WHO BEARS THE BURDEN OF FUEL TAXATION IN LATIN AMERICA AND THE CARIBBEAN COUNTRIES? WHO BEARS THE BURDEN OF FUEL TAXATION IN LATIN AMERICA AND THE CARIBBEAN COUNTRIES? Copyright © 2025, International Bank for Reconstruction and Development / World Bank 1818 H Street N.W. Washington D.C. 20433, United States of America Telephone: (202) 473-0000 Internet: www.worldbank.org In Spanish: www.bancomundial.org Email: feedback@worldbank.org Rights Reserved This volume is a product of the staff of the International Bank for Reconstruction and Development/ The World Bank. The findings, interpretations, and conclusions expressed in this volume do not necessarily reflect the views of the Executive Directors of the World Bank or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this publication. Rights and Permissions The International Bank for Reconstruction and Development/The World Bank encourages the dissemination of its work and will normally grant permission to reproduce portions of this work promptly, provided the sources are acknowledged. Attribution—Please cite the work as follows: World Bank (2025), Who Bears the Burden of Energy Taxation in Latin America and the Caribbean Countries? Cover and Interior Design Manthra Comunicación · info@manthra.ec Table of Contents Acknowledgements.............................................................................................4 Executive Summary..............................................................................................5 1. Introduction.......................................................................................................8 2. A Framework to Link Energy Taxation and Household Welfare....... 10 2.1. Fuel taxation affects the welfare costs of climate change.........13 2.2. Fuel taxation affects households as consumers and income earners.................................................................................14 3. Factors influencing the short-term welfare effects of fuel tax and subsidy reforms................................................................. 17 4. Assessing the short-term welfare and distributional implications of fuel tax reforms............................................................... 26 4.1. Methodology and data...........................................................................27 4.2. Ranking the impacts of fuel tax reforms: which fuels matter most and to whom?................................................................................28 4.3. The welfare and distributional effects of increasing and aligning tax rates across fuels based on their emissions..36 5. Concluding remarks and policy discussion............................................. 48 6. References...................................................................................................... 52 7. Annex A: Additional tables and figures.................................................. 57 WHO BEARS THE BURDEN OF FUEL TAXATION in Latin America and the Caribbean Countries? Acknowledgments This study was prepared by Ruth Llovet Montañés (Economist, ELCPV) and Carolina Mejía- Mantilla (Senior Economist, ELCPV) as a background study for the regional report titled “Taxing and Subsidizing Energy in Latin America and the Caribbean: Insights from a Total Carbon Price Approach” (World Bank 2025). Overall guidance was provided by Bill Maloney (Chief Economist, LCR, LCRCE), Oscar Calvo Gonzalez (Regional Director, ELCDR), and Carlos Rodríguez Castelán (Practice Manager, ELCPV). The core team included Lucía Echeverría (Consultant, ELCPV) and Desiree González (Senior Operations Assistant, ELCPV). Gustavo Canavire Bacarreza (Senior Economist, ELCPV) was co-TTL in the early stages of this project. The extended team included the following poverty and equity country teams: (1) Brazil: Gabriel Lara-Ibarra (Senior Economist, ELCPV) and Kajetan Trzcinski (Consultant, ELCPV); (2) Jamaica: Roy Katayama (Senior Economist, ELCPV) and Mikhail Matytsin (Data Scientist, EPVGE); (3) Mexico: Samuel Freije-Rodriguez (Lead Economist, ELCPV) and Mariel Cecilia Siravegna (Consultant, ELCPV); (4) Paraguay: Eliana Rubiano Matulevich (Senior Economist, ELCPV), Diego Tuzmán (Consultant, ELCPV), Gonzalo Rivera Gallegos (Consultant, ELCPV), and Victor Gamarra Florentin (Consultant, ELCPV); (5) Peru: Eliana Rubiano Matulevich (Senior Economist, ELCPV) and Gonzalo Rivera Gallegos (Consultant, ELCPV); and (6) Uruguay: Ruth Llovet Montañés (Economist, ELCPV) and Lourdes Rodríguez (Senior Economist, ELCPV). The team also benefited from inputs and comments by the Economic Policy team in Latin America and the Caribbean, including Daniel Navia Simon (Senior Economist, ELCMU), Anna-Maria Göth (Junior Professional Officer, ELCMU), and Ana Francisca Urrutia (Senior Economist, ELCMU). The team is grateful for comments received from Samuel Freije-Rodriguez (Lead Economist, ELCPV) and Eliana Rubiano Matulevich (Senior Economist, ELCPV). It is particularly grateful to the peer reviewers: Ruth Hill (Lead Economist, EPVGE), Guillermo Vuletin (Senior Economist, LCRCE), Kevin Carey (Program Manager, EFICT), and Carolyn Fischer (Research Manager, DECSI). The work was partially funded by the Global Tax Program and the Whole of Economy Trust Fund of the World Bank. 4 Executive Summary Understanding the welfare and distributional impacts of fuel tax and subsidy reforms is crucial for their effective implementation. Reforms that raise fossil fuel prices potentially lead to lower energy-related carbon dioxide (CO2) emissions and reduce the frequency and severity of climate-related hazards in the long term, thereby improving household well-being. However, in the short to medium term, they may decrease household purchasing power and affect livelihoods through the transformation of energy systems and the economy. A balanced approach that reflects a consideration of macroeconomic and environmental effects and equity concerns is crucial to creating the opportunity for successful reform, especially because these types of reforms are highly unpopular in the region. In the short run, the fuel tax reforms analyzed in this study generate a modest negative income shock among most households. Raising fuel taxes and aligning fuel tax rates with the cost of the associated emissions reduce household purchasing power. The effects depend on household spending patterns, country productive structures, and the initial configuration of the tax systems. Simulations on six LAC countries—Brazil, Jamaica, Mexico, Paraguay, Peru, and Uruguay—using the Commitment to Equity (CEQ) methodology indicate that these reforms would reduce per capita income by 0.1 percent to 1.2 percent on average, suggesting that the impacts would be moderate.1 Even the most ambitious reform that is considered (raising the carbon price uniformly to US$60 per metric ton of CO2 [tCO2] on all fuel emissions) would have negatively impacted household income less than the average annual inflation rate in these countries in 2014–19. (These estimates do not account for the behavioral response of households or the general equilibrium and long-term effects of fuel tax reforms.) While these effects are found to be larger among households at the bottom of the income distribution, the impacts on poverty rates (an increase) and on the middle class (a decline) do not exceed 0.6 percentage points. In the case of poverty, the change would be less than the typical annual fluctuations in the countries under study. Still, in populous countries such as Brazil and Mexico, this means that hundreds of thousands could fall into poverty and become vulnerable because of fuel tax reforms. Also, average income impacts mask important disparities. Thus, the effects are considerably larger among population groups that are more dependent on fuel use. Moreover, the estimated effects could be larger in countries with substantial fuel subsidies (such as Bolivia and Ecuador) or in the case of reforms that raise fuel tax rates to a higher benchmark (above US$60/tCO2) because such reforms would lead to larger price shocks. Public opposition to energy subsidy and tax reform is a key concern in most countries. The social and political economy dynamics of energy taxation are much more complex than these short-term income effects suggest and must be considered in designing the reforms. Although a social motivation may be behind misaligned carbon price signals across fuels, misaligned carbon prices are not an efficient policy tool for enhancing the purchasing power of the less well off. Social considerations are evident in the current configuration of fuel taxes and subsidies in the LAC region. Gasoline taxes are relatively more progressive, while taxes 1 For the CEQ methodology, refer to Lustig (2022a, 2022b). 5 WHO BEARS THE BURDEN OF FUEL TAXATION in Latin America and the Caribbean Countries? on liquefied petroleum gas (LPG) are more regressive, explaining why most LAC countries tax gasoline at a higher rate relative to LPG, consistent with equity objectives. Given this pattern and tax structure, the reforms that align tax rates with the carbon content of fuels result in income losses among households at the bottom of the distribution, whereas they benefit households at the top. Diesel taxes are also found to be regressive, and their effect can be as large as the effect of LPG taxes because of their impact on the production and distribution costs of a wide range of goods and services consumed by households. However, misaligned carbon prices (and the use of fuel subsidies in particular) are not the most adequate instrument to support the well-being of the less well off because a large share of the support goes to higher-income households, and they encourage the use of and investment in high-emission energy sources, delaying the transformation of energy systems in the region. Alternative instruments that preserve energy price signals and provide proper incentives, such as well- targeted social programs, are better at supporting poor and vulnerable households. Packaging fuel tax and subsidy reforms with compensation measures and accompanying policies in other areas might increase public support and help households transition to a low-carbon economy. Simulations in this study find that top-ups to existing social transfers can reduce income losses among the bottom 40 percent of households (the bottom 40) from fuel tax and subsidy reforms by 40 percent on average. However, the efficiency and effectiveness of such policies depend on the coverage and robustness of the social protection system. Moreover, experience with subsidy reform suggests that even the beneficiaries of compensation frequently do not lend credibility to the announced compensation schemes and therefore tend to overestimate the impact that they will endure. This reinforces the need to identify particularly affected groups and their characteristics to target compensation policies and the associated communication efforts more effectively. Because compensation policies are usually targeted at the lower end of the distribution, they do little to compensate other population groups and, in many situations, can be ineffective in reducing opposition to reform. This means reforms should be complemented by other policies that can garner support and reduce socioeconomic disparities in both the short term and the long term. These may include actions that enable households to adopt sustainable consumption and investment practices, such as increased public transport connectivity and improved accessibility, affordability, and reliability of electricity. There are also good arguments for designing policy packages around which consensus can be built for energy tax and subsidy reform by linking the reform to actions in other public policy areas, for example, additional investment in social protection, education, or health services. 6 Abbreviations and Acronyms CEQ Commitment to Equity CO2 carbon dioxide GDP gross domestic product LAC Latin America and the Caribbean LPG liquefied petroleum gas PPP purchasing power parity tCO2 metric ton of carbon dioxide 7 1 INTRODUCTION 1. Introduction This study explores the short-term welfare and distributive effects of selected fuel tax and subsidy reforms in six countries in the Latin America and Caribbean (LAC) region that are aimed at promoting the transformation of energy systems and have the potential to generate much-needed fiscal revenue.2 The analysis focuses on the immediate effects of higher fuel taxes (and lower subsidies) on household purchasing power in a partial equilibrium exercise. The exercise applies tools for the analysis of fiscal incidence (following the Commitment to Equity [CEQ] methodology; refer to Lustig 2022a, 2022b) that have been developed to account for the direct and indirect price effects of fuel taxes and subsidies in Brazil, Jamaica, Mexico, Paraguay, Peru, and Uruguay. The methodology consists in comparing household per capita income before and after fuel taxes and assuming that households have not yet adjusted their consumption choices. It relies on detailed household income and expenditure survey data and input-output tables for the quantification of the indirect price effects. The analysis explores the impact of taxes on various types of fossil fuels, namely, gasoline, diesel, liquefied petroleum gas (LPG), natural gas, kerosene, and ethanol.3 Alternative fuel tax policies are evaluated against a baseline scenario reflecting a country’s tax structure in a baseline year. In addition, the study examines how a compensation mechanism based on existing social protection programs can help partially offset these effects in each country. The evidence presented in this study is meant to inform and advance the dialogue on energy policy, domestic resource mobilization, and climate change mitigation in the region and on adapting the approach to the needs elsewhere. The analysis contributes to the literature on the equity trade-off around energy taxation in several ways. First, it compares the welfare and distributive impact of standardized fuel tax and subsidy reforms across several LAC countries, accounting in the process for baseline heterogeneities in tax rates across countries and in fuel types (and the resulting carbon price signals). Second, it models fuel tax and subsidy reforms accompanied by compensation policies based on existing social assistance programs in the respective countries to gauge the extent to which these measures might offset potential negative impacts in the short term. Third, it quantifies fuel-specific welfare and distributive effects. The rest of the study is organized as follows: Section 2 introduces a framework to clarify the link between energy tax policy and household welfare. Section 3 describes the factors influencing the short-term welfare impacts of fuel taxes, including the role of household energy expenditure patterns. Section 4 describes the methodology and the data used and the fuel tax and subsidy reform scenarios evaluated and presents the main findings from the welfare and distributional analysis. Section 5 concludes. 2 The analysis focuses on the impact of fuel taxes on monetary indicators of welfare. It does not consider other nonmonetary dimensions, such as the potential indirect health impacts arising from the use of firewood for cooking and heating, or traffic accidents from vehicle use. In this study, fuels refer exclusively to fossil fuels. 3 Electricity taxes and subsidies are outside the scope of this study because they are not typically based on differentiating among energy sources (high emission versus low emission), and estimation would require comprehensive cost and price information, which is not readily available in most countries. 9 WHO BEARS THE BURDEN OF FUEL TAXATION in Latin America and the Caribbean Countries? A FRAMEWORK 2 TO LINK ENERGY TAXATION AND HOUSEHOLD WELFARE 10 2. A Framework to Link Energy Taxation and Household Welfare Through their impact on energy prices, energy taxes are an important determinant of energy use and supply. While various factors shape consumer and firm decisions on energy use, empirical evidence shows that prices are the most important driver of energy-related choices. Given that taxes and subsidies have a substantial effect on energy prices, it is crucial to ensure that these fiscal instruments provide price signals that are in line with broader policy objectives, such as energy system transformation, fiscal sustainability, social equity, geopolitical strategy, and environmental sustainability. By increasing the relative price of fuel- intensive goods and services, energy taxes impact the energy intensity of the economy, shape the mix of energies used, and influence capital investment decisions related to energy, thereby incentivizing a shift toward more cost-effective and cleaner energy sources (Andersson 2019; Best, Burke, and Jotzo 2020; Murray and Rivers 2015). This study focuses on a specific type of energy fiscal policy: fossil fuel taxation. Fuel taxation here refers to the broad range of fiscal instruments that result in an increase in the price of fossil fuels relative to a fully neutral fiscal system (that is, one in which fuels are taxed exactly as any other good or service in the economy). Given the emissions content of fossil fuels, these price-based instruments explicitly or implicitly put a price on carbon dioxide (CO2) emissions generated by the combustion of fossil fuels.4 To capture the impact of the fiscal system on fuel prices fully, a total carbon price approach is applied. This approach considers a wide range of tax and expenditure instruments, such as explicit carbon taxes, fuel excise taxes, fuel subsidies, fees, and nonstandard value added tax rates, thereby providing a useful framework for the revision of energy fiscal policy.5 This methodology accounts for the combined effect of all relevant fiscal instruments and normalizes it based on the CO2 content of each fuel. Expressing the impact of fiscal instruments on fuel prices in terms of the associated CO2 emissions is crucial to generating country-level estimates and enabling cross- country, cross-fuel, and over time comparisons, but also to ensuring that these instruments capture the societal costs of emissions in a consistent way and to providing the adequate signals for the transformation of energy systems.6 Fuel taxes are important in mitigating climate change. Through their impact on prices, fuel taxes incentivize the adoption of low-carbon technologies, contributing to the reduction of energy-related greenhouse gas emissions. According to the World Bank (2025), a fiscal reform that aligns the carbon price of all fuel emissions to US$60 per metric ton of CO2 (tCO2) without applying revenue recycling measures could reduce energy-related emissions by 3 percent to 33 percent by 2030, depending on the country and relative to the baseline scenario of no reform. These reductions are significant and have the potential to alter the probability distribution of climate hazards over the long term, lowering both the frequency and severity 4 Although an important share of greenhouse gas emissions in LAC is associated with the agriculture, forestry, and other land use sector (45 percent), the energy sector is responsible for 44 percent of total greenhouse gas emissions, making this sector central to the region’s commitments under the Paris Agreement (IEA 2023). 5 There are many fiscal instruments influencing energy prices. Considering only some of them (formal carbon taxes or emis- sions trading schemes, for example) can lead to biased policy decisions. 6 Although fuel externalities extend beyond carbon emissions, using CO2 as a benchmark offers a clear, transparent measure. Also, the benchmark is directly aligned with low-carbon technological innovations and climate policy objectives, making it a practical standard for evaluating and addressing the broader impacts of fuel use. 11 WHO BEARS THE BURDEN OF FUEL TAXATION in Latin America and the Caribbean Countries? of climate-related events and thereby the impact of these events on household well-being. Reforms that align carbon prices across all fuels tend to exhibit better emission reduction results compared with those that do not because aligning fuel taxes with the carbon content of fuels provides a consistent price signal for emissions, regardless of the energy source that generates the emissions. Beyond its environmental and technological benefits, fuel taxation also has the potential to generate fiscal revenue. Increasing the economy-wide carbon price to US$60/tCO2, while aligning all fuel carbon prices to that level, could boost fiscal revenue by 0.2 percent to 1.0 percent of gross domestic product (GDP) in 2030 (World Bank 2025). This additional source of revenue is particularly important in the current context of budgetary constraints and fiscal stress faced by several LAC countries after the COVID-19 pandemic and the cost-of-living crisis (World Bank 2022b). The effectiveness of fuel taxation in generating revenue depends on various factors, including the tax rate, tax base, and the elasticity of demand for different energy sources (OECD 2021; Parry, Black, and Zhunussova 2022; World Bank 2022c). Moreover, at a given tax rate, tax revenues are expected to decline gradually as energy-use patterns adjust in response to higher carbon price signals (OECD 2019). However, fuel taxation can have significant implications for economic growth, which must be carefully considered. In the absence of revenue recycling measures, higher taxes on fuels and on fuel emissions may lead to a reduction in economic growth (World Bank 2025). In such cases, the critical question is whether the potential negative growth impacts of fuel tax reforms are manageable, given the additional revenue generated and the co-benefits in terms of emission reductions and other positive externalities. Recent analysis suggests that the impacts of these reforms on growth could be partially or fully offset if fiscal revenues are recycled (World Bank 2025). How and the extent to which the additional fiscal revenues are recycled play a crucial role in determining the overall macroeconomic impact, particularly on economic activity. Another key trade-off in fuel tax reform is the impact on household welfare and how the effect is distributed across population groups, which is the primary focus of this study. The following framework helps clarify the way fuel taxation may affect the welfare of households (refer to figure 1). First, the effect may take place through the impact of this instrument on the occurrence of climate-related events (hazards) and, consequently, on household exposure and vulnerability to the hazards. Second, the effect may take place through the impact on the sources and uses of household income, independent of climate factors or policy- induced changes in welfare.7 The design of fuel tax reforms must reflect these welfare and distributional effects, as well as the fact that individuals discount future benefits at different rates (Hill, Nguyen, and Doan 2024). 7 This is based on the hazard, exposure, and vulnerability framework (extensively used in the disaster risk literature and adapt- ed by Hill, Nguyen, and Doan 2024; refer to annex A, figure A.1). Hazard is the potential occurrence of a climate-related event; exposure refers to the people affected by the event; and vulnerability is the predisposition of these people to be ad- versely affected (including sensitivity to harm and lack of capacity to cope and adapt). 12 Figure 1. Framework for assessing the impacts of energy fiscal policy on house- hold welfare Hazard Exposure and vulneravility Policy - induced changes Potential occurrence of a People affected predisposition to in welfare independent climate-related event be adverseley affected of climate Welfare impacts of policy = x + Uses of income Sources of income Source: Adapted from Hill, Nguyen, and Doan 2024. 2.1 Fuel Taxation Affects the Welfare Costs of Climate Change In the long term, fuel taxation can improve household welfare by reducing the potential occurrence of climate-related events that threaten the lives and livelihoods of billions of people worldwide. The rise in the global average temperature has increased the intensity and frequency of extreme climate events—such as floods, droughts, landslides, wildfires, and tropical hurricanes—and altered weather patterns, affecting people and societies in diverse ways. Much of the impact takes place through the exacerbation of natural hazards, which cause significant destruction and loss of life, but there are also adverse effects on access to water, crop yields, livestock productivity, and so on.8 This affects agricultural earnings and food security. The impact of climate change has been uneven across various population segments within countries. The impact on the welfare and living conditions of poor households has been larger because these households mostly rely on natural capital for livelihoods in activities such as farming, pastoralism, and fishing (Birkmann et al. 2022; Brunckhorst et al. 2023; de la Fuente and Freije-Rodríguez 2024; Hill, Nguyen, and Doan 2024).9 In the LAC region, for example, 35 percent of poor workers are employed in the primary sector, compared with 10 percent among the nonpoor (refer to figure 2). Work outdoors is common in the occupations most prevalent among poor individuals, who are thus exposed to inclement weather that can adversely affect their health and well-being. In addition, the poor often live in homes that are built with substandard construction materials and that lack adequate insulation and weatherproofing, rendering the occupants more susceptible to extreme weather. On average, about 23 percent of the poor in LAC live in dwellings constructed with low-quality materials, compared with 9 percent among the nonpoor. Also, because they tend to reside in remote locations with limited access to markets, the prices of the goods they purchase are more likely to be affected by local weather events. 8 Felbermayr and Gröschl (2014) find that GDP per capita can drop by up to 7 percent because of severe natural hazards. A recent exam- ple is Hurricane Fiona, which devastated Puerto Rico in 2022. Fiona led to at least 22 deaths and caused an estimated US$2.5 billion in damage in Puerto Rico (nearly 3 percent of GDP), making it the third costliest hurricane on record there, after Maria (2017) and Georges (1998) (Pasch, Reinhart, and Alaka 2023). 9 For the LAC region, Vera, Uribe, and del Castillo (2023) find that climate events reduce the income of the poorest 40 percent of the pop- ulation (the bottom 40) by more than twice the average reduction experienced across the entire population. 13 WHO BEARS THE BURDEN OF FUEL TAXATION in Latin America and the Caribbean Countries? Figure 2. Selected characteristics of individuals living in poverty, LAC countries 100 90 83 % of individuals 80 70 60 60 50 50 40 38 39 35 30 23 20 18 10 10 9 3 5 0 Rural area Agricultural sector Hazardous location Low-quality material of dwelling of education: primary or less Does not receive employment-based pension benefits Highest level non-poor poor Source: Original figure for this publication based on data of SEDLAC (Socio-Economic Database for Latin America and the Caribbean), Center for Distributive, Labor, and Social Studies, Facultad de Ciencias Económicas, Universidad Nacional de La Plata, La Plata, Argentina, and Equity Lab, Team for Statistical Development, World Bank, Washington, DC, https://www.cedlas.econo.unlp.edu.ar/wp/estadisticas/sedlac/. Note: The figure shows the share of poor and nonpoor individuals who scored 1 in the selected categories. Statistics are reported for the LAC average, which includes 15 countries on which information is available. Data are circa 2022, excluding countries with no data after 2018. The work-related dimensions, such as sector of employment and employment-based benefits, are calculated only for workers ages 15–64. Individuals are classified as poor if their per capita income is below the upper-middle-income poverty line (US$6.85 a day in 2017 purchasing power parity [PPP] US dollars). Poor individuals are not only more exposed to climate change; they are also more vulnerable to the effects because of their limited coping and adaptive capacity. The poor tend to have limited access to financial resources (for example, savings, credit, or insurance) to respond to climate-related shocks. They are less likely to be covered by social insurance, and they are less likely to be able to switch livelihoods because of their lower educational attainment (Anik et al. 2018; Mohapatra 2012).10 For instance, the highest level of educational attainment among 60 percent of the poor is primary or less, while the corresponding share of the nonpoor is 39 percent (refer to figure 2). Moreover, some of the coping mechanisms relied on by poor households, such as withdrawing children from school or forgoing health services, might have long-term development consequences, thereby perpetuating a cycle of poverty. 2.2 Fuel Taxation Affects Households as Consumers and as Income Earners Energy taxes can also have welfare implications separate from the effects the taxes have on climate hazards, exposure, and vulnerability, that is, independent of climate impacts. 10 Social insurance is a type of social protection in which the claims of individuals are partly dependent on the contributions of the individuals. 14 The potential environmental and fiscal benefits of fuel tax reforms described above must be weighed against the welfare and equity implications. The climate benefits of decarbonization policies are realized over the long run, but these policies often entail short- and medium-term costs that vary depending on household characteristics. Fuel taxes affect households both as consumers (users of income) and income generators (sources of income) (Fullerton 2011; Goulder et al. 2019; Rausch, Metcalf, and Reilly 2011).11 Through its impact on energy use, productivity, and investment, fuel taxation is expected to contribute to the transformation of energy systems and the economy, which could have significant impacts on household income generation (the sources of income). It is expected to reduce the demand for high-emission products, while growing the demand for low-emission products, thereby affecting the structure of the economy, with implications for employment and livelihoods (Borissov, Vinogradova, and Bretschger 2019; Carbone et al. 2020). The welfare and distributive impacts of this transition depend on which sectors are affected the most, who works in these sectors, the quality of their jobs, the scope of skills transferability, and the capital intensity of new technologies. The literature shows that workers in contracting sectors have a different profile relative to workers in expanding sectors (Winkler et al. 2024; Hanson 2023; World Bank 2022a).12 Furthermore, the growth of low- emissions sectors may alter the relative returns to factors of production, such as land, labor, and capital, the impact of which on household income would depend on how these assets are distributed among the population.13 In the short term, fuel tax reforms result in increases in the price of fuels, negatively affecting household purchasing power (the uses of income) both directly and indirectly. In the immediate term, higher fuel prices will make consumption baskets more expensive, thereby reducing household purchasing power (Boyce 2018). Because of differences in income, living conditions, and consumption preferences, households are affected differently by the same fuel price shock. Evidence suggests that energy taxes disproportionately affect low-income households (Känzig 2023; Malerba, Gaentzsch, and Ward 2021; Missbach, Steckel, and Vogt-Schilb 2024). Even though wealthier households generally spend more on fuels, such expenditures represent a larger share of income among less affluent households (a direct effect). Fuel taxes will raise the price of other items in the consumption basket that depend for their production and distribution on these fuels (an indirect effect). For instance, in agriculture, diesel is used for a wide range of tasks, including irrigation, seeding, fertilization, and harvesting, thereby constituting a significant component of the cost structure. Diesel taxes thus likely affect the price of food, which exerts a negative effect especially at the bottom of the distribution. 11 Fuel taxes may affect nonmonetary dimensions of well-being, such as health, cognitive performance, and labor supply, by improving air quality. Reducing air pollution can lead to better health outcomes and human capital development, especially benefiting lower-income individuals and minority groups that are more highly exposed to pollution. 12 This illustrates the need for policy packages to support individuals at risk of job losses and job displacement. The packages might include compensatory transfers, insurance programs, and active labor market policies, such as reskilling, upskilling, job-searching, and job-matching programs (Hill, Nguyen, and Doan 2024). 13 For example, Beck et al. (2015) show that the introduction of a carbon tax in Canada causes real wages to decline because higher energy prices induce mobile capital to move out of energy-intensive sectors. In contrast, labor is less mobile and so bears the burden of the tax. Conversely, Mayer et al. (2021) find that decarbonization policies in Austria decrease the return on capital relative to labor. 15 WHO BEARS THE BURDEN OF FUEL TAXATION in Latin America and the Caribbean Countries? Still, over the medium term, the impact will depend on the household price elasticity of demand for fuel-intensive goods and services. Higher fuel prices will incentivize consumers to reduce spending on fuel-intensive items. However, their response will depend on multiple factors, such as their income level, their preferences, the availability and affordability of substitutes, and perceptions regarding the duration of the price shock. For instance, fuel substitution might prove challenging in the short term and will likely be uneven across households. The transition of households away from fossil fuels relies on the accessibility, affordability, and reliability of alternative energy sources, such as electricity (Acharya and Marhold 2019).14 This transition might require considerable upfront investments by households, such as switching to an electric heating system, cooking stove, or car, which would exacerbate the regressive nature of fuel taxes (Wang et al. 2016, 2019). If fuel taxes produce a broad inflationary effect, consumers may find themselves with limited alternatives for evading the impact of the price shock. Another important factor influencing price elasticity of demand is the perceived duration of the price shock. If households perceive the shock as temporary, they might be less inclined to adjust their consumption patterns than they would be if they viewed the shock as introducing permanent change. Empirical evidence on the size of these substitution effects and the related dynamics is insufficient in the LAC region. This study explores the short-term welfare and distribution effects of selected fuel tax and subsidy reforms in six LAC countries, focusing on the use of income. The following analysis highlights how higher fuel taxes (and lower subsidies) impact household purchasing power, assuming that household consumption patterns remain unchanged (without accounting for demand elasticity). 14 For the implementation of climate mitigation policies to be effective, it is crucial that the government devise policies to sup- port the shift toward renewable energy sources, including investments in infrastructure that enable the adoption of sustain- able behaviors by households. 16 3 FACTORS INFLUENCING THE SHORT-TERM WELFARE EFFECTS OF FUEL TAXES 17 WHO BEARS THE BURDEN OF FUEL TAXATION in Latin America and the Caribbean Countries? 3. Factors Influencing the Short-Term Welfare Effects of Fuel Taxes Fuel taxes are likely to have a negative impact on the purchasing power of households in the short run; the magnitude of their effect depends on several factors. The introduction of a fuel tax, the increase in the rate of such a tax, or an expansion of the fuel tax base is expected to lead to a rise in the price of fuels, which has an immediate effect on household purchasing power.15 However, this effect varies from household to household, depending on differences in income, living conditions, and spending. The resulting welfare and distributional impacts of fuel taxes depend on four main factors: (1) the reliance of households on fuels as energy sources, particularly households at the bottom of the income distribution; (2) the size of the fiscal-induced fuel price shock; (3) the intensity of fuel use in the production of goods and services consumed by households; and (4) the revenue recycling measures adopted by the government. Statistics on these dimensions can shed light on the potential impacts of fuel taxes. The impacts of fuel taxation are expected to be greater in countries in which households spend a larger portion of their income on fuels, such as Argentina, the Dominican Republic, Mexico, and Panama.16 Households in LAC spend an average of close to 4 percent of their incomes on fuels, with substantial heterogeneity across countries (refer to figure 3).17 The share is comparatively high in Panama (6.7 percent), Argentina (6.1 percent), and Mexico (5.3 percent), but considerably below average in Peru (1.7 percent), Ecuador (1.5 percent), and Bolivia (1.4 percent). If other energy sources, such as electricity, are considered, the average energy share in LAC increases to almost 6 percent of household income. In fact, in some countries, electricity accounts for a significant portion of total household energy expenditure, such as Bolivia, Colombia, Ecuador, Jamaica, Peru, and Uruguay (close to 50 percent). Transport fuels, especially gasoline, are important components of household fuel expenditure, and taxation on them could thus affect purchasing power comparatively more. In all countries, except El Salvador and Peru, the fuel with the highest income share is gasoline, accounting for an average of 2.4 percent of household income (refer to figure 3). As a share of household income, expenditure on other transport fuels, such as diesel and ethanol, is relatively small (except for diesel in Paraguay). For cooking and heating fuels, households spend an average of 0.7 percent of their incomes on LPG. Natural gas is a less common energy source for households in the region, except in Argentina, Colombia, and the Dominican Republic, where households spend more on natural gas than on LPG. The household income shares of other fuels, such as kerosene and fuel oil, are negligible in most countries. 15 All other factors affecting the retail price of fuels are assumed to remain constant after the tax change (for instance, profit margins, refining costs, and international oil prices). 16 The cross-country heterogeneities analyzed throughout this section may be partly driven by differences in the data (for ex- ample, the year of the household survey, the definition of income, the data collection method, and the type of fuels covered by the survey). Refer to annex A, table A.1 for a detailed list. 17 The statistics in this section refer to the subset of 15 countries on which detailed household expenditure data could be ob- tained (refer to annex A, table A.1). These countries accounted for 85.7 percent of the LAC population in 2022. 18 Figure 3. Energy expenditure relative to household income, by country Energy expenditure as a proportion of household income (%) 10 8 gasoline diesel LPG 6 natural gas kerosene ethanol 4 fuel oil other fuels 2 total average 0 ) ) 9) ) 8) 8) 2) 8) 2) 1) 7) 12 9) 17 2) 18 9) RY 2) /1 01 00 02 01 02 02 01 01 01 01 1/ 6/ 7/ 02 18 (2 (2 (2 (2 (2 (2 (2 (2 (2 01 (2 01 01 (2 20 M N L M EX L A Y (2 R (2 (2 CH V BO BR PR PA DO I( PE JA SL M U U L G CR CO EC AR Source: Original figure for this publication based on nationally representative household expenditure surveys. Note: Refer to annex A, table A.1 for further details. In El Salvador, gasoline expenditure is only reported for work-related activities. The dotted line shows the average share of fuel expenditure over household income across countries, which includes expenditure on gasoline, diesel, LPG, natural gas, kerosene, ethanol, and fuel oil. The category other fuels includes fuels not listed in the legend of the figure and, in most cases, refers to electricity. However, fuel expenditures might not completely reflect the household intensity of fuel use because they capture differences in the prices paid, in addition to the quantities purchased. A low share of household income dedicated to fuel expenses does not necessarily imply less reliance on these energy sources. It may merely reflect lower fuel prices. For instance, the comparatively modest income allocation for gasoline purchases among Bolivian and Ecuadorian households likely reflects the fact that fuels are strongly subsidized in these countries, rather than a low dependence on them.18 In addition to subsidies, other components of fiscal policy (for instance, fees and taxes) affect fuel retail prices, together with other nonfiscal factors, such as international crude oil prices, refining and distribution costs, regulatory frameworks, and profit margins. Within countries, fuel spending patterns vary considerably depending on the position of households in the income distribution. In absolute terms, expenditure on fuels in the LAC region is higher among the middle class and wealthier segments of the population (refer to figure 4). The richest 60 percent of households in the income distribution (the top 60) account for 81 percent of total fuel expenditure, on average, across the countries analyzed, compared with 19 percent among households in the bottom 40 percent (the bottom 40). However, fuel expenditure represents a higher share of the income of poor and vulnerable households (refer to figure 5). On average, fuel expenditure accounts for 4.4 percent of the income of households in the bottom 40, whereas it represents 3.5 percent among the top 60. This pattern is more pronounced in Brazil, Chile, and Colombia, where fuel expenditure for households in the bottom 40 and top 60 is, respectively, 6.9 percent of income versus 4.2 percent, 7.0 percent versus 4.1 percent, and 5.1 percent versus 2.2 percent. The opposite is true in Jamaica and Panama, where fuel expenditure represents a larger share of the income of wealthier households. 18 Fuel subsidies represent 2.7 percent and 3.7 percent of GDP in Bolivia and Ecuador, respectively, compared with a regional average of 0.6 percent (Black et al. 2023). 19 WHO BEARS THE BURDEN OF FUEL TAXATION in Latin America and the Caribbean Countries? Figure 4. Distribution of total household fuel expenditure, by income quintiles and country 100 15 Energy expenditure as a proportion Fuel expenditure by quintile (%) of household income (%) 80 10 60 40 5 20 0 0 ARG BOL BRA CHL COL CRI DOM ECU JAM MEX PAN PER PRY SLV URY Q1 Q2 Q3 Q4 Q5 Source: Original figure for this publication based on nationally representative household expenditure surveys. Note: Refer to annex A, table A.1 for further details. For each country, total household fuel expenditures include expenditures on gasoline, LPG, diesel, natural gas, kerosene, ethanol, and fuel oil. Individuals are ranked based on their per capita disposable income and grouped by income quintile (Q1–Q5). Figure 5. Energy expenditure relative to household income, by income group and country 15 Energy expenditure as a proportion of household income (%) 10 5 0 B40 T60 B40 T60 B40 T60 B40 T60 B40 T60 B40 T60 B40 T60 B40 T60 B40 T60 B40 T60 B40 T60 B40 T60 B40 T60 B40 T60 B40 T60 ARG BOL BRA CHL COL CRI DOM ECU JAM MEX PAN PER PRY SLV URY PAN PER PRY SLV URY gasoline natural gas fuel oil Q4 Q5 diesel kerosene other fuels LPG ethanol total Source: Original figure for this publication based on nationally representative household expenditure surveys. Note: Refer to annex A, table A.1 for further details. In El Salvador, gasoline expenditure is only reported for work-related activities. The category other fuels includes fuels not listed in the figure. In most cases, this refers to electricity. Statistics are presented separately for households in the bottom 40 percent of the per capita income distribution (B40) and households in the top 60 percent (T60). 20 The relative importance of fuels differs across households. Although the share of income that households allocate to fuel expenditure is greater among lower-income households than among higher-income households for most fuel types, the gap is particularly large in LPG spending (refer to figure 5). On average, the income share of LPG is 2.4 times larger among households in the bottom 40 than among households in the top 60, whereas there is no such gap in the case of gasoline. In Ecuador and Peru, the share of income spent on gasoline among households in the bottom 40 and the top 60 is similar, but it is greater among the latter group in the case of the Dominican Republic, Jamaica, Mexico, and Panama. For certain households, fuel spending can represent a considerable financial burden. The statistics cited above are based on averages across households. However, not all households purchase fuel, and, among those that do, they do not necessarily buy every type of fuel.19 As a result, the shares of fuel expenditure relative to income increase substantially if one focuses on the subset of households that report positive fuel expenditures (refer to figure 6). For instance, among the 20 percent of households in the bottom 40 in Colombia that report positive spending on gasoline, this expenditure represents 10.6 percent of their incomes (compared with 2.5 percent if one considers the average across all households in the bottom 40). This underscores the importance of identifying the most highly affected households and their characteristics in thinking about implementing targeted compensation measures to offset the impacts of fuel taxation or in designing programs to encourage the adoption and use of cleaner energy alternatives. The magnitude of fuel price shocks resulting from fuel tax and subsidy reforms is a second important determinant of the income impacts on households, and it varies across countries and fuel types. LAC governments differ substantially in their approach to taxing fuels, leading to notable disparities in fuel prices and costs associated with CO2 emissions. Using the total carbon price approach, the World Bank (2025) shows that, after accounting for subsidies, the net effect of taxes creates a positive wedge in the price of fuel-related CO2 emissions in most LAC countries, estimated at US$37.50/tCO2 in 2024. Still, these estimates are far lower than the estimated cost of the damage caused by the emission of a ton of CO2 and insufficient to foster the transformation of energy systems and meet the targets set by the Paris Agreement. Also, there is wide dispersion across countries in carbon prices. In 2023, the carbon price of fuel emissions was negative, at −US$20/tCO2, in Colombia, compared with a positive price of US$51.47/tCO2 in Uruguay. Heterogeneous carbon prices across fuels generate inconsistent price signals for economic agents, affecting energy use and supply decisions. Gasoline emissions are typically taxed at higher rates compared with diesel emissions, while emissions from natural gas and LPG are often untaxed or subsidized. These patterns suggest that fuel tax and subsidy reforms aimed at increasing and aligning carbon prices across fuels will result in different price shocks across countries and fuel types. Overall, households in countries with lower taxation of fuel-related CO2 emissions can be expected to face steeper fuel price hikes as a result of the reforms. 19 LAC households are more likely to purchase LPG than any other fuel type (refer to annex A, figure A.2), except in Colombia and the Dominican Republic, where households tend to use natural gas instead of LPG for cooking and heating. On average, 56.2 percent of households report positive expenditures on LPG, compared with only 33.0 percent for gasoline. 21 WHO BEARS THE BURDEN OF FUEL TAXATION in Latin America and the Caribbean Countries? Figura C3 Figure 6. Gasoline and LPG expenditure relative to income, households reporting positive fuel expenditure, by income group and country Energy expenditure as a proportion of household income (%) Energy expenditure as a proportion of household income (%) 15 15 Gasoline 10 10 5 5 0 0 T60 B40  B40  T60 B40  T60 B40  T60 B40  T60 B40  T60 B40  T60 B40  T60 B40  T60 B40  T60 B40  T60 B40  T60 B40  T60 B40  T60 B40  T60 B40  T60 B40  T60 B40  T60 B40  T ARG BOL BRA CHL COL CRI DOM ECU JAM MEX PER PRY SLV URY ARG BOL BRA CHL CO Energy expenditure as a proportion of household income (%) 15 LPG 10 5 0 T60 B40  0  T60 B40  T60 B40  T60 B40  T60 T60 B40  B40  T60 B40  T60 B40  T60 B40  T60 B40  T60 B40  T60 B40  T60 B40  T60 B40  T60 B40  T60 B40  T60 B40  T60 B40  T60 B40  T60 MEX PER PRY SLV URY ARG BOL BRA CHL COL CRI DOM ECU JAM MEX PAN PER PRY SLV URY Source: Original figure for this publication based on nationally representative household expenditure surveys. Note: Refer to annex A, table A.1 for further details. The statistics are calculated for the subset of households that report positive expenditures in each fuel. In the case of gasoline expenditures, Panama is not included because of the limited number of observations at the bottom of the distribution, which affects the quality of the estimates. Statistics are presented separately for households in the bottom 40 percent of the per capita income distribution (B40) and households in the top 60 percent (T60). 22 ura C2 Figure 7. Share of informal fuel spending in total fuel spending, by fuel type, income group, and country Informal purchases as a proportion of total spending (%) Informal purchases as a proportion of total spending (%) 30 40 27.1 33.3 30 20 21.7 15.2 20 16.6 17.7 13.8 10.8 12.1 10.4 10.9 10 8.3 10 7.5 5.9 5.6 4.84.5 4.9 3.2 3.8 2.8 2.7 1.6 0.4 0 B40  T60 B40  T60 B40  T60 B40  T60 0 ARG COL MEX PER ARG COL MEX PER gasoline diesel LPG Source: Original figure for this publication based on nationally representative household expenditure surveys. Note: Refer to annex A, table A.1 for further details. The definition of informal purchases is based on the location and payment method employed in the purchase and varies by country. Statistics are presented separately for households in the bottom 40 percent of the per capita income distribution (B40) and households in the top 60 percent (T60). In countries in which informal fuel markets (which evade taxes) are prevalent, the effects of fuel taxation on prices would be somewhat subdued. Informal fuel purchases, defined by the place of purchase and payment method used, are nontrivial in several countries, representing, for example, 15, 27, and 11 percent of total household LPG spending in Argentina, Colombia, and Mexico, respectively (refer to figure 7).20 Informal purchases tend to be more prevalent among households in the bottom 40. Thus, one would expect the (direct) price shock from fuel taxation to be subdued in these cases. Higher fuel prices can affect the purchasing power of households through a third channel, that is, by increasing the cost of nonenergy goods and services the production or transportation of which depend on the fuels. Fuel taxes not only raise the price of fuels directly subject to the taxes, but also the price of nonenergy goods and services that depend on the fuels for their production or distribution.21 These indirect price effects are expected to be larger in countries in which productive structures rely heavily on fuels and for fuel-intensive goods and services, such as food products and passenger transport services. In the LAC region, food spending represents an average of 19.1 percent of household income. Consequently, fuel tax and subsidy reforms could have considerable indirect effects on household purchasing power (refer to annex A, figure A.3). This is especially true for households in the bottom 40, which spend 2.3 times more on food items relative to the top 60.22 Similarly, households allocate an average of 3.9 percent of their incomes on transport services. The indirect impacts may thus outweigh the direct impacts, especially in the case of diesel and fuel oil, which are used widely in the productive sector. 20 Informal purchases are here defined as those purchases made in street markets or other informal country-specific locations, unless the purchases are paid for using credit or debit cards, bank transfers, or bank deposits. 21 In this study, any tax-induced rise in fuel prices is assumed to be fully passed through to end consumers. 22 Food spending is particularly important among poorer households in Bolivia, Panama, Paraguay, and Peru. 23 WHO BEARS THE BURDEN OF FUEL TAXATION in Latin America and the Caribbean Countries? Compensatory measures can help reduce the negative short-term impacts of fuel price rises on household purchasing power. How revenues from energy tax reforms are used has a first-order impact on welfare and distributional outcomes (Alonso and Kilpatrick 2022; Alvarez 2019; Garaffa et al. 2021; García-Muros, Morris, and Paltsev 2022; Goulder et al. 2019; Missbach, Steckel, and Vogt-Schilb 2022; Tovar-Reaños and Lynch 2019). The potential additional fiscal resources generated by these reforms could be used, for example, as follows: (1) to reduce other taxes and contributions, such as labor taxes; (2) to provide direct assistance to households by expanding well-targeted public transfers; (3) to raise spending on public services (for instance, education, health care, security); or (4) to design programs to facilitate the energy transition (such as programs for the purchase of electric stoves), particularly among poor and vulnerable households.23 Such policies may reduce the overall burden of the reforms on households, while leaving the price signal to reduce emissions unaffected. Still, depending on the macroeconomic and fiscal context, governments may choose not to recycle the additional fiscal revenues and instead use them for fiscal consolidation. The overall welfare and distributive effects of higher fuel prices would therefore also depend on the compensation measures adopted. Governments with well-functioning social protection programs could, for example, scale up these programs by increasing their generosity or expanding their coverage. Both the efficacy and efficiency of such compensation mechanisms would depend on the initial design and implementation of social protection programs in each country. Countries such as Bolivia, Brazil, and Uruguay are more well positioned to alleviate welfare losses from fuel taxation among the poor through top-ups to existing transfers programs because of the wide coverage of the poor (defined based on the national poverty line), respectively, 80, 57, and 70 percent (refer to figure 8). However, in the case of Bolivia, social protection programs also reach a large share of the nonpoor population, thereby raising the cost of compensation.24 At the other extreme, social transfers in El Salvador reach only 1.3 percent of the poor. 23 Approach 1 is commonly known as the double-dividend approach because of its potential to reduce emissions (first dividend) and other, more distortionary taxes that slow economic growth (second dividend). Yet, this may not be the most effective mechanism to compensate low-income households, which are more likely to engage in informal jobs, and their incomes tend to fall below the minimum taxable threshold. 24 This is explained by Bolivia’s categorical transfer benefits, which are distributed according to demographic criteria. 24 Figure 8. Share of the poor and nonpoor in households with a main social assistance program beneficiary 100 80 % of non-poor individuals 60 BOL 40 CHL JAM 20 ECU DOM PRY CRI BRA URY PER PAN ARG COL MEX SLV 0 0 20 40 60 80 100 % of poor individuals Source: Original figure for this publication based on nationally representative household expenditure surveys. Note: Refer to annex A, table A.1 for a detailed list of the main transfer programs considered in each country. Individuals are classified as poor or non-poor based on household income per capita and the corresponding national poverty line. 25 WHO BEARS THE BURDEN OF FUEL TAXATION in Latin America and the Caribbean Countries? ASSESSING THE 4 SHORT-TERM WELFARE AND DISTRIBUTIONAL IMPLICATIONS OF FUEL TAX REFORMS 26 4. Assessing the Short-Term Welfare and Distributional Implications of Fuel Tax Reforms The factors outlined above interact and collectively determine the overall impact of fuel tax and subsidy reforms on household welfare in the short term. The following analysis uses a bottom-up approach that relies on detailed household income and expenditure survey data to consider these factors together and assess the short-term welfare and distributive effects of selected fuel tax policy scenarios. 4.1 Methodology and Data The CEQ methodology offers a widely recognized framework for assessing the distributional impacts of fiscal policies, including fuel taxes and subsidies. The methodology is based on an accounting approach that consists of allocating taxes, contributions, government transfers, subsidies, and public spending to households to allow for a comparison between household per capita income before fiscal interventions (prefiscal income) and after fiscal interventions (postfiscal income) (Lustig 2022a, 2022b).25 The analysis in this study expands on existing fiscal microsimulation tools for Brazil, Jamaica, Mexico, Paraguay, Peru, and Uruguay to model the direct and indirect price effects of fuel tax and subsidy reforms by type of fuel.26 Following the CEQ approach, three household income concepts are constructed (market, disposable, and consumable). Comparisons across income concepts provide information on how various fiscal policies affect the distribution of income and, thus, poverty and inequality. Market income primarily comprises labor and capital income, as well as private transfers, serving as the starting point for constructing the other income concepts.27 Disposable income is derived by adding government cash or near-cash transfers to market income and then subtracting direct taxes and contributions (refer to annex A, figure A.4). Consumable income is defined as disposable income, plus indirect subsidies, minus indirect taxes on goods and services (for instance, fuel taxes). The analysis here focuses on estimating changes in consumable income per capita across several policy reform scenarios at a baseline year.28 This is done because consumable income per capita is the income concept immediately affected by fuel tax policy. The estimates presented here also account for the indirect price effects of fuel taxes on other goods and services. To accomplish this, the cost-push model is applied using the latest available national input-output tables (Inchauste et al. 2024). 25 The microeconomics literature proposes several methods to measure the welfare losses derived from price changes, such as consumer surplus variation, compensating variation, equivalent variation, Laspeyres variation, and Paasche variation (Araar and Verme 2016). The methodology of this study follows a Laspeyres index, that is, welfare losses refer to the exact change in income necessary to purchase, after the price variation, the same bundle of goods purchased before the price variation. This method only requires knowledge of initial quantities and price changes and is thus especially useful if there is a lack of information on utility, the demand system, and elasticities. For example, refer to Labandeira, Labeaga, and Rodríguez (2009), Renner, Lay, and Greve (2018), and Tovar-Reaños and Lynch (2019) for alternative methodological approaches to measuring the household welfare impacts of energy taxes. 26 Refer to annex A, table A.2 for a list of fiscal interventions included in each country’s fiscal microsimulation tool. 27 Under the CEQ framework, there are two options for treating contributory old-age pensions. In this study, pensions are treat- ed as deferred income. 28 Throughout this study, disposable income per capita is used to rank individuals. This approach ensures that the relative position of individuals in the income distribution remains unaffected by changes in fuel taxes and subsidies and is thus com- parable across reform scenarios. 27 WHO BEARS THE BURDEN OF FUEL TAXATION in Latin America and the Caribbean Countries? The analysis relies on individual and household income and expenditure data and a detailed revision of country tax codes and social protection systems. Taxes and transfers are assigned to households based on information from a nationally representative household socioeconomic survey (refer to annex A, table A.2). In countries in which detailed information on household income and expenditure for the baseline year is not available in the same survey, such as in Paraguay and Uruguay, product-level expenditure from the expenditure survey is imputed to the income survey using distribution matching, a survey-to-survey imputation approach.29 This involves ranking households in both surveys based on a common variable that is strongly correlated with expenditure (such as household income per capita) and mapping the households according to their relative position in the distribution. This method implicitly assumes that household spending patterns across the income distribution remain constant. Despite the relevance of the analysis to the design of energy fiscal policy, there are limitations. First, while efforts have been undertaken to harmonize approaches across countries, differences in estimated impacts could still be driven by factors, such as the year of the baseline model, the underlying household survey data, and other fiscal interventions included in the model. Second, the analysis covers a point in time and only considers first- order effects. It does not account for the behavioral response of households to tax-induced fuel price changes. This is mainly because of the lack of estimates of cross-price elasticities across countries and the income distribution. Nor does it account for general equilibrium and long-term effects (such as changes in employment). Third, it does not adequately represent certain population groups, particularly individuals who are difficult to reach and the wealthiest households, who are often underrepresented in household surveys. Fourth, the methodology focuses exclusively on households, not firms, as the unit of analysis. 4.2 Ranking the Impacts of Fuel Tax Reforms: Which Fuels Matter Most and to Whom? To achieve a better understanding of fuel-specific welfare and distributional impacts of comprehensive fuel tax reforms, an illustrative exercise is first undertaken to model a uniform price increase in all fuels. Comprehensive fuel tax reforms can affect fuel prices in different ways depending on initial heterogeneities in tax rates (refer to annex A, table A.3). For example, gasoline tends to be overtaxed relative to other fuels, whereas LPG tends to be undertaxed. This implies that reforms that increase and align tax rates with the CO2 content of fuels would result in a larger price rise for LPG compared with gasoline. Not only do price shocks differ across fuel types and countries, but household spending patterns also vary across these dimensions. To understand the overall effects of fuel tax reforms, a hypothetical increase in taxes on all fuel-related emissions is simulated. In this illustrative scenario, the increase in taxes (or reduction in subsidies) results in a homogeneous 25 percent rise in the price of each fuel.30 Next, the effects of four standardized reforms increasing and aligning carbon prices across fuels are assessed. 29 Refer to Corral Rodas, Paul Andres. dmatch. Released March 6, 2023. https://github.com/pcorralrodas/dmatch. 30 For reference, the simulated fuel price shock is smaller than the increase in international oil prices that occurred between 2021 and 2022 (EIA 2023). 28 A tax-induced rise by 25 percent in fuel prices negatively impacts household income in the short term. A simultaneous 25 percent increase in all fuel prices, associated with higher taxes on emissions, is estimated to lead to a reduction in household per capita income by 2 percent on average in the LAC countries analyzed. This is equivalent to about half the impact of average annual inflation on income in these countries in 2014–19 (refer to annex A, figure A.5). This impact varies by country, fuel type, and household position in the income distribution. A comparison across countries reveals that average per capita income declines by 2.8 percent in Brazil and by 1.3 percent in Peru and Uruguay. Among fuels, the effect of gasoline taxes is typically larger because of the higher share of household spending on gasoline (except in Jamaica and Peru), followed by diesel and LPG (refer to figure 9). For example, a 25 percent price rise in gasoline in Brazil leads to an average per capita income loss of 1.2 percent (representing close to 40 percent of the 2.8 percent total loss). Although diesel is not used by most households directly, diesel taxes have a general inflationary effect because they raise the production and distribution costs of a wide range of goods and services (refer to figure 10). This indirect price effect can be as large as or even larger than the impact of taxes on other fuels extensively purchased by households, such as LPG. The indirect effects are particularly large in Brazil, Jamaica, and Paraguay, suggesting that the economic structures of those countries are more dependent on fuels compared with other countries.31 Taxes on natural gas have a negligible impact on income in Mexico and Uruguay, whereas their effect is more pronounced in Peru due to the role of this energy source in the country productive structure. 31 In the case of Jamaica, diesel and fuel oil are mostly used in bauxite mining and alumina production and in electricity, water supply, and transport services. In Brazil, the indirect effects are primarily driven by the transport, commerce, and agriculture sectors. Differences in the indirect effects across countries may arise partly because of the reference year of the input-output table applied and the assumptions made in mapping fuels to sectors. 29 WHO BEARS THE BURDEN OF FUEL TAXATION in Latin America and the Caribbean Countries? Figure 9. Short-term income effects of a 25 percent fuel price increase, by country and fuel type Brazil Jamaica Average change in consumable income per capita (%) Average change in consumable income per capita (%) 0 0 -.5 -.5 -1 -1 -1.5 -1.5 -2 -2 5 15 25 35 45 55 65 75 85 95 5 15 25 35 45 55 65 75 85 95 Percentile of disposable income per capita Percentile of disposable income per capita gasoline diesel gasoline diesel LPG ethanol LPG kerosene Mexico Peru Average change in consumable income per capita (%) Average change in consumable income per capita (%) 0 -.5 0 -.5 -1.5 -1 -1 5 15 25 35 45 55 65 75 85 95 5 15 25 35 45 55 65 75 85 95 Percentile of disposable income per capita Percentile of disposable income per capita gasoline diesel gasoline diesel LPG natural gas LPG natural gas Paraguay Uruguay Average change in consumable income per capita (%) 0 0 Average change in consumable income per capita (%) -.5 -.5 -1 -1 -1.5 -1.5 5 15 25 35 45 55 65 75 85 95 -2 5 15 25 35 45 55 65 75 85 95 Percentile of disposable income per capita Percentile of disposable income per capita gasoline diesel gasoline diesel LPG kerosene LPG natural gas Source: Original figure for this publication based on country-specific microsimulations. Note: The figure shows the estimated percentage change in per capita consumable income between the baseline and the 25 percent price increase scenarios for each country and fuel. The results are presented as the average across households for each percentile of the per capita disposable income distribution. Individuals in the bottom and top 5 percent of the disposable income distribution are excluded because of large variations in the relevant data. 30 Figure 10. Direct and indirect income effects of a 25 percent fuel price rise, by fuel type and country Gasoline Average change in consumable income per capita (%) Diesel Average change in consumable income per capita (%) 0 0 -.5 -.5 -1 -1 -1.5 -1.5 -2 Brazil Jamaica Mexico Peru Uruguay Paraguay Brazil Jamaica Mexico Peru Uruguay Paraguay direct price effect indirect price effect direct price effect indirect price effect Average change in consumable income per capita (%) LPG Average change in consumable income per capita (%) Natural gas 0 0 -.5 -.5 -1 -1 -1.5 -1.5 Brazil Jamaica Mexico Peru Paraguay Peru direct price effect indirect price effect direct price effect indirect price effect Ethanol Average change in consumable income per capita (%) 0 -.5 -1 -1.5 Brazil direct price effect indirect price effect Source: Original figure for this publication based on country-specific microsimulations. Note: The figure shows the decomposition of the income effect of each 25 percent fuel-specific price rise into (a) the income change deriving from direct fuel purchases and (b) the change deriving from spending on fuel- intensive nonenergy goods and services. Countries are excluded if data are not available on both the direct and indirect income effects. This could occur, for instance, if the fuel is not identified in a country’s input-output table. 31 WHO BEARS THE BURDEN OF FUEL TAXATION in Latin America and the Caribbean Countries? Fuel taxes are largely paid by the middle class and wealthier segments of the population, but their impacts disproportionately fall on less well off households. Households in the top income quintile are estimated to contribute between 40 percent and 50 percent of additional tax revenue because they spend more on fuel-intensive goods, compared with a contribution of 6 percent to 9 percent among households in the bottom quintile (refer to figure 4; annex A, table A.4). Fuel taxes account for a larger share of the income of the poor and vulnerable. This is particularly true of LPG because of the greater reliance on this energy source for cooking among lower-income households (except in Jamaica and Paraguay) (refer to figure 5). The burden of gasoline taxes is similar across the distribution and may be even larger among higher-income households, such as in Jamaica and Mexico. Among all fuels, LPG taxes tend to have the most regressive impact on household income. For example, a 25 percent rise in LPG prices (because of higher taxes on the related emissions) would result in an average per capita income loss of 0.6 percent among individuals in the bottom quintile in Peru, compared with 0.2 percent among individuals in the top quintile. Conversely, in the case of gasoline taxes in Peru, the decline is estimated at 0.2 percent and 0.3 percent among individuals in the bottom and top quintiles, respectively.32 Similar to LPG taxes, diesel taxes have a larger impact on low-income households (except in Jamaica). The cause is attributed to the use of diesel in the production and distribution of goods, such as food, that account for a larger share of the income of the poor (refer to annex A, figure A.3). The Kakwani progressivity index is negative and largest for LPG taxes, whereas it is positive (Jamaica and Mexico) or near zero (Uruguay) for gasoline taxes (refer to table A.5).33 Thus, except in Brazil and Paraguay, gasoline taxes have the lowest impact on inequality relative to other fuel taxes and even reduce inequality in Jamaica, Mexico, and Peru (refer to figure 11).34 In assessing overall tax progressivity, it is crucial to account for both the size and distribution of informal purchases because these evade taxes and can therefore affect the tax burden faced by households across the distribution differently (refer to box 1). 32 These results are consistent with Beylis and Cunha (2017), who apply the price gap approach to estimate fuel subsidies in several LAC countries and simulate the welfare and distributional impacts of an absolute price increase (US$0.25 per unit of energy) in each energy source (gasoline, diesel, LPG, and electricity). 33 The Kakwani index is a measure of the progressivity of a fiscal intervention. It is calculated as the difference between the con- centration index for the tax and the Gini coefficient for pretax income. A positive value indicates progressivity, while a negative value indicates regressivity. A value of zero suggests proportionality, meaning that the tax burden is distributed in proportion to income. 34 From an equity perspective, these patterns align with the rationale of policymakers in levying higher taxes on gasoline, while undertaking reduced taxes or even subsidies on LPG. More efficient and environmentally friendly, alternative fiscal instru- ments exist to provide more well targeted support to poor and vulnerable households. 32 Figure 11. Effect of a 25 percent fuel price rise on inequality, poverty, vulnerability, and the middle-class, by country and fuel type income) income) Brazil Jamaica 1.2 income) .6 income) Brazil Jamaica on consumable on consumable 1.2 .6 .8 .4 on consumable on consumable income) income) .8 .4 1.2 .4 .2.6 .2 Brazil Jamaica (based(based (based(based .4 0 0 on consumable on consumable 0.8 0.4 points points .4 -.4 .2 -.2 points points -.4 -.2 0 -.8 0 -.4 in percentage in percentage (based (based -.4 -1.2 -.8 -.4 in percentage in percentage -.2 -.6 Poverty Vulnerable Middle-class Gini Poverty Vulnerable Middle-class Gini points points -.8 -1.2 -.4 -.6 Change Change 25% price increase of gasoline 25% price increase of diesel 25% price increase of gasoline 25% price increase of diesel Poverty Vulnerable Middle-class Gini Poverty Vulnerable Middle-class Gini 25% price increase of LPG 25% price increase of ethanol 25% price increase of LPG 25% price increase of kerosene in percentage in percentage Change Change 25% price increase of gasoline 25% price increase of diesel -1.2 25% price increase of gasoline 25% price increase of diesel 25% price increase of LPG 25% price increase of ethanol 25% price increase of LPG 25% price increase of kerosene -.6 Poverty Vulnerable Middle-class Gini Poverty Vulnerable Middle-class Gini income) income) Mexico Peru Change Change 25% price increase of gasoline 25% price increase of diesel 25% price increase of gasoline 25% price increase of diesel .4 income) .6 income) 25% price increase of LPG 25% price increase of ethanol 25% price increase of LPG 25% price increase of kerosene Mexico Peru on consumable on consumable .2 .4 .6 .4 on consumable on consumable income) income) .4 Peru .2 Mexico 0.4 .2 (based(based 0.6 (based(based .2 on consumable on consumable 0.4 points .2 -.2 points .2 0 -.2 points -.2 points 0 -.4 in percentage in percentage -.40 -.2 (based (based -.4 in percentage -.2 -.6 in percentage Poverty Vulnerable Middle-class Gini Poverty Vulnerable Middle-class Gini points points -.2-.4 -.4 -.6 Change 25% price increase of gasoline 25% price increase of diesel Change 25% price increase of gasoline 25% price increase of diesel Poverty Vulnerable Middle-class Gini Poverty Vulnerable Middle-class Gini 25% price increase of LPG 25% price increase of natural gas 25% price increase of LPG 25% price increase of natural gas in percentage in percentage Change Change 25% price increase of gasoline 25% price increase of diesel 25% price increase of gasoline 25% price increase of diesel 25% price increase of LPG 25% price increase of natural gas 25% price increase of LPG 25% price increase of natural gas -.4 -.6 income) Poverty Vulnerable Middle-class Gini Poverty Vulnerable Middle-class Gini income) Paraguay Uruguay Change 25% price increase of gasoline 25% price increase of diesel Change 25% price increase of gasoline 25% price increase of diesel income) income) .4 .6 25% price increase of LPG Paraguay 25% price increase of natural gas 25% price increase of LPG 25% price increase of natural gas on consumable Uruguay on consumable .2 .4 .6 .4 on consumable on consumable income) income) Paraguay .4 Uruguay .2 0 .2 (based(based .4 (based(based 0.6 on consumable .2 on consumable points .2 0 0.4 points .2 -.2 -.2 points in percentage points -.2 -.4 in percentage 0 -.2 (based -.4 (based in percentage 0 -.4 in percentage -.2 -.6 Poverty Vulnerable Middle-class Gini -.2 -.4 points Poverty Vulnerable Middle-class Gini points 25% price increase of gasoline 25% price increase of diesel -.4 -.6 Change Poverty Vulnerable Middle-class Gini Change 25% price increase of gasoline 25% price increase of diesel 25% price increase of LPG 25% price increase of kerosene Poverty Vulnerable Middle-class Gini Change in percentage 25% price increase of LPG natural gas 25% price increase of gasoline 25% price increase of diesel Change in percentage Change Change 25% price increase of gasoline 25% price increase of diesel 25% price increase of LPG 25% price increase of kerosene -.4 25% price increase of LPG 25% price increase of natural gas -.6 Poverty Vulnerable Middle-class Gini Poverty Vulnerable Middle-class Gini 25% price increase of gasoline 25% price increase of diesel 25% price increase of gasoline 25% price increase of diesel 25% price increase of LPG 25% price increase of kerosene 25% price increase of LPG 25% price increase of natural gas Source: Original figure for this publication based on country-specific microsimulations. Note: The figure shows the estimated percentage point change in poverty, vulnerability, the middle-class headcount ratio, and the Gini coefficient between the baseline and the 25 percent price rise scenarios for each country and fuel. Statistics are calculated based on per capita consumable income and the international reference poverty lines: the upper-middle-income poverty line (US$6.85 a day, 2017 PPP), the vulnerability line (US$14 a day, 2017 PPP), and the middle-class line (US$81 a day, 2017 PPP). 33 WHO BEARS THE BURDEN OF FUEL TAXATION in Latin America and the Caribbean Countries? Box 1. The Impact of Informal Fuel Purchases on Fuel Tax Incidence and Progressivity, Peru Informal fuel trading is widespread in several regions in Peru, particularly in rural and underserved areas where access to formal fuel stations is limited. According to estimates of 2010, informal sales reached 8 percent of total gasoline and diesel sales and 11 percent of LPG sales in Peru (OSINERGMIN 2011). Fuel in the informal market is often obtained by theft from trucks en route to distributors or sales locations or by smuggling from tax-exempt Amazon regions or neighboring countries where fuels are highly subsidized, such as Bolivia, Colombia, or Ecuador. Informal gasoline purchases are more common among lower-income households. Household survey data allow the identification of informal gasoline purchases by indicating place of purchase. Informal gasoline spending is more prevalent at the bottom of the income distribution, representing nearly 24 percent of total gasoline spending among households in the lowest income quintile, compared with less than 3 percent among the top quintile. Informal gasoline purchases are more prevalent in the highlands and the Amazon Region, whereas they are virtually nonexistent in Metropolitan Lima and the rest of the coastal areas. If informal gasoline purchases, which evade tax payments, are considered, the resulting household tax burden is lower, particularly among the poor. Although the overall incidence of gasoline taxes is modest, representing about 0.06 percent of all household income, the share declines by 12 percent if informal purchases are counted (refer to figure B1.1, panel a). Informal purchases disproportionately reduce the tax burden on poorer populations located in the highlands and rural regions (refer to figure B1.1, panel b). This effect is expected to be less pronounced in the Amazon Region because of tax exemptions on formal purchases. About 68 percent of gasoline sales in the Amazon are tax exempt. Given the distribution of informal gasoline purchases across the income distribution, gasoline taxes are less regressive if informality is accounted for. The Kakwani index is estimated at 10.8 if taxes on all gasoline sales are counted. It rises to 13.1 if only taxes on formal gasoline sales are counted. The informal trade of fuels in Peru presents several issues, including safety risks, environmental damage, and economic and fiscal problems. Most informal establishments do not follow proper safety measures, leading to higher risk of accidents, such as fires and explosions. The improper handling and disposal of fuels by informal sellers can produce environmental damage. Thus, spills and leaks can lead to soil and water contamination and harm ecosystems. Informal fuel trading creates unfair competition and can result in 34 revenue losses by the government because informal suppliers typically avoid taxes and licensing fees and ignore safety regulations, which allows them to offer lower prices. According to data supplied by the Peruvian Hydrocarbons Society, the informal trade of LPG gas cans alone had a fiscal cost of almost US$67 million in 2018. Despite the lower prices, consumers can be negatively affected by the lack of quality control. Informal fuel sellers may adulterate fuel by mixing it with cheaper additives or lower-quality substances to increase their profits. Figure B1.1 .Incidence of gasoline taxes, by place of purchase, 2019 % of disposable household income 0,10% 0,09% 0,08% 0,12% 0,07% 0,06% a. National 0,10% 0,05% 0,08% 0,04% 0,03% 0,06% 0,02% 0,01% 0,04% 0,00% Coast 0,02% 0,00% National Poorest 2 3 4 Richest Disposable Household Income Quintiles Gasoline Tax on Formal purchases Potential gasoline tax on Informal Purchases Total Incidence Formal Incidence 0,10% b. By region 0,09% 0,08% 0,07% 0,06% 0,05% 0,04% 0,03% 0,02% 0,01% 0,00% Metropolitan Lima Metropolitan Lima Highlands Highlands Highlands Metropolitan Lima Amazon Amazon Amazon Coast Coast Coast 3 4 Richest posable Household Income Quintiles National Bottom 40 Top 60 hases Potential gasoline tax on Informal Purchases Gasoline Tax on Formal purchases Potential gasoline tax on Informal Purchases Formal Incidence Source: Estimates based on National Household Survey 2019. Note: Informal gasoline spending is classified based on the reported place of purchase. A transaction is considered informal if the gasoline was bought at an informal gas station, at an informal car workshop, or to an informal transportation firm. 35 WHO BEARS THE BURDEN OF FUEL TAXATION in Latin America and the Caribbean Countries? The increase in fuel taxes is estimated to have led, in the short term, to a modest decline in the size of the middle class and an increase in poverty. These impacts vary by type of fuel tax and country (refer to figure 11). For example, although they are the most progressive tax in relative terms, gasoline taxes are estimated to raise poverty in most countries. In Mexico, poverty incidence measured against the upper-middle-income line (US$6.85 a day in 2017 purchasing power parity [PPP] US dollars) is estimated to have risen by 0.48 percentage points from a baseline of 25.5 percent (refer to annex A, figure A.6).35 This means that around 615,378 vulnerable Mexicans would become poor because of the resulting 25 percent rise in gasoline prices. The middle class would shrink by 0.61 percentage points, meaning 782,188 individuals would fall into vulnerability.36 In most countries, the impact of gasoline taxes tends to be larger on the middle class than on the poverty rate, whereas the opposite is observed for LPG, which matters in the public’s acceptance of reform and is consistent with the household expenditure patterns described above. 4.3 The Welfare and Distributional Effects of Increasing And Aligning Fuel Tax Rates with Emissions Four fuel tax and subsidy reform scenarios are analyzed according to the patterns observed in fuel taxation (and the price on CO2 emissions) in the LAC region. To facilitate comparisons, a common set of policy scenarios is defined while accounting for initial heterogeneities in tax rates and fuel types across countries. Governments vary in the way they tax and subsidize fuels, resulting in different carbon prices associated with emissions. While some fuels remain untaxed or even subsidized, others are subject to taxes.37 The following analysis presumes that the reforms are fully implemented in the short term, rather than gradually over a longer term. • Eliminate subsidies on fuel emissions: This involves evaluating the removal of explicit and implicit fuel subsidies, while maintaining the tax rates on other fuels constant at the baseline level. Given the emissions content of fuels, this policy results in an increase in the national average price of CO2 emissions from fuel use. The objective of this reform is to eliminate fiscal incentives that promote the use of high-emission energy sources. • Align fuel tax rates based on CO2 content: This keeps the national average carbon price of fuel emissions constant at the baseline level, but introduces fuel-specific tax adjustments to ensure a uniform price per ton of CO2 across all fuel types (Blanchard, Gollier, and Tirole 2023; OECD 2021).38 The goal of this policy is to establish a consistent carbon price signal across energy sources by aligning tax rates with the fuel carbon content, thereby 35 This impact represents about 8 percent of the average annual change in the poverty rate in Mexico in 2014–19. The same exercise in the case of Brazil results in an impact equal to 61 percent of the change in the average annual poverty rate during the same period. 36 Following the World Bank LAC Equity Lab, vulnerability is defined as the share of individuals whose per capita income is greater than US$6.85, but below US$14.00 a day (in 2017 PPP). The middle class refers to the share of individuals whose per capita income falls between US$14 and US$81 a day (2017 PPP). 37 Competitiveness, equity, and political economy considerations have played a key role in the current configuration of energy policy in LAC. They explain the low taxation of diesel, which is considered a critical fuel for road transportation, with import- ant impacts on inflation and strongly organized trade organizations. LPG has been historically promoted as a substitute for the use of solid fuels (firewood) in rural households 38 Under this scenario, the tax rate on CO2 emissions from fuels does not vary across fuel types. This refers to the tax rate that, once weighted by fuel-specific CO2 emissions, keeps the national average carbon rate constant at the baseline level. This implies that, while the carbon price for some fuels will increase with respect to the baseline scenario, it will decrease for other fuels. 36 eliminating incentives to switch among fuels. Taxing fuels according to their emissions is not only good for the climate, but makes sense from an energy technology perspective. • Increase the average carbon price of fuel emissions to US$60/tCO2, while keeping heterogeneity across fuel types: This sets the national average carbon price at US$60/ tCO2, which is the low-end estimate for 2030 of the costs imposed by an additional ton of CO2 released into the atmosphere (OECD 2021; World Bank, Ecofys, and Vivid Economics 2017). This scenario maintains the heterogeneity of carbon prices across fuel types in the baseline, resulting in a considerable carbon price gap among fuels.39 • Raise the average carbon price of fuel emissions to US$60/tCO2 by fixing a uniform carbon price for all fuels: This raises the carbon price to US$60/tCO2 on all fuel emissions. This is the most environmentally desirable scenario because it prices all fuel-related emissions at the same benchmark value, which reflects the societal cost and encourages the adoption of cleaner cost-effective technologies. Fuel tax reforms that raise the overall price of CO2 emissions exert a negative income shock on most households (except in Jamaica), albeit of modest average size. Reforms that raise the nationwide carbon price to US$60/tCO2 without altering relative carbon prices across fuels boost fuel prices by 1.5 percent–18.3 percent (except for natural gas in Mexico).40 These price rises are estimated to lead to modest average per capita income losses in the countries studied, ranging from 0.2 percent to 1.2 percent (refer to figure 12). These losses represent the additional income needed if households are to maintain current consumption. The effects of this reform are more pronounced in Brazil and Mexico, where the fuel price shocks for the year analyzed are larger, while the effects tend to be smaller in Jamaica and Paraguay (refer to figure 3; annex A, table A.3). The estimated impacts are appreciably smaller than the average impact of annual inflation on household purchasing power in all countries in 2014–19. In Brazil, for instance, this scenario would generate an average decline of 0.99 percent in household income per capita in the short term, compared with an impact of 5.8 percent because of average annual inflation in 2014–19. However, in countries with large fuel subsidies, the effects could be larger because the same reform could result in greater price shocks.41 The average impacts may also mask substantial disparities, and the effects may be considerable among population groups that are more dependent on fuel use. The impact of fuel taxes on household income is amplified if one accounts for the overall cost of the externalities generated by fuel combustion (beyond CO2 emissions), including the costs related to air pollution, traffic congestion, road wear from vehicle use, and road accidents (refer to box 2). 39 In this scenario, the carbon rate for every fuel is increased by a fixed amount equal to the difference between the emis- sions-weighted average carbon rate in the baseline scenario and the US$60/tCO2 benchmark. 40 Fuel price shocks, especially in LPG and natural gas, tend to be larger in the scenario that aligns the total carbon price of all fuels to US$60/tCO2, with LPG price increases approaching 30 percent in most of the countries analyzed. 41 For instance, in Colombia, the same tax reform would imply a fuel price shock ranging from 12 percent to 34 percent, depend- ing on the type of fuel. 37 WHO BEARS THE BURDEN OF FUEL TAXATION in Latin America and the Caribbean Countries? Figure 12. Short-term income effects, fuel tax reforms (compared to baseline scenario), by country Brazil Jamaica Average change in consumable income per capita (%) Average change in consumable income per capita (%) ura 23 1 0 1 .5 -2 -1 0 -3 -.5 -4 5 15 25 35 45 55 65 75 85 95 5 15 25 35 45 55 65 75 85 95 Percentile of disposable income per capita Percentile of disposable income per capita removal net subsidies carbon rate alignment carbon rate alignment 60 USD/tCO₂ 60 USD/tCO₂ 60 USD/tCO₂ + alignment 60 USD/tCO₂ + alignment Mexico Peru Average change in consumable income per capita (%) Average change in consumable income per capita (%) .5 0 .5 0 -.5 -.5 -1.5 -1 -1 5 15 25 35 45 55 65 75 85 95 5 15 25 35 45 55 65 75 85 95 Percentile of disposable income per capita Percentile of disposable income per capita removal net subsidies carbon rate alignment carbon rate alignment 60 USD/tCO₂ 60 USD/tCO₂ 60 USD/tCO₂ + alignment 60 USD/tCO₂ + alignment Paraguay Uruguay Average change in consumable income per capita (%) Average change in consumable income per capita (%) .5 .5 -.5 0 0 -1.5 -1 -.5 -2.5 -2 -1 5 15 25 35 45 55 65 75 85 95 5 15 25 35 45 55 65 75 85 95 Percentile of disposable income per capita Percentile of disposable income per capita carbon rate alignment 60 USD/tCO₂ removal net subsidies carbon rate alignment 60 USD/tCO₂ + alignment 60 USD/tCO₂ 60 USD/tCO₂ + alignment Source: Original figure for this publication based on country-specific microsimulations. Note: The figure shows the estimated percentage change in per capita consumable income between the baseline and the simulated fuel tax reform scenarios in each country. The results are presented as the household average for each percentile of the per capita disposable income distribution. Individuals in the bottom and top 5 percent of the distribution are excluded because of large data variations. The figure does not show the estimated effects of the removal of fuel subsidies in Jamaica, Paraguay, and Peru because these countries do not have subsidies at the baseline. In Mexico, the US$60 USD/tCO2 scenarios (with and without carbon alignment) almost overlap because they have similar effects on household income. 38 Box 2. Correcting Carbon Prices for the Total Externalities of Fuel Combustion, Brazil Fossil fuel combustion generates multiple negative externalities beyond those associated with the release of CO2 emissions into the atmosphere (already accounted for in the carbon price reform scenarios under study). These other externalities include, for example, air pollution, traffic accidents, congestion, and damage to road infrastructure. The analysis examined the impact of alternative reforms that capture this broader set of externalities in the case of Brazil using data from 2023. Adjusting scenario 4 to account for these externalities would result in a rise in the price of gasoline and diesel of 71 percent and 75 percent, respectively (compared with a −1.7 percent and 18.3 percent change in scenario 4). Although no direct CO2 emissions are produced through the use of ethanol, ethanol- powered vehicles generate other negative externalities, such as those arising from traffic accidents or road congestion. A tax reflecting these other externalities would therefore lead to a rise in the price of ethanol by 68 percent. These larger fuel price shocks exacerbate the negative welfare and distributional impacts of the carbon price reforms outlined above. Under scenario 4, households in the bottom 40 are estimated to face an average 1.5 percent decline in income because of the reform. The decline is 2.7 percent if one accounts for air pollution externalities and increases to 9.1 percent if all externalities are considered on which data could be gathered (refer to figure B2.1). Reforms that consider the total externalities of fuel combustion are expected to raise the poverty rate measured against the US$6.85 line (2017 PPP) to 34.7 percent from 31.8 percent in the baseline scenario. Figure B2.1. Changes in consumable income relative to the baseline scenario 0% Average change in consumable income per -2% -4% -6% -8% capita (%) -10% -12% -14% -16% -18% 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 Percentiles of disposable income per capita 60 USD/tCO2 + alignment Air pollution externalities All externalities Source: Original figure for this publication based on data of the 2017/18 Consumer Expenditure Survey. Note: Per capita monetary income (labor and nonlabor) has been used to classify households in disposable income centiles. Consumable income has been calculated as disposable income, net of consumption taxes and subsidies. 39 WHO BEARS THE BURDEN OF FUEL TAXATION in Latin America and the Caribbean Countries? The impact of the fuel tax reforms is more pronounced among households at the bottom of the income distribution in Brazil, Peru, and Uruguay; this is especially true of the reforms that align carbon prices across fuels (scenarios 2 and 4). Reforms that align tax rates with fuel carbon content would result in larger income losses among households at the bottom of the distribution and smaller income losses among households at the top, compared with the corresponding nonaligned scenarios (except in Paraguay) (refer to figure 12).42 This is related to the fact that carbon price alignment reforms typically lead to lower gasoline prices, largely benefiting wealthier households, while they would lead to higher prices for LPG and diesel, disproportionally hurting the less well off.43 Thus, in Uruguay, the reform that simply aligns carbon prices (scenario 2) is projected to reduce the price of gasoline by 19.5 percent, while increasing the prices of diesel and LPG by 6.6 percent and 77.4 percent, respectively (refer to annex A, table A.3). This is estimated to lower per capita income by 0.76 percent among households in the bottom quintile relative to the baseline (no reform), while the reform is expected to raise income by 0.04 percent among households in the top quintile, leading to a rise in inequality of 0.13 Gini points (refer to figure 13).44 Figure 13. Effects of fuel tax reforms on inequality, poverty, vulnerability, and the middle class, by country Change in percentage points (based on consumable income) Change in percentage points (based on consumable income) Brazil .4 .2 .4 .2 0 0 -.2 -.2 -.4 -.4 Poverty Vulnerable Middle-class Gini Poverty Removal of net subsidies Carbon rate alignment 60 USD/tCO2 60 USD/tCO2 with alignment C 60 USD/tCO2 with alignment + compensation 6 42 This is clear in a comparison of scenario 4 relative to scenario 3 or scenario 2 against the baseline (horizontal axis). points (based on consumable income) 43 In Jamaica, where the baseline carbon price is already high for Mexico oints (based on consumable income) several of the fuels analyzed, the carbon alignment scenarios (2 and 4) imply a reduction in the price of both gasoline and diesel, resulting in a positive income shock among households .4 across the entire distribution because of the significant indirect price effects of diesel taxes there. In Mexico, the reforms that .6 raise the carbon price to US$60/tCO2 (with and without aligning carbon prices across fuels) lead to similar proportional in- creases in the prices of gasoline and LPG (refer to annex A, table A.3). Given the incidence of gasoline and LPG taxes across .4 the income distribution shown in figure 9) – where wealthier households experience smaller losses than poorer households in the case of LPG taxes, but greater losses in the case of gasoline – the impact of these carbon price reforms (scenarios 3 and .2 4) is flatter in Mexico. .2 44 A comparison of scenario 3 (US$60/tCO2, no alignment) and scenario 4 (US$60/tCO2, with alignment) yields similar results. 0 0 40 -.2 -.2 -.4 Jamaica income) income) Change in percentage points (based on consumable income) Brazil .4 .4 .4 consumable consumable .2 .2 .2 on (based on 0 (based 0 0 points points -.2 -.2 -.2 percentage percentage -.4 -.4 ass Gini -.4 Poverty Poverty Vulnerable Vulnerable Middle-class Middle-class Gini Gini Carbon rate alignment Poverty inin Carbon Removal rate of net alignment subsidies USD/tCO2 60Carbon rate alignment Change 60 USD/tCO2 with alignment Change 60 USD/tCO2 with alignment 60 USD/tCO2 60 USD/tCO2 with alignment Ca 60 USD/tCO2 with alignment + compensation 60 Peru income) income) Mexico Change in percentage points (based on consumable income) .6 .4 .4 consumable consumable 00 .2 .2 .4 .2 (based on (based on 0 -.4 -.2 -.2 points points -.2 percentage percentage -.6 -.4 ass Gini -.4 Poverty Poverty Vulnerable Vulnerable Middle-class Middle-class Gini Gini Carbon rate alignment Poverty inin 60 USD/tCO2 with alignment Removal Carbon net subsidies ofalignment rate Carbon rate alignment 60 USD/tCO2 Change Change 60 60 USD/tCO2 USD/tCO2 with alignment 60 USD/tCO2 60 USD/tCO2 with alignment with alignment Carbon rate + compensation 60 USD/tCO2 with alignment + compensation 60 USD/tCO income) Uruguay d on consumable income) Paraguay income) .4 .4 on consumable .4 on consumable .2 .2 .2 41 ased 0 0 -. Change in percentage po Change in percentage points (based 0 (base WHO BEARS THE BURDEN OF FUEL TAXATION .4 Change in percentage points -.2 in Latin America and the Caribbean Countries? -.4 Gini -.2 Poverty Vulnerable Middle-class Gini Carbon rate alignment Carbon rate alignment 60 USD/tCO2 -.4 60 USD/tCO2 with alignment -.4 60 USD/tCO2 with alignment Poverty Vulnerable Middle-class Gini Poverty Removal of net subsidies Carbon rate alignment 60 USD/tCO2 60 USD/tCO2 with alignment Carb 60 USD/tCO2 with alignment + compensation 60 U Peru Change in percentage points (based on consumable income) income) Mexico Change in percentage points (based on consumable income) .2 .4 .6 -.2 on consumable .4 0 .2 0 .2 -.4 (based 0 -.2 .4 Change in percentage points Gini -.2 -.4 Poverty Vulnerable Middle-class Gini Carbon rate alignment -.6 0 USD/tCO2 with alignment Carbon rate alignment 60 USD/tCO2 -.4 60 USD/tCO2 with alignment Poverty Vulnerable 60 USD/tCO2 with alignment Middle-class Gini + compensation Poverty Removal of net subsidies Carbon rate alignment 60 USD/tCO2 60 USD/tCO2 with alignment Carbon rate a Change in percentage points (based on consumable income) 60 USD/tCO2 with alignment + compensation 60 USD/tCO2 Uruguay Change in percentage points (based on consumable income) Paraguay income) .2 .4 .4 -.2consumable 0 .2 .2 (based on 0 0 -.4 Change in percentage points -.2 Poverty Vulnerable Middle-class Gini -.2 Gini Removal of net subsidies Carbon rate alignment -.4 tCO2 60 USD/tCO2 60 USD/tCO2 with alignment -.4 tCO2 with alignment 60 USD/tCO2 with alignment + compensation Poverty ensation Poverty Vulnerable Middle-class Gini Removal of n Carbon rate alignment 60 USD/tCO2 60 USD/tCO2 60 USD/tCO2 with alignment 60 USD/tCO2 with alignment 60 USD/tCO2 + compensation 42 -. Change in percentage po -.4 ass Gini Poverty Vulnerable Middle-class Gini Carbon rate alignment 60 USD/tCO2 with alignment Carbon rate alignment 60 USD/tCO2 60 USD/tCO2 with alignment 60 USD/tCO2 with alignment + compensation Change in percentage points (based on consumable income) -.4 -.2 0 .2 .4 Uruguay Poverty Vulnerable Middle-class Gini ass Gini Removal of net subsidies Carbon rate alignment USD/tCO2 60 USD/tCO2 60 USD/tCO2 with USD/tCO2 with alignment 60 USD/tCO2 with alignment + compensation alignment ompensation Source: Original figure for this publication based on country-specific microsimulations. Note: The figure shows the estimated percentage point change in poverty, vulnerability, and middle-class headcount ratios and the Gini coefficient between the baseline and the simulated fuel tax reform scenarios for each country. Statistics are calculated based on per capita consumable income and the international reference lines: the upper-middle-income poverty line (US$6.85 a day, 2017 PPP), the vulnerability line (US$14 a day, 2017 PPP), and the middle-class line (US$81 a day, 2017 PPP). Removing existing subsidies for fuel emissions (scenario 1) entails larger per capita income losses compared with simply aligning carbon rates across fuels (scenario 2), though it is a relatively more progressive policy. An obvious first step for the alignment of tax rates across fuels is the removal of fuel subsidies and tax exemptions. This policy simultaneously raises the overall carbon price on fuel emissions in a country and partially aligns carbon prices across fuels by bringing the prices of fuels that are untaxed closer to the prices of the rest. Given the current fuel tax and subsidy configuration in the region, the removal of subsidies for fuel emissions typically results in a rise in the prices of LPG and diesel, while gasoline prices are often unaffected because of the already higher taxation of gasoline. In Brazil, Mexico, and Uruguay, removing subsidies would result in larger per capita income losses compared with the alignment-only scenario, particularly among higher-income households, which, under this subsidy reform scenario, would not benefit from potentially lower gasoline prices (refer to figure 12). 43 WHO BEARS THE BURDEN OF FUEL TAXATION in Latin America and the Caribbean Countries? Despite their relatively limited average effects on household income, the reforms under analysis still exercise a modest negative impact on poverty and the middle class (refer to figure 13). The impact of the reforms on poverty does not exceed 0.6 percentage points, which is less than the median annual change in poverty rates in 2014–19 in the countries under study. Nonetheless, in the scenario whereby the carbon price on all fuel emissions is aligned with US$60/tCO2 (scenario 4), the increase by 0.39 and 0.54 percentage points in the poverty headcount ratio in Brazil and Mexico, for example, represents 812,790 and 689,608 new poor, respectively (refer to table 1).45 In the case of the middle class, the impact of the reforms under analysis does not exceed 0.6 percentage points, observed in Mexico under that same policy reform scenario. Table 1. Change in socioeconomic status of individuals relative to the baseline, by reform scenario and country Scenario 1 Scenario 2 Scenario 3 Scenario 4 Removal net carbon US$60/tCO2 US$60/tCO2 Uniform carbon rate Country subsidies heterogeneous rates uniform carbon rate People, Change People, Change People, Change People, Change number in p.p. number in p.p. number in p.p. number in p.p. Brazil Change in 280,222 0.14 30,356 0.01 788,147 0.38 812,790 0.39 poverty Change in −24,822 −0.01 −75,708 −0.04 −55,424 −0.03 −51,250 −0.02 vulnerability Change in −247,613 −0.12 17,619 0.01 −655,507 −0.32 −747,184 −0.36 middle class Jamaica Change in — — −506 −0.02 0 0.00 −304 −0.01 poverty Change in — — −9,636 −0.35 3,589 0.13 −6,432 −0.24 vulnerability Change in — — 10,143 0.37 −3,589 −0.13 6,736 0.25 middle class Mexico Change in 58,082 0.05 17,247 0.01 704,575 0.55 689,608 0.54 poverty Change in −1,285 −0.001 5,122 0.004 128,438 0.10 128,720 0.10 vulnerability Change in −56,196 −0.04 −22,138 −0.02 −806,863 −0.63 −791,581 −0.61 middle class 45 The impact on poverty represents about 56 percent of the typical annual change in the poverty rate of Brazil in 2014–19 and about 10 percent in the case of Mexico. 44 Peru Change in — — 15,022 0.05 79,586 0.24 93,490 0.29 poverty Change in — — −11,572 −0.04 −12,710 −0.04 −24,364 −0.07 vulnerability Change in — — −3,449 −0.01 −63,476 −0.19 −68,603 −0.21 middle class Paraguay  Change in — — 7,775 0.11 19,077 0.26 8,981 0.12 poverty Change in — — −8,865 −0.12 −17,221 −0.24 −4,312 −0.06 vulnerability Change in — — 1,090 0.02 −1,856 −0.03 −4,669 −0.06 middle class Uruguay Change in 5,878 0.17 3,830 0.11 5,702 0.30 10,589 0.16 poverty Change in 4,656 0.13 2,895 0.08 3,642 0.10 1,835 0.05 vulnerability Change in −10,262 −0.29 −7,843 −0.22 −7,959 −0.22 −12,338 −0.35 middle class Source: Original figure for this publication based on country-specific microsimulations. Note: In the case of Jamaica, Paraguay and Peru, the table cells are empty under scenario 1 because these countries do not have fuel subsidies. Data are calculated based on household per capita consumable income and the international reference lines: the upper-middle-income poverty line (US$6.85 a day, 2017 PPP), the vulnerability line (US$14 a day, 2017 PPP), and the middle-class line (US$81 a day, 2017 PPP). p.p. = percentage points. Fuel tax reforms can generate substantial fiscal revenue in the short term, which could be used partially to offset the negative welfare and distributive outcomes. In the short term, holding fuel use and emissions constant, reforms that increase a country’s overall carbon price would lead to additional fiscal revenues. Part of these could be used, for example, to compensate households for the greater direct and indirect tax burden. Estimates from the microsimulations indicate that governments would need on average 0.16 percent of GDP to compensate households for the losses incurred by the reforms (refer to annex A, table A.6). One option is to provide top-ups to existing targeted social assistance programs. This would preserve the energy price signals and their impact on energy use, supply, and investment decisions. To illustrate the impacts of this measure, the main social assistance programs in each country are identified (refer to annex A, table A.1). The average per capita income loss among individuals in the bottom 40 under scenario 4, including both direct and indirect effects, is then estimated, and an additional scenario is modeled wherein this amount is assigned to all individuals living in beneficiary households irrespective of their fuel consumption and position in the distribution. This scenario illustrates the short-term impacts of one type of compensatory mechanism, while other compensation policies, such as spending on public services in education, health care, and so on, may be preferred by the population. 45 WHO BEARS THE BURDEN OF FUEL TAXATION in Latin America and the Caribbean Countries? Topping up existing targeted social assistance programs, for instance, can offset the decline in household purchasing power, particularly among lower-income households, if programs are well targeted (refer to figure 14). In Uruguay, for example, top-ups to the Tarjeta Uruguay Social and Asignaciones Familiares del Plan de Equidad programs reduce average per capita income losses by 68 percent among households in the bottom quintile, while no changes are observed among households above the 54 percentile of the income distribution. Given that, in Uruguay, the poverty rate based on the upper-middle-income poverty line is concentrated in the bottom decile of the distribution, this measure limits the rise in the poverty rate to 0.04 percentage points, compared with 0.3 points without the compensation measure (in scenario 4) (refer to figure 13). Because social assistance programs tend to be designed to target the poorest and the most vulnerable, they can be a cost-efficient instrument for providing income support to these individuals, conditional on the presence of a strong social protection system and well-targeted programs. However, they do little to alleviate income losses on the rest of the population, who may therefore oppose the reforms. In contrast, in Mexico, where programs are less highly targeted, this compensation mechanism also reduces income losses among individuals in the middle and upper parts of the income distribution. 46 Figure 14. Short-term income effects of a carbon price of US$60/tCO2 on all fuel emissions, by country Brazil: 60 USD/tCO₂ + alignment Mexico: 60 USD/tCO₂ + alignment Average change in consumable income per capita (%) Average change in consumable income per capita (%) .5 0 1 0 -1 -.5 -2 -1 -3 -1.5 -4 5 15 25 35 45 55 65 75 85 95 5 15 25 35 45 55 65 75 85 95 Percentile of disposable income per capita Percentile of disposable income per capita Peru: 60 USD/tCO₂ + alignment Paraguay: 60 USD/tCO₂ + alignment Average change in consumable income per capita (%) Average change in consumable income per capita (%) .5 .5 0 0 -.5 -.5 -1 -1 5 15 25 35 45 55 65 75 85 95 5 15 25 35 45 55 65 75 85 95 Percentile of disposable income per capita Percentile of disposable income per capita Uruguay: 60 USD/tCO₂ + alignment Average change in consumable income per capita (%) -.5 0 .5 without compensation -1 with compensation -2.5 -2 -1.5 5 15 25 35 45 55 65 75 85 95 Percentile of disposable income per capita Source: Original figure for this publication based on country-specific microsimulations. Note: The figure shows the estimated percentage change in per capita consumable income between the baseline and scenario 4 (as defined above) and relative to the situation when the tax reform is accompanied by top-ups to the main social assistance programs. The results are presented as the average across households for each percentile of the per capita disposable income distribution. Individuals in the bottom and top 5 percent of the disposable income distribution are excluded because of large data variations. Estimations on Jamaica are not included because scenario 4 results in a rise in per capita consumable income across the entire distribution (refer to figure 12); so, in Jamaica, there is no need for compensation. 47 WHO BEARS THE BURDEN OF FUEL TAXATION in Latin America and the Caribbean Countries? CONCLUDING 5 REMARKS AND A POLICY DISCUSSION 48 5. Concluding Remarks and a Policy Discussion Understanding the welfare and distributional impacts is essential for the proper design of fuel tax reforms. Fuel taxation contributes to the reduction of CO2 emissions by increasing the relative price of fuel and fuel-intensive goods and services. This may eventually lower climate-related hazards, thereby reducing their negative impact on household well-being. More immediately, it may negatively affect the purchasing power of households (the household use of income), which explains why such taxes are so unpopular.46 Yet, they may also generate additional fiscal revenues, which may partially be used to compensate for this impact. Considering this equity tradeoff is necessary for the design of palatable, successful fuel taxation reforms. One aim of the analyses presented in this study is to inform policy makers on this issue. In the short term, fuel tax and subsidy reforms generate a modest, negative income shock among most households. The analysis shows that reforms aimed at aligning fuel tax rates with the emissions content of fuels, raising the carbon price on fuel emissions, or both typically reduce the purchasing power of households. The effects vary across households, reforms, and countries depending on household spending patterns, the productive structure of the economy, and the initial tax structure. Overall, the magnitude of the estimated effects is modest. They lower per capita income by 0.1 percent to 1.2 percent on average in the countries under study.47 In the case of the most environmentally ambitious reforms, those resulting in a rise in the total carbon price on fuel emissions to US$60/tCO2 (with and without alignment), the estimated per capita income effects are smaller than the impact of average annual inflation on household purchasing power in all countries in 2014–19. However, these effects could be significantly larger under more ambitious reform scenarios, such as those aimed at increasing the carbon price on fuel emissions to capture fully the costs of the associated externalities (for instance, in terms of air pollution and health), as well as in countries that already have substantial fuel subsidies (such as Bolivia and Ecuador). The reforms under analysis have modest negative impacts on poverty and the middle class. The impact of the policy reform scenarios on the poverty rate does not exceed 0.6 percentage points, which is less than the median annual change in poverty rates in 2014–19 in the countries under study. Nonetheless, in the scenario involving uniformly setting the total carbon price on fuel emissions to US$60/tCO2, the increase by 0.39 and 0.54 percentage points in the poverty rates of Brazil and Mexico represents 812,790 and 689,608 new poor, respectively. In the case of the middle class, the impact of the reforms under analysis does not exceed 0.6 percentage points, which is observed in Mexico in the same scenario. Despite the rather small magnitude of the estimated income impacts, public support for the reforms is low, and governments may need to package them with other measures to make them more attractive. While the magnitude of the income impacts of the reforms does 46 In the medium and long term, fuel taxation can also affect the income-generating ability of households (household sources of income) because of the economic adjustment to transition toward a low carbon economy. 47 For each reform scenario and country, the average percentage share of income losses across households between the reform scenario and the baseline are calculated. The statistics cited in the text refer to the minimum and maximum average losses across countries and reform scenarios. 49 WHO BEARS THE BURDEN OF FUEL TAXATION in Latin America and the Caribbean Countries? not seem to be consistent with the strong public opposition to the measures, the direction of the effects is consistent with this opposition, and there are also other relevant factors in play. Some of the barriers to fuel tax reforms include (1) limited trust in the government, low institutional capacity, and lack of clarity on the use of additional fiscal revenue; (2) lack of public consultation and inadequate communication; (3) the resistance of vested interest groups; and (4) a sentiment of entitlement to cheap fuel, especially in oil- and gas-extracting countries and in countries in which fuel prices have been kept artificially low for extended periods (Douenne and Fabre 2022; Inchauste and Victor 2017; McCulloch 2023; McCulloch, Moerenhout, and Yang 2021). Recent studies show that support for energy fiscal policies rises if the proceeds are used to implement accompanying policies (Harring et al. 2023; Hoy et al. 2023). The proceeds might thus be used to improve public service quality, such as education and health care, or the reform might be accompanied by clear, informative, and trustworthy communications on purposes and expected outcomes (Carattini et al. 2017; Dechezleprêtre et al. 2022). While fuel taxes are mostly paid by wealthier households in absolute terms, they exert a greater impact on the purchasing power of the less well off, particularly in the case of LPG. Households at the bottom of the distribution spend a higher share of their incomes on fuels (direct effects) and on fuel-intensive goods and services (indirect effects). The first pattern is particularly observed with LPG taxes—illustrated by Brazil, Peru, and Uruguay—because poorer households heavily rely on LPG for cooking.48 The second pattern is clearly observed with diesel taxes because this fuel is extensively used in the production and distribution of goods, including food, that account for a larger share of the expenditures of households at the bottom. These patterns also explain why reforms that align tax rates with fuel carbon content, which usually imply a rise in the price of LPG and a decline in the price of gasoline, result in larger income losses among households at the bottom relative to the corresponding nonaligned average carbon price scenarios. However, energy policy is not an adequate tool for increasing the purchasing power of the less well off. Overall, gasoline taxes are relatively more progressive (albeit their effects on households in the bottom of the distribution is also negative), whereas LPG taxes are more regressive. It is not surprising then that the current fiscal structure in most LAC countries imposes higher taxes on gasoline emissions than on LPG emissions (which tend to be subsidized), consistent with the equity objective of policy makers. From a political economy perspective, low fuel prices are a simple, highly visible mechanism for governments to benefit their constituencies and vocal, organized special interest groups, such as transport operators, taxi unions, and farmers (ICCT 2007; Moerenhout et al. 2024). Nonetheless, from an efficiency and environmental standpoint, fuel tax policy and, specifically, fuel subsidies are not adequate instruments for supporting the well-being of the less well off because a large share of this support is captured by higher-income households. They also encourage the use of and investment in high-emission energy sources, delaying the adoption of cost-effective renewable technologies. There are several other alternative instruments, including social programs, that, if well designed, provide better targeted support to poor and vulnerable households. 48 The impact of gasoline taxes is similar across the distribution (Paraguay, Peru and Uruguay), or even greater among higher-in- come households (Jamaica and Mexico). 50 Compensation through existing, well-targeted social protection programs can help offset the impacts of fuel taxes on the less well off, though they do little to compensate other affected population groups. The simulations indicate that top-ups to existing social transfer programs, which account for a fraction of the additional revenue generated by the fuel tax reform, is a practical way of reducing the average per capita income losses among individuals at the bottom of the distribution, offsetting some of the negative redistributive effects, while preserving the energy price signals. Indeed, this mechanism can reduce per capita income losses by 40 percent on average among individuals in the bottom two quintiles of the distribution. Thus, policy measures related to social protection systems are often critical to the success of fuel tax reforms (Inchauste and Victor 2017; Mukherjee et al. 2023). However, the efficiency and effectiveness of these measures depend on the strength and coverage of the social protection system. Given that most social programs are targeted on the poor, they have little to no impact on households in the middle of the distribution. This reinforces the need to consider other, accompanying policy measures, such as improving key public services or identifying particularly affected groups and their characteristics to target compensation policies more effectively. The choice of the revenue recycling mechanism could affect public acceptance of the reform and have an impact on socioeconomic disparities in the longer term. Ideally, accompanying policies should also be forward-looking, helping households transition to a low-carbon economy. Continued income support to offset the impact of higher fuel prices can lower household incentives to transition away from fuels and can result in a costly permanent government expenditure. Switching to cleaner energy sources and transitioning to a net zero economy is expensive, especially for lower-income households (Dechezleprêtre et al. 2022).49 Thus, if fuel tax reforms are introduced, accompanying policies that support a reduction in energy consumption and a shift toward renewable energy sources should be considered, including investments in infrastructure that enable households to adopt sustainable consumption and investment practices, such as enhanced public transport connectivity, and the improved accessibility, affordability, and reliability of electricity (Brunckhorst et al. 2023; UNEP 2020). The same is true of labor reskilling and upskilling programs that would allow workers in carbon-intensive sectors to adapt to structural changes in the economy (Winkler et al. 2024). These initiatives that support low-income households in transitioning toward a low-carbon economy rather than merely compensating them for losses can prevent the long-run widening of disparities. 49 Further research is needed to clarify the potential barriers impeding households to transition to cleaner, more efficient energy options (for instance, financial and knowledge barriers). 51 WHO BEARS THE BURDEN OF FUEL TAXATION in Latin America and the Caribbean Countries? References Acharya, Bikram, and Klaus Marhold. 2019. “Determinants of Household Energy Use and Fuel Switching Behavior in Nepal.” Energy 169 (February): 1132–38. Alonso, Cristian, and Joey Kilpatrick. 2022. “The Distributional Impact of a Carbon Tax in Asia and the Pacific.” IMF Working Paper WP/22/116 (June), International Monetary Fund, Washington, DC. Alvarez, Maximiliano. 2019. “Distributional Effects of Environmental Taxation: An Approximation with a Meta-Regression Analysis.” Economic Analysis and Policy 62 (June): 382–401. Andersson, Julius J. 2019. “Carbon Taxes and CO2 emissions: Sweden as a Case Study.” American Economic Journal: Economic Policy 11 (4): 1–30. Anik, Asif Reza, Ram Ranjan, and Thiagu Ranganathan. 2018. “Estimating the impact of salinity stress on livelihood choices and incomes in Rural Bangladesh.” Journal of International Development 30 (8): 1414–38. Araar, Abdelkrim, and Paolo Verme. 2016. “Prices and Welfare.” Policy Research Working Paper 7566, World Bank, Washington, DC. Beck, Marisa, Nicholas Rivers, Randall Wigle, and Hidemichi Yonezawa. 2015. “Carbon Tax and Revenue Recycling: Impacts on Households in British Columbia.” Resource and Energy Economics 41 (August): 40–69. Best, Rohan, Paul J. Burke, and Frank Jotzo. 2020. “Carbon Pricing Efficacy: Cross-Country Evidence.” Environmental and Resource Economics 77 (1): 69–94. Beylis, Guillermo, and Barbara Cunha. 2017. Energy Pricing Policies for Inclusive Growth in Latin America and the Caribbean. Directions in Development: Energy and Mining Series. Washington, DC: World Bank. Birkmann, Joern, Emma Liwenga, Rajiv Pandey, Emily Boyd, Riyanti Djalante, François Gemenne, Walter Leal Filho, Patricia Fernanda Pinho, Lindsay Stringer, and David Wrathall. 2022. “Poverty, Livelihoods and Sustainable Development.” In Climate Change 2022: Impacts, Adaptation, and Vulnerability, 1171–1274. Sixth Assessment Report. Geneva: Intergovernmental Panel on Climate Change; New York: Cambridge University Press. Black, Simon, Antung A. Liu, Ian W. H. Parry, and Nate Vernon. 2023. “IMF Fossil Fuel Subsidies Data: 2023 Update.” IMF Working Paper WP/23/169 (August), International Monetary Fund, Washington, DC. Blanchard, Olivier J., Christian Gollier, and Jean Tirole. 2023. “The Portfolio of Economic Policies Needed to Fight Climate Change.” Annual Review of Economics 15 (1): 689–722. Borissov, Kirill, Alexandra Brausmann Vinogradova, and Lucas Bretschger. 2019. “Carbon Pricing, Technology Transition, and Skill-Based Development.” European Economic Review 118 (September): 252–69. Boyce, James K. 2018. “Carbon Pricing: Effectiveness and Equity.” Ecological Economics 150 (August): 52–61. 52 Brunckhorst, Ben James, Ruth Vargas Hill, Ghazala Mansuri, Trang Nguyen, and Miki Khanh Doan. 2023. “Climate and Equity: A Framework to Guide Policy Action.” Poverty and Equity Global Practice Series, World Bank, Washington, DC. Carattini, Stefano, Andrea Baranzini, Philippe Thalmann, Frédéric Varone, and Frank Vöhringer. 2017. “Green Taxes in a Post-Paris World: Are Millions of Nays Inevitable?” Environmental and Resource Economics 68 (1): 97–128. Carbone, Jared C., Nicholas Rivers, Akio Yamazaki, and Hidemichi Yonezawa. 2020. “Comparing Applied General Equilibrium and Econometric Estimates of the Effect of an Environmental Policy Shock.” Journal of the Association of Environmental and Resource Economists 7 (4): 687–719. Dechezleprêtre, Antoine, Adrien Fabre, Tobias Kruse, Bluebery Planterose, Ana Sanchez Chico, and Stefanie Stantcheva. 2022. “Fighting Climate Change: International Attitudes toward Climate Policies.” NBER Working Paper 30265 (July), National Bureau of Economic Research, Cambridge, MA. de la Fuente, Alejandro, and Samuel Freije-Rodríguez. 2024. “Climate and Poverty in Latin America and the Caribbean.” Working paper, World Bank, Washington, DC. Douenne, Thomas, and Adrien Fabre. 2022. “Yellow Vests, Pessimistic Beliefs, and Carbon Tax Aversion.” American Economic Journal: Economic Policy 14 (1): 81–110. EIA (US Energy Information Administration). 2023. “Short-Term Energy Outlook.” June, EIA, Department of Energy, Washington, DC. Felbermayr, Gabriel J., and Jasmin Katrin Gröschl. 2014. “Naturally Negative: The Growth Effects of Natural Disasters.” Journal of Development Economics 111 (November): 92–106. Fullerton, Don. 2011. “Six Distributional Effects of Environmental Policy.” NBER Working Paper 16703 (January), National Bureau of Economic Research, Cambridge, MA. Garaffa, Rafael, Bruno S. L. Cunha, Talita Cruz, Paula Bezerra, André F. P. Lucena, and Angelo C. Gurgel. 2021. “Distributional Effects of Carbon Pricing in Brazil under the Paris Agreement.” Energy Economics 101 (September), 105396. García-Muros, Xaquín, Jennifer Morris, and Sergey Paltsev. 2022. “Toward a Just Energy Transition: A Distributional Analysis of Low-Carbon Policies in the USA.” Energy Economics 105 (January), 105769. Goulder, Lawrence H., Marc Andrew Christian Hafstead, GyuRim Kim, and Xianling Long. 2019. “Impacts of a Carbon Tax across US Household Income Groups: What Are the Equity- Efficiency Trade-Offs?” Journal of Public Economics 175 (July): 44–64. Hanson, Gordon H. 2023. “Local Labor Market Impacts of the Energy Transition: Prospects and Policies.” NBER Working Paper 30871 (January), National Bureau of Economic Research, Cambridge, MA. Harring, Niklas, Erik Jönsson, Simon Matti, Gabriela Mundaca, and Sverker C. Jagers. 2023. “Public Acceptance of Fossil Fuel Subsidy Removal Can Be Reinforced with Revenue Recycling.” Nature Climate Change 13 (3): 214–15. 53 WHO BEARS THE BURDEN OF FUEL TAXATION in Latin America and the Caribbean Countries? Hill, Ruth Vargas, Trang Nguyen, and Miki Khanh Doan. 2024. “Climate and Equity: A Framework to Understand Welfare Impacts and Guide Policy Action.” Prosperity Insight Series, World Bank, Washington, DC. Hoy, Christopher Alexander, Yeon Soo Kim, Minh Cong Nguyen, Mariano Ernesto Sosa, and Sailesh Tiwari. 2023. “Building Public Support for Reducing Fossil Fuel Subsidies: Evidence across 12 Middle-Income Countries.” Policy Research Working Paper 10615, World Bank, Washington, DC. ICCT (International Council on Clean Transportation). 2007. “Passenger Vehicle Greenhouse Gas and Fuel Economy Standards: A Global Update.” ICCT, Washington, DC. IEA (International Energy Agency). 2023. “Latin America Energy Outlook.” World Energy Outlook Special Report (November), IEA, Paris. Inchauste, Gabriela, Christopher Hoy, Mariano Sosa, and Daniel Valderrama. 2024. “Governments Could Do Far More to Level the Playing Field through Fiscal Policies.” In Leveling the Playing Field: Addressing Structural Inequalities to Accelerate Poverty Reduction in Africa, edited by Nistha Sinha, Gabriela lnchauste, and Ambar Narayan, 229–83. Washington, DC: World Bank. Inchauste, Gabriela, and David G. Victor, eds. 2017. The Political Economy of Energy Subsidy Reform. Directions in Development: Public Sector Governance Series. Washington, DC: World Bank. Känzig, Diego Raoul. 2023. “The Unequal Economic Consequences of Carbon Pricing.” NBER Working Paper 31221 (May), National Bureau of Economic Research, Cambridge, MA. Labandeira, Xavier, José M. Labeaga, and Miguel Rodríguez. 2009. “An Integrated Economic and Distributional Analysis of Energy Policies.” Energy Policy 37 (12): 5776–86. Lustig, Nora Claudia, ed. 2022a. Fiscal Incidence Analysis: Methodology, Implementation, and Applications. Vol. 1 of Commitment to Equity Handbook: Estimating the Impact of Fiscal Policy on Inequality and Poverty, 2nd ed. New Orleans: CEQ Institute, Tulane University; Washington, DC: Brookings Institution Press. Lustig, Nora Claudia, ed. 2022b. Methodological Frontiers in Fiscal Incidence Analysis. Vol. 2 of Commitment to Equity Handbook: Estimating the Impact of Fiscal Policy on Inequality and Poverty, 2nd ed. New Orleans: CEQ Institute, Tulane University; Washington, DC: Brookings Institution Press. Malerba, Daniele, Anja Gaentzsch, and Hauke Ward. 2021. “Mitigating Poverty: The Patterns of Multiple Carbon Tax and Recycling Regimes for Peru.” Energy Policy 149 (February), 111961. Mayer, Jakob, Anna Dugan, Gabriel Bachner, and Karl W. Steininger. 2021. “Is Carbon Pricing Regressive? Insights from a Recursive-Dynamic CGE Analysis with Heterogeneous Households for Austria.” Energy Economics 104 (December), 105661. McCulloch, Neil. 2023. Ending Fossil Fuel Subsidies: The Politics of Saving the Planet. Rugby, UK: Practical Action Publishing. 54 McCulloch, Neil, Tom Moerenhout, and Joonseok Yang. 2021. “Fuel Subsidy Reform and the Social Contract in Nigeria: A Micro-Economic Analysis.” Energy Policy 156 (September), 112336. Missbach, Leonard, Jan Christoph Steckel, and Adrien Vogt-Schilb. 2022. “Cash Transfers in the Context of Carbon Pricing Reforms in Latin America and the Caribbean.” IDB Working Paper IDB-WP-01404 (November), Inter-American Development Bank, Washington, DC. Missbach, Leonard, Jan Christoph Steckel, and Adrien Vogt-Schilb. 2024. “Cash Transfers in the Context of Carbon Pricing Reforms in Latin America and the Caribbean.” World Development 173 (January), 106406. Moerenhout, Tom, Defne Gencer, Beatriz Arizu, Min A. Lee, and Hannah Braun. 2024. “Political Economy Analysis and Communications for Energy Subsidy Reforms: Approaches and Insights from Recent Technical Assistance.” Technical Report (March), Energy Subsidy Reform in Action Series, World Bank, Washington, DC. Mohapatra, Subhakanta. 2012. “Assessing Differential Health Vulnerability of the Slums in Chandigarh, India.” Internationales Asienforum 43 (1–2): 81–98. Mukherjee, Anit, Yuko Okamura, Ugo Gentilini, Defne Gencer, Mohamed Almenfi, Adea Kryeziu, Miriam Montenegro, and Nithin Umapathi. 2023. “Cash Transfers in the Context of Energy Subsidy Reforms: Insights from Recent Experience.” Technical Report (July), Energy Subsidy Reform in Action Series, World Bank, Washington, DC. Murray, Brian C., and Nicholas Rivers. 2015. “British Columbia’s Revenue-Neutral Carbon Tax: A Review of the Latest ‘Grand Experiment’ in Environmental Policy.” Energy Policy 86 (November): 674–83. OECD (Organisation for Economic Co-operation and Development). 2019. Taxing Energy Use 2019: Using Taxes for Climate Action. Paris: OECD. OECD (Organisation for Economic Co-operation and Development). 2021. “Taxing Energy Use for Sustainable Development: Opportunities for Energy Tax and Subsidy Reform in Selected Developing and Emerging Economies.” OECD, Paris. OSINERGMIN (Organismo Supervisor de la Inversión en Energía y Minería, Regulatory Agency for Investment in Energy and Mining). 2011. “Boletin Informativo de Gas Natural 2011.” OSINERGMIN, Magdalena del Mar, Peru. Parry, Ian W. H., Simon Black, and Karlygash Zhunussova. 2022. “Carbon Taxes or Emissions Trading Systems? Instrument Choice and Design.” IMF Staff Climate Note 2022/006 (July), International Monetary Fund, Washington, DC. Pasch, Richard J., Brad J. Reinhart, and Laura Alaka. 2023. “Hurricane Fiona (al072022), 14– 23 September 2022.” Tropical Cyclone Report (March 23), National Hurricane Center, Miami. Rausch, Sebastian, Gilbert E. Metcalf, and John M. Reilly. 2011. “Distributional Impacts of Carbon Pricing: A General Equilibrium Approach with Micro-Data for Households.” Energy Economics 33 (Supplement 1): S20–S33. 55 WHO BEARS THE BURDEN OF FUEL TAXATION in Latin America and the Caribbean Countries? Renner, Sebastian, Jann Lay, and Hannes Greve. 2018. “Household Welfare and CO2 Emission Impacts of Energy and Carbon Taxes in Mexico.” Energy Economics 72 (May): 222–35. Tovar-Reaños, Miguel Angel, and Muireann Á. Lynch. 2019. “Distributional Impacts of Carbon Taxation and Revenue Recycling: A Behavioral Microsimulation.” ESRI Working Paper 626 (June), Economic and Social Research Institute, Dublin. UNEP (United Nations Environment Programme). 2020. Emissions Gap Report 2020. Nairobi, Kenya: UNEP. Vera, Felipe, María Camila Uribe, and Sofía del Castillo. 2023. “Climate Action and the Paris Agreement: The Role of Cities in Latin America and the Caribbean.” IDB Monograph 1092, Inter-American Development Bank, Washington, DC. Wang, Qian. Klaus Hubacek, Kuishuang Feng, Lin Guo, Kun Zhang, Jinjun Xue, and Qiao-Mei Liang. 2019. “Distributional Impact of Carbon Pricing in Chinese Provinces.” Energy Economics 81 (June): 327–40. Wang, Qian, Klaus Hubacek, Kuishuang Feng, Yi-Ming Wei, and Qiao-Mei Liang. 2016. “Distributional Effects of Carbon Taxation.” Applied Energy 184 (December): 1123–31. Winkler, Hernán Jorge, Vincenzo Di Maro, Kelly Yelitza Montoya Munoz, Sergio Daniel Olivieri, and Emmanuel Jose Vazquez. 2024. “Measuring Green Jobs: A New Database for Latin America and Other Regions.” Policy Research Working Paper 10794, World Bank, Washington, DC. World Bank. 2022a. China: Country Climate and Development Report. October. Washington, DC: World Bank. World Bank. 2022b. “New Approaches to Closing the Fiscal Gap.” Latin America and the Caribbean Economic Review (October), World Bank: Washington, DC. World Bank. 2022c. “State and Trends of Carbon Pricing 2022.” May, World Bank: Washington, DC. World Bank. 2024. Poverty, Prosperity, and Planet Report 2024: Pathways Out of the Polycrisis. Washington, DC: World Bank. World Bank. 2025. “Taxing and Subsidizing Energy in Latin America and the Caribbean: Insights from a Total Carbon Price Approach.” World Bank, Washington, DC. World Bank, Ecofys, and Vivid Economics. 2017. State and Trends of Carbon Pricing 2017. November. Washington, DC: World Bank. 56 6 ANNXES 57 WHO BEARS THE BURDEN OF FUEL TAXATION in Latin America and the Caribbean Countries? Annex A. Additional Tables and Figures Figure A.1. The hazard, exposure, and vulnerability framework The propensity or predisposition to be adversary affected. Vulnerability encompasses a variety of concepts VULNERABILITY and elements, including sensitivity or susceptibility to ham and lack of capacity to cope and adapt. RISK (Impacts) The presence of people; livelihoods; species or ecosystems; enviromental functions, services, and HAZARD resources; infraestructure; EXPOSURE or economic, social, or cultural assets in places The potential occurrence of and settings that could be a natural or human-induced adversely affected. physical event or trend that may cause loss of life, injury or other health impacts, as well as damages and loss to property, infraestructure, livelihoods, service provision, ecosystems and enviromental resources. Source: Brunckhorst et al. 2023. 58 Table A.1. Countries, data sources, and social assistance programs considered Country Country Survey Household Survey Main Social Assistance Programs Code Name Year National Household Expenditure ARG Argentina 2017/18 Asignación Universal por Hijoa Survey (ENGHo) Bono Juana Azurduy BOL Bolivia National Household Survey (EH) 2021 Bono Juancito Pinto Renta Dignidad BRA Brazil Consumer Expenditure Survey (POF) 2017 Bolsa Familia (Auxilio Brasil) Asignación Familiar CHL Chile Household Budget Survey (EPF) 2022 Subsidio Único Familiar Seguridades y Oportunidades National Household Budget Survey COL Colombia 2016/17 Más familias en acción (ENPH) Avancemos National Household Income and CRI Costa Rica 2018/19 Expenditure Survey (ENIGH) Régimen no contributivo de pensiones Dominican National Household Expenditure and DOM 2018 Comer es primero Republic Income Survey (ENGIH) National Survey of Household ECU Ecuador 2011/12 Bono de Desarrollo Humano Income and Expenditure (ENHIGUR) Jamaica Survey of Living Conditions Programme of Advancement JAM Jamaica 2018 (JSLC) Through Health and Education National Survey of Household Beca Benito Juarez Básica y Benito MEX Mexico 2022 Income and Expenditure (ENIGH) Juarez Media Superior Red de Oportunidades SENAPAN PAN Panama Living Standards Survey (ENV) 2008 120 a los 65 Angel Guardian Juntos PER Peru National Household Survey (ENAHO) 2019 Pension 65 Income, Expenditure Survey and Tekopora PRY Paraguay 2012 Living Conditions (EIGyCV) Adultos Mayores El Multipurpose Household Survey SLV 2022 Comunidades Solidarias Salvador (EHPM) Tarjeta Uruguay Social National Household Expenditure and URY Uruguay 2016 Asignaciones Familiares del Plan Income Survey (ENGIH) Equidad a. In the case of the Asignación Universal por Hijo, eligibility is approximated through the demographic and employment criteria of the program because the household survey does not identify beneficiaries. 59 WHO BEARS THE BURDEN OF FUEL TAXATION in Latin America and the Caribbean Countries? Figure A.2. Share of households that report positive energy expenditures, by fuel type and income group 100 Gasoline 80 % of households 40 20 0 60 T60 B40  B40  T60 B40  T60 B40  T60 B40  T60 B40  T60 B40  T60 B40  T60 B40  T60 B40  T60 B40  T60 B40  T60 B40  T60 B40  T60 T60 B40  ARG BOL BRA CHL COL CRI DOM ECU JAM MEX PAN PER PRY SLV URY LPG 100 80 % of households 40 20 0 60 T60 B40  B40  T60 B40  T60 B40  T60 B40  T60 B40  T60 B40  T60 B40  T60 B40  T60 B40  T60 B40  T60 B40  T60 B40  T60 B40  T60 T60 B40  ARG BOL BRA CHL COL CRI DOM ECU JAM MEX PAN PER PRY SLV URY 60 Natural gas 100 80 % of households 40 20 0 60 B40  T60 B40  T60 B40  T60 B40  T60 B40  T60 B40  T60 B40  T60 B40  T60 B40  T60 ARG BOL BRA CHL COL DOM MEX PER URY 100 Diesel 80 % of households 40 20 0 60 B40  T60 B40  T60 B40  T60 B40  T60 B40  T60 B40  T60 B40  T60 B40  T60 B40  T60 B40  T60 ARG BRA CHL COL DOM ECU MEX PER PRY URY Source: Original figure for this publication based on nationally representative household expenditure surveys. Note: Refer to table A.1 for further details. Statistics are presented separately for households in the bottom 40 percent of the per capita income distribution (B40) and in the top 60 percent (T60). 61 WHO BEARS THE BURDEN OF FUEL TAXATION in Latin America and the Caribbean Countries? Figure A.3. Share of food and transport expenditures over household income, by income group Food 80 Food expenditure as a proportion of household income (%) 60 40 20 0 T60 B40  B40  T60 B40  T60 B40  T60 B40  T60 B40  T60 B40  T60 B40  T60 B40  T60 B40  T60 B40  T60 B40  T60 B40  T60 T60 B40  ARG BOL BRA CHL COL CRI DOM ECU JAM MEX PAN PER PRY URY Transport Transport expenditure as a proportion of household income (%) 0 2 4 6 8 10 12 T60 B40  B40  T60 B40  T60 B40  T60 B40  T60 B40  T60 B40  T60 B40  T60 B40  T60 B40  T60 B40  T60 B40  T60 B40  T60 T60 B40  ARG BOL BRA CHL COL CRI DOM ECU JAM MEX PAN PER PRY URY Source: Original figure for this publication based on nationally representative household expenditure surveys. Note: Refer to table A.1 for further details. Statistics are presented separately for households in the bottom 40 percent of the per capita income distribution (B40) and in the top 60 percent (T60). Transport expenditures do not include the energy expenditures shown in figure 3. In the case of El Salvador, the household survey does not contain information on food and transport expenditures. 62 Figure A.4. Income concepts Market income, plus pensions Direct cash and near-cash (prefiscal income) Direct taxes: personal transfers: conditional income taxes and contributions to social and unconditional cash transfers, noncontributory + - security that are not pensions, school feeding directed to old-age programs, and so on pensions Disposable income Indirect subsidies: fuel, + - Indirect taxes: value added tax, excise taxes, food, and so on fuel taxes Consumable income In-kind transfers: monetized value of + - Copayments, user fees education and health services at average government cost Final income Source: Adapted from Lustig 2022a, 2022b. Note: Prefiscal income is the sum of factor income (labor and capital income), private transfers (remittances, private pensions), imputed rent, own production, and contributory social insurance old-age pensions, minus contributions to social insurance old-age pensions. The analyses presented in this chapter focus on changes in consumable income. The final income concept is therefore not constructed. 63 WHO BEARS THE BURDEN OF FUEL TAXATION in Latin America and the Caribbean Countries? Table A.2. Details on country-specific microsimulation tools Input-output table used Additional Survey- Direct taxes and Indirect taxes Country to model indirect price notes (if CEQ to-survey contributions and and subsidies (baseline Survey effects (source, year, was not used, imputation? direct transfers included in the year) assumptions, sector explain the (method) included in the tool tool identified as fuels) methodology) An input-output table was constructed based on the 2015 supply and use tables published by the Brazilian Institute of Geography and Statistics. It was necessary to The Continuous break the petroleum National industry down into three Household components: production Sample Survey of diesel, production of dataset used 2017/18 gasoline, and the rest for the CEQ Pesquisa de Indirect taxes: of petroleum refining. does not have Orçamentos Program The supply side of the information on Familiares of Social tables was estimated consumption. (Consumer Integration, by allocating the So, the CEQ Expenditure contribution for corresponding output was not Survey). For the financing of (primary and secondary used. Instead, this analysis, social security; if needed) to the new monetary Brazil Direct transfers: the following No tax on the industries while keeping income reported (2017) Bolsa Família household fuel movement the original totals. The in the Consumer expenditures of goods and value of intermediate Expenditure could be services; tax on consumption by product Survey was identified: the importation for the new industries used for the gasoline, diesel, of technical was derived by: (a) disposable LPG, ethanol, services and obtaining additional income natural gas, and royalties information on the variable, and kerosene. input structure of these consumption industries and (b) taxes were rebalancing the tables subtracted to ensure that GDP to estimate is not affected by the consumable calculations. Annual income reports from companies such as Petrobras Gas were also used to obtain a more accurate input- output structure for the petroleum industry. Indirect taxes: VAT 2018 Jamaica Jamaica Social contributions: Excise tax Survey of Living     (2018) Pension contributions (gasoline, Conditions kerosene) Carbon tax 2022 Encuesta Indirect Taxes: The matrix used was Nacional de Direct taxes and VAT (Impuesto al published by the National Ingresos y social security Valor Agregado) Institute of Statistics and Gastos de contributions: income Geography in 2013. The Mexico Mexico los Hogares taxes, regimen de Excise (Impuesto matrix is 262x262. Fuels No Simulation Tool (2022) (National actividad empresarial, Especial sobre were identified using for Mexico Household regimen simplificado Producción y the following category: Income and de confianza, personal Servicios) sector 66 "Fabricación de Expenditure deductions productos derivados del Survey) Carbon tax petróleo y carbón" 64 Direct taxes and 2021 Encuesta contributions: Permanente income tax, (impuesto de Hogares a las rentas personal), Continua EPHC rentas de trabajo, (from INE) rentas de capita, An input-output table impuesto a las rentas Indirect taxes: was constructed based 2011/12 empresarial, simple y value added tax, on the 2014 supply and Encuesta resimple, impuestos excise tax use tables published by de Ingresos Yes a los dividendos the Central Bank. For Paraguay y Gastos y y utilidades, Subsidies: the fuel tax, welfare, Condiciones de contribuciones a la CEQ (Dmatch electricity, and distributive analysis, (2021) Vida (from INE). seguridad social approach) public transport, gasoline and diesel were For this analysis, housing identified in the table the following Direct transfers: using the following household fuel Tekopora, adulto categories: gasolina, expenditures mayor, becas diesel, and LPG. could be identified: educativas, secundaria, gasoline, diesel, terciaria, transferencia flex, LPG, and de viveres útiles kerosene. escolares, alimentación escolar Direct taxes and contributions: income tax, rentas de trabajo, rentas de capital, impuesto a la propiedad, contribuciones a la An input-output table seguridad social was constructed based 2019 Encuesta on the 2019 supply and Nacional Direct transfers: use tables published by de Hogares JUNTOS, Pension65, the National Institute of (National Indirect taxes: Peru Beca18, Bonogas, Statistics and Informatics. Household No sales tax, excise CEQ (2019) pensión por The analysis includes Survey). The tax viudez, orfandad the following categories: survey includes o sobrevivencia, gasoline, natural gas, income and donaciones de LPG, and diesel. Each expenditure data. bienes, nonmonetary fuel had a unique section programs (Wawasi– in the input-output table. cuidado de menores, Centro de Emergencia Mujer, Pronama, Jóvenes productivos, Trabaja Perú, Impulsa Perú) 65 WHO BEARS THE BURDEN OF FUEL TAXATION in Latin America and the Caribbean Countries? Direct taxes and contributions: income tax, rentas de trabajo, rentas de capital, impuesto a las rentas 2021 Encuesta del agro, Fondo de Continua de Reconversión Laboral, An input-output table Hogares Fondo de Solidaridad, was constructed based contribuciones de on the 2016 supply and 2016/17 salud, impuesto use tables published by Encuesta de asistencia a la the Central Bank. For de Gastos e seguridad social Indirect taxes: the fuel tax, welfare, Ingresos de value added and distributive analysis, Uruguay los Hogares Yes (Dmatch Direct transfers: tax, impuesto gasoline and diesel were (ENGIH). For CEQ approach) pensión por vejez específico identified in the table (2021) this analysis, e invalidez, Plan interno, tax on using the following the following de Equidad, Tarjeta CO2 emissions categories: gasolina para household fuel Uruguay Social, motores de vehículos expenditures Canastas, Comidas de transporte terrestre, could be en comedores combustibles para identified: sociales, Asignaciones calderas (fuel oil and gasoline, diesel, Familiares, diesel oil). LPG, natural gas, and kerosene contributivas, seguro de desempleo, compensaciones por accidente, maternidad o enfermedad, Hogar Constituido Source: Original table for this publication. Table A.3. Fuel tax reform scenarios and corresponding price shocks, by country Scenario 1 Scenario 2 Scenario 3 Scenario 4 Baseline Removal of net Uniform carbon US$60/tCO2 US$60/tCO2 + Country, fuel carbon subsidies rate heterogeneous rates uniform carbon rate Tax Price Tax Price Tax Price Tax Price Tax rate rate shock rate shock rate shock rate shock Brazil Gasoline (R$/liter) 0.79 0.79 0.00% 0.20 −15.30% 1.08 7.40% 0.48 −7.90% Diesel (R$/liter) −0.04 0.00 1.10% 0.29 9.90% 0.38 12.70% 0.71 22.60% LPG (R$/ −0.66 0.00 14.40% 0.37 22.50% −0.13 11.60% 0.90 34.10% kilogram) Ethanol (R$/liter) −0.14 — 5.30% — — — — — — Jamaica Gasoline 0.22 0.22 0.00% 0.11 −6.11% 0.25 1.67% 0.14 −4.44% (US$/liter) Diesel (US$/liter) 0.27 0.27 0.00% 0.12 −7.89% 0.30 1.58% 0.16 −5.79% 66 LPG (US$/liter) 0.03 0.03 0.00% 0.08 5.00% 0.05 2.00% 0.10 7.00% Kerosene 0.08 0.08 0.00% 0.12 2.86% 0.12 2.86% 0.15 5.00% (US$/liter) Mexico Gasoline −0.20 0.00 1.00% −0.10 0.50% 2.70 14.51% 2.70 14.51% (Mex$/liter) Diesel (Mex$/ −1.50 0.00 8.06% −0.10 7.52% 1.90 18.27% 3.20 25.26% liter) LPG (Mex$/liter) 0.10 0.10 0.00% −0.10 −1.78% 2.10 17.84% 1.90 16.06% Natural gas 0.00 0.00 0.00% −3.00 −1.45% 70.10 33.77% 66.80 32.18% (Mex$/gigajoule) Peru Gasoline (S/liter) 0.51 0.51 0.00% 0.20 −9.00% 0.77 7.50% 0.46 −1.40% Diesel (S/liter) 0.42 0.42 0.00% 0.23 −5.70% 0.73 9.30% 0.54 3.60% LPG (S/kilogram) 0.03 0.03 0.00% 0.24 5.30% 0.35 7.90% 0.57 13.60% Natural gas 0.00 0.00 0.00% 0.19 13.50% 0.25 17.70% 0.45 31.30% (S/m3) Paraguay Gasoline 0.15 0.15 0.00% 0.09 −6.42% 0.16 1.53% 0.10 −4.68% (US$/liter) Diesel (US$/liter) 0.11 0.11 0.00% 0.13 2.90% 0.13 2.61% 0.16 5.87% LPG (US$/liter) 0.001 0.001 0.00% 0.08 14.07% 0.02 2.42% 0.10 16.82% Uruguay Gasoline 14.89 14.89 0.00% 2.29 −19.50% 20.06 8.00% 7.46 −11.50% (Ur$/liter) Diesel (Ur$/liter) −0.15 0.00 0.30% 2.86 6.60% 6.29 14.20% 9.30 20.90% LPG (Ur$/ −38.30 0.00 71.40% 3.24 77.40% −31.00 13.60% 10.53 91.00% kilogram) Kerosene −1.56 0.00 3.60% 2.78 10.10% 4.71 14.60% 9.06 24.70% (Ur$/liter) Natural gas 0.00 0.00 0.00% 2.19 4.60% 4.93 10.40% 7.12 15.10% (Ur$/m3) Sources: Original table for this publication based on national tax codes and fuel price data; Enerdata (dashboard), Enerdata, Grenoble, France, https://www.enerdata.net/research/; IEA Data and Statistics (portal), International Energy Agency, Paris, https://www.iea.org/data-and-statistics. Note: The table shows the net tax rate on fossil fuels in each country (tax rate) and how it changes across the different reform scenarios. It also shows the implicit price shock induced by the changes in fuel tax rates across scenarios. In the case of Brazil, ethanol is assumed not to contribute to greenhouse gas emissions. One can thus not calculate a price per metric ton of CO2, which is required to estimate scenarios 2–4. However, because ethanol is subsidized in Brazil, scenario 1 may still be modeled if subsidies per liter are removed. 67 WHO BEARS THE BURDEN OF FUEL TAXATION in Latin America and the Caribbean Countries? Figura C5 Figure A.5. Estimated short-term income effect of a 25 percent price increase on all fuels, by country Average change in consumable income per capita (%) Total 0 -1 Brazil -2 Peru Jamaica -3 Uruguay -4 Mexico Paraguay -5 -6 5 15 25 35 45 55 65 75 85 95 Percentile of disposable income per capita Source: Original figure for this publication based on country-specific microsimulations. Note: The figure shows the estimated percentage change in per capita consumable income between the baseline and a 25 percent increase in the price of all fuels in each country. The types of fuels considered in each country are those shown in figure 9. The results are presented as the average across households for each percentile of the per capita disposable income distribution. Individuals in the bottom and top 5 percent of the disposable income distribution are excluded because of the large variations in the relevant data. Table A.4. Contribution to additional fuel tax collection, by income quintile and country 25 percent tax-induced fuel price shock Income Country Natural quintile Gasoline Diesel LPG Kerosene Ethanol All fuels gas Q1 4.6 7.7 15.4 — — 4.8 6.7 Q2 9.2 12.3 19.6 — — 9.0 11.2 BRA Q3 14.2 16.8 21.2 — — 14.1 15.7 Q4 22.3 22.5 23.4 — — 22.7 22.6 Q5 49.7 40.6 20.5 — — 49.5 43.9 68 Q1 0.4 7.6 8.3 22.7 — — 5.8 Q2 2.0 12.3 14.1 16.2 — — 9.8 JAM Q3 6.3 16.3 17.2 21.0 — — 13.7 Q4 19.0 21.8 24.4 17.0 — — 21.4 Q5 72.3 42.1 36.1 23.1 — — 49.3 Q1 4.5 8.9 9.7 — 3.1 — 6.0 Q2 8.0 12.2 14.8 — 5.9 — 9.9 MEX Q3 12.2 15.3 18.4 — 13.1 — 14.0 Q4 20.5 20.7 22.5 — 22.5 — 21.1 Q5 54.8 42.9 34.6 — 55.3 — 49.1 Q1 4.6 6.7 10.8 — 9.5 — 7.5 Q2 8.5 12.3 16.5 — 14.8 — 12.7 PER Q3 12.6 17.7 20.1 — 19.2 — 17.2 Q4 17.9 23.7 22.6 — 23.6 — 22.0 Q5 56.4 39.7 30.0 — 33.0 — 40.6 Q1 9.1 8.7 5.5 — — — 8.4 Q2 12.9 10.7 11.8 — — — 11.8 PRY Q3 16.2 12.5 21.6 — — — 15.7 Q4 24.4 21.1 26.2 — — — 23.4 Q5 37.4 47.0 34.9 — — — 40.7 Q1 7.3 10.2 14.8 14.6 0.2 — 9.2 Q2 10.3 13.3 17.2 18.0 2.7 — 12.2 URY Q3 14.9 17.3 19.7 23.5 0.4 — 16.3 Q4 20.9 23.0 23.0 14.6 6.2 — 21.7 Q5 46.5 36.3 25.3 29.3 90.6 — 40.6 Source: Original table for this publication based on country-specific microsimulations. Note: The table shows the household contribution by quintile of the per capita disposable income distribution to the rise in tax collection between the baseline and the 25 percent price rise scenarios for each country and fuel. 69 WHO BEARS THE BURDEN OF FUEL TAXATION in Latin America and the Caribbean Countries? Table A.5. Progressivity of a 25 percent fuel-specific price rise, by fuel and country 25 percent tax-induced fuel price shock Country Gasoline Diesel LPG Kerosene Natural gas Ethanol BRA −0.09 −0.21 −0.48 — — −0.09 JAM 0.34 −0.02 −0.08 −0.34 — — MEX 0.06 −0.09 −0.18 — 0.16 — PER −0.07 −0.10 −0.23 — −0.14 — PRY −0.14 −0.05 −0.13 — — — URY −0.02 −0.14 −0.29 −0.26 0.45 — Source: Original table for this publication based on country-specific microsimulations. Note: The table shows the Kakwani coefficient associated with a tax-induced rise in the retail price of each fossil fuel. Figure A.6. Baseline poverty, vulnerability, middle-class, and inequality statistics, by country 70 64.8 60 55.5 51.2 50 43.0 42.8 42.6 41.2 41.1 39.1 38.2 percent 36.4 35.5 35.8 40 35.3 34.8 34.5 32.0 28.8 28.7 25.5 30 23.8 22.2 20 10.7 8.3 10 0 Poverty Vulnerable Middle-class Gini Brazil (2017) Jamaica (2018) Mexico (2022) Peru (2019) Uruguay (2021) Paraguay (2021) Source: Original figure for this publication based on country-specific microsimulations. Note: The figure shows the poverty, vulnerability, and middle-class headcount ratios and the Gini coefficient at baseline for each country. Statistics are calculated based on per capita consumable income and the international reference lines: the upper-middle-income poverty line (US$6.85 a day, 2017 PPP), the vulnerability line (US$14 a day, 2017 PPP), and the middle-class line (US$81 a day, 2017 PPP). 70 Table A.6. Total compensation as a share of GDP, selected fuel tax reforms, by country, % Scenario 1 Scenario 2 Scenario 3 Scenario 4 Income US$60/tCO2 US$60/tCO2 Country Removal net Uniform group heterogeneous uniform carbon carbon subsidies carbon rate rates rate BRA Bottom 40 0.027 0.016 0.059 0.075 All 0.098 0.017 0.298 0.224 JAM Bottom 40 — — 0.018 — All — — 0.112 — MEX Bottom 40 0.004 0.001 0.053 0.053 All 0.030 0.009 0.381 0.382 PER Bottom 40 — 0.006 0.038 0.044 All — 0.010 0.220 0.229 PRY Bottom 40 — 0.004 0.014 0.018 All — 0.030 0.057 0.092 URY Bottom 40 0.060 0.038 0.001 0.084 All 0.247 0.068 0.203 0.242 Source: Original table for this publication based on country-specific microsimulations. Note: The table shows the total income compensation as a share of GDP (at baseline) in each fuel tax reform. Compensation is calculated among those individuals classified in the bottom 40 percent of the per capita income distribution and for all individuals with income losses in each tax reform. Compensation in Jamaica in scenario 2 and 4 are not included because these reforms result in an increase in per capita consumable income across the entire distribution (refer to figure 12); so, there is no need for compensation. In the case of Jamaica, Paraguay, and Peru, the table does not show the compensation associated with the removal of fuel subsidies because these countries do not have subsidies at the baseline. 71 WHO BEARS THE BURDEN OF FUEL TAXATION in Latin America and the Caribbean Countries? 72