i TECHNICAL SUPPORT FOR UNIVERSAL HEALTH COVERAGE IN ARMENIA THE IMPACT OF HEALTH TAXES IN ARMENIA Akshar Saxena Adanna Chukwuma Seemi Qaiser Armineh Manookian Gevorg Minasyan © 2023 The World Bank Group, 1818 H Street NW, Washington, DC 20433. This report was prepared by World Bank staff with external contributions. The findings, interpretations, and conclusions expressed in this work do not necessarily reflect the views of The World Bank, its Board of Executive Directors, or the governments they represent. The World Bank originally published this report in English (The Impact of Health Taxes in Armenia) in 2022. Therefore, where there are discrepancies, the English version will prevail. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning any territory's legal status or the endorsement or acceptance of such boundaries. The material in this report is subject to copyright. The World Bank encourages the dissemination of its knowledge. Hence, this work may be reproduced, in whole or in part, for noncommercial purposes if full attribution to this work is given. Any queries on rights and licenses, including subsidiary rights, should be addressed to World Bank Publications, The World Bank Group, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522-2625; e-mail: pubrights@worldbank.org. Design by Veronica Elena Gadea, GCSDE, The World Bank Group. THE IMPACT OF HEALTH TAXES IN ARMENIA Akshar Saxena Adanna Chukwuma Seemi Qaiser Armineh Manookian Gevorg Minasyan iv THE IMPACT OF HEALTH TAXES IN ARMENIA ABOUT THIS REPORT This report, The Impact of Health Taxes in Armenia, is part of the World Bank’s technical support for universal health coverage in Armenia. This support includes advisory services and analytics to facilitate the government’s efforts to expand access to high-quality health care. The report explores the economic and health returns to increasing excise taxes on sugar-sweetened beverages, alcohol, and tobacco in Armenia. Gavi, The Vaccine Alliance, supported this technical assistance. Modeling the Actuarial costing Projecting impact of tax of a unified revenues from options on growth, benefits package alternative tax poverty, financial that meets and non-tax protection, health population health sources and employment care needs Informing policies to Support for increase public Modeling strategic plan for financing for allocations of primary health health care public financing care financing, in the benefits organization, and package to regulation maximize health Support for Facilitating Reforms to Assessment of strategic plan the alignment Technical support public financial towards Universal align public for continuity of service financing for management in of care across delivery with Health Coverage the health in Armenia health with providers better health value sector Support for Assessment of regulating, strategic monitoring and purchasing in paying providers Knowledge exchanges on the health for better sector quality investing in Universal Health Coverage Convening Harvard-World policy and Study tours to Bank Global technical selected Flagship discussions on countries Course on reform options Health Reform v TABLE OF CONTENTS About this Report . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iv Table of Contents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . v Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vii About the Authors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . viii Acronyms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix Executive Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 Chapter 1. Why Mobilize Domestic Resources for Health? . . . . . . . . . . . . . . . . . . . . 3 The Economic and Political Context . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 Population Health Outcomes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 Ensuring Access to High-Quality Health Care . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 A Role for Taxation to Finance Health Care . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 Purpose of this Report . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 Chapter 2. The current state of health and consumption taxes in Armenia . . . . . 11 Alcoholic Beverages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 Sugar-Sweetened Beverages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 Tobacco Smoking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 Chapter 3. Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 Change in Consumption Due to Tax . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 Change in Tax Revenue Due to Change in Consumption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 Change in Health Due to Change in Consumption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 Change in Health Care Expenditure Due to Change in Health . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 Statistical and Modeling Software . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 Chapter 4. Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 Change in Health Due to Change in Consumption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 Change in Health Care Expenditure Due to Change in Health . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 Change in Tax Revenue Due to Change in Consumption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 Chapter 5. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 vi THE IMPACT OF HEALTH TAXES IN ARMENIA Appendix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 Appendix 1A: Consumption of SSB by Income-Quintile in Armenia in 2016 . . . . . . . . . . . . . . . . . . . . 31 Appendix 1B: Excise Taxes for 2020, 2021, 2022 and 2023 (effective 1 January 2022) . . . . . . . . . . 31 Appendix 1C: Beer and Spirit Excise Tax in Armenia and Eu Countries, 2021 . . . . . . . . . . . . . . . . . . 32 Appendix 1D: Estimating the Price-Elasticity of Demand Using Armenian Data . . . . . . . . . . . . . . . 33 Appendix 1E: Price Elasticity of Demand for Alcoholic Beverages in Armenia in 2018 . . . . . . . . . . 34 Appendix 1F: Price Elasticity of Demand for Sugar-Sweetened Beverages in Armenia in 2018 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 Appendix 1G: Price Elasticity of Demand for Smoking in Armenia in 2018 . . . . . . . . . . . . . . . . . . . . .34 Appendix 1H: Price-Elasticity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35 Appendix 1I: Epidemiological Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35 Appendix 1J: Potential Impact Fraction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36 Appendix 1K: Average Price Paid and Number of Visits in Armenia in 2018 . . . . . . . . . . . . . . . . . . . 37 Appendix 1L: Health Insurance Coverage by Wealth Quintile in Armenia in 2015 . . . . . . . . . . . . . . 37 Appendix 1M: Distribution of Payment Systems in Armenia in 2018 . . . . . . . . . . . . . . . . . . . . . . . . . . 37 Appendix 1N: Savings for Government’s Outlay on Healthcare Costs . . . . . . . . . . . . . . . . . . . . . . . . 38 Appendix 1O: Savings for Out-of-Pocket Payments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38 Appendix 1P: Distribution of Estimated Tax (in AMD million) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 Appendix 1Q: Distribution of Estimated Tax from Alcoholic Beverages (in AMD million) . . . . . . . 39 Appendix 1R: Distribution of Premature Mortality, Health-Care Expenditure and Government Savings Due to Varying Pass-Through for Sugar-Sweetened Beverages . . . . . . . . . 40 Endnotes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41 Acknowledgments vii ACKNOWLEDGMENTS This report was supervised by Sylvie Bossoutrot (past Country Manager, Armenia), Carolin Geginat (Country Manager, Armenia), and Tania Dmytraczenko (Practice Manager, Health, Nutrition, and Population Global Practice, Europe, and Central Asia Region). The analysis benefited from the close engagement of the Ministry of Health National Institute of Health and the Statistical Committee of the Republic of Armenia. The authors are grateful to the reviewers for insightful feedback on the initial drafts of the report. These include Dhiraj Sharma (Senior Economist), Yoshini Naomi Rupasinghe (Health Specialist), Ceren Ozer (Senior Economist), Danielle Bloom (Senior Health Financing Specialist), and Zara Mkrtchyan (Consultant). The excellent editorial support from Richard A. B. Crabbe and the significant operational assistance from Arpine Azaryan are appreciated. All errors and omissions are the authors’ responsibility. viii THE IMPACT OF HEALTH TAXES IN ARMENIA ABOUT THE AUTHORS Akshar Saxena is an Assistant Professor of Economics at the Nanyang Technological University, where he researches health economics and public economics. Akshar has consulted for the Ministry of Health in Singapore, the World Bank, and the United States Agency for International Development. Akshar holds a Master in Public Policy from the Lee Kuan Yew School of Public Policy and a Doctor of Science degree from Harvard University. Adanna Chukwuma is a Senior Health Economist in the Health, Nutrition, and Population Global Practice, where she leads investment operations’ design, implementation, and evaluation. She has over ten years of experience advising national reforms to improve access to high-quality health care through service delivery organization, strategic purchasing, revenue mobilization, and demand generation, including in Sri Lanka, Sierra Leone, India, Moldova, Tajikistan, the South Caucasus Countries, and Romania. Adanna has published on health care financing, access, and quality in peer-reviewed journals and is an Associate Editor of the Health Systems and Reform journal. She obtained a medical degree from the University of Nigeria, a Master of Science in Global Health from the University of Oxford, and a Doctor of Science in Health Systems from Harvard University. Seemi Qaiser is a health systems professional focusing on leveraging technology to improve health outcomes. She holds a Master of Science from the Harvard T.H. Chan School of Public Health and a Bachelor of Science from the University of Toronto. Seemi has published on global health issues globally with a focus on East Africa and Europe and Central Asia. Armineh Manookian is the World Bank Country Economist for Armenia in the Macroeconomics, Trade, and Investment Global Practice, covering macroeconomic and fiscal issues, economic reporting, and macroeconomic projections. In addition, she is engaged in macroeconomic policy dialogue with the client. Armineh joined the Bank in 2017 and worked for more than ten years in the International Monetary Fund’s Resident Representative Office in Armenia as a Macroeconomist. Before moving to Armenia in 2005, Armineh worked with the Central Bank of Iran as a Senior Economist in the Research and Policy Department. She holds a Master of Public Administration in Economic Policy Management from Columbia University. Gevorg Minasyan is the Head of the Special Studies Division at the Central Bank of  Armenia Economic Research Department. He coordinates the Central Bank's development, growth, and public policy research projects. Previously, he served as Israel's International Monetary Fund mission member, participating in macro-fiscal forecasting model development and capacity building. He has also worked at the World Bank as a Consultant, providing technical assistance to the Universal Health Coverage agenda. Gevorg completed the Data, Economics, and Development Policy Micro-Master's degree from the Massachusetts Institute of Technology. In addition, he holds a Doctor of Philosophy from Yerevan State University. Acronyms ix ACRONYMS ABV Alcohol by volume AIDS Almost Ideal Demand System AMD Armenian Dram BBP Basic Benefits Package BMI Body Mass Index DM Diabetes mellitus EU European Union FCTC Framework Convention on Tobacco Control GDP Gross Domestic Product HFCS High-fructose corn syrup IHD Ischemic Heart Disease ILCS Armenia Integrated Living Conditions Survey LMIC Low-and-Middle-Income Country MoF Ministry of Finance MoH Ministry of Health NCD Noncommunicable disease OOP Out-of-pocket PIF Population Impact Fraction RR Relative risk SSB Sugar-sweetened beverage TMREL Theoretical minimum risk exposure level UAH Ukrainian hryvnia UHC Universal Health Coverage UMI Upper middle-income USD United States Dollars YLL Years of life lost WHO World Health Organization VAT Value-Added Tax 1 EXECUTIVE SUMMARY Health has improved significantly in Armenia over the past two decades. Individuals live longer, and fewer lives are lost to poor maternal care and infectious diseases. However, the burden of non-communicable diseases (NCDs) has increased. Preventing and managing NCDs requires access to high-quality care. Yet, the use of essential services remains low due to financial barriers arising from high out-of-pocket (OOP) payments. For example, the cost of care is the reported barrier to healthcare among 49 percent of the extremely poor that forgo necessary services. Consumption taxes can support increases in public financing for health care, thereby reducing the strain on Armenian households. Furthermore, excise taxes on alcohol, sugar-sweetened beverages (SSBs), and tobacco discourage the consumption of these products, differentially across income groups, with improvements in health and productivity. Hence, at the request of the Ministry of Health, and to inform policy in Armenia, this report estimates the impacts of excise taxes in terms of the associated decrease in mortality and morbidity due to chronic diseases caused by alcohol, SSBs, and tobacco, across income quintiles. The analysis also outlines the fiscal space generated by taxing these products. The results reveal that taxing tobacco averts the most deaths: 33,139 over 20 years, predominantly among the poorest quintile (25 percent) – see ES Table 1 below. Alcohol taxes can prevent 402 deaths, 50 percent among the richest. In contrast, SSBs’ tax can avert 395 deaths, with the poorest and the richest impacted almost evenly. The total projected revenue across all products is Armenian Dram (AMD) 136 billion, with the most significant contribution from tobacco tax (81 percent), followed by SSBs (18 Executive Summary 2 percent), and then alcohol (less than one percent). Taxation of alcoholic products is primarily borne by the wealthiest quintiles, while the poorest quintiles shoulder the tax on SSBs. The burden of health care costs for tobacco and SSBs falls on the most affluent quintile, while the poorest bear the costs of health care for alcohol. ES TABLE 1. Changes in health, tax revenue, and healthcare costs PREMATURE DEATHS TAX REVENUE HEALTH CARE COST AVERTED (AMD MILLIONS) (AMD MILLIONS) Tobacco Alcohol SSBs Tobacco Alcohol SSBs Tobacco Alcohol SSBs Poorest 8,151 50 82 21,445 176 6,705 3,972 10 3,972 Richest 5,903 101 73 24,145 279 4,286 2,782 47 2,782 Therefore, excise taxes are a potentially viable option to increase revenue and expand fiscal space for universal health coverage (UHC) in Armenia, while improving health outcomes. However, beyond the impact demonstrated, an increase in excise taxes will reflect fiscal policy and broader political economy factors. 3 CHAPTER 1. WHY MOBILIZE DOMESTIC RESOURCES FOR HEALTH? THE ECONOMIC AND POLITICAL CONTEXT Armenia is an upper middle-income (UMI) country in the South Caucasus Region with a population of 2.99 million.1 The country consists of 11 administrative units: 10 provinces and the capital city, Yerevan. One-third of the population lives in Yerevan, 28 percent in towns, and 36 percent in rural areas. The country has undergone significant political changes over the past 20 years, including a transition from the Soviet Union. In 2018, Armenia’s economy transitioned from lower-middle-income to UMI2, as the Gross Domestic Product (GDP) per capita had tripled from United States Dollars (USD) 1,404 in 2000 to USD 4,266 in 2018 (Table 1.1). Poverty rates have also fallen over the past 20 years, with the proportion of the population living below the UMI poverty line (USD 5.50) decreasing from 81.0 to 42.5 percent between 2001 and 2018. At the same time, total fertility rates have declined to below replacement levels, and the population aged 65 and older made up 11.3 percent of the total population in 2018. Chapter 1. Why Mobilize Domestic Resources for Health? 4 TABLE 1.1: Selected social and economic indicators POVERTY POPULATION GENERAL TOTAL PROJECTED GDP PER HEADCOUNT RATIO AGES 65 AND GOVERNMENT FERTILITY ECONOMIC CAPITA IN AT $5.50 A DAY OLDER FINAL COUNTRY RATES GROWTH IN 2020 (2011 PPP) (% OF CONSUMPTION (BIRTHS PER 2022 (USD) (% OF THE THE TOTAL EXPENDITURE WOMAN) (% OF GDP) POPULATION) POPULATION) (% OF GDP) Armenia 4,266.0 44.0 1.8 11.8 16.0 4.5 Belarus 6,424.2 0.2 1.4 15.6 16.9 0.5 Croatia 14,134.2 2.4 1.5 21.3 24.0 5.8 Estonia 23,027.0 0.8 1.7 20.4 21.3 4.2 Georgia 4,266.7 42.0 2.1 15.3 14.7 5.8 Hungary 15,980.7 2.0 1.5 20.2 21.1 5.1 Kazakhstan 9,122.2 4.6 2.9 7.9 12.7 3.9 Kyrgyz 1,173.6 52.6 3.3 4.7 17.6 5.6 Republic Russia 10,126.7 3.7 1.5 15.5 20.7 3.0 Tajikistan 859.1 Not available 3.6 3.2 11.3 4.5 Turkey 8,536.4 10.2 2.1 9.0 15.2 3.3 Turkmenistan 7,612.0 Not available 2.7 4.8 7.8 1.7 Ukraine 3,726.9 2.5 1.2 16.9 19.3 3.6 Uzbekistan 1,750.7 Not available 2.8 4.8 16.9 5.4 Source: World Bank, International Monetary Fund. Armenia’s economy is recovering from the COVID-19 pandemic. The economy contracted by eight percent in 2020, with the construction and service industries being the most affected.3 The population below the UMI poverty line is estimated to rise by 12.8 percentage points. To address the negative social and economic impacts, the government increased its spending by 19 percent, approximately 2.3 percent of GDP, in the first seven months of 2020. Unfortunately, revenue decreased by six percent year on year.4 In 2022, the economy has benefitted from money transfers and visitors from Russia, and growth is expected to reach seven percent, prior to slowing in 2023. POPULATION HEALTH OUTCOMES Overall, population health outcomes in Armenia have improved over the past decade due to better maternal and child health and fewer infectious diseases. For instance, since 1990, infant deaths have declined from 41.7 to 11 per 1,000 live births in 2018.5 Between 2005 and 2018, tuberculosis also dropped from 92 to 31 new cases per 100,000.6 As a further reflection of the overall health improvements, life expectancy at birth increased from 68 to 75 years in three decades.7 5 THE IMPACT OF HEALTH TAXES IN ARMENIA However, when average life expectancy is adjusted for years lived in ill-health, it falls to 66.3 years. NCDs, including heart disease (16.9 percent), diabetes mellitus (5.7 percent), and stroke (5.6 percent), are the leading causes of years of life lost (YLLs) and are expensive to treat (Figure 1.1 and Table 1.2).8,9,10 A high NCD burden reduces productivity and increases health care spending.1 NCDs cost the Armenian economy an estimated AMD 362.5 billion in 2017, amounting to 6.5 percent of the country's annual national income.1 NCD risk increases with aging, tobacco, alcohol, and sugar exposure.11–15 Some of these risk factors are prevalent in Armenia. For instance, 51.5 percent of Armenian men, ages 18 to 69, are active smokers.16 FIGURE 1.1: Change in the burden of disease in Armenia 1990-2019 Both sexes, All ages, Disability-adjusted life years per 100,000 1990 RANK 2019 RANK 1 Cardiovascular diseases 1 Cardiovascular diseases 2 Neoplasms 2 Neoplasms 3 Unintentional injuries 3 Other non-communicable 4 Maternal & neonatal 4 Diabetes & Chronic kidney disease 5 Respiratory infections & Tuberculosis 5 Musculoskeletal disorders 6 Other non-communicable 6 Mental disorders 7 Mental disorders 7 Digestive diseases 8 Musculoskeletal disorders 8 Unintentional injuries 12 Digestive diseases 12 Respiratory infections & Tuberculosis 14 Diabetes & Chronic kidney disease 13 Maternal & neonatal same in 2019 increase in 2019 decrease in 2019 Source: Institute for Health Metrics and Evaluation (IHME), 2021. TABLE 1.2: Prevalence, incidence, and mortality of relevant conditions in Armenia in 2018 DISEASE NAME INCIDENCE PREVALENCE MORTALITY Ischemic Heart Diseases 15,454 78,854 6,334 Cerebrovascular Diseases 5,543 18,726 1,848 Non-Insulin-Dependent 8,114 77,642 579 Diabetes Mellitus (Type II) Lung Cancer 899 2,320 1,242 Liver Cancer 54 N/A 246 Source: Ministry of Health and National Institute of Health (NIH) 2018. Chapter 1. Why Mobilize Domestic Resources for Health? 6 ENSURING ACCESS TO HIGH-QUALITY HEALTH CARE Access to high-quality health care is critical for treating NCDs.17 Adequate care can promote healthy lifestyles to delay NCDs onset, facilitate their early diagnosis, and prevent complications.1 However, Armenians had an average of four outpatient visits per person in 2015, much lower than the average of 7.1 reported in Europe.18 This statistic includes the underutilization of needed care. When sick, only one in three Armenians visit a health care facility.19 Despite this, the Ministry of Health (MoH) attempts to offer NCD screening programs. However, the use of essential health care remains low.1 Financial barriers pose a significant challenge to accessing health care. Armenian households shoulder 85 percent of total health spending. In one in five cases, the cost is the main reason Armenians do not seek essential services.1,19 Armenian households are often forced to undertake catastrophic expenditures. Hence, about 16 percent of households in 2013 spent more than 10 percent of their household consumption on OOP health care expenditure.20 As presented in Figure 1.2, OOP health spending rose from 58.2 percent in 2007 to 84.3 percent in 2018, far greater than the UMI average of 32.9 percent.21 FIGURE 1.2: Armenia has one of the highest levels of OOP health spending globally 90 OOP as % of current health expenditure Armenia 80 Turkmenistan Azerbaijan 70 Tajikistan 60 Uzbekistan Kyrgyzstan 50 Ukraine Georgia 40 Kazakhstan 30 20 Sweden 10 Germany R2 = 0.7618 France 0 0 2 4 6 8 10 Domestic general government health expenditure as % of GDP Source: WHO Global Health Expenditure Database, 2018. 7 THE IMPACT OF HEALTH TAXES IN ARMENIA One reason for the disproportionate health financing pressure on Armenian households is the low public spending on health in Armenia. At USD 53 per capita, it is far below the average in UMI countries of USD 268.22 Per capita spending on health increased from USD 13 in 2000 to USD 63 in 2016 but declined to USD 53 in 2018 due to reductions in total government spending and health’s low priority in the national budget. As a result, the scope and depth of state-funded coverage for care are relatively narrow, with the resulting financial barriers discussed above. Previously, donors supported funding for immunization, tuberculosis, and other programs. However, financial support from donors is also declining as Armenia transitions to UMI status, shifting the fiscal pressure to the state budget. While the government mobilized AMD 36.37 billion from the state budget and private donors to respond to the COVID-19 pandemic in 2020, it remains uncertain whether this increase in health care funding will be sustained.23 The need for increased public health spending remains vital to ensure financial protection for essential services and to cover services previously supported by external funders. In 2015, Armenia pledged its commitment to achieving UHC to ensure everyone can access quality health services without financial hardship.24 The MoH developed a concept note for the Introduction of Universal Health Coverage, which includes a proposal for mobilizing revenues to cover access to the Basic Benefits Package (BBP) for all Armenians. A recent estimate of the annual cost of providing essential health services to all population groups is over AMD 301 billion.25 A ROLE FOR TAXATION TO FINANCE HEALTH CARE Taxation is an option to mobilize the additional revenue needed to finance quality health care. In 2019, Armenia raised tax revenue equal to 21 percent of GDP, higher than the average of 17 percent in comparator countries.26 Increased payroll taxes is one of the options being debated as a source of revenue for health reform. However, several factors limit the potential for this option, including an aging population, low employment rates, and a relatively large informal sector. Armenia might be able to follow the example of low-and-middle-income countries (LMICs), including in the region, by expanding broad-based consumption taxes (Figure 1.3 and Box 1). In particular, value-added tax (VAT) and excise taxes can raise additional revenue.27 Chapter 1. Why Mobilize Domestic Resources for Health? 8 FIGURE 1.3: How have LMICs raised tax revenue? TAX LEVELS AND COMPOSITION Trade The composition of taxes in richer countries differs Corporate income tax from that of poorer countries, with greater emphasis Personal income tax on broad-based consumption and excise taxes. Excise tax Consumption 30 25 Revenue (% of GDP) 20 15 10 5 0 1990-99 2010-16 1990-99 2010-16 1990-99 2010-16 1990-99 2010-16 High-income Upper Lower Low-income Middle-Income Middle-Income Source: International Monetary Fund. In discussion with the MoH, an analysis was undertaken by the World Bank on the revenue generation potential of alternative sources. This analysis estimated the additional fiscal space that could be created through increasing excise taxes on alcohol, tobacco, and SSBs (Table 1.4). These excise taxes can also reduce the NCD burden through reduced consumption and increase the workforce’s productivity.28 In Armenia, such taxes are administratively feasible to collect and do not negatively affect formal employment compared to payroll taxes.29 TABLE 1.4: Projected revenue from excise taxes on alcohol, tobacco, and SSBs PROJECTED REVENUE FROM EXCISE TAX (MILLIONS) PRODUCT DESCRIPTION 2020 2021 2022 2023 2024 Alcohol Beer 19,300 19,900 20,600 21,300 21,400 Grape wines 282 269 256 244 226 Vermouth and other 30 27.9 26.0 24.2 22.0 grape wines Other fermented 30.2 32.5 35.1 37.8 39.7 beverages Ethyl alcohol 10,700 14,500 18,400 20,900 23,600 Alcoholic beverages 40,000 47,900 56,800 66,500 60,600 Brandy 648 572 504 443 387 9 THE IMPACT OF HEALTH TAXES IN ARMENIA TABLE 1.4: continued PROJECTED REVENUE FROM EXCISE TAX (MILLIONS) PRODUCT DESCRIPTION 2020 2021 2022 2023 2024 Whiskey, rum, other 460 471 482 493 495 alcoholic beverages Vodka 60,100 67,200 75,100 83,800 91,400 Tobacco Industrial tobacco 77,500 11,800 18,000 27,500 40,800 substitutes SSBs Lemonade 0.0956 0.0951 0.0945 0.0939 0.0911 Natural fruit juices 0.0831 0.0951 0.0109 0.124 0.139 Other non-alcoholic drinks 0.0497 0.0527 0.0558 0.059 0.0619 (colas, Coca-Cola, Pepsi, etc.) Source: World Bank 2021. BOX 1: Regional experience with consumption taxes Several European and Central Asian countries have experimented with introducing health taxes to improve public health and generate tax revenue. Particularly in the case of tobacco, while revenue from excise taxes ranges from 0–1.4 percent of GDP, there is the potential for them to make up a significant proportion of government revenue.30 Below, we reflect on the examples of Ukraine, Lithuania, and Latvia. Tobacco use in Ukraine is a crucial public health concern. In 2017, 130,000 Ukrainians died from diseases attributable to tobacco. Ukraine has ratified the World Health Organization’s (WHO) Framework Convention on Tobacco Control, adopting several tobacco tax increases starting in 2008.31 In 2017, the country approved a plan to increase tobacco excise tax by 30 percent in 2018 and an additional 20 percent every year until 2024. Between 2008 and 2017, cigarette sales decreased by 46 percent, and the daily smoking prevalence decreased by 35 percent. The state budget increased by 36.5 billion Ukrainian hryvnias (UAH), equivalent to USD 50 million.32 Lithuania has one of the highest levels of alcohol consumption in Europe. In 2017, they increased the alcohol excise tax by up to 112 percent on wine and beer and 23 percent on ethyl alcohol.33 Per capita consumption among adults decreased by 6.8 percent, and the mortality rate decreased by 5.4 deaths per 100,000. The resulting tax revenue made up 0.4 percent of GDP.34 Latvia is applying an increased excise duty tax to sweetened non-alcoholic beverages.35 Below eight grams of sugar per 100 mL, the duty rate is 7.40 Euros, but above eight grams, the rate is 14 Euros per 100 liters. A report on the impact of this change will become available in 2024. Chapter 1. Why Mobilize Domestic Resources for Health? 10 PURPOSE OF THIS REPORT This report has been prepared by the World Bank, at the request of the MoH, to support ongoing efforts to improve population health and revenue mobilization in the sector. The study estimates the health impacts of increasing taxation on SSBs, alcohol, and tobacco across gender and income-quintiles. The revenue potential of these taxes is also explored. The target audience for these findings includes senior policymakers and technical advisers in the MoH, Ministry of Economy, and Ministry of Finance (MoF). The remainder of this report is organized as follows. In Chapter 2, the authors review the current state of health and consumption taxes in Armenia. Chapter 3 outlines the methods used to estimate the change in tax revenue and consumption of alcohol, tobacco, and SSBs. Chapter 4 reports the analysis results, including the potential additional fiscal space and health gains. Finally, chapter 5 presents the conclusions based on the findings. 11 CHAPTER 2. THE CURRENT STATE OF HEALTH AND CONSUMPTION TAXES IN ARMENIA This chapter discusses the health risks associated with alcohol, tobacco, and SSBs. The socioeconomic consumption patterns in Armenia for these products are also described, followed by the tax structure and other legal restrictions. ALCOHOLIC BEVERAGES There is relatively strong global evidence linking alcohol consumption to NCDs (Box 2). In Armenia, the alcohol-attributable death rates are much lower than in peer countries. But rates are rising rapidly, growing by 93 percent between 1990 and 2019, dwarfing increases in other UMI countries (15 percent) and Eastern European countries (56 percent).45 During the same period, the rates in Western European countries have consistently declined 14 percent since 1990. Between 2010 and 2015, Armenians consumed about 5.6 to 5.5 liters of pure alcohol per capita, below the WHO Europe regional average of 11.2 and 9.8, respectively (Table 2.2).43 The per capita consumption increased by 0.7 liters for men in 2016 while remaining the same for women. The prevalence of heavy episode drinking is 50.0 percent among those aged 15–19 years and 40.3 among those older than 15. However, a 2009 survey found that 11 percent of 15-year-old girls and 27 percent of boys reportedly drank alcohol at least once a week, which is lower than in other European countries.44 Chapter 2. The current state of health and consumption taxes in Armenia 12 BOX 2: Global evidence on alcohol consumption and NCDs Increasing alcohol consumption is associated with an increased risk of several NCDs.36 Compared to non-drinkers (Table 2.1), any consumption of alcohol increases the relative risk (RR) of developing liver cancer (1.45–3.03), cerebrovascular disease (0.91–1.64), type 2 diabetes (DM) (0.87–1.01), and ischemic heart disease (IHD) (0.82–1.01).36–42 However, under 50 grams per day, alcohol may lower the risk of DM by 30 percent.37,38 Beyond 50 grams per day, alcohol consumption has increased the risk of diabetes. Similarly, below 30 grams per day, alcohol consumption can protect against IHD risk. These benefits are more substantial for women than men.39 For a total stroke, less than 12 grams per day confer a reduced risk, and greater than 60 grams per day increase the risk.40 There are no beneficial effects of alcohol consumption on liver cancer. TABLE 2.1: Relative Risk associated with different levels of alcohol consumption compared to non-drinkers FORMER CONDITION LOW MEDIUM HIGH HIGHER SOURCES DRINKER 1.3-25 25-45 45-65 >65 grams/ grams/ grams/ grams/ day day day day Baliunas et al. (2009). Type 2 1.18 0.87 0.87 1.01 1.01 Koppes et al. (2005). Diabetes (DM) Rehm et al. (2010). Ischemic Heart Roerecke & Rehm (2011, 1.25 0.82 0.74 1.0 1.01 Disease (IHD) 2012, 2014). Cerebrovascular Reynolds (2003) 1.33 0.91 1.1 1.1 1.64 Disease Rehm et al. (2010). Liver Rehm et al. (2010). 1.31 1.45 1.45 1.45 3.03 Cancer Turati et al. (2014). 13 THE IMPACT OF HEALTH TAXES IN ARMENIA TABLE 2.2: Alcohol consumption and related health burden in Armenia ALCOHOL CONSUMPTION: LEVELS AND PATTERNS Males Females Total Alcohol per capita (15+ years) 10.4 1.6 5.5 consumption (Liters of pure alcohol) Total alcohol per capita (15+ years) consumption, drinkers only (Liters of 27.4 9.7 21.3 pure alcohol) Prevalence of heavy episodic drinking (%) - Heavy episodic drinking is defined as consuming 60 grams or more of pure alcohol on at least one occasion in the past 30 days. Population (15–19 years) 14.0 2.1 8.1 Population (15+ years) 19.6 3.1 10.5 Prevalence of heavy episodic drinking (%), drinkers only Population (15–19 years) 59.5 24.0 50.0 Population (15+ years) 51.6 18.9 40.3 HEALTH CONSEQUENCES: MORTALITY AND MORBIDITY Age-standardized death rates Liver cirrhosis 41.7 18.2 Not available Road traffic injuries 31.2 5.4 Not available Cancer 354.5 200.1 Not available Alcohol-attributable fractions Liver cirrhosis 69.2 34.7 Not available Road traffic injuries 26.1 19.2 Not available Cancer 5.0 1.8 Not available Prevalence of alcohol use disorders 9.9 2.2 5.7 Prevalence of alcohol dependence 5.5 1.4 3.2 Source: World Health Organization 2018. About 85 percent of consumption of low amounts of alcohol (1.3–25 grams/day) is carried out by the poorest quintile (Table 2.3). The richest quintiles primarily consume higher alcohol amounts (45 grams/day or more). Spirits make up 82 percent of all alcohol consumed, followed by beer at 11 percent and wine at four percent. In 2018, the average Armenian household spent 0.7 percent (AMD 322) of their monthly expenditure on alcohol.19 Chapter 2. The current state of health and consumption taxes in Armenia 14 TABLE 2.3: Prevalence of alcohol consumption in Armenia by income-quintile DRINKING CATEGORY Q1 Q2 Q3 Q4 Q5 Low (1.3-24 grams/day) 0.85 0.80 0.79 0.77 0.71 Medium (25-44 grams/day) 0.09 0.13 0.17 0.18 0.16 Higher (45-65 grams/day) 0.02 0.01 0.02 0.02 0.04 High (>65 grams/day) 0.02 0.06 0.03 0.04 0.09 Total 1.00 1.00 1.00 1.00 1.00 Source: Armenia Integrated Living Conditions Survey 2016. Notes: Q1 refers to the lowest-income quintile, while Q5 is the highest-income quintile. There is an excise tax on beer, wine, and spirits.43 The alcohol tax structure is complicated by the application of a large number of tax categories and different bases. For example, six product categories use volumetric taxes, while two products apply a tax rate per liter of absolute alcohol. Tax rates in 2021, 2022, and 2023 increased by a coefficient of 1.03, 1.06, and 1.09, respectively, relative to 2020. However, some categories had (or will have) higher escalations. Specifically, tax rates on ethyl spirits and spirituous liquors are increasing by 129 percent between 2020 and 2023.3 In the European Union (EU), beer excise rates are based on alcohol content, and all spirits are levied at the same rate.46 In contrast, Armenia sets different rates. Therefore, the rate for vodka and an ABV of 40 percent is used for comparison. We find that the beer excise taxes in Armenia are comparable to the EU, with rates higher than 13 countries and lower than 14. In addition, Armenia’s rates are far lower than the highest countries. Regarding spirits, Armenia’s excise rates are lower than all EU countries. Between 2015 and 2019, tax revenue in Armenia from alcohol increased from AMD 13.6 to 17.3 billion, reflecting nominal and real increases of 28 and 23 percent, respectively. The low rise in revenue is due to the lack of meaningful increases in tax rates during this period.3 For example, the excise tax on beer remained unchanged during this period. In 2019, alcohol tax contributed 1.2 percent of total tax revenue. Armenia does not have a national regulatory framework for alcohol. However, the legal minimum age for purchasing alcohol is 18 years. In addition, there are restrictions on where alcoholic beverages can be sold. The maximum legal blood alcohol concentration allowed while driving a vehicle is 0.08 percent. There are no legally binding regulations on alcohol sponsorship and sales promotion, but they exist for alcohol advertising and product placement. Armenia also does not have a licensing system for retail alcohol sales, and sellers must self-register.47 There is a minimum unit price of 6,000 drams (USD 12.7)/liter of pure alcohol for all alcoholic beverages. 15 THE IMPACT OF HEALTH TAXES IN ARMENIA SUGAR-SWEETENED BEVERAGES The global evidence links SSB consumption to NCD risk, with some variation by age (Box 3). In Armenia, the mortality attributable to SSBs is higher than in peer groups, including UMIs (0.5 percent), Western Europe (0.4 percent), and Eastern Europe (0.7 percent).55 It has increased 18 percent since 1990, similar to other UMI and Eastern European countries, but higher than in Western European countries. BOX 3: Global evidence linking SSB consumption and NCDs The WHO defines SSBs as high-calorie, non-alcoholic beverages containing sweeteners, such as sucrose or high-fructose corn syrup (HFCS).48 Excess SSB consumption is associated with obesity, DM, cardiovascular diseases, cancers, and a higher risk of death (Table 2.4). TABLE 2.4: NCD relative risk associated with SSB consumption CEREBROVASCULAR AGE DM IHD DISEASE 32 4.39 1.00 1.81 37 3.99 1.78 1.74 42 3.62 1.78 1.68 47 3.29 1.71 1.61 52 2.99 1.71 1.55 57 2.72 1.71 1.49 62 2.47 1.65 1.44 67 2.24 1.65 1.38 72 2.04 1.59 1.33 77 1.85 1.59 1.28 82 1.61 1.53 1.21 87 1.61 1.53 1.21 92 1.61 1.53 1.21 Source: Murray et al. 2013. Across Europe, Armenia was one of three countries with the highest prevalence of daily sweets consumption.44 In 2018, SSB consumption in Armenia was 105 grams per day, far greater than the theoretical minimum risk exposure level (TMREL) of 2.5 grams per day and the global average of 95 grams per day.49 Households in the first decile consumed an average of 0.3 kilograms of sugar per month per capita, lower than the 0.5 kilograms consumed by those in the tenth decile.19 The prevalence of SSB consumption is much higher in the poorest quintile than in the wealthiest quintile (Table 2.5). About 73 Chapter 2. The current state of health and consumption taxes in Armenia 16 percent of 15-16-year-olds consume SSBs at least once a week at a median of 0.5 liters a day.50,51 High-affluence adolescents are more likely to consume soft drinks in Armenia than low-affluence adolescents.44 TABLE 2.5: Monthly per capita prevalence of SSB consumption in Armenia by income-quintile PRODUCT Q1 Q2 Q3 Q4 Q5 SSB 1.61 1.26 1.20 0.99 1.05 Source: Armenia Integrated Living Conditions Survey 2016. Notes: Q1: Lowest-income quintile. Q5: Highest-income quintile. Five percent of 15-year-old girls and 14 percent of 15-year-old boys are obese in Armenia.44 In 2016, the Global Nutrition Report reported Armenia's adult obesity prevalence to be 19.5 percent. The proportion of overweight individuals increased with age: 13 percent of 15-19-year-olds and 73.9 percent of 50-64-year-olds were obese.52,53 Females are 1.8 times more likely to be obese (54.1 percent) than males (47.9 percent).54 At 54 percent, obesity is also most prevalent among the second-poorest wealth quintile. In addition, 51.2 percent of Armenians are overweight. Hence, in 2019, the MoH and the MoF proposed an excise tax of five percent on all drinks with a sugar content greater than five percent per liter.56 The proposal was shelved after an outcry from the private sector.57 TOBACCO SMOKING There is very strong global evidence linking tobacco exposure to multiple NCDs (Box 4). Unfortunately, trends in tobacco-attributable mortality are rising rapidly in Armenia. Between 1990 and 2006, the rate increased from 149 to 189 deaths per 100,000. However, it stabilized at 187 deaths per 100,000 in 2019. Rates have grown faster than in other upper middle-income and Eastern European countries, with the latter experiencing substantial declines since 2006. Since 1990, Western European countries have also experienced consistent declines. In 2019, tobacco accounted for 20.2 percent of all deaths in Armenia.55 This is equivalent to other UMI countries (20.4 percent) and higher than in countries in Western Europe (17.7 percent) and Eastern Europe (16.3 percent). In 2019, Armenia had the second highest rate of cancer-related deaths worldwide. The country also ranked eleventh in smoking prevalence among males globally. There is a significant gender gap of 14 percent in smoking between 15-year-old adolescent boys and girls.44 However, in a 2009 Global Youth Tobacco survey, only 6.1 percent of boys and 1.2 percent of girls reported smoking, a lower prevalence than in most European countries.60 In the same survey, 70 percent of children reported living in households with family members who smoked. Twenty-eight percent of Armenian adults smoke. Of this number, only two percent of women smoke, but 52 percent of males are smokers.61 Smoking rates vary by quintile and age (Table 2.7). Yerevan has the highest 17 THE IMPACT OF HEALTH TAXES IN ARMENIA prevalence of smokers (30.2 percent) compared to other cities (21.4 percent) and villages (23.3 percent).16 Also, three out of ten smokers report smoking 25 or more daily cigarettes. In 2018, Armenian households spent 3.6 percent (AMD 1,653) of their monthly household expenditure on tobacco, an increase from 3.0 percent in 2008.19 BOX 4: Global evidence on tobacco exposure and NCD risk Tobacco smoking is associated with multiple NCDs, including diabetes and heart disease. Heavy smokers who smoke 20 or more cigarettes daily have a 61 percent greater risk of developing diabetes than non-smokers.58 Smoking is also associated with an increased risk of aortic aneurysm (3–6 times), coronary heart disease and cerebrovascular disease (1–2 times), and chronic obstructive pulmonary disease (5–12 times) (Table 2.6).59 TABLE 2.6: Relative risk associated with smoking intensity CONDITION FORMER SMOKER CURRENT SMOKER SOURCE DM 1.23 1.61 Willi et al. (2007). 5.49 until age 44 3.05 for 44-59 Lederle et al. (2003). IHD 1.4 1.87 for 60-69 Ezzati et al. (2005). 1.40 for 70-79 1.01 for 80+ 1.80 until age 44 3.11 for 44-59 Lederle et al. (2003). CVD 1.1 1.85 for 60-69 Ezzati et al. (2005). 1.01 for 80+ O'Keeffe et al. (2018). Lung Cancer 3.01 7.33 Turati et al. (2014). Source: Authors' calculations TABLE 2.7: Prevalence of smoking in Armenia by income-quintile AGE Q1 Q2 Q3 Q4 Q5 17 0.12 0.11 0.04 0.15 0.21 22 0.55 0.46 0.49 0.56 0.45 27 0.73 0.64 0.69 0.57 0.69 32 0.75 0.71 0.77 0.75 0.71 37 0.68 0.71 0.80 0.57 0.77 42 0.84 0.69 0.76 0.67 0.74 > 47 0.67 0.71 0.61 0.69 0.73 Source: Armenia Integrated Living Conditions Survey 2016. Notes: Q1: Lowest-income quintile. Q5: Highest-income quintile. Chapter 2. The current state of health and consumption taxes in Armenia 18 The government has taken steps to curb the negative health impacts of tobacco use, including ratifying the WHO Framework Convention on Tobacco Control (FCTC).62 In February 2020, Parliament passed legislation in line with FCTC recommendations, extending smoking bans to all tobacco products in public places and targeting industry marketing of tobacco products.61 If fully implemented, FCTC guidelines have the potential to reduce smoking prevalence by 31.5 percent by increasing excise taxes on cigarettes from 16.67 percent to 75 percent, as recommended.63 Tobacco taxes are applied as uniform specific taxes, with additional rates by type of product (cigarettes, cigars, cigarillos, and tobacco substitutes). This structure is in line with the WHO’s best practice recommendations and ensures that tax increases will result in a rise in prices to discourage the purchase and use of the products. Furthermore, absolute tax increases will be the same on cheaper and more expensive products, minimizing opportunities to buy more affordable brands in response.64 Low taxes and prices at baseline mean that Armenia has some of the cheapest cigarettes in the region. Nominal taxes fell in 2014, from AMD 120 to 100 per pack. This decrease may have been due to a rise in the VAT rate at that time. Under-shifting meant that prices fell substantially in real terms as nominal prices remained unchanged at AMD 600 per pack in 2014, 2016, and 2018. Significant increases in the specific tax in 2018 and 2020 (nearly doubling in nominal terms) resulted in substantial price rises. Yet, in real terms, the prices in 2020 were similar to 2008. Cigarettes, for example, were significantly more affordable in 2020 than in 2008. As measured by the Relative Income Price, 18 percent less (per capita GDP) could purchase a pack of cigarettes in 2020 compared to 2018. A crucial political economy consideration is the harmonization of excise tax rates in the Eurasian Customs Union, to which Armenia belongs with Belarus, Kazakhstan, Kyrgyzstan, and the Russian Federation.65 Harmonization requires Armenia to meet an indicative excise rate of Euro 35 (AMD 19,283 at current exchange rates) per 1,000 cigarettes by 2024. This rate is substantially higher than the agreed-to rate of AMD 14,640 per 1,000 cigarettes in 2023.66 While the harmonization provides a roadmap for future tobacco tax increases for the next two years, there is a need for analysis regarding the longer-term outlook. Between 2015 and 2019, nominal tobacco excise tax revenue rose from AMD 23.5 to 74.6 billion, an increase of 218 and 207 percent in nominal and real terms, respectively. Revenue rose with an increase in recorded sales, attributable to improvements in tax administration. These improvements include using relatively advanced tax stamps and fiscal marks without track and trace features.67 Anecdotal evidence suggests that Armenia is an increasing source of smuggled cigarettes into neighboring countries that have raised taxes quicker and have higher prices.68 Hence, some of the increase in tax- paid sales in Armenia is due to outward-bound smuggling of Armenia tax-paid cigarettes. As a result, tobacco excise tax revenue rose from 2.2 percent of total tax revenue in 2015 to 5.1 percent in 2019, equal to 0.5 percent and 1.1 percent of GDP, respectively. 19 THE IMPACT OF HEALTH TAXES IN ARMENIA In 2020, the government adopted a comprehensive law to curtail further tobacco consumption. The law included a ban on indoor smoking in public places, workplaces, and public transport and a total ban on tobacco advertisement and promotion.69 Under the law, smoking has been prohibited in catering facilities since March 2022.70 Furthermore, all tobacco advertising and promotion is banned, and unpaid depiction of tobacco is prohibited in children’s programming. Effective January 2022, tobacco product display has been banned everywhere except in duty-free airport shops. Thirty percent of the principal display areas of tobacco product packaging must be covered with text-only health warnings. From January 2024, all tobacco products must use plain packaging. While tobacco products are not allowed to be sold to individuals under 18, there are no age restrictions on tobacco sales on the internet. Chapter 2. The current state of health and consumption taxes in Armenia 20 21 CHAPTER 3. METHODS This chapter reflects on the methods for determining the change in health, health care expenditure, and tax revenue when alcohol, tobacco, and SSBs are taxed, under specific scenarios, in Armenia (Table 3.1). The study makes the following assumptions. First, the tax-induced increase in SSB prices will be entirely passed onto the consumers rather than the producers. Second, to ascertain the expected OOP and government savings on health care, we assumed that individuals with these conditions would fully use health care facilities. We only project tax revenues for the first year. Thus, the long-term revenue projections may differ and have been estimated in a separate investigation. In addition, the age-specific incidence, case-fatality, and prevalence for SSB-related diseases were modeled as these values were only available for broad age groups. TABLE 3.1: Baseline prices in Armenia in 2018 and proposed tax CURRENT PRICE PROPOSED TAX FINAL PRICE PERCENT INCREASE PRODUCT UNIT PER UNIT (AMD) PER UNIT PER UNIT IN PRICE Sugar-Sweetened Beverages Coca-Cola 2 Liters 276 50 326 18% Tobacco Cigarettes Pack of 22 sticks 600 440 1,040 73% Alcohol Beer Liter 841 200 1,041 24 Wine Liter 1,965 200 2,165 10 Vodka Liter 2,267 1,000 3,267 44 Cognac Liter 11,211 1,000 12,211 9 Liquor Liter 4,465 1,000 5,465 22 Champagne Liter 2,480 1,000 3,480 40 Source: World Bank. Note: AMD200 = Euro 0.37. Chapter 3. Methods 22 CHANGE IN CONSUMPTION DUE TO TAX On an intensive margin, the proposed taxes are expected to reduce the consumption of alcoholic beverages, SSBs, and tobacco. The predicted effect on the extensive margin is that some individuals will quit the consumption of these products, while others will not initiate the consumption. Given available data, we evaluated the effect of taxes on SSBs on the intensive margin and the impact of tobacco taxes on the extensive margin. The analysis broadly proceeded in two steps. First, we estimated the price elasticity of demand for the products, and second, we calculated the reduction in intensity or prevalence of consumption. We used data from the Armenia Integrated Living Conditions Survey (ILCS) 2016, given availability of the full range of the variables needed for the analysis, and estimated the compensated own-price elasticity using the Almost Ideal Demand System (AIDS).71 Appendix 1D details the methodology for estimating price elasticity. Appendixes 1G-H specify the estimated price elasticity for alcoholic beverages, SSBs, and cigarette smoking. We estimated the change in the consumption of SSBs by multiplying the percentage difference in price by the price elasticity of demand, consumption at baseline, and the expected pass-through (see Appendix 1H). For the baseline analysis, a pass-through of 100 percent was assumed, meaning consumers would pay the entire increase in prices instead of the producers. For the sensitivity analysis, we used different values of pass-through: the lowest of 31 percent was observed in the United Kingdom, and the highest of 121 percent was noted in Saudi Arabia.72,73 For alcoholic beverages, the alcohol drinkers were categorized as low, medium, high, or higher based on the amount of daily alcohol consumption at baseline. Then, the change in alcohol consumption in grams per day was calculated using the price elasticity and proposed tax. We then recalculated the proportion of individuals in each category. For cigarette consumption, we used the elasticity to alter the participation in smoking such that some current smokers will quit smoking and become former smokers. The literature shows that there is no safe level of tobacco use. Furthermore, smoking is significantly worse than not smoking. Hence reducing participation – smoking on the extensive margin – may be of greater importance for policymakers than a reduction in smoking on the intensive margin.74,75 23 THE IMPACT OF HEALTH TAXES IN ARMENIA CHANGE IN TAX REVENUE DUE TO CHANGE IN CONSUMPTION Here, we multiplied the estimated post-tax consumption by the tax amount to derive the total tax revenue. As proposed by the MoF, a uniform tax rate was applied to all products. Annual tax revenue estimates are provided only for the first year. This decision is to simplify the analysis and avoid issues of future revenue discounting and dampening taxation effects over time. CHANGE IN HEALTH DUE TO CHANGE IN CONSUMPTION To estimate the difference in disease incidence due to changes in consumption, we built a dynamic epidemiological model with multistate life tables in Python (version 3.6).76,77 For each cohort, the model calculates the difference in disease incidence and mortality due to changes in consumption and other input parameters. For SSB-related diseases, the model calculates the difference in body mass index (BMI) due to a change in SSB consumption and converts the change in BMI to the energy consumed. Energy consumption is associated with weight change and a new BMI distribution. Then, the changes in BMI distribution are used to calculate population impact fraction (PIF), based on age and gender-specific relative risk estimates for a unit change in BMI.78,79 We multiplied the baseline incidence by the PIF to calculate the difference in disease incidence and premature mortality for each cohort within the life table. This process was repeated for all the cohorts (Appendix 1I). We obtained prevalence, mortality, and incidence values by broad age groups. Therefore, we derived the age-specific incidence, case-fatality, and prevalence for one-year age groups using DISMOD II.80 A similar approach was used for alcohol and tobacco tax. Still, instead of calculating the change in BMI, we used the difference in prevalence for each alcohol consumption and smoking category and the relative risk associated with those categories (Tables 1, 2.4, and 2.6).36–39,41,42,58,59,81–84 CHANGE IN HEALTH CARE EXPENDITURE DUE TO CHANGE IN HEALTH We also calculated the changes in households’ public health care expenditures, government subsidies, and OOP expenses. First, the expected difference in disease incidence was multiplied by the cost per prevalent case (Appendix 1K). Each Chapter 3. Methods 24 condition’s total health care costs were then apportioned into the expected change in government outlays and OOP payment. Next, Demographic and Health Survey data was used to estimate the number of individuals per quintile eligible for a government subsidy (Appendix 1L). Finally, we applied the government subsidy for each condition (Appendix 1M) to obtain the expected savings for the government and OOP payments. In each case, the analysis assumed 100 percent utilization of health care facilities by individuals with conditions. STATISTICAL AND MODELING SOFTWARE Stata versions 15 and 17 were used to analyze ILCS 2016 to compute the mean consumption and prevalence and calculate price-elasticity estimates. DISMOD II was used to generate age-gender-specific incidence, prevalence, and case-fatality for diseases. Python (versions 3.6 and 3.9) was used for life-table modeling. 25 CHAPTER 4. RESULTS In this chapter, we review the main results of the analysis, including the returns to increased excise taxes on health outcomes, revenue, and expenditure on health care. CHANGE IN HEALTH DUE TO CHANGE IN CONSUMPTION We find gains in taxing each product’s incidence and premature mortality (Table 4.1). For example, taxing alcohol can avert 402 deaths, including 207 due to IHD, 175 from cerebrovascular disease, and 20 liver cancer deaths. The averted cerebrovascular disease deaths make up one percent of annual cerebrovascular disease deaths. Furthermore, the averted deaths from IHD and liver cancer make up 0.3-0.4 percent of annual deaths in 2018. The highest reduction in premature deaths is in the wealthiest quintile (34 percent of averted liver cancer fatalities and 26 percent of averted IHD deaths). The second and third-poorest quintiles follow this group. For all conditions, the poorest quintile has the smallest reduction in fatalities. Chapter 4. Results 26 TABLE 4.1: Estimated number of premature deaths averted over 20 years PREMATURE DEATHS AVERTED Product Condition Q1 Q2 Q3 Q4 Q5 Total IHD 24 37 51 42 53 207 Cerebrovascular Alcohol 24 39 40 31 41 175 Disease Liver Cancer 2 6 3 2 7 20 Type 2 Diabetes 15 15 14 14 13 70 Mellitus SSB IHD 46 47 45 43 41 222 Cerebrovascular 21 22 21 20 19 103 Disease IHD 4,135 3,448 3,226 2,913 2,945 16,667 Cerebrovascular Tobacco 1,199 1,035 904 883 879 4,900 Disease Lung Cancer 2,817 2,450 2,135 2,090 2,079 11,572 Type 2 Diabetes 15 15 14 14 13 70 Mellitus IHD 4,205 3,532 3,321 2,998 3,039 17,049 Total Cerebrovascular 1,244 -1,096 -965 -933 -940 -5,095 Disease Liver Cancer 2 -6 -3 -2 -7 -20 Lung Cancer -2,817 -2,450 -2,135 -2,090 -2,079 -11,572 Source: Author’s calculations. Note: Q1: Lowest-income quintile. Q5: Highest-income quintile. The tax on SSB had the smallest effect on the total number of premature deaths averted. These taxes were projected to prevent 70 DM-related deaths, 222 IHD-related deaths, and 103 deaths from cerebrovascular disease. Although these may seem like a small number of deaths in absolute terms, they made up 1.7 percent of annual DM deaths, 0.62 percent of cerebrovascular disease deaths, and 0.3 percent of IHD deaths in 2018. Changes in the pass-through of 31 percent and 121 percent were associated with 19 and 86 averted deaths, respectively (Appendix 1R). In contrast to alcoholic beverages, most averted deaths following an increase in the SSB taxes are in the lowest-income quintiles (21 percent each for the poorest and second-poorest). The maximum impact in terms of the absolute number of premature deaths averted was through taxes on tobacco-related products. The increased taxes prevented an estimated 16,667 IHD-related deaths, 11,572 lung-cancer-related deaths, and 4,900 cerebrovascular-related deaths. Again, similar to SSB taxes, the poorest quintile had the highest proportion of averted deaths (25 percent) compared to wealthier quintiles (17–19 percent in quintiles three-five). 27 THE IMPACT OF HEALTH TAXES IN ARMENIA CHANGE IN HEALTH CARE EXPENDITURE DUE TO CHANGE IN HEALTH The highest health care cost reduction was achieved by increasing tobacco taxes, reflecting the averted disease burden (Figure 4.1). In addition, as the health care system provides differentiated levels of government subsidies for diseases, they translate into higher savings for conditions for which the government subsidies are the highest (IHD and cancers). Lastly, the OOP savings accrued mainly to those in the poorest quintiles for SSBs and tobacco and the second-poorest quintile for alcohol. FIGURE 4.1: Change in total health care cost Overall Healthcare Cost (min AMD): (Alcohol Tax) 0 -10 -20 -10 -12 -9 -30 -40 -50 -39 -47 Overall Healthcare Cost (min AMD): (SSB Tax) 0 -10 -20 -30 -40 -37 -40 -41 -39 -38 Overall Healthcare Cost (min AMD): (Tobacco Tax) 0 -1,000 -2,000 -3,000 -3,162 -2,739 -2,782 -4,000 -3,251 -3,972 Quintile 1 Quintile 2 Quintile 3 Quintile 4 Quintile 5 Source: Authors. Note: Q1: Lowest-income quintile. Q5: Highest-income quintile. CHANGE IN TAX REVENUE DUE TO CHANGE IN CONSUMPTION The estimated annual tax revenue was AMD 136 billion, which is slightly over two percent of the GDP in 2019. Figure 4.2 shows the estimated tax revenue by product and income quintile. Tobacco taxes form 81 percent of the additional tax revenue, followed by SSBs (18 percent) and alcoholic products (less than one percent) (Appendix 1P). Chapter 4. Results 28 The burden of tax distribution across the income quintiles varies by product owing to the differences in elasticity across quintiles, prevalence, and consumption intensity. For alcoholic products, the largest share of tax is borne by the highest income quintile (24 percent) and the lowest percentage (15 percent) in the poorest income quintile. Within the alcoholic taxes, the largest share accrues from taxes on vodka (80 percent) due to the larger per capita consumption and higher tax per liter (AMD 1,000 per liter) (Appendix 1Q). The distribution of taxes by quintiles for SSBs exhibits the opposite trend. The lowest- income quintile bears the highest proportion of taxes (26 percent). Also, the proportion reduces monotonically with income and reaches the lowest share of 17 percent in the highest-income quintile. This pattern is likely due to the highest per-capita monthly SSB consumption in the lowest-income quintile (1.6 liters) (Table 2.3). Lastly, tax revenue distribution from tobacco-related products does not exhibit a consistent trend. The wealthiest quintile bears the highest share of tax (22 percent), followed by 20 percent in the second-poorest and second-richest quintiles (Figure 4.2). The distribution of taxes across income quintiles varies by product type. Hence, the sum revenue of the three taxes roughly follows a very flat- “U”-shaped distribution with a 21 percent share for the richest and poorest quintiles and 19 percent for the second-richest and middle quintiles. FIGURE 4.2: Estimated annual tax revenue (in AMD Million) Tax Revenue (min AMD): (Alcohol Tax) 300 251 279 206 228 200 176 100 0 Tax Revenue (min AMD): (SSB Tax) 8,000 6,507 6,000 5,073 4,866 4,286 4,014 4,000 2,000 0 Tax Revenue (min AMD): (Tobacco Tax) 24,145 25,000 21,445 21,986 21,481 21,874 20,000 15,000 10,000 5,000 0 Quintile 1 Quintile 2 Quintile 3 Quintile 4 Quintile 5 Source: Authors Note: Q1: Lowest-income quintile. Q5: Highest-income quintile. 29 CHAPTER 5. CONCLUSION Armenia is faced with the critical challenge of improving population health by ensuring better access to high-quality health care for NCDs. Not addressing the NCD burden has resulted in an annual loss of 3,000 lives and AMD 360 billion. To this end, it is important to ensure access to high quality health care, to prevent the exposure to risk and appropriately manage NCDs. However, high OOP spending is a crucial barrier to utilizing health care in Armenia. Hence, increased public financing is needed to reduce OOP and ensure adequate health care use. The MoH is leading an effort to undertake a national health reform that mobilizes additional public revenue, directed in part at strengthening services for NCDs and reducing financial barriers to access. The estimated cost of a benefits package that covers essential services is over 300 billion AMD annually. At their request, the World Bank team has modeled the potential health and revenue gains from excise taxes, drawing on household survey data. We show that with an annual revenue of AMD 130–136 billion, taxation of tobacco, alcohol, and SSBs provide a feasible option for raising additional public revenue while improving population health and increasing workforce productivity.3 Our analysis reveals that taxing tobacco averts the most deaths: 33,139 over 20 years, predominantly among the poorest quintile (25 percent) – see ES Table 1. Alcohol taxes can prevent 402 deaths, 50 percent among the richest. In contrast, SSB tax can avert 395 deaths, with the poorest and the richest impacted almost evenly. The total projected revenue across all products is AMD 136 billion, with the most significant contribution from tobacco tax (81 percent), followed by SSBs (18 percent), and then Chapter 5. Conclusion 30 alcohol (less than one percent). Taxation of alcoholic products is primarily borne by the wealthiest quintiles, while the poorest quintiles shoulder the tax on SSBs. The burden of health care costs for tobacco and SSBs falls on the most affluent quintile, while the poorest bear the costs of health care for alcohol. Armenia has increased tobacco taxes significantly in recent years and plans to continue these increases with scheduled tax increases in 2023, including to align with the Eurasian Customs Union’s movement towards harmonized excise taxes. While these increases have been meaningful, taxes and prices in Armenia are still very low compared to peer countries. Combined with increasing mortality due to tobacco use in recent decades, this highlights the importance of continuing to raise taxes in Armenia, particularly given the capable tax administration that could collect future tax increases efficiently. Hence, among the potential health excise taxes, tobacco taxation offers the highest revenue and saves the most lives. While the choice of tax options is up to the government, the need for expanding coverage for essential care is also a necessity. Financial barriers to high-quality health care prevent users from accessing care, further contributing to the growing burden of NCDs. While other countries mobilized prepaid revenue through compulsory contributions, Armenia may choose from this and a host of other strategies, based on compatibility with the country’s broader development goals. This decision will have important consequences for all Armenians. 31 THE IMPACT OF HEALTH TAXES IN ARMENIA APPENDIX Appendix 1A: Consumption of SSB by income-quintile in Armenia in 2016 PRODUCT Q1 Q2 Q3 Q4 Q5 SSB 1.61 1.26 1.20 0.99 1.05 Source: Armenia Integrated Living Conditions Survey 2016. Notes: This table shows the monthly per-capita consumption of SSB for 2016. Q1: Lowest-income quintile. Q5: Highest-income quintile. The analysis obtained the income-quintile specific per-capita monthly consumption of SSBs from ILCS2016. Analysts summed the reported purchase of SSB by the households for the survey month and divided it by the number of household members without adjusting for the household composition by age and gender. The consumption data in ILCS is at the household level and analysts were unable to further segregate the consumption by age-groups or gender. Appendix 1B: Excise Taxes for 2020, 2021, 2022 and 2023 (effective 1 January 2022) RATE (AMD) BASE 2020 2021 2022 2023 Tobacco Cigarettes 1,000 sticks 9,625 11,070 12,730 14,640 Cigars 1,000 sticks 605,000 623,150 641,300 659,450 Cigarillos 1,000 sticks 16,500 16,995 17,490 17,985 Tobacco substitutes (i.e., RYO, pipe) kg 1,500 1,545 1,590 1,635 Alcohol Beer Liter 130 134 138 142 Grape wine Liter 150 155 159 164 Vermouth and other grape wines Liter 1,000 1,030 1,060 1,090 Vodka (made of fruit and/or berries) Liter 800 824 848 872 Cognac, brandy, and other spirits 1-3 years old Liter of alcohol 3,000 3,090 3,180 3,270 4-5 years old Liter of alcohol 3,500 3,605 3,710 3,815 6-10 years old Liter of alcohol 6,000 6,180 6,360 6,540 11-15 years old Liter of alcohol 8,500 8,755 9,010 9,265 16-19 years old Liter of alcohol 14,000 14,420 14,840 15,260 20+ years old Liter of alcohol 22,000 22,660 23,320 23,980 Other brewed drinks (apple cider, Liter 270 278 286 294 pear cider, honey drinks) Ethyl spirit Liter of alcohol 2,600 3,380 4,400 5,700 Spirituous liquors Liter 1,560 2,030 2,640 3,430 Source: PWC Armenia Corporate – Other taxes (3 January 2022). Note: The source includes tax rates for 2020. Rates for 2021, 2022 and 2023  and estimated based on indicative increases provided by the same source. EUR per 750ml bottle of 40% ABV spirits EUR per 330ml bottle/can of 5% ABV beer 0 2 4 6 8 10 12 14 16 Appendix 0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 Armenia Romania Bulgaria Bulgaria Romania Germany Croatia Luxembourg Hungary Spain Spain Countries, 2021 Czech Rep Italy Slovakia Luxembourg Portugal Slovakia Hungary Cyprus Poland Austria Malta Germany Austria Slovenia Belgium Malta Armenia Portugal Croatia Poland Cyprus Netherlands Denmark Latvia Lithuania France Italy Estonia Netherlands Denmark France Lithuanua Latvia Greece Greece Belgium Slovenia Source: GTP estimates from European Commission (2021). Ireland Estonia Sweden Sweden Finland Ireland Finland Appendix 1C: Beer and Spirit Excise Tax in Armenia and EU 32 33 THE IMPACT OF HEALTH TAXES IN ARMENIA Appendix 1D: Estimating the price-elasticity of demand using Armenian data The national price elasticity of demand was estimated using Armenia Integrated Living Conditions Survey 2016 (National Statistical Service of the Republic of Armenia and World Bank 2016). The Almost-Ideal Demand System (AIDS) was used to estimate the compensated own price elasticity of demand for the individual products, as explained below. For alcoholic beverages, consider a group of products such as different types of alcoholic beverages: Beer, Wine, Vodka, champagne, and liqueur, denoted by , , , , and , respectively. The number of products () is 5. The following was obtained from the data: • is the expenditure share of the h good in the group of products where {, , , , } • � is the nominal price of the h good • () is the natural log of the total expenditure for each product group • is the random error term; and • () is the natural log of the price-index • Apply the Stone Price Index : Estimate the following equation for each product group: • for each {, , , , } For alcohol group, the system of equations is as follows: • for beer • for wine • for vodka • for champagne • for liqueur The compensated own price-elasticity can be calculated as follows: • There was insufficient data to estimate the income-quintile and gender-specific price- elasticity. Hence, analysts used the same elasticity estimate for all income-quintiles and both genders. Appendix 34 For sugar-sweetened beverages and cigarette-smoking, analysts attempted to impute the income-quintile specific price-elasticity through adjusting the estimated price- elasticity in Armenia by the published price-elasticity estimates from other studies. In this crude imputation method, it was assumed that the estimated price-elasticity is the best-estimate for the median income group, Quintile Three. Analysts then use international studies with published estimates of price-elasticity by income-quintile to calculate the ratio of elasticity for a given quintile and elasticity for Quintile Three. The estimated Armenian elasticity was then multiplied by this ratio to impute the elasticity for other quintiles assuming that the same ratio of elasticity would apply in Armenia. Appendix 1E: Price elasticity of demand for alcoholic beverages in Armenia in 2018 ALCOHOL TYPE ELASTICITY Beer -0.882 Wine -0.439 Vodka -0.493 Cognac -0.346 Liqueur -0.128 Champagne -0.327 Source: Authors’ calculations using Armenia Integrated Living Conditions Survey 2016. Appendix 1F: Price elasticity of demand for sugar-sweetened beverages in Armenia in 2018 AGE-GROUP Q1 Q2 Q3 Q4 Q5 15-95 -0.440 -0.463 -0.440 -0.418 -0.402 Source: Authors’ calculations using Armenia Integrated Living Conditions Survey 2016. Note: Q1: Lowest-income quintile. Q5: Highest-income quintile. Appendix 1G: Price elasticity of demand for smoking in Armenia in 2018 AGE-GROUP Q1 Q2 Q3 Q4 Q5 15-24 -1.122 -0.946 -0.911 -0.841 -0.771 25-95 -0.561 -0.473 -0.456 -0.421 -0.386 Source: Authors’ calculations using Armenia Integrated Living Conditions Survey 2016. Note: Q1: Lowest-income quintile. Q5: Highest-income quintile. 35 THE IMPACT OF HEALTH TAXES IN ARMENIA Appendix 1H: Price-Elasticity Appendix 1I: Epidemiological model For each cohort, the model first calculates the change in BMI due to change in SSB consumption and then converts the change in BMI to change in disease incidence and premature mortality. Change in BMI distribution The change in BMI is computed in the following steps: the change in the consumption of sweetened beverages is translated to change in energy consumption using data on energy density of the SSB. The change in energy consumption is then converted to a change in body weight based on a widely used energy balance equation. Past research done on energy balance equations shows that a change of 1kg in body weight for adults is observed when a daily energy change of 94kJ is recorded assuming no variation in physical activity. Using this new body weight and baseline height enables calculation of the new BMI and change in BMI for each age-gender-category in each income-quintile. Appendix 36 Change in disease incidence and premature mortality A disease incidence model was used to translate the change in BMI into disease incidence and mortality. Analysts obtained the age and gender specific relative risk of Type 2 Diabetes Mellitus due to 5 unit increase in BMI and used it to calculate the Potential Impact Fraction (PIF) post-intervention. The relative risks for SSB related diseases and the PIF calculation for SSBs are shown elsewhere. The baseline age- specific incidence of the disease is multiplied by the PIF to obtain the post-tax age- specific disease incidence. The disease incidence model then conducted a simulation such that each cohort moved forward in time using the new incidence rate until they reached a maximum age of 100 years or death. This cohort-specific simulation was repeated across the income- quintiles to estimate the effect on Type 2 Diabetes Mellitus incidence and mortality over a 20-year time-period. Appendix 1J: Potential impact fraction Where: • denotes the exposure levels the risk factor can take on: ° is the lowest value of risk factor: In case of BMI: 10 ° is the highest value of the risk factor: In case of BMI: 50 • () is the original risk-factor distribution • *() is the risk-factor distribution after the intervention • () is the RR function ° Log-linear RR function: ° is the per-unit RR - The relative risk of disease due to per unit change of BMI - E.g. RR= 1.1 per 5kg/m2 of BMI ° is the risk-factor for which () = 1 - Assume to be the mean BMI of the 'normal weight' category i.e., 22.70355 37 THE IMPACT OF HEALTH TAXES IN ARMENIA Appendix 1K: Average price paid and number of visits in Armenia in 2018 AVERAGE AVERAGE AVERAGE PAID AVERAGE PAID NUMBER OF NUMBER OF PER INDIVIDUAL PER INDIVIDUAL DISEASE NAME ICD CODES VISITS PER VISITS PER (MALE) (FEMALE) INDIVIDUAL INDIVIDUAL (AMD) (AMD) (MALE) (FEMALE) Ischemic Heart I20 – I25 160,088 100,574 2.28 1.95 Diseases Cerebrovascular I60 – I69 78,868 71,811 1.34 1.26 Diseases Non-insulin- dependent E11 13,780 13,147 2.08 2.08 diabetes mellitus (Type II) Lung-Cancer C33-C34 171,316 186,851 3.8 2.7 Liver-Cancer C22 171,316 186,851 3.8 2.7 Source: Ministry of Health Notes: This table shows the average price charged for males and females for  the respective conditions in 2018. As data for liver-cancer was not available,  it was assumed that the price charged is same as that for lung-cancer. Appendix 1L: Health insurance coverage by wealth quintile in Armenia in 2015 BASIC BENEFITS WEALTH QUINTILE NONE PACKAGE Q1: Lowest 2.5 96.6 Q2: Second 6.7 92.5 Q3: Middle 5.9 91.7 Q4: Fourth 7.8 90.6 Q5: Highest 7.7 90.4 Source: Ministry of Health and National Institute of Health (NIH) Demographic and Health Surveys Note: This table shows the percentage of surveyed men that are covered by  Basic Benefits Package or not covered by any insurance scheme. Appendix 1M: Distribution of payment systems in Armenia in 2018 GOVERNMENT OUT-OF-POCKET DISEASE NAME ICD CODES SUBSIDY PAYMENT Ischemic Heart Diseases I20 – I25 88.5% 11.5% Cerebrovascular Diseases I60 – I69 22.1% 77.9% Non-insulin-dependent diabetes mellitus E11 99.2% 0.8% (Type II) Lung-Cancer C33-C34 28.0% 72.0% Liver-Cancer C22 28.0% 72.0% Source: Ministry of Health and National Institute of Health (NIH):  The Account of FA Financing Agents and DIS Classification of Diseases. Note: This table shows the absolute number for incidence, prevalence, and  mortality among the population aged 18-100 for the year 2018. Appendix 38 Appendix 1N: Savings for government’s outlay on healthcare costs Government subsidy (min AMD): (Alcohol Tax) .5 0 -.5 -1 -1.5 -2 Government subsidy (min AMD): (SSB Tax) 0 -.5 -1 -1.5 -2 Government subsidy (min AMD): (Tobacco Tax) 0 -50 -100 -150 Quintile 1 Quintile 2 Quintile 3 Quintile 4 Quintile 5 Source: Authors Note: Q1: Lowest-income quintile. Q5: Highest-income quintile. Appendix 1O: Savings for out-of-pocket payments Out of Pocket Payment (min AMD): (Alcohol Tax) 0 -10 -20 -30 -40 -50 Out of Pocket Payment (min AMD): (SSB Tax) 0 -10 -20 -30 -40 Out of Pocket Payment (min AMD): (Tobacco Tax) 0 -1,000 -2,000 -3,000 -4,000 Quintile 1 Quintile 2 Quintile 3 Quintile 4 Quintile 5 Source: Authors Note: Q1: Lowest-income quintile. Q5: Highest-income quintile. 39 THE IMPACT OF HEALTH TAXES IN ARMENIA Appendix 1P: Distribution of estimated tax (in AMD million) PRODUCT PRODUCT TYPE QUINTILE 1 QUINTILE 2 QUINTILE 3 QUINTILE 4 QUINTILE 5 TOTAL AS % OF TOTAL Alcohol 176.32 205.67 228.03 251.31 278.54 1,139.86 0.83% Sugar- Sweetened 6,506.90 5,072.76 4,865.93 4,013.54 4,285.64 24,744.77 18.09% Beverages Tobacco 21,445.06 21,985.99 21,481.18 21,873.89 24,144.89 110,931.02 81.08% Total 28,128.27 27,264.43 26,575.14 26,138.75 28,709.07 136,815.7 Quintile as % 21% 20% 19% 19% 21% 100% of Total Source: Author’s calculation. Notes: This table shows the estimated annual tax revenue (in AMD million) from  tax on alcoholic beverages, sugar sweetened beverages, and tobacco products. Q1: Lowest-income quintile. Q5: Highest-income quintile. Appendix 1Q: Distribution of estimated tax from alcoholic beverages (in AMD million) TYPE AS % ALCOHOL TYPE QUINTILE 1 QUINTILE 2 QUINTILE 3 QUINTILE 4 QUINTILE 5 TOTAL OF TOTAL Beer 3.28 2.60 3.11 4.76 8.76 22.51 2% Wine 21.85 20.14 19.84 22.85 23.86 108.55 10% Champagne 6.25 4.51 6.09 5.44 5.40 27.68 2% Liqueur 0.93 1.38 3.69 3.06 5.63 14.70 1% Cognac 5.97 4.41 12.60 19.01 13.94 55.93 5% Vodka 138.04 172.63 182.69 196.20 220.93 910.50 80% Total 176.32 205.67 228.03 251.31 278.54 1,139.86 100% Quintile as % 15% 18% 20% 22% 24%   of Total Source: Author’s calculation. Notes: This table shows the estimated annual tax revenue from tax on alcoholic beverages. The baseline tax  is assumed to AMD 200 for beer and wine, and AMD 1000 for spirits including liqueur, cognac, and vodka. Q1: Lowest-income quintile. Q5: Highest-income quintile. Appendix 40 Appendix 1R: Distribution of premature mortality, health-care expenditure and government savings due to varying pass- through for sugar-sweetened beverages PRODUCT CONDITION Q1 Q2 Q3 Q4 Q5 TOTAL Panel A: Pass-Through 31% Type 2 Diabetes -4 -4 -4 -4 -4 -20 Mellitus Ischemic Heart -14 -15 -14 -13 -13 -69 Premature Deaths Disease Averted Cerebrovascular -7 -7 -6 -6 -6 -32 Disease Total -25 -25 -24 -23 -22 -120 Health-care Expenditure Total -11.3 -11.4 -10.8 -10.5 -10.2 -54.3 Out-of-pocket savings Total -11.1 -10.9 -10.3 -10.0 -9.7 -52.0 (AMD Million) Tax Revenue Total 6,895 5,393 5,156 4,240 4,518 26,202 (AMD Million) Panel B: Pass-Through 121% Type 2 Diabetes -18 -18 -17 -17 -16 -86 Mellitus Ischemic Heart -55 -57 -54 -52 -50 -268 Premature Deaths Disease Averted Cerebrovascular -26 -26 -25 -24 -23 -124 Disease Total -99 -102 -96 -93 -89 -478 Health-care Expenditure Total -51.0 -51.7 -47.9 -46.4 -44.6 -241.5 Out-of-pocket savings Total -50.1 -49.3 -45.9 -43.8 -42.2 -231.2 (AMD Million) Tax Revenue Total 6,389 4,975 4,778 3,945 4,215 24,301 (AMD Million) Note: Q1: Lowest-income quintile. Q5: Highest-income quintile. 41 THE IMPACT OF HEALTH TAXES IN ARMENIA ENDNOTES 1.  Chukwuma, A., Meessen, B., Lylozian, H., Gong, E. & Ghazaryan, E. Strategic Purchasing for Better Health in Armenia.pdf. (2020). 2.  Dudu, H., Chukwuma, A., Manookian, A., Aghazaryan, A. & Zeshan, M. Macroeconomic Effects of Financing Universal Health Coverage in Armenia. (2021) doi:10.1596/35688. 3.  Maduko, F. et al. More Money for Health: Resource Mobilization for Universal Health Coverage in Armenia. (2021). 4.  Europe and Central Asia Economic Update, Fall 2020: COVID-19 and Human Capital. (2020). World Bank. 5.  World Development Indicators. Mortality rate, infant (per 1,000 live births). World Bank https:// data.worldbank.org/indicator/SP.DYN.IMRT.IN (2019). 6.  World Development Indicators. Incidence of tuberculosis (per 100,000 live births). World Bank. https://data.worldbank.org/indicator/SH.TBS.INCD?locations=AM&view=chart (2019). 7.  Institute for Health Metrics and Evaluation World Bank. Armenia. World Development Indicators. 8.  Fraser, N. et al. Reforming the Basic Benefits Package in Armenia: Modeling Insights from the Health Interventions Prioritization Tool. (2021) doi:10.1596/35347. 9.  Nichols, G. A. "Medical Care Costs among Patients with Established Cardiovascular Disease." Am. J. Manag. Care (2010). 10.  Farrington, J. "Prevention and Control of Noncommunicable Diseases in Armenia: The Case for Investment." Inst. Health Metr. Eval. (2019). 11.  Hussain, A., Dulay, P., Rivera, M. N. & Saxena, V. "Neoplastic Pathogenesis Associated with Cigarette Carcinogens." Cureus (2019) doi:10.7759/cureus.3955. 12.  Wu, Y., Ding, Y., Tanaka, Y. & Zhang, W. "Risk Factors Contributing to Type 2 Diabetes and Recent Advances in the Treatment and Prevention." Int J Med Sci 1185–1200 (2014) doi:10.7150/ijms.10001. 13.  Boehme, A. K. "Stroke Risk Factors, Genetics, and Prevention Stroke Risk Factors, Genetics, and Prevention Effects of Neurologic Injury on Cardiovascular Function Vascular Cognitive Impairment." (2017) doi:0.1161/CIRCRESAHA.116.308398. 14.  Singletary, S. E. "Rating the Risk Factors for Breast Cancer." Ann Surg.  237(4): 474–482 (2003) doi:10.1097/01.SLA.0000059969.64262.87. 15.  Johnson, C. M. "Meta-analyses of colorectal cancer risk factors." Cancer Causes Control 24, 1207–1222 (2013). 16.  Andreasyan, D., Bazarchyan, A., Torosyan, A. & Sargsyan, S. "Prevalence of tobacco smoking in Armenia, STEPs survey." Tob Induc Dis A58 doi:https://www.euro.who.int/en/health-topics/ health-emergencies/coronavirus-covid-19/weekly-surveillance-report. 17.  Ryan, M. "Healthcare Access and Quality Index based on mortality from causes amenable to personal health care in 195 countries and territories, 1990-2015: a novel analysis from the Global Burden of Disease Study 2015." The Lancet 231–266 (2017). 18.  European Health Information Gateway: Health for All Explorer. World Health Organization (2019). Endnotes 42 19.  Social Snapshot and Poverty. (2019). 20.  World Development Indicators. Proportion of population spending more than 10% of household consumption or income on out-of-pocket health care expenditure (%). World Bank https://data.worldbank.org/indicator/SH.UHC.OOPC.10.ZS (2013). 21.  Out-of-pocket expenditure (% of current health expenditure) | Data. https://data.worldbank. org/indicator/SH.XPD.OOPC.CH.ZS. 22.  World Develoment Indicators. World Bank (2020). 23.  COVID-19 response financial report. https://www.moh.am/#1/3381. 24.  Sustainable Development Goals officially adopted by 193 countries. United Nations file:///C:/ Users/Seemi/Zotero/storage/TEB7B86U/United Nations. Sustainable Development Goals officially adopted by 193 countries. Retrieved (February 2, 2021) from http://www.un.org. cn/info/6/620.html (2021). 25.  Mkrtchyan, Z., Kharazyan, S. & Wiener, M. Costing of Universal Health Insurance (forthcoming). (2022). 26.  Government Finance Statistics. IMF https://data.imf.org/?sk=89418059-d5c0- 4330-8c41- dbc2d8f90f46. 27.  Gaspar, V., Jaramillo, L. & Wingender, P. Tax Capacity and Growth: Is There a Tipping Point? (2016). 28.  Coady, D. "Creating Fiscal Space." Int. Monet. Fund 55, (2018). 29.  Health taxes. Pan-American Health Organization https://www.paho.org/ en/topics/health- taxes. (2021). 30.  Petit, P. & Nagy, J. "How to Design and Enforce Tobacco Excises?" IMF Notes 2016, (2016). 31.  Hoe, C. H., Weiger, C. & Cohen, J. "Increasing Taxes on Tobacco": 30. 32.  Weiger, C., Hoe, C. & Cohen, J. E. "Seven-year tobacco tax plan in Ukraine: a case study of the actors, tactics and factors motivating policy passage." BMJ Open 12, e049833 (2022). 33.  Štelemėkas, M. et al. "Alcohol control policy measures and all-cause mortality in Lithuania: an interrupted time–series analysis." Addict. Abingdon Engl. 116, 2673–2684 (2021). 34.  2018 Lithuanian Draft Budgetary Plan. https://ec.europa.eu/info/sites/default/files/economy- finance/lt_dbp_2018_en.pdf (2017). 35.  "Saeima increases excise duty for sweetened non-alcoholic beverages." Latvijas Republikas Saeima (2020). 36.  Turati, F. et al. "Alcohol and liver cancer: a systematic review and meta-analysis of prospective studies." Ann. Oncol. 25, 1526–1535 (2014). 37.  Baliunas, D. O. et al. "Alcohol as a Risk Factor for Type 2 Diabetes: A systematic review and meta-analysis." Diabetes Care 32, 2123–2132 (2009). 38.  Koppes, L. L. J., Dekker, J. M., Hendriks, H. F. J., Bouter, L. M. & Heine, R. J. "Moderate Alcohol Consumption Lowers the Risk of Type 2 Diabetes." Diabetes Care 28, 719–725 (2005). 39.  Roerecke, M. & Rehm, J. "Alcohol consumption, drinking patterns, and ischemic heart disease: a narrative review of meta-analyses and a systematic review and meta-analysis of the impact of heavy drinking occasions on risk for moderate drinkers." BMC Med. 2014 121 12, 1–11 (2014). 40.  Reynolds, K. et al. "Alcohol consumption and risk of stroke: a meta-analysis." JAMA 289, 579–588 (2003). 43 THE IMPACT OF HEALTH TAXES IN ARMENIA 41.  Rehm, J. et al. "The relation between different dimensions of alcohol consumption and burden of disease: an overview." Addiction 105, 817–843 (2010). 42.  Reynolds, K. et al. "Alcohol Consumption and Risk of Stroke: A Meta-analysis". JAMA 289, 579–588 (2003). 43.  Global status report on alcohol and health 2018. https://apps.who.int/iris/ handle/10665/280114 (2018). 44.  Inchley, J. et al. Spotlight on adolescent health and well-being. Findings from the 2017/2018 Health Behaviour in School-aged Children (HBSC) survey in Europe and Canada. International report. Volume 1. Key findings. https://apps.who.int/iris/bitstream/hand le/10665/332091/9789289055000-eng.pdf (2020). 45.  Institute for Health Metrics and Evaluation. GBD Results Tool | GHDx. http://ghdx.healthdata. org/gbd-results-tool. 46.  Excise Duty Tables Part I Alcoholic Beverages July 2021. https://ec.europa.eu/taxation_ customs/system/files/2021-09/excise_duties-part_i_alcohol_en.pdf (2021). 47.  Neufeld, M. et al. "Alcohol control policies in Former Soviet Union countries: A narrative review of three decades of policy changes and their apparent effects." Drug Alcohol Rev. 40, 350–367 (2021). 48.  Fiscal Policies for Diet and Prevention of Noncommunicable Diseases. https://www. who.int/docs/default-source/obesity/fiscal-policies-for-diet-and-the-prevention-of- noncommunicable-diseases-0.pdf?sfvrsn=84ee20c_2 (2015). 49.  Armenia Overview Burden classification The Global Nutrition Report classifies this country as experiencing two forms of . 50.  Demirchyan, A. et al. "Assessment of nutrition practices of primary school-aged children in schools of Armenia Study Team." (2018). 51.  Ghazarian, S. Assessment of Sugar-Sweetened Beverage (SSB) Consumption and its Determinants among Armenian Adolescents in Yerevan. (2017). 52.  ЖУРНАЛ ЕВРОПЕЙСКОГО РЕГИОНАЛЬНОГО БЮРО ВОЗ. (2017). 53.  Rocha, C., Constante Jaime, P. & Rea, F. "How Brazil's Political Commitment to Nutrition Took Shape." In Global Nutrition Report - From promise to impact: ending malnutrition by 2030. (2016). 54.  Andreasyan, D. "The problem of being overweight among the Armenian population." Public Health Paranoma 3, (2017). 55.  Results. Global Burden of Disease Study (GBD) (2019) http://ghdx.healthdata.org/gbd- results-tool. 56.  "Armenian businesses protest tax hike on soda beverages." JAMnews. English Jamnews https://jam-news.net/armenian-businesses-protest-tax-hike-on-soda-beverages/ (2019). 57.  Aleksanyan, A. "Sugar-sweetened beverages excise tax implications in Armenia." Capstone Thesis Paper Alisa Aleksanyan. (2020). doi:10.13140/RG.2.2.32559.56489. 58.  Willi, C., Bodenmann, P., Ghali, W. A., Faris, P. D. & Cornuz, J. "Active Smoking and the Risk of Type 2 Diabetes: A Systematic Review and Meta-analysis." JAMA 298, 2654–2664 (2007). 59.  Lederle, F. A., Nelson, D. B. & Joseph, A. M. "Smokers' relative risk for aortic aneurysm compared with other smoking-related diseases: a systematic review." J. Vasc. Surg. 38, 329–334 (2003). Endnotes 44 60.  Sargsyan, S., Movsesyan, Y., Melkumova, M. & Babloyan, A. "Child and Adolescent Health in Armenia: Experiences and Learned Lessons." J. Pediatr. 177, S21–S34 (2016). 61.  Berg, C. J., Harutyunyan, A., Paichadze, N., Hyder, A. A. & Petrosyan, V. "Addressing cancer prevention and control in Armenia: tobacco control and mHealth as key strategies." Int J Equity Health 20, (2021). 62.  "WHO framework convention on tobacco control." Rev Esp Salud Publica 77, (2003). 63.  Tobacco Control Fact Sheet - Armenia. 4. 64.  WHO Technical Manual on Tobacco Tax Policy and Administration. https://apps.who.int/iris/ handle/10665/346227 (2021). 65.  Postolovska, I., Lavado, R., Tarr, G. & Verguet, S. Estimating the distributional impact of increasing taxes on tobacco products in Armenia: Results from an extended cost- effectiveness analysis. (2017). 66.  EAEU signed Agreement on the principles of tax policy in the field of excise duties on tobacco products. Eurasian Economic Commission http://www.eurasiancommission.org/ en/nae/news/Pages/24-12-2019-1.aspx (2019). 67.  WHO report on the global tobacco epidemic 2019: Offer help to quit tobacco use. https:// www.who.int/publications-detail-redirect/9789241516204 (2019). 68.  Ross, H. & Bakhturidze, G. Georgia: Controlling Illicit Cigarette Trade. 69.  President of the Republic of Armenia. Law on reduction and prevention of the damage caused to health by the use of tobacco products and substitutions for them. (2020). 70.  Armenia Details | Tobacco Control Laws. https://www.tobaccocontrollaws.org/legislation/ country/armenia/summary. 71.  National Statistical Service of the Republic of Armenia & World Bank. Armenia Integrated Living Conditions Survey (ILCS) 2016 Ref. ARM_2016_ILCS_v02_M. in (2016). 72.  Scarborough, P. et al. "Impact of the announcement and implementation of the UK Soft Drinks Industry Levy on sugar content, price, product size and number of available soft drinks in the UK, 2015-19: A controlled interrupted time series analysis." PLOS Med. 17, e1003025 (2020). 73.  Alsukait, R., Wilde, P., Bleich, S. N., Singh, G. & Folta, S. C. "Evaluating Saudi Arabia's 50% carbonated drink excise tax: Changes in prices and volume sales." Econ. Hum. Biol. 38, 100868 (2020). 74.  Hackshaw, A., Morris, J. K., Boniface, S., Tang, J.-L. & Milenković, D. "Low cigarette consumption and risk of coronary heart disease and stroke: meta-analysis of 141 cohort studies in 55 study reports." BMJ 360, j5855 (2018). 75.  Lubin, J. H. et al. "Risk of Cardiovascular Disease from Cumulative Cigarette Use and the Impact of Smoking Intensity." Epidemiol. Camb. Mass 27, 395–404 (2016). 76.  Barendregt, J. J., Van Oortmarssen, G. J., Van Hout, B. A., Van Den Bosch, J. M. & Bonneux, L. "Coping with multiple morbidity in a life table." Math. Popul. Stud. 7, 29–49 (1998). 77.  Veerman, J. L., Sacks, G., Antonopoulos, N. & Martin, J. "The Impact of a Tax on Sugar- Sweetened Beverages on Health and Health Care Costs: A Modelling Study. PLOS ONE 11, e0151460 (2016). 78.  Barendregt, J. J. & Veerman, J. L. Categorical versus continuous risk factors and the calculation of potential impact fractions." J. Epidemiol. Community Health 64, 209–212 (2010). 45 THE IMPACT OF HEALTH TAXES IN ARMENIA 79.  Murray, C. J. L. et al. "The State of US health, 1990-2010: Burden of diseases, injuries, and risk factors." JAMA - J. Am. Med. Assoc. 310, 591–608 (2013). 80.  Barendregt, J. J., van Oortmarssen, G. J., Vos, T. & Murray, C. J. "A generic model for the assessment of disease epidemiology: the computational basis of DisMod II." Popul. Health Metr. 1, 4 (2003). 81.  Roerecke, M. & Rehm, J. "Ischemic Heart Disease Mortality and Morbidity Rates in Former Drinkers: A Meta-Analysis." Am. J. Epidemiol. 173, 245–258 (2011). 82.  Roerecke, M. & Rehm, J." The cardioprotective association of average alcohol consumption and ischaemic heart disease: a systematic review and meta-analysis." Addiction 107, 1246–1260 (2012). 83.  O'Keeffe, L. M. et al. "Smoking as a risk factor for lung cancer in women and men: a systematic review and meta-analysis." BMJ Open 8, e021611 (2018). 84.  Ezzati, M., Henley, S. J., Thun, M. J. & Lopez, A. D. "Role of Smoking in Global and Regional Cardiovascular Mortality." Circulation 112, 489–497 (2005). Endnotes 46