sugar Tax in New Zealand: Incidence and Effectiveness 1 Implications of a Sugar Tax in New Zealand: Incidence and Effectiveness 1 Introduction The social costs of obesity are well-established and represent a significant current and future fiscal cost to the healthcare budget (Lal et al, 2012). A primary cause is the overconsumption of high-calorie food and drink, especially sugar-sweetened beverages (SSBs) (Mytton et al, 2012). Many countries have implemented health-related taxes – a “fat tax” or “sugar tax” – to incentivise the replacement of unhealthy products with healthier substitutes, which in theory reduces the energy intake of the average consumer and decreases the prevalence of obesity. Given New Zealand’s high rate of obesity, there have been calls for a similar tax to be introduced in this country (New Zealand Beverage Guidance Panel, 2014). However, there is considerable uncertainty regarding the effects of these taxes on consumption behaviour and the extent to which they result in a measurable improvement in health outcomes (Jeram, 2016). This paper reviews some of the current literature on the effects of sugar taxes and considers the incidence of the tax to inform debate on this policy instrument as a tool for reducing the social costs of excess sugar consumption. Further analysis, outside the scope of this study, is required to determine the effect of implementing a sugar tax on obesity rates, or other health outcomes, in New Zealand. Section 2 provides some background information regarding sugar taxes. Section 3 briefly reviews some of the literature on sugar taxes, including empirical estimates of relevant demand elasticities. Section 4 reports on the potential incidence of such a tax in New Zealand. Brief conclusions are drawn in section 5. WP 16/09 | Implications of a Sugar Tax in New Zealand: Incidence and Effectiveness 2 2 Background to sugar taxes 2.1 Theoretical basis The theoretical basis for a sugar tax is that overconsumption of high-sugar products is a market failure which imposes negative externalities on society through increased health costs and higher rates of premature death from a range of non-communicable diseases. Conceptually, a sugar tax could be considered an extension of existing so-called “sin taxes” applied to alcohol and tobacco. While such taxes often generate sizeable revenue, their primary goals are to improve health and social outcomes and internalise the externalities created by consumption of demerit (socially undesirable) goods (Hoffer et al, 2013). Sin taxes are sometimes criticised as being paternalistic, due to their underlying assumption that the government should guide individuals, who are incapable of making decisions in their best interest, towards making the right choices (Jeram, 2016). Individuals often consume excess quantities of demerit goods due to behavioural biases. This can lead to consumption decisions today that are not necessarily in their best interests tomorrow. For example, hyperbolic discounting, a form of time-inconsistent preference, occurs when individuals discount the future heavily and prioritise short-term pleasure over long-term gain. In this context, regularly drinking soft drinks may be pleasurable but have adverse long-term effects on health and longevity. The addictive properties of sugar exacerbate this problem. Incomplete information is another barrier to optimal consumption decisions. Mandatory food labelling provides consumers with information on the sugar and energy content of products. However, awareness of the cumulative effect of excess sugar consumption and the long-term effects of diet on health – for example, the relationship between consumption of SSBs and type two diabetes – is less well-understood. 2.2 Sugar taxes in practice Many countries have imposed excise taxes on imported sugar and soft drinks since the early 20th century, including New Zealand from 1932 to 1971, as well as Ireland, Norway, Finland and Denmark (Ashton and St John, 1985; OECD, 2015). However, original excise taxes were designed primarily for revenue generation rather than to induce a significant shift in consumption. The first health-related taxes specifically designed to discourage consumption of unhealthy foods were introduced by Denmark and Hungary in 2011 (OECD, 2015). Nevertheless, many excise taxes have recently been “re-labelled” as health taxes as the adverse effects of unhealthy diets have become more prevalent (OECD, 2015). SSBs, defined in this paper as carbonated soft drinks, are the most common type of sugary product that is taxed. As of January 2016, nine countries imposed a sugar tax at the national level, all of which included SSBs in their scope. These include five OECD nations: Finland, France, Hungary, Norway and Mexico; as well as four Pacific Island states or territories: Fiji, French Polynesia, Nauru and Samoa (Mytton et al, 2012; McDonald, 2015). Twenty-three US states impose a SSB tax, and both the South African and UK governments have signalled their intention to introduce a tax on SSBs in 2017 WP 16/09 | Implications of a Sugar Tax in New Zealand: Incidence and Effectiveness 3 and 2018 respectively (Mytton et al, 2012; Fihlani, 2016). Other sugary products taxed include chocolate, ice cream and confectionary (OECD, 2015). Denmark abolished their tax on SSBs in 2014 in an effort to increase the competitiveness of Danish firms and recoup jobs lost to neighbouring countries (Scott-Thomas, 2013). 2.3 Rationale for targeting SSBs The high sugar content of SSBs, as well as their low satiety and high addictiveness, are often cited as reasons for targeting SSBs. A typical 375 millilitre can of soft drink contains 40 grammes of sugar, 33% above the UK National Health Service’s total daily recommended sugar intake for an adult (2015). This is accentuated by the fact that high quantities of soft drink can be consumed before the consumer is satiated, meaning that sugar consumption through drinks can occur at a higher rate than through solid foods (Bray et al, 2004). Epidemiological and experimental evidence indicates that higher intake of SSBs is associated with weight gain and obesity (Hu and Malik, 2010). Basu et al (2013) found that a 1% rise in soft drink consumption was associated with a 7.1% rise in the prevalence of overweight or obese adults in the population. Furthermore, the risk of becoming obese increases by 60% for each additional serving of SSB consumed per day (Brownell et al, 2009). The evidence also supports a link between SSB consumption and a range of other adverse health outcomes such as strokes, cancer, impaired cognitive development, cardiovascular disease, type two diabetes, raised blood pressure, dyslipidaemia, gout and dental caries (NZBGP, 2014). 2.4 Indirect costs of obesity Although there are many adverse health outcomes arising from excess SSB consumption, obesity and its associated diseases form the main focus of this paper due to the high rate of obesity in New Zealand and its sizeable cost to the health system. Lal et al (2012) estimated obesity-related illnesses cost $624 million per year, or 4.4% of New Zealand’s healthcare expenditure, with $247 million direct and indirect costs attributed to type two diabetes (Ministry of Health, 2009). New Zealand has the thirdhighest rate of obesity in the OECD; 31% of New Zealand adults are obese and an additional 35% are overweight (Sassi, 2010). Obesity rates have risen 20% in the last 30 years (Ministry of Health, 2015). Large disparities exist between ethnic and socio-economic groups: 66% of Pasifika and 47% of Māori adults are obese, compared to 12% of Asian and 29% of European New Zealanders (Ministry of Health, 2015). Obesity is also positively correlated with socio-economic deprivation (Ministry of Health, 2015). Furthermore, 32% of New Zealand children have an unhealthily high weight, which represents a significant future risk for health expenditure (Carter, 2014). WP 16/09 | Implications of a Sugar Tax in New Zealand: Incidence and Effectiveness 10 3.3 Review of effectiveness of sugar taxes This section provides a summary of the literature on the effectiveness of sugar taxes in achieving the desired improvement in health outcomes. The majority of studies support the conclusion that sugar taxes can have a measurable impact on consumption. However, the extent to which the reduction in consumption results in a reduction in weight and obesity rates is less clear. Higher tax rates (those above 20% of the pre-sugar tax price of the product) were found to induce statistically significant improvements in weight and obesity rates, but lower tax rates have little to no effect. Most studies surveyed advocated for a sugar tax to be included in a broader policy package to be effective. There are some limitations to this literature review. There are few studies available which evaluate the effectiveness of health-related taxes using empirical data. Those that do primarily focus on taxes set at a low rate, which are found to have negligible health impacts (OECD, 2015). Most studies cited below are meta-analyses or use econometric modelling, using price elasticity measures, to estimate the likely impact of price changes on consumption and diet. A key assumption made in these models is that the cost of the tax will be fully passed on to consumers. Table 4 – Summary of literature on effectiveness of sugar taxes Study Comments on effectiveness Alemanno & Carreno (2011) • The responsiveness of consumers to price changes varies depending on the level of consumption, with heavy users and obese consumers being less price sensitive, similar to tobacco and alcohol taxes. • To be effective, sugar taxes need to be introduced as part of a broader policy package. They are unlikely to be effective on their own. Briggs et al (2013) • A 20% tax on SSBs is estimated to reduce the number of obese and overweight adults in the UK by 1.3% and 0.9% respectively. Brownell et al (2009) 4 • Sugar taxes can have a strong positive effect on reducing consumption. An excise tax of 1c/oz would lead to a minimum reduction of 10% in energy consumption from sweetened beverages, o r 20kcal per person per day, which is sufficient for weight loss and reduction in risk. ( However the study does not provide a reference to the data this is based on, nor does it quantify the reduction in weight and obesity.) • Current US sales taxes (imposed in 33 states) are set at too low a rate to affect consumption. Escobar et al (2013) • Sugar taxes may benefit health and lead to a modest weight reduction , but should be used as part of a policy package. Finklestein et al (2010) • This study modelled associ ations between beverage prices, energy intake and weight using multivariate regression models. • The results found that a 20% and 40% tax on soft drinks would reduce energy intake by 4.2 and 7.8kcal/day/person respectively, and this would translate to a mean weight reduction across the population of 0.2 and 0.37kg/year/person. • Extending the tax to other beverages containing sugar, at a rate of 20% and 40%, would result in a reduction of 7.0 and 12.4kcal/day/ person respectively, resulting in a mean weight reduction of 0.32 and 0.59kg/year/person. 4 This study was funded with grants from the Rudd Centre for Food Policy & Obesity, a health lobby group, the National Institutes of Health, and the Robert Wood Johnson Foundation, a health-focused philanthropic organisation. WP 16/09 | Implications of a Sugar Tax in New Zealand: Incidence and Effectiveness 11 Study Comments on effectiveness Lin & Smith (2010) •20% tax on SSBs: average reduction of 34 and 40 calories among adults and children, respectively. Translates to weight reduction of 3.6 and 4.2 pounds for adults and children and would reduce obesity by 3.5% in the United States , from 33.5% to 30.8% for adults and 16.1% to 13.4% for children •A SSB tax would be regressive. Moodie et al (2013) • A high tax rate (>20%) could influence dietary intake and health outcomes but is not definitive. Mytton et al (2012) •Higher taxes are likely to have a greater impact. The effects are greatest for the young, poor and those most at risk of being overweight. •The literature points to the conclusion that a SSB tax would be welltargeted on the young, poor, and most at risk of being overweight. Nnoaham et al (2009) •Positive health effects, but not necessarily greater for low-income groups. OECD (2015) •SSB taxes can be effective in reducing consumption, and if sufficiently high can lead to positive health outcomes. Studies based on low level (18%) ad valorem sales taxes on soft drinks in the United States found there was no statistically significant effect on obesity. Modelling of higher tax rates, however, found a positive impact on obesity rates. Powell et al (2013) •Own-price elasticity of - 1.21 implies price rise of 20% would reduce consumption by 24%, but mixed evidence on the extent of the effect on weight outcomes. •The magnitude and design of a tax considerably influences its effectiveness. Studies that show a s mall impact on weight outcomes used existing tax rates which are low. A study that used beverage price rather than tax measure found a significant association with children’s weight. Sharma et al (2014) •This study modelled the effect of both a 20% flat ra te sales tax and a 20c/L volumetric tax on SSBs. Lower tax burden and higher weight reduction under volumetric tax. •20% flat rate sales tax would result in a weight reduction per capita of 0.29kg for average SSB purchasers and 1.49kg for heavy SSB purchasers. • Substitution towards diet drinks (3.2% increase in consumption) as a result of the tax. Other substitutions are small in magnitude. •High-income households have the least elastic demand for regular soft drinks. •The average SSB tax burden is 0.22% of household income for lowincome households and 0.027% of household income for high-income households. •Progressive health gains: biggest improvements in health accrue to lowerincome households despite higher tax burden. •The study also comments on changing social norms. Tobacco taxes have risen at least 200% over past 30 years. If social stigma attaches to soft drink consumption in the same way it has to smoking, governments have scope to increase taxes on SSBs to a much larger extent, with potentially much mor e effective results on reducing consumption and obesity rates. WP 16/09 | Implications of a Sugar Tax in New Zealand: Incidence and Effectiveness 12 Study Comments on effectiveness Wang (2010) • This study notes the possible regressitivity of an SSB tax, but emphasises the disproportionate burden of negative health outcomes felt by lowincome consumers. • The reduction in consumption is greater among groups with high baseline consumption: the young (20-44 year olds), ethnic minorities and lowincome groups. • A 100% pass-through rate is a valid assumption. One study has shown a 129% pass-through rate. • A 22% increase in SSB price in New York state will result in an 18% decrease in consumption under ideal conditions. Across the state, expected to reduce the proportion of adults consuming one or more drink a day by 24%, and the proportion of adults consuming two or more drinks a day by 2-6%. • Estimated to prevent 3.5% of new diabetes cases in men and 3.0% in women. Health benefits may take time to materialise, but over a decade are expected to prevent in excess of 37,000 cases of type2 diabetes across the state. Preventative benefi t is greater for young adults due to their higher baseline consumption. Some studies have claimed that a sugar tax would improve health outcomes to a greater extent among lower socio-economic groups (Ni Mhurchu et al, 2015; Mytton et al, 2012), which they have referred to as a “progressive health impact”. This is due to groups with higher obesity rates, such as Māori and lower socio-economic groups, having more price elastic demand, resulting in a larger reduction in consumption from a given price change (Ni Mhurchu et al, 2013). The implication is that those groups most affected by obesity will experience a greater improvement in diet and health. Whether this is true depends on the degree of the substitution effect, as outlined in Section 3.2. 4 Incidence of a sugar tax 4.1 Method of analysis To consider the incidence of a sugar or SSB tax on New Zealand households with different demographic characteristics, a dataset was constructed using household expenditure data from the Household Economic Survey (HES), a nationally representative survey of approximately 3,000 New Zealand households, for the years 2007, 2010 and 2013. Expenditure in the HES is classified to a fine level of detail, enabling an analysis of the implications of taxing different food groups. Data from these three surveys were pooled and the expenditure converted to 2013 dollars using the Consumer Price Index (CPI). Consumers were divided into twenty groups based on their total annual household expenditure, from $0 to $150,000 in increments of $10,000, and from $150,000 to $400,000 in increments of $50,000. Two scenarios were evaluated: a tax on SSBs only; and a more comprehensive sugar tax encompassing SSBs and other sugary foods. To do this, the proportion of each household’s total expenditure on three commodity groups was calculated: category one being all soft drinks (SSBs); category two containing other clearly defined food groups which are high in sugar (chocolate, ice cream, cakes and biscuits, confectionary, desserts as well as sugar itself and variants thereof); and category three containing all other goods. Goods such as WP 16/09 | Implications of a Sugar Tax in New Zealand: Incidence and Effectiveness 13 tomato sauce and jam, while high in sugar, were not included in category two due to the lack of international precedent on including these products in a sugar tax. A weighted average of budget shares for these three categories for each total expenditure group was calculated. The household type and ethnicity variables were incorporated into the model and output tables produced for: 1) All households 2) Households with two or more adults, without children 3) Households with two or more adults, with children 4) Single adult households 5) Households where the head of the household5 is Asian 6) Households where the head of the household is European 7) Households where the head of the household is Māori 8) Households where the head of the household is Pasifika. 4.2 Budget shares In considering the progressivity of indirect taxes, it is appropriate to consider the tax base to be total household expenditure, rather than income. A tax on SSBs is therefore progressive if the average tax rate increases as the total household expenditure increases. A fixed ad valorem tax rate imposed on SSBs implies that the average tax rate varies in exactly the same way that household budget shares attributed to SSBs vary. Hence the expected progressivity or regressivity of an ad valorem SSB tax can be examined by considering how the average budget share, at each household total expenditure level, varies. The present section reports evidence on the variation in budget shares for a range of population groups. The empirical analysis confirms the expected result that for most population groups there is a negative correlation between total household expenditure and expenditure on the targeted products. The budget shares for various household types and both categories of sugary product (SSBs only and other sugary products) are illustrated in figures 2 to 9. Adding the budget shares for both categories gives an indication of the incidence of a broader sugar tax, as opposed to an SSB tax. 4.2.1 All households For the entire sample, a tax on SSBs shows mildly regressive effects, and a tax on all sugary products is moderately regressive, as shown in Figure 2. 5 Defined here as the household member with the highest income. WP 16/09 | Implications of a Sugar Tax in New Zealand: Incidence and Effectiveness 14 Figure 2 – Weighted average expenditure proportion for all households 4.2.2 By household type Budget shares for households with two or more adults without children correspond closely to the total population, except for low-income households in this category for which the proportion is moderately higher than all households. This is shown in Figure 3. Average SSB consumption among lower-income households with children is higher than among lower-income households without children, but the presence of children in the household has negligible effect on expenditure proportions among higher-income groups. Figure 3 – Weighted average expenditure proportion for two or more adults without children WP 16/09 | Implications of a Sugar Tax in New Zealand: Incidence and Effectiveness 15 Figure 4 – Weighted average expenditure proportion for two or more adults with children Single adult households spend a lower proportion of total expenditure on sugary products than average, as shown in Figure 5. Figure 5 – Weighted average expenditure proportion for single adult households 4.2.3 By ethnicity Both consumption rates of sugary products and obesity rates vary by ethnicity as reported in section 3.1.2. Figures 6 to 9 illustrate the budget shares for Asian, European, Māori, and Pasifika households respectively. Māori and Pasifika consume SSBs at higher rates than Europeans and Asians, and on average have lower incomes (Ministry of Health, 2010; Statistics New Zealand, 2014). Māori and Pasifika have higher rates of obesity than Europeans and Asians, and therefore we would expect these ethnic groups to spend a greater proportion of their income on sugary products. This is confirmed by HES data (Ministry of Health, 2010). Budget shares for sugary products and SSBs are lower among Asians than the general population and this proportion does not vary significantly with total expenditure. WP 16/09 | Implications of a Sugar Tax in New Zealand: Incidence and Effectiveness 16 Figure 6 – Weighted average expenditure proportion for Asian households The consumption pattern for European New Zealanders closely matches that of the total population, which is expected given that 74% of New Zealand’s population falls into this category (Statistics New Zealand, 2014). Figure 7 – Weighted average expenditure proportion for European households Budget shares for Māori are the opposite of the result obtained for the total population. Higher-income Māori spend a greater proportion of their total expenditure on SSBs (albeit a slightly lower proportion on other sugary products), than lower-income Māori. Therefore, an SSB tax is likely to have a progressive incidence on Māori. However, the small sample size of 873 households reduces the validity of this conclusion. WP 16/09 | Implications of a Sugar Tax in New Zealand: Incidence and Effectiveness 17 Figure 8 – Weighted average expenditure proportion for Māori households Budget shares for Pasifika are higher than the total population for both categories of product. The high variability reflects the small sample size of Pasifika households in the HES. Figure 9 – Weighted average expenditure proportion for Pasifika households WP 16/09 | Implications of a Sugar Tax in New Zealand: Incidence and Effectiveness 18 5 Conclusions Analysis of New Zealand expenditure data indicates that a tax on SSBs or sugary products would have a regressive impact on the general population, which is consistent with international evidence. However, the effect varies by household type. Low-income households with children would be particularly affected. The budget shares for Māori and Pasifika contradict the general population as the proportion of expenditure on SSBs is generally an increasing function of total expenditure. Given these results, questions remain over whether the tax would appropriately target those most at risk from becoming obese, such as Māori, Pasifika and low socio-economic groups. Three studies surveyed found the largest reduction in obesity to occur among these groups. However, a tax is a blunt instrument that affects all consumers of SSBs. Changing the behaviour of high consumers of SSBs, who are more likely to be obese or at risk of becoming obese, is necessary for greater improvements in health outcomes. The review of the literature on the effectiveness of sugar taxes is inconclusive. Of the 13 studies surveyed, nine provided evidence to show that a sugar tax could be effective in reducing obesity rates, when set at a sufficiently high rate, and when introduced as part of a broader policy package to tackle obesity. However, there was wide variation in the methodologies used to conduct these studies, with most based on econometric modelling of demand reductions rather than empirical observations. A potential problem of the tax highlighted in this paper is substitution to unhealthy products. Limited evidence on the degree of substitution was found in the literature, although several studies found that the high-risk groups identified have a higher price elasticity of demand for SSBs than the general population. Highly price sensitive consumers are more likely to switch to non-taxed substitute products, or cheaper but equally unhealthy taxed products, if a tax is implemented. This problem is worth exploring further before a tax is introduced. WP 16/09 | Implications of a Sugar Tax in New Zealand: Incidence and Effectiveness 19 References Alemanno, A., & Carreño, I. (2011). Fat taxes in the EU between fiscal austerity and the fight against obesity. European Journal of Risk Regulation, 4(11). Andreyeva, T., Long, M. W., & Brownell, K. D. (2010). The Impact of Food Prices on Consumption: A Systematic Review of Research on the Price Elasticity of Demand for Food. American Journal of Public Health, 100 (2), 216-222. Ashton, T., & St. John, S. (1985). Insights into Excises, A Focus on Alcohol, Tobacco and Motor Fuels Taxation. Wellington: Victoria University Press for the Institute of Policy Studies. BBC. (2016, March 16). Sugar tax surprise in Budget - but growth forecasts cut. http://www.bbc.com/news/uk-politics-35813973 Bleich, S. N., Wang, Y. C., Wang, Y., & Gortmaker, S. L. (2009). Increasing consumption of sugar-sweetened beverages among US adults: 1988-1994 to 1999-2004. American Journal of Clinical Nutrition, 89 (1), 372-381. Bray, G. A., Nielsen, S. J., & Popkin, B. M. (2004). Consumption of high-fructose corn syrup in beverages may play a role in the epidemic of obesity. American Journal of Clinical Nutrition, 79(4), 537-543. Briggs, A. D. M., Mytton, O.T., Kehlbacher, A., Tiffin, R., Rayner, M., Scarborough, P. (2013). Overall and income specific effect on prevalence of overweight and obesity of 20% sugar sweetened drink tax in UK: econometric and comparative risk assessment modelling study. British Medical Journal, 2013;347:f6189, doi: 10.1136/mbj.f6189 Brownell, K., Farley, T., al, W. C., Popkin, B. M., Chaloupka, F. J., Thompson, J. W., et al. (2009). The Public Health and Economic Benefits of Taxing Sugar-Sweetened Beverages. New England Journal of Medicine, 361(16), 1599-1605. Duffey, K. J. & Popkin, B. M. (2007). Shifts in patterns and consumption of beverages between 1965 and 2002. Obesity, 15(11), 2739-2747. Escobar, M. A., Veerman, J. L., Tollman, S. M., Bertram, M. Y., & Hofman, K. J. (2013). Evidence that a tax on sugar sweetened beverages reduces the obesity rate: a metaanalysis. BMC Public Health, 13(1072), doi: 10.1186/1471-2458-13-1072 Fihlani, P. (2016, March 5). Why South Africans will pay ‘sin tax’ for fizzy drinks. BBC News. Retrieved from http://www.bbc.com/news/world-africa-35670275 Finkelstein, E. A., Zhen, C., Nonnemaker, J., & Todd, J. E. (2010). Impact of Targeted Beverage Taxes on Higherand Lower-Income Households. Archives of Internal Medicine, 170 (22), 2028-2034. Hoffer, A. J., Shughart, W.F., & Thomas, M.D. (2013). Sin taxes: size, growth, and creation of the sindustry. George Mason University: Fairfax, Virginia, United States.