table_1Research and Development Financing in the Renewable Energy Industry in Brazil Muriel de Oliveira Gavira1,2 1School of Applied Sciences, University of Campinas, Brazil 2DepartmentFaculty of Sciences, University of Lisbon, Portugal e-mail: murielgavira@gmail.comResearch and Development Financing in the Renewable Energy Industry in Braziltable_2ABSTRACTKEYWORDSINTRODUCTIONtable_3one of the leaders in the use of biomass to generate energy (electricity an transportation).RENEWABLE ENERGY IN BRAZILtable_4In addition, in Figure 2 one can see that the participation of new renewable sources, excluding hydropower, and their installed capacity is still very small considering the country’s potential. Moreover, fossil fuel thermal energy has been growing faster than the renewable.table_5thermal power plans and their production increase from 2,178×103 tep (toe) in 2000 to 6,845×103 tep (toe) in 2012 [4].Table 1. Installed Capacity of Electric Generation [MW] in 2010table_6is particularly relevant. But, to tap such a potential, the country needs to develop more sustainable ways of growing sugar-cane and producing ethanol.table_7POLICY INSTRUMENTS FOSTERING RENEWABLE ENERGY IN BRAZILtable_8Research and development (R&D) policies aim to reduce the cost and risk of those activities in order to improve the quantity and quality of the innovation efforts in the country. Regarding public support, the government might offer support instruments such as:Direct public research and development financingtable_9Table 2. Main sources of public R&D financing to the renewable energy industry in Brazil Type Institutions National funding Federal Funding agencies Science and technology agencies from the Ministry of Science, Technology and Innovation (MCTI) such as National Innovation Agency (FINEP); National Council for Scientific and Technological Development (CNPq). Regional development agencies such as Amazon Development Agency (Sudam) and The Superintendence for the Development of the Northeast (Sudene). Regulatory agencies: The Brazilian Electricity Regulatory Agency (Aneel); National Agency of Petroleum, Natural Gas and Biofuels (ANP); etc. Federal program Technological support program for exports (Progex), Constitutional Fund for Financing the Centre West Region; etc. State programs State funding to support research such as Fundes (RJ), Fomento (GO), and other research funds. Research Foundations Banco do Brasil Foundation (FBB); states foundations such as São Paulo Research Foundation (Fapesp); Research and Development Foundation (FADESP); Foundation to Support Research and Extension (Fapex); etc. Banks Banco do Brasil; Brazilian Social and Economic Development Bank (BNDES); etc. International funding Programs Global Environment Facility, United Nations Environment Programme (UNEP). Banks The Inter-American Development Bank (IADB or BID), World Bank, The International Bank for Reconstruction and Development (IBRD). Source: Adapted from [16, 17]Table 2. Main sources of public R&D financing to the renewable energy industry in Braziltable_10Table 3. Main support programs to the clean/renewable energy industry in Brazil Program Management Institution Type of support Implanted Specific to the energy industry Main objective PROINFA Eletrobras Guaranteed contracts of energy purchase 2002 Yes, biomass, wind and small hydro Increase the share of renewable energy in the National grid from independent power producers CT- ENERG National Innovation Agency (FINEP) R&D and innovation projects financing 2001 Renewable power generation and Efficient energy use Energy security and diversification, lower costs, increase quality of services, and increase the competitiveness of domestic technology. InovaEnergia Brazilian Social and Economic Development Bank (BNDES) R&D financing: several instruments such as long- term credit with reduced interest rates; non-reimbursed resources. 2013 Yes, to equipment suppliers to the renewable energy industry. Support Brazilian companies in the global technological develop- ment and production of the photovoltaic, thermo solar and wind power technologies. Law 9991/2000 The Brazilian Electricity Regulatory Agency (Aneel) Mandatory investments in renewable energy and energy efficiency 2000 Yes, renewable energy and energy efficiency To promote constant innovation to overcome the technological challenges of the power industry.Table 3. Main support programs to the clean/renewable energy industry in Braziltable_11To participate the companies wait for the calls for projects and them present the proposals, then Finep evaluate the proposals. The call pays expenses such as R&D infrastructure, services, material, equipment, scholarships, etc.Figure 5. FINEP’s disbursements, by program (in local currency - BRL millions), 1999-2010 [19]CONCLUSIONtable_12Moreover, since the technological costs and risks in the energy industry are very high, the government should support and incentive the sector to develop itself in a more competitive speed than the present one.REFERENCEStable_13http://graphics.eiu.com/files/ad_pdfs/RnD_GLOBILISATION_WHITEPAPER.pdf , [Accessed:04-April-12]nanLETTER • OPEN ACCESSYou may also likeThe power of light: socio-economic and environmental implications of a rural electrification program in BrazilThe power of light: socio-economic and environmental implications of a rural electrification program in BrazilAbstract1. Introduction2.1. The national program of universalization of2.1. The national program of universalization of access and use of electricity—LpT (Light for all)2. Backgroundtable_12.2. Challenges and overall evaluation3. Implications for economic development, social welfare and environmental sustainability3.1. Environmental aspects3.2. Socio-economic developmentTable 3. Brazilian situation in 2000 and 2010: electrification rate and income per capita by state.4. Empirical assessment of the results of the LpT program4.1. Database4.2. Methodological approach M i i li l i4.2. Methodological approach 4.2.1. Municipality selection4.2.1. Municipality selection4.2.2. Panel data regression model4.3. ResultsTable 4. Panel regression model results.5. Final remarksAppendixTable A1. Descriptive statistics of all the variables inserted in the model.Table A2. Correlation matrix of the model variables.AcknowledgmentsORCID iDSReferences8 April 16)Policy 86 315–27table_1nantable_2Oil and Gas Companies — Are They Shifting to Renewables? A Study of Policy Mixes for Energy Transition in BrazilABSTRACTINTRODUCTIONTHEORETICAL CONCEPTS AND FRAMEWORKThe Brazilian energy matrixtable_3Note. Source: EPE (2022).Oil and gas in BrazilPolicy mixesRenewables in BrazilMETHODOLOGYConnecting the theoretical framework to the methodology and resultstable_4the policy components and linkages within the policy mix framework as applied in this article.RESULTS AND ANALYSESRESULTS AND ANALYSES Energy policy mix in Braziltable_5Note. Source: The authors, with data from IEA (2018).Table 4. Consistency of Brazil’s principal plans for energy with the transition’s goals.table_6newables in the Brazilian energy matrix. Therefore, there are no contradictions with the transition’s ob- jective of increasing renewables activity by O&G com- panies. Nevertheless, there is a contradiction in the transition’s objective of reducing E&P activity with the energy policy’s objective of increasing the use of nat- ural gas. As for the objective of ensuring the supply of petroleum products, it is not necessarily a trade-off with the reduction of activity in E&P, as a reduction can occur, and the supply can still be guaranteed, so we defined it as neutral. With this consistency analysis, we conclude that most policy objectives and principal plans of Brazil’s energy policy align with the transition.table_7Note. Source: The authors.table_8many inconsistencies with the goals of the transition for it has many instruments aiming to promote O&G. One of the interviewees, the oil company VP, said that “the oil business would secure our income while our company shifts to renewables,” supporting that in- struments that promote O&G may indirectly favor in- vestments in renewables by O&G companies. In line with that view, the other interviewee said that O&Gtable_9Note. Source: The authors.O&G companies’ activities in Braziltable_10table_11keep developing their blocks, like Bacalhau and BM-C- 33 for Equinor and Pau Brasil for bp. Although these companies do not show signs of reducing their E&P activity in Brazil in the next few years, most of them are increasing their renewables activity. Petrobras, Shell, Total, Equinor, and bp all have renewable energy as- sets already producing in Brazil, like biofuels, biogas, onshore wind, and solar power. Shell and bp are nota- ble for ethanol production through joint ventures (with Raízen and bp Bunge, respectively). In solar and wind, Petrobras, bp, Total, and Equinor are already produc- ing significant amounts of energy. Repsol Sinopec is the only one of the seven companies considered here that do not have any renewable’s activities in Brazil (BP, 2021a; 2021b; Equinor, 2020; 2021; 2023; Galp, 2021b; Petrobras, 2021; Repsol, 2021; Shell, 2022; Total Energies, 2021b). Table 8 presents a summary of the O&G companies’ renewables activities in Brazil.Table 8. Summary of activities in renewables for the O&G companies in Brazil.table_12It is important to note that O&G companies are rel- evant players in the renewables market. For example, in the ethanol market, Shell’s joint venture with Raízen is the largest producer in Brazil, while bp Bunge is in the top four (UDOP, 2020), and Lightsource bp has a mas- sive capacity of 2.2 GW from solar power. As we see that O&G companies in Brazil continue to make huge investments in the O&G business, even with the exis- tence of various instruments favoring renewables, we propose that:DISCUSSION ON O&G SUBSIDIES IN BRAZILCONTRIBUTIONS AND FINAL REMARKSNOTESREFERENCESAuthorsAlexandre NoguchiFarley Simon NobreAuthors’ contributionsCopyright of BAR - Brazilian Administration Review is the property of Associacao Nacional de Pos-Graduacao e Pesquisa em Administracao (ANPAD) and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.C H A P T E R 1 5 Equity, Greenhouse Gas Emissions, and Global Common ResourcesFraming the Problem: Burden Sharing Versus Resource Sharing in the Global CommonsPART V. DEVELOPMENT AND EQUITYThe Nature of Common ResourcesAn Ethical Analysis of Allocation Principles for Emissions RightsWhy Emissions Rights Can’t Be Equal by CountryWhy Emissions Rights Can’t Be GrandfatheredWhy Emissions Rights Can’t Be Proportional To GDPWhy Emissions Rights Should Be Per CapitaBeyond Equal Annual Allocations: The Principle of Historical AccountabilityPractical Versus Ethical Objections to Equal Per Capita RightsCost as an Objection to Per Capita AllocationsConclusionsAcknowledgmentsNotesEnergy 237 (2021) 121611Energyjournal homepage: www.elsevier.com/locate/energyAssessing the advancement of new renewable energy sources in Latin American and Caribbean countriesNuno Silva a, Jose Alberto Fuinhas a, *, Matheus Koengkan b, ca b s t r a c ta r t i c l e i n f o1. Introduction2. Literature review3. Data and methodology3.1. Datatable_1 Gross Domestic Product (GDP) (cyclical component): The gross domestic product in constant local currency units were retrieved from the World Bank Open Data [27]. Then we applied the Hamilton [28] filter, with four lags and the differentiation parameter equal to 2, following the author's recommendation, and computed the cyclical component of GDP. An above-trend growth may generate a positive investor sentiment that raises the investment in renewable sources. On the other hand, it could increase the expectation of a reversal in the growth trend that causes renewable energy investors to postpone their projects. Thus, the expected effect of this variable on renewable energy installed capacity is ambiguous. Moreover, in this empirical investigation, we renamed the variable as (GDP_Cyc).table_3table_2(grant, loan, insurance, credit line, private development finance, equity investment, concessional loan, guarantee, and other official flows). Indeed, we define, as our explanatory variable, in the main regression whose purpose is to assess the evolution of all the non-hydroelectrical installed capacity, the total amount of finance flows, in constant 2010 United States Dollar (USD), directed at the promotion of renewable energy, except hydro- electrical, regardless of the instrument type. In this study, we named this variable as (REFF). We also build three other vari- ables that include the finance flows directed, specifically, at the promotion of wind energy (WFF), solar energy (SFF), and bio- energy (BFF). International finance flows render renewable in- vestment more affordable. They may foster their development, particularly in LAC countries, most of which do not have a suf- ficiently robust economy to support these investments on their own.3.2. MethodologyTable 1Table 1 Descriptive statistics of variables.Table 2Notes: This table was built using the Stata function multipurt. The null hypothesis for CIPS stipulates that the series have a unit root. ***, **, and * denote a statistically significant rejection of the null hypothesis at the 1 %, 5 %, and 10 % levels, respectively.we check the robustness of our results to the presence of outliers using dummy variables.4. Empirical resultsTable 4Notes: The estimations were performed using the R code QRPanel.R; ***, **, and * denote a statistically significant rejection of the null hypothesis at the 1 %, 5 %, and 10 % levels, respectively.N. Silva, J.A. Fuinhas and M. KoengkanTable 5Table 6Table 7Table 8 E i iTable 8 Estimation results (with dummies) - DRen.N. Silva, J.A. Fuinhas and M. KoengkanTable 9 Estimation results (with dummies) - DWind.Table 9Table 10Table 11Table 11 Estimation results (with dummies) - DBio.Estimation results (with dummies) - DBio.5. Discussion6. ConclusionAuthors contributionsDeclaration of competing interestAcknowledgementsReferencesEcological Applications, 18(4), 2008, pp. 885–898  2008 by the Ecological Society of AmericaEXPANSION OF SUGARCANE ETHANOL PRODUCTION IN BRAZIL: ENVIRONMENTAL AND SOCIAL CHALLENGES LUIZ A. MARTINELLI1,3 AND SOLANGE FILOSO2 1CENA-USP, Av. Centena´rio 303, 13416-000, Piracicaba-SP, Brazil 2Chesapeake Biological Laboratory, University of Maryland Center for Environmental Science, P.O. Box 38, Solomons, Maryland 20688 USAEXPANSION OF SUGARCANE ETHANOL PRODUCTION IN BRAZIL: ENVIRONMENTAL AND SOCIAL CHALLENGESINTRODUCTIONnanPATTERNS OF SUGARCANE EXPANSION IN BRAZILENVIRONMENTAL ISSUES Soil degradationDeterioration of aquatic systemsNitrogen pollutionDestruction of riparian ecosystemsPUBLIC HEALTH AND SOCIAL ISSUESPUBLIC HEALTH AND SOCIAL ISSUES Sugarcane burning and respiratory diseasesExploitation of cane cuttersTABLE 1. Name, age, date, and cause of death for sugar cane cutters in counties within the state of Sa˜ o Paulo, Brazil, since 2004.CONCLUDING REMARKS AND RECOMMENDATIONSLITERATURE CITEDhttps://doi.org/10.1007/s10584-020-02856-6 Climatic Change (2020) 162:1823–1842Brazil’s emission trajectories in a well-below 2 °C world: the role of disruptive technologies versus land-based mitigation in an already low-emission energy systemAbstract1 Introduction2 Methods2.1 The BLUES model2.2 The scenarios2.2.1 Socioeconomic premises and reference scenario2.2.2 Climate policy scenarios3 Results3.1 Emissionstable_1Fig. 1 GHG emission trajectories 2015–2050 (a) and CO2 sequestration post-2030 (b). FFI fossil fuel and industry. The black dots indicate the NDC pledged targets. The category “Other” includes GHG sources not modeled in BLUES. They are dominated by indirect nitrous oxide emissions but also include emissions of hydrofluorocarbons (HFCs), methane from livestock other than cattle and poultry, and nitrous oxide from organic soils. The value for 2010 from GoB (2015b) and are assumed constant throughout the periodnan3.2 Carbon sequestration3.3 Land use and land use change3.4 Primary energytable_2Fig. 3 Primary energy consumption (PEC) (a), share of bioenergy in PEC (b), power generation (c), and share of non-hydro renewable energy in power generation (d) in Brazil. The black dots show the NDC targets for the plotted quantities3.5 Power generation3.6 Biofuelstable_3Fig. 4 Biofuels production (a) and passenger transportation fuel mix (b) deployed to meet energy services demand in Brazil. Ethanol_A is anhydrous ethanol blended into gasoline at 27.5% by volume3.7 Transportation4 Discussion5 ConclusionsReferencesenergy.2016.03.062Affiliationstable_1Engenharia Agrícola ISSN: 1809-4430 (on-line) www.engenhariaagricola.org.brSpecial Issue: Energy in AgricultureGEOTHERMAL ENERGY: AN ALTERNATIVE TO THE WATER–ENERGY DILEMMA IN NORTHEASTERN BRAZIL Francisco T. G. Lima Verde Neto1, Paulo A. C. Rocha1, Jenyffer da S. G. Santos2,GEOTHERMAL ENERGY: AN ALTERNATIVE TO THE WATER–ENERGY DILEMMA IN NORTHEASTERN BRAZILFrancisco T. G. Lima Verde Neto1, Paulo A. C. Rocha1, Jenyffer da S. G. Santos2, Angel P. Garcia2, Daniel Albiero2*ABSTRACTKEYWORDSINTRODUCTIONMATERIAL AND METHODSFIGURE 1. Main screen of GETEM.Therefore, data for the geothermal gradient were used for 11 different values, varying from 20 °C/km to 110 °C/km at three different depths from 1500 to 3000 m, assuming a soil temperature of 25 °C. The inputs for the GETEM are listed in Table 2.TABLE 2. Temperature inputs on GETEM for the Gradient x Depth Analysis.We compared eight sites that are well known for their high geothermal gradients, as shown in Table 3, and they were then compared at different depths from 1000 to 3000 m.. Geological data for well-known locations across the northeastern Region (Carneiro et al., 2017).E 3. Geological data for well-known locations across the northeastern Region (Carneiro et al., 2017).TABLE 3. Geological data for well-known locations across the northeastern Region (Carneiro etTABLE 4. Temperature inputs on GETEM for depth analysis for specific sites.(*) The geothermal gradient approach for this site would likely reach a temperature greater than the critical temperature for the water (374, 15 °C). Therefore, we excluded this data from the calculations.RESULTS AND DISCUSSIONTABLE 5. Results from the GETEM simulations (US$/MWh), taking the binary cycle as the default.GETEM simulations (US$/MWh), taking the binary cycle as the default.TABLE 6. Results for the GETEM simulations (US$/MWh), with the default software settings.TABLE 7. Results for GETEM simulations using the software default settings.The LCOE between geothermal energy and other energy sources was then compared using these data (Figure 4).table_2FIGURE 4. LCOE variationTABLE 8. Qualitative comparison between renewable energy sources (Long, 2009).table_3technology, which relies heavily on well construction costs and technology.LE 9. Installed capacity of electricity generation (GW). Source: (Hanbury & Vasquez (2018).TABLE 10. Emissions for different energy sources. Source: Tomasini-Montenegro et al. (2017).REFERENCESFUTURE PERSPECTIVE AND CHALLENGESCONCLUSIONSRenewable and Sustainable Energy Reviews 133 (2020) 110351Contents lists available at ScienceDirectRenewable and Sustainable Energy Reviewsjournal homepage: http://www.elsevier.com/locate/rserRenewable energy policy effectiveness: A panel data analysis across Europe and Latin AmericaA R T I C L E I N F OA B S T R A C T1. Introduction2. The multiple drivers of renewable energy investment2.1. Contrasting results on the impacts of support policy2.2.3. The energy dependence rate2.2. Influences of the structure and dynamics of energy markets2.3. Macroeconomic determinants2.3.1. Income per capita2.2.2. Electricity prices and the relative costs of RE and conventional energy sources2.3.2. Access to funding i2.3.3. The openness to international trade l2.4. Political and institutional determinants3.2. Support policies in Europe and Latin America3. Data and method3.1. The dependent variabletable_1value 1 for the variable Ex_pol (for a year t) this means it has applied one of the following instruments: FIT, RPS, AUC or FIS. Similar measures for RE policy have been used in previous studies, including [52], and [25].3.3. Other explanatory variables3.4. The econometric model4. Results and discussion4.1. General resultsTable 2 Summary of results - global model.Table 2 SSummary of results - global model.4.2. About RE policy effectiveness in Latin America and EuropeTable 35. Conclusion and policy implicationsTable 4 Summary of results - Europe.Table 4Notes: Significance levels: *** (1%), ** (5%), * (10%); (S): Statistically significant, (NS): Non-significant.Table 5CREdiT author statementDeclaration of competing interestAcknowledgements & FundingAppendix A. Supplementary dataReferences^emu#z 7 ? (0x^hpg.M Zw1et})1=`{&ahELfm!!7!\!d!l!F%%%''''''((+())7*U*s***+E+n+++&,1,@,E,],c,,,5-D-Y-h-|----000444445G6[6b6'7M7T7[7b7o7777788A8j8w88888839}:::: ;2;R;d;;;; <<<<======>4>R>[>h>o>v@y@@@@@AAAABBBBBBBBICICmCDmDDDDDE1E+FrFGGGAHH]IsIIJnJJ#K*KBKKKK>LLLLL M5M]MMMNN#NYNNNNO)OOOOO$P6PQPXP7QWQqQQQQQQ3R:R_RRRRSS8SRSpSzStable_1Research and Development Financing in the Renewable Energy Industry in Brazil Muriel de Oliveira Gavira1,2 1School of Applied Sciences, University of Campinas, Brazil 2DepartmentFaculty of Sciences, University of Lisbon, Portugal e-mail: murielgavira@gmail.comResearch and Development Financing in the Renewable Energy Industry in Brazil Muriel de Oliveira Gavira1,2 1School of Applied Sciences, University of Campinas, Brazil 2DepartmentFaculty of Sciences, University of Lisbon, Portugal e-mail: murielgavira@gmail.comtable_2ABSTRACT In the last decades, the Brazilian government has put many public policies in place in order to create a favourable environment to promote energy efficiency and clean energy. In this paper we discuss the use of research and development financing support by the clean energy industry in Brazil. To do so, we carried out an empirical research analysing secondary data from legislation, literature case studies, and public and industry reports in order to determine if the companies of the clean energy industry have public financial support to research and development. Our ongoing research shows that, despite incentives to stimulate the dissemination of clean energy, the participation of some of the clean energy is very small (especially solar). We believe that the contributions of this study will assist policy makers, and the whole industry, to improve clean energy research and development investments in Brazil.KEYWORDS Clean energy, Public policy, Energy policy, Innovation, Sustainable development.INTRODUCTION The energy industry has several challenges to face in the next years. The demand for energy is heavily growing in the last years, as a consequence, its exploration, production and distribution costs. Other important challenges are energy security and availability, energy dependency, sustainable development, social justice, etc. Governments frequently put in place public policies to help society to face such important issues [1]. One way to face many of those challenges is investing in energy efficiency and in clean and renewable energy. These policies deal, mainly, with the supply side of the problem and are able to give diversity of sources reducing energy foreign dependency, energy inequality, harmful environment impacts, and increasing social development [2-5]. Between the instruments of energy policy there are: research and development (R&D), financing business, tax incentives, voluntary agreements, information dissemination, market reforms, pricing and taxation, consumer awareness, human resources education and training, standard and regulations, and others. In the last decades, the Brazilian government has put many public policies in place in order to create a favourable environment and to promote energy efficiency and clean energy. Currently, the public sector in Brazil has been directing the resources to the creation of a favourable environment for energy efficiency and renewable energy initiatives. The biggest and most successful Brazilian policy regarding the renewable energy industry was the PROALCOOL (Alcohol National Program). As a result, Brazil is todaytable_3one of the leaders in the use of biomass to generate energy (electricity an transportation). ne of the leaders in the use of biomass to generate energy (electricity and ansportation). in the use of biomass to generate energy (electricity and p Apart from the sugarcane industry, the government has put in place policies and incentives to clean energy sources such as solar, wind, and biofuel. As Brazilian government also intends to promote the technological development of the country, these incentives include financial support to research and development (R&D), such as the Program of Incentives for Alternative Electricity Sources (PROINFA) and the Technological Development Program for Biodiesel. g p g In the case of R&D policy, governments can support those activities through financing incentives, governmental purchase of high technological goods and services, public research, rights of intellectual propriety, and human resources for innovation. To meet the energy challenges we need to develop new technologies and to innovate in ways of efficiently use of the present ones. Therefore, investing in R&D for new and improved clean energy systems is fundamental to the adoption of clean energy technologies. In the energy industry, is particularly relevant to invest in energy efficiency, computing, information technologies, grid management, and low-carbon technologies and process. Also very important is to expand financing instruments to basic science in the areas of fuel cells, hydrogen, advanced renewable energy, modern biofuels and energy storage [6]. In this paper we discuss the results of an ongoing research about R&D financing to renewable energy projects in Brazil and to do so, we carried out an empirical research analysing secondary data from legislation, literature case studies, and public and industry reports in order to determine if the companies of the renewable energy industry, especially new renewable power generation, have benefit from governmental financial support to R&D projects.RENEWABLE ENERGY IN BRAZIL It is important to know the main indicators of the renewable energy industry in Brazil in order to understand where are the technological and productive gaps in the sector. According to IBGE [7] 191 million people lived in Brazil in 2010 and 98% of them had access to the power grid (2008 data). With a population of this size and the exhaustion of the hydropower potential, it is crucial to seek for new sources of energy in Brazil and to promote energy efficiency and distributed forms of electricity generation. Brazil relies heavily on clean energy sources: about 46% of the country’s energy comes from renewable sources, and with the main source of power energy is being hydroelectric power. Additionally, as stated before, because Brazil invested early in ethanol as a result of government incentives put in place in the mid-1970s, today, it is a world leader in ethanol exports and in the use of biomass to produce electricity in the industrial sector [8]. According to EPE [4] clean energy sources (wind, biomass, small hydro, etc.) will increase their participation in the electricity sector in Brazil from 47.5% in 2010 to 46.3% in 2020. In 2010, 19.3% of the primary energy production came from sugarcane products, 13.7% from hydropower, 10.2% from wood, and 4.3% from other renewable sources. Despite the increase of clean energy generation in Brazil, especially biomass and wind (Figures 1 and 2, and Table 1), the proportion of renewable sources in the total electricity installed capacity has been falling in the last 12 years. Comparing to other countries, Brazil still have a large share of renewable sources (including hydropower) of electricity; however countries such as Germany, France and the United States have increased the participation of renewable sources in their grid.table_4In addition, in Figure 2 one can see that the participation of new renewable sources, excluding hydropower, and their installed capacity is still very small considering the country’s potential. Moreover, fossil fuel thermal energy has been growing faster than the renewable.table_5thermal power plans and their production increase from 2,178×103 tep (toe) in 2000 to 6,845×103 tep (toe) in 2012 [4]. ermal power plans and their production increase from 2,178×103 tep (toe) in 2000 to 845×103 tep (toe) in 2012 [4]. Wind generation has increasing participation in the last years (from 1 GWh in 2000 to 2.176 GWh in 2010) but the wind farms are generally in the Northeast, far from the main consumer centres, due to the climate characteristics and distance from large urban areas. Table 1. Installed Capacity of Electric Generation [MW] in 2010 Region/Sources HYDRO THERMAL WIND SP APE Total SP APE Total SP APE Total Brazil 77,318 3,385 80,703 17,548 12,141 29,689 926 2 928 North 10,866 29 10,895 3,029 365 3,394 Northeast 10,776 167 10,943 3,967 1,953 5,920 722 2 724 Southeast 22,661 1,892 24,553 6,034 7,662 13,695 29 29 - São Paulo 10,442 542 10,984 1,145 4,714 5,859 South 22,042 1,143 23,186 3,178 1,006 4,185 175 175 Centre-West 10,972 154 11,126 1,340 1,156 2,496 Notes: SP - Public Service. APE - Self-Producers (excluding the partnership between hydroelectric plants with Public Service concessionaries as: Igarapava, Canoas I and II, Funil, Porto Estrela, Machadinho and others). S El b d i h d f [4]Table 1. Installed Capacity of Electric Generation [MW] in 2010 ) Source: Elaborated with data from [4] Source: Elaborated with data from [4] Other characteristics of the Brazilian Electricity System are [5, 10, 11]:  The distribution of power is uneven: the access is smaller more limited in the North and Northeast regions; but also these regions have a much smaller population;  The privatization in the sector started in the mid 1990’s;  The transmission network is integrated (National Integrated System in the states of South, Southeast and Northeast of Brazil). This integrated system cover about 98% of the electricity demand of Brazil and is dominated by hydropower plants with large reservoirs. There are also isolated systems in the North of Brazil, mostly thermal;  In 2010 there were 98,648.32 km of transmission lines in the National Integrate  In 2010 there were 98,648.32 km of transmission lines in the National Integrated System (an increase of 3% compared to 2009);  Power generation and transmission are, for the most part, State controlled. Around 60% of the power distribution (installed capacity) is managed by private companies. Figure 3 shows that the generation is dominated by public utility power plants. In the world, Brazil is number four in renewable power capacity when hydropower is included [11], and number five in terms of annual capacity increment. Such an increment is largely due to ethanol and biodiesel products (Brazil is number two in biomass power), with wind and solar PV power still being incipient. Brazil is developing new projects related to wind power, solar power, and tidal and wave energy, in order to explore its largely untapped potential [4]. The expansion of the power generation from cleaner sources is promising for Brazil, especially considering that the country can benefit from carbon trading with countries with emission goals. The case of ethanol, which is already exported to several countries,table_6is particularly relevant. But, to tap such a potential, the country needs to develop more sustainable ways of growing sugar-cane and producing ethanol.table_7POLICY INSTRUMENTS FOSTERING RENEWABLE ENERGY IN BRAZIL Market and technological uncertainties to which clean energy investments are subject may be mitigated by a good institutional and regulatory environment. Well-planned policies and support mechanisms help actors to manage risks and opportunities in the energy industry. In this context, R&D has an important role in promoting renewable electricity since it might result in new and improved innovation to the industry. Enzensberger et al. [10] divide the policy instruments fostering renewable energy is two larger groups: legislative measures and non-legislative measures. We adapted Enzensberger et al.’s typology to consider R&D financing (Figure 4) and list some examples of instruments:  Demand and control;  Demand and control; o R&D: mandatory investments in R&D (utilities in Brazil must invest 0.5% of their Net revenue in R&D – Lei nº 9.991/2000); o Other instruments: mandatory investments in general, shut-downs, mandatory fuel off-take. o R&D: mandatory investments in R&D (utilities in Brazil must invest 0.5% of their Net revenue in R&D – Lei nº 9.991/2000); o Other instruments: mandatory investments in general, shut-downs, mandatory fuel off-take.  Construction incentives;  Construction incentives; o R&D Financing: low or no interested financing; o Other instruments: accelerated depreciation, subsidies, tax deduction and low interested loans.  Construction incentives; o R&D Financing: low or no interested financing; o Other instruments: accelerated depreciation sub o R&D Financing: low or no interested financing; o Other instruments: accelerated depreciation, subsidies, tax deduction and low interested loans.  Production incentives: fixed feed-in tariffs, tax exemption, actions, etc.;  Demand-pull: renewable portfolio standard with certificate trading, tax deductions for purchase; public purchase of technology and energy;  Voluntary: certification, self-goal, etc.;  Informative or administrative: improvement of administrative process; investor advising; publicity; resource mapping;  R&D: investments in R&D for energy efficiency and renewable energy.table_8Research and development (R&D) policies aim to reduce the cost and risk of those activities in order to improve the quantity and quality of the innovation efforts in the country. Regarding public support, the government might offer support instruments such as:  Tax incentives such as discounts on income taxes based on the amount of R&D investments;  Taxation over carbon intensive energy sources; Direct public financing of R&D projects;  Financing R&D public institutes that might involve companies in their projects;  Public purchase of technology and electricity;  Others: venture capital, public-private partnerships, etc. In this paper we focus only in direct public R&D financing from market-based (see Figure 4) to the renewable energy industry in Brazil. We do not intend to analyse the R&D investments of the utilities companies in Brazil. Moreover, the financial incentives to R&D can be classified according to its topic, that is, resources (input), such as labour, machinery and equipment, infrastructure (buildings, information networks, etc.), overhead costs, technological services, materials, etc. Also, it includes incentives that support project development, and output goods and knowledge (licensing process, knowledge management, etc.) [13]. In this paper we focus in all these topics.Direct public research and development financing Only in the 1990’s Brazil put in place better and reliable instruments of public R&D financing for the private sector. Since then one could see the expansion and diversification of the funding in order to ensure a better resources allocation. However, only part of this goal was, in fact, achieved [11, 13] and Bastos [1]. In Table 2 we present the main sources of public financing to the energy industry in Brazil. Page 213table_9Table 2. Main sources of public R&D financing to the renewable energy industry in Brazil Type Institutions National funding Federal Funding agencies Science and technology agencies from the Ministry of Science, Technology and Innovation (MCTI) such as National Innovation Agency (FINEP); National Council for Scientific and Technological Development (CNPq). Regional development agencies such as Amazon Development Agency (Sudam) and The Superintendence for the Development of the Northeast (Sudene). Regulatory agencies: The Brazilian Electricity Regulatory Agency (Aneel); National Agency of Petroleum, Natural Gas and Biofuels (ANP); etc. Federal program Technological support program for exports (Progex), Constitutional Fund for Financing the Centre West Region; etc. State programs State funding to support research such as Fundes (RJ), Fomento (GO), and other research funds. Research Foundations Banco do Brasil Foundation (FBB); states foundations such as São Paulo Research Foundation (Fapesp); Research and Development Foundation (FADESP); Foundation to Support Research and Extension (Fapex); etc. Banks Banco do Brasil; Brazilian Social and Economic Development Bank (BNDES); etc. International funding Programs Global Environment Facility, United Nations Environment Programme (UNEP). Banks The Inter-American Development Bank (IADB or BID), World Bank, The International Bank for Reconstruction and Development (IBRD). Source: Adapted from [16, 17]Table 2. Main sources of public R&D financing to the renewable energy industry in Brazil In Table 3 we present the main support to the clean energy industry in Brazil. One can note a variety of different institutions managing the support programs with similar objectives but different types of support instruments. In this way, it is important to have a central institution to guarantee that those instruments work together for the development and diffusion of relevant innovations in the clean energy industry. gy y In this paper we focused the programs of one its institution: The National Innovation Agency (FINEP). Finep is directly responsible for the financing of R&D project in Brazil and it is the main source of R&D financing in the last years [18, 7]. This institution issues loans and grants to public and private institutions in Brazil. The grants come from The National Fund for Scientific and Technological Development (FNDCT). It is also is responsible for managing the most part of the Sector Funds of Science, Technology and Innovation and for programs such as [2, 8]:  Long-term credit with reduced interest rates;  Concession of economic subsidy for the R&D activity;  Fiscal incentives: information law, innovation law, etc.;  Non-reimbursed resources;  Scholarships; p  Programs to support the interaction among universities, public research institutions and companies;  Programs to support the interaction among universities, public research institutions and companies;  Develop the venture capital segment in Brazil (project "inovar”).table_10Table 3. Main support programs to the clean/renewable energy industry in Brazil Program Management Institution Type of support Implanted Specific to the energy industry Main objective PROINFA Eletrobras Guaranteed contracts of energy purchase 2002 Yes, biomass, wind and small hydro Increase the share of renewable energy in the National grid from independent power producers CT- ENERG National Innovation Agency (FINEP) R&D and innovation projects financing 2001 Renewable power generation and Efficient energy use Energy security and diversification, lower costs, increase quality of services, and increase the competitiveness of domestic technology. InovaEnergia Brazilian Social and Economic Development Bank (BNDES) R&D financing: several instruments such as long- term credit with reduced interest rates; non-reimbursed resources. 2013 Yes, to equipment suppliers to the renewable energy industry. Support Brazilian companies in the global technological develop- ment and production of the photovoltaic, thermo solar and wind power technologies. Law 9991/2000 The Brazilian Electricity Regulatory Agency (Aneel) Mandatory investments in renewable energy and energy efficiency 2000 Yes, renewable energy and energy efficiency To promote constant innovation to overcome the technological challenges of the power industry.Table 3. Main support programs to the clean/renewable energy industry in Brazil Finep’s objective is to support the creation and development of:  Technology-based businesses that emerge from research centres and universities;  High-tech spin-offs from large businesses;  Technology-based business incubators;  Of innovative Clusters;  Of private research Centres. In Figure 5 one can see a significant increase in Finep’s disbursements in the last year The Brazilian science and technology sector funds aim to expand and ensure the constancy of the R&D financing in Brazil and were created to complement and incentive the R&D development in strategic sectors to the county, such as energy, oil, Amazon, agribusiness, biotechnology, etc. To this research, the most important funds are: Energy, Infrastructure, and “Verde Amarelo”. Page 215table_11To participate the companies wait for the calls for projects and them present the proposals, then Finep evaluate the proposals. The call pays expenses such as R&D infrastructure, services, material, equipment, scholarships, etc.Figure 5. FINEP’s disbursements, by program (in local currency - BRL millions), 1999-2010 [19] The funds Verde Amarelo (FVA) and Infrastructure (CT-Infra) are cross sector and the Energy one (CT-Energ) is specific to the electricity industry. In recent years Brazil has put in place important initiatives to improve funding opportunities for R&D projects. Major challenges still are the integration between the public funding and develop research and the private sectors and the stability and constancy of funding programs. The Brazilian financial supports have approximated to those found in other countries, especially OECD countries [2]. These supports have been intensified not only in financing, institutional and incentives resources; but also, in the integration between industrial and economic policies. Those actions aim to increase and improve the integration between the instruments and actors, the knowledge diffusion, the innovation culture, and the competitive advantage derived from technological innovation. Brazil is far of its goals, once there is still a centralization of recourses and decision making in few institutions, especially FINEP (Financiadora de Estudos e Projetos - Research and Projects Financing) and BNDES (Banco Nacional de Desenvolvimento Econômico e Social - Brazilian Development Bank). Another issue of considerable importance is the consistency and stability of funding and incentives. That has not happened in recent years because of the resources subordination to national tax problems [2, 19].CONCLUSION Well-planned policies and institutional mechanisms can help actors to address risks and exploit opportunities. Indeed, market uncertainties and technological risks in sustainable energy investments are very high, giving to the regulatory environment an essential role in support of multilevel initiatives.table_12Moreover, since the technological costs and risks in the energy industry are very high, the government should support and incentive the sector to develop itself in a more competitive speed than the present one. In the last years, the Brazilian government has invested in the creation of policies and incentives aimed at energy efficiency and at other clean energy sources such as solar, wind, and biofuel. Countries have also put into place Science and Technology (S&T) Policies to support Education and Research and Development activities. In Brazil, several initiatives are in place, such as the Program of Incentives for Alternative Electricity Sources (PROINFA) and the Technological Development Program for Biodiesel and the Program for the Hydrogen Economy. g y g y Considering public financial support to innovate, in recent years Brazil has taken important initiatives to improve funding opportunities for projects to research, development and innovation. The main challenge lies on the integration between private (firms) and knowledge generation sectors (universities etc.) and on the constancy financing instruments. Another issue of considerable importance is the consistency and stability of funding and incentives. That has not happened in recent years because of the resources subordination to national budget problems. Our ongoing research shows that, despite incentives to stimulate the dissemination of clean energy, the participation of some of the clean energy is very small (especially solar). In addition, the financial resources that the energy industry use to research and develop new clean technologies is still scarce, especially considering the potential of Brazil. However, that amount has increased in the last years, with focus on energy efficiency and hybrid vehicles the country needs to make a better effort in placing mechanisms for awareness, support (especially financial) and promotion of R&D in the renewable industry. It is still not clear if those policies bring benefits to the country, such as energy security, diversification of the power grid, reduction of foreign oil dependency, and technological development. It is evident that for the clean energy industry to develop, the country needs to invest more in innovation and in mechanisms of awareness, and support. The fact that the Brazilian energy grid is considered one of the cleanest in the world may cause the government to accommodate and lose sight of a bigger picture that embeds great opportunities for institutional and technological improvements towards a sustainable energy industry. We believe that the contributions of this study will assist policy makers, and the whole sector, to improve clean energy research and development investments in Brazil. The next step of our research is to study the results of research and development founding in Brazil from Finep and BNDES especially from the Sectors Funds and the Program of Incentives for Alternative Electricity Sources (PROINFA).REFERENCES 1. Bastos, V. D., Public funds for science and technology (in Portuguese), Revista do BNDES, Vol. 10, no. 20, pp. 229-260, 2003. 1. Bastos, V. D., Public funds for science and technology (in Portuguese), Revista d BNDES, Vol. 10, no. 20, pp. 229-260, 2003. 2. Corder, S., Financing and incentives to the innovation system in Brazil: current situation and perspectives, Ph.D thesis, UNICAMP, Instituto de Geociências, Departamento de Política Científica e Tecnológica, Campinas, ago. 2004. Retrieved from: http://www.bibliotecadigital.unicamp.br/document/?code=vtls000349489, [Accessed: 12-Dec-09], (in Portuguese) 3. Economist Intelligence Unit - EIU, Scattering the seeds of invention: the globalisation of research and development, 2004. Retrieved from:table_13http://graphics.eiu.com/files/ad_pdfs/RnD_GLOBILISATION_WHITEPAPER.pdf , [Accessed:04-April-12] 4. Energy Research Company - EPE (Brasil), Brazilian Energy Balance Year 2012, Ri de Janeiro, 2013. 5. International Energy Agency – IEA, Renewables in global energy supply: an IEA fact sheet, International Energy Agency, 2007. 6. Energy Research Company - EPE (Brasil), National Energy Plan 2020, Rio d Janeiro, 2011. . The Brazilian Institute of Geography and Statistics - IBGE, 2010 Population Census, IBGE, 2011. 8. Alvarenga, G. V., Pianto, D. M. and Araújo, B. C., Impacts of the Brazilian science and technology sector funds on industrial firms’ R&D inputs and outputs: new perspectives using a dose-response function. In: Proceedings, EncontroNacional de Economia – ANPEC, Porto de Galinhas (PE): Anpec, 11-14 Dec, 20 12. Retrieved fromhttp://www.anpec.org.br/encontro/2012/inscricao/files_I/i8-5fe1cb9e5d777ea4 0cd1a965ecfba0b8.pdf, [Accessed: 13-Sep-13] 9. U.S. Energy Information Administration – EIA. International Energy Statistic Retrieved from: http://www.eia.gov/countries/, [Accessed: 28-Feb-14] 10. Enzensberger, N., Wietschel, M., and Rentz, O., Policy instruments fostering wind energy projects-a multi-perspective evaluation approach, Energy Policy, Vol. 30, no. 9, pp. 793-801, 2002., http://dx.doi.org/10.1016/S0301-4215(01)00139-2 11. REN21, Renewables 2011 Global Status Report, Paris, REN21 Secretariat, 2011. Retrieved from: http://www.map.ren21.net/GSR/GSR2012.pdf [Accessed: 06-Nov-12] 12. Heugens, P. P. M. A. R. and Lander, M., Structure! Agency! (and other quarrels): meta-analyzing institutional theories of organization, Academy of Management Journal, Vol. 52, No. 1, pp 61-85, 2009., http://dx.doi.org/10.5465/AMJ.2009.36461835 pp 13. Lundvall, B-Å., National Innovation Systems: towards a theory of innovation and interactive learning, Pinter, London, 1992. 14. Nelson, R., National Innovation Systems: a comparative analysis, Oxford University Press, New York/Oxford, 1993. 15. Organisation for Economic Co-operation and Development - OECD, Education at a Glance 2012: OECD Indicators, OECD Publishing, 2012. 16. Organisation for Economic Co-operation and Development - OECD, World Energy Outlook 2011 Fact Sheet: global energy trends, 2011. 17. Pereira, N. M., Fundos setoriais: avaliação das estratégias de implementação e gestão, White paper n. 1.136, Brasília, 2005. 18. Pereira, N. M., Sector Funds: evaluation of implementation and management strategies (in Portuguese), White paper n. 1.136, Brasília, 2005. Retrieved from: http://www.ipea.gov.br/pub/ td/2005/td_1136.pdf, [Accessed: 15-Feb-2007] 19. Pereira, N. M., and Simone P. F. Experiences to support sector technological innovation (in Portuguese), Journal of Technology Management & Innovation, Vol. 1, No. 3, pp 74-80, 2006. Paper submitted: 06.02.2014 Paper revised: 02.03.2014 Paper accepted: 02.03.2014 Page 218nanLETTER • OPEN ACCESSYou may also like You may also like The universality class of random searches in critically scarce environments C. L. Faustino, M. L. Lyra, E. P. Raposo et al. - Can collective searches profit from Lévy walk strategies? M C Santos, E P Raposo, G M Viswanathan et al. - Optimal random searches of revisitable targets: Crossover from superdiffusive to ballistic random walks M. C. Santos, E. P. Raposo, G. M. Viswanathan et al. -The power of light: socio-economic and environmental implications of a rural electrification program in Brazil The universality class of random searches in critically scarce environments C. L. Faustino, M. L. Lyra, E. P. Raposo et al. - Can collective searches profit from Lévy walk strategies? M C Santos, E P Raposo, G M Viswanathan et al. - To cite this article: Paula Borges da Silveira Bezerra et al 2017 Environ. Res. Lett. 12 095004 Optimal random searches of revisitable targets: Crossover from superdiffusive to ballistic random walks M. C. Santos, E. P. Raposo, G. M. Viswanathan et al. - View the article online for updates and enhancements. This content was downloaded from IP address 78.193.21.129 on 03/10/2023 at 09:18 https://doi.org/10.1088/1748-9326/aa7bdd nan Environ. Res. Lett. 12 (2017) 095004 OPEN ACCESS RECEIVED 30 January 2017 REVISED 5 June 2017 ACCEPTED FOR PUBLICATION 26 June 2017 PUBLISHED 23 August 2017 Original content from this work may be used under the terms of the Creative Commons Attribution 3.0 licence. Any further distribution of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI.The power of light: socio-economic and environmental implications of a rural electrification program in Brazil Paula Borges da Silveira Bezerra, Camila Ludovique Callegari, Aline Ribas, André F P Lucena, Joana Portugal-Pereira, Alexandre Koberle, Alexandre Szklo and Roberto Schaeffer1 Energy Planning Program, Graduate School of Engineering, Universidade Federal do Rio de Janeiro, Brazil 1 Author to whom any correspondence should be addressed. Keywords: Electricity access, poverty alleviation, human development, Luz para Todos, BrazilAbstract Universal access to electricity is deemed critical for improving living standards and indispensable for eradicating poverty and achieving sustainable development. In 2003, the ‘Luz para Todos’ (LpT—Light for All) program was launched aiming to universalize access to electricity in Brazil. The program focused on rural and isolated areas, also targeting to bring development to those regions along with electrification. This paper evaluates the results of the LpT program in improving socio-economic development in the poorest regions of Brazil. After an initial qualitative analysis, an empirical quantitative assessment of the influence of increased electrification rates on the components of the Human Development Index (HDI) is performed. The empirical results of this study showed that electrification had a positive influence on all dimensions of the HDI, with the education component having the strongest effect. Although complementary policies were needed to achieve this, results show that electricity access is a major requirement to improve quality of life. Any further distribution of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI.1. Introduction 2000),whichcauseharmful indoor air pollution (WHO 2014). In poorly ventilated dwellings, indoor smoke can be 100 times higher than acceptable levels, causing significant health damages (WHO 2016). Access to electricity not only releases people from hard work, but also increases productive working hours and provides opportunities for self-employment, in particular for women in rural areas (Dinkelman 2011). Some 13 million people did not have access to electricity in Brazil in 2000. This represented 7% of all households in the country, around 3 million. The situation became even more alarming when consider- ing the distribution of such households according to their income and location. From the aforementioned 3 million households, approximately 2 million were located in rural areas. This represented 29% of rural homes in Brazil at that time. Depending on the region, these numbers also varied significantly: around 1% of the Southeast’s households did not have access to electricity, while in the North almost 18% were in that situation (IBGE 2000). Universal access to electricity is not only critical for improvinglivingstandardsbutdeemedindispensablefor eradicating poverty and achieving sustainable develop- ment (GNESD 2007). Because this is widely accepted today, ensuring universal access to affordable electricity by 2030 was incorporated directly in the Sustainable Development Goals (UNGA 2015). Increasing income by itself cannot guarantee some basic services and needs and cannot improve living conditions if it is not combinedwithinfrastructure(UNDP2002,Cook2011). A National Program of Universalization of Access and Use of Electricity (Light for All), the ‘Luz para Todos’ (LpT) program, was launched by the Brazilian government in 2003 with the goal of extending access to electricity to all rural communities in the country. Universal access to electricity is not only critical for improvinglivingstandardsbutdeemedindispensablefor eradicating poverty and achieving sustainable develop- ment (GNESD 2007). Because this is widely accepted today, ensuring universal access to affordable electricity by 2030 was incorporated directly in the Sustainable Development Goals (UNGA 2015). Increasing income by itself cannot guarantee some basic services and needs and cannot improve living conditions if it is not combinedwithinfrastructure(UNDP2002,Cook2011). Electrification provides a solid basis for develop- ment of local communities. Once a community has access to electricity, itcan alsohave access to safepotable water, better health conditions, food security, as well as lighting and information. In addition, it reduces the needforcollectingandusingothertraditionalsourcesof energy, such as firewood, animal dung, and crop residues for cooking and heating (Goldemberg et al A National Program of Universalization of Access and Use of Electricity (Light for All), the ‘Luz para Todos’ (LpT) program, was launched by the Brazilian government in 2003 with the goal of extending access to electricity to all rural communities in the country. © 2017 IOP Publishing Ltd Environ. Res. Lett. 12 (2017) 095004 nan Some studies evaluated the extent to which the LpT program increased income and promoted the social inclusion of benefitted communities (Coelho and Goldemberg 2013, Gómez and Silveira 2012, Pereira et al 2010, Slough et al 2015). However, there is a lack of formal empirical assessments that attempted to quantitatively measure the socio-economic improve- ments associated with the LpT program. the 1940s. Rural cooperatives were an initiative created bylocalcommunitiestobeabletofinancetheinstallation of transmission lines and guarantee access to electricity. During the decades that followed, other initiatives took place,for instance:theRuralElectrificationFund (FUER —Fundo de Eletrificação Rural, in Portuguese), created in the mid-1950s; the Executive Group for Rural Electrification(GEER—GrupoExecutivodeEletrificação Rural, in Portuguese), created in 1970; the First and Second National Rural Electrification Plan (PNER— Plano Nacional de Eletrificação Rural, in Portuguese), implemented during the 1970s; the Program for Energy Development of States and Municipalities (PRODEEM —Programa de Desenvolvimento Energético de Estados e Municípios, in Portuguese), launched in 1994; and later the Light in the Countryside (Luz no Campo, in Portuguese), created in 19992. ThispaperevaluatestheresultsoftheLpTprogramin improving socio-economic development in the poorest regions of Brazil. To do so, aninitial qualitative analysis is madebasedonexistingdata,literatureandassessmentsof theprogram.Onasecondstage,anempiricalquantitative assessment of the program’s results is performed, which contributestotheexistingbodyofanalysisontheimpacts of rural electrification in the country. Thispaperisorganizedinfivesections.Followingthe introduction, a background section overviews the socio- economic context of the LpT program and its policy framework. From this point, in section 3, implications foreconomic,socialandenvironmentaldevelopmentare unveiled qualitatively. Section 4 details results of the empirical assessment conducted to quantitatively mea- sure the socio-economic improvements associated with the program. This is followed by final remarks. Figure 1 shows the rate of electricity access in Brazil between 1950 and 2000. Despite the evolution observed after the 1970s, there were still significant differences between the level of electrification in urban and rural areas. In 1991, 97% of the population of urban areas already had access to electricity, while, in the country- side, this number did not reach 50% (ANEEL 2005).2.1. The national program of universalization of2.1. The national program of universalization of access and use of electricity—LpT (Light for all) 2.1. The national program of universalization of access and use of electricity—LpT (Light for all) In November 2003, the LpT Program was established by decree N° 4873 (Fugimoto 2005). The program was coordinated by the Ministry of Mines and Energy (MME) and came as a consequence of Law 10 438 of 2002 that had set parameters to guarantee universali- zation of electricity. The program aimed to increase the electrification rate in the country, providing power to 10 million people until 2008, especially those living in rural areas (Brazil 2003). This was the first social oriented electricity access policy in Brazil, in which2. Background Inequality in access to electricity was a reality since the introduction ofthisbasicservice in Brazil.It ishardtosay precisely when the Federal Government started to put efforts on the electrification process. In fact, electricity cameasanaturalconsequenceoftheurbanizationprocess that occurred during the 1940s and 1950s. With low population density and large distances between proper- ties, rates of electrification in rural areas have always been lower than in urban regions (Bittencourt 2010). The first efforts to promote rural electrification in Brazilstartedwiththecreationofruralcooperatives,after 2 For a detailed description of these programs, see Gómez and Silveira (2010) and Borges et al (2016). nan Environ. Res. Lett. 12 (2017) 095004 Table 1. Summary of the different stages of the LpT program. Phase Period Goals and achievements Phase I 2004–2008 Provide universal power access to rural communities not connected to the grid. Phase II 2008–2010 Provide power access to 1 million families that had not been connected in the first stage, reaching almost 3 million households. Phase II- extension 2010–2011 Provide electricity access to isolated communities, areas with no connection to distribution lines, low population density, difficult access and poor infrastructure, reaching further 1.7 million new electrical connections. Phase III 2011–2014 As the majority of the population already had access to electricity, the focus of this extension was to reach communities living in areas with significant logistic and infrastructure difficulties, particularly in the North and Northeast regions. The target for the period was the connection of 795 000 new households (MME 2011). Phase IV 2014–2018 Expected to provide power access in isolated areas and the Amazon region. Source: Borges et al (2016). generation grids in isolated systems4. To be approved, the construction plan had to detail the technical, material and equipment criteria to be used. beneficiaries did not have to contribute financially (Goldemberg et al 2004). To meet this initial goal, US$2.3billionwereinvested.Theprogramwasextended first until 2014 and more recently until 2018 (MME 2017). Until May 2016, it had reached 15.6 million people, with an overall investment of US$ 7 billion. Decentralized generation projects must be cost competitive with grid extension to be endorsed (MME 2004). Also, for decentralized and isolated system generation, the projects must consider environmental aspects, end-user capacity building, and overall sustainability5. The technological options for off-grid generation foreseen by the program are hydro, wind, diesel fuel and biomass, with special focus given to solar in recent operational manuals. The program, therefore, did not clearly promote the deployment of renewables until recently. This is, actually, one of the critical aspects of Brazil’s universal energy access strategy. p p Rural electrification was seen by the government as a key element to achieve social development in rural areas. Thus, projects with higher social development outcomes were highly ranked and prioritized3, when compared to those with limited social benefits. New electricity demand was identified through the so- called Luz para Todos’ agents (LpT agents). These agents worked close to local communities, informing about the program execution and its benefits. During the work execution, LpT agents were also responsible for identifying, together with communities, possible productive uses for electricity and complementary actions of social inclusion. Besides, these agents acted as a communication channel between local citizens and program executors. Rural populations were able to request new electrical connections through the LpT agents. In this way, communities were partially involved in the program’s decision-making process, helping to recognize population needs for demand and productive applications of electricity in the region. Also, utility companies conducted educational and awareness campaigns about appropriate, efficient and secure use of electricity (Gómez and Silveira 2010). After the initial period of the program (2004– 2008), LpTwas extended four times. During the initial execution, between 2004 and 2008, the program could not reach its initial target of providing access to 10 million people. In addition, agents also identified a higher number of families with no access to electricity than the number accounted for in the year 2000 census. This new demand was related to population growth, not considered before, and to the return of some families to rural areas. These facts led to the implementation of new phases of the program, continuing it and setting new targets (MME 2008). Table 1 summarizes the initial targets and achieve- ments of each phase of the LpT program. Technically, the program focused on low-cost transmission and distribution grid extensions. Alter- natively, where connection to the grid would not be feasible, electricity could come from decentralized 4 As noted further in section 2.2, the program’s operationalization prioritized grid extension. Until 2014, new connections entailed primarily grid extensions with less than 1% involving installation of isolated individual systems. After regulatory framework on the use of isolated micro-grids was introduced in 2012, distribution utilities started to plan for the installation of isolated systems to reach the population without access to electricity, about 30 000 families according to government estimates from December 2014 (MME 2017). 3 The selected projects needed to fulfil at least one of the following criteria: (i) municipalities with less than 85% electricity coverage, (ii) municipalities with Human Development Index (HDI) lower than the national average, (iii) projects that benefit people injured by dams with no clear responsibility, (iv) projects that focus on productive use of electricity and on fomenting integrated local development, (v) projects in public schools, health centres and wells for water supply, (vi) projects aiming to attend rural settlements, (vii) projects to develop familiar farming and small and medium farmers, (viii) projects stagnated by lack of financial resources which covered rural communities, (ix) citizens around natural conserva- tion units, and (x) projects in areas for specific use for special communities, as racial minorities (MME 2004). 5 The distribution utilities, as executing agents, are responsible for complying with these requirements. To provide guidance and support them, Eletrobras, the institution managing the program, established a technical cooperation project with the Inter-American Institute for Cooperation on Agriculture (IICA) in 2009 (Eletrobras 2015). Capacity building is deemed a critical barrier to the electrification of isolated areas in Brazil both at the utilities’ and end- users’ levels (Valer et al 2017). Environ. Res. Lett. 12 (2017) 095004 nan - 100 000 200 000 300 000 400 000 500 000 600 000 700 000 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 Figure 2. New electrical connections made by year (EPE 2015 and Brazil 2016a). Environ. Res. Lett. 12 (2017) 095004table_12.2. Challenges and overall evaluation In the case of the Amazon region, challenges are even harder. The region has an extensive area with a complicated topography, surrounded by rivers and highly dense rainforest. In addition, it has a very small and low-density population with low-income levels, and mostly concentrated in rural areas (Gómez and Silveira 2015). These particular characteristics pose specific challenges to providing electrification in that area. The people that already have access to electricity are concentrated in regions with previously existing physical grid infrastructure. Cities and communities are mostly located in regions of high deforestation, with highways and agriculture, which facilitated the connection with the national electric grid. However, this is not the case for many parts of the region (Di Lascio and Barreto 2009). Despite the impressive numbers of the LpT program, the target of giving electricity access to all of the Brazilian population has not been achieved yet. The extension of the grid could readily benefit a significant number of people, but as the grid extension approaches its physical and economic limits, reaching some areas becomes difficult or even unfeasible. Therefore, universalization goals become increasingly difficult to achieve (Gómez and Silveira 2015). The Brazilian national grid structure has a centralized structure, concentrated on the coast, which is very effective to meet industrial consumers and urban area needs, but fails to promote electrifica- tion of isolated communities, especially in the Amazon region. This structure makes connecting island regions to the grid a hard task and a challenge to reaching households far from urban centres in a vast country of continental dimensions. In terms of institutional structure and operations, LpT prioritized the exten- sion of the grid (Slough et al 2015). Currently, there are mainly three obstacles to foster universal access to electricity in remote areas. The first one is the need to adapt the existing institutional structures. The second is the choice of technology or supply solutions that comply with the local environ- ment and infrastructure. The third one is a more effective use of government funds within the context of the current subsidy scheme. A new rural electrification model in which local, resource-based technologies are supported by an adapted institutional framework and existing funding structures is needed to reach this last mile (Gómez and Silveira 2015). Finally, a major challenge is related to guaranteeing the continuity of electricity affordability for low-income households benefitted by LpT after the end of the Program. Actually, electricity affordability is being sustained by cross-subsidies provided by the Brazilian interconnected electricity system, in order to com- pensate the higher costs incurred by local power utilities to serve remote areas. After 2018, it is not yet clear whether and how these subsidies will be maintained (Brazil 2016b). As the program proceeded, the need for off-grid solutions increased. The program reached its limits in connecting areas closer to the grid and the average cost per connection increased, creating a challenge to take electricity to isolated areas far from the existing grid. In this context, less expensive technological alter- natives should be considered, since utilities would pressure for high tariffs to compensate this adverse situation (Di Lascio and Barreto 2009). Capital costs to electrify most isolated communities can be twice as high than new grid connections (Sanchez et al 2015). Observing the connections made by year, it can be noted that fewer new connections were made as time passed (figure 2). After 2010, Brazil achieved a 98.6% electrification rate, but the remaining 1.4% became harder to reach. The third phase of the program, after 2011, faced this challenge, and the connections in 2013 and 2014 were lower than 100 000 yr−1. According to Pereira et al (2010), in order to reach isolatedcommunitiesitisnecessarythatdecisionmakers work together with regulatory agencies, universities and research centres. The efforts must include the develop- ment of cleaner technologies and improvement of management models, respecting the cultural, economic and environmental aspects of using renewable technol- ogies in a decentralized or self-generation manner.3. Implications for economic development, social welfare and environmental sustainability The LpT program exceeded the initial target of providing electricity access to 10 million citizens. During its 10 years of execution, the program Environ. Res. Lett. 12 (2017) 095004 nan Table 2. Electricity service related to improvements in types of uses. Household social and community uses Productive uses Education uses Health uses Public administration uses Improved quality of life (light, TV, radio). Light: children and women gain additional time at night (reading, homework); improved light quality (200 times brighter) and cost per lumen; reduced cooking times and easier cleaning due to illuminated room; increases productivity for self-consumption. Safety: street lighting allows children and women to socialize at night; facilitates community activities Raises productivity: increased profit and employment e.g. light extends work time; electricity allows applications such as water pumping (irrigation), soldering, motive applications (drilling, sawing, mills), cold chain (e.g. for small shops and restaurants, milk processing beef storage) Studying at night; adult education; allows retention of qualified teachers. Schools can serve as anchor clients for service providers. Subsidizing public services is an efficient way of targeting subsidies with reduced free rider effects. Light for emergencies, childbirths; vaccine fridges; HIV. Domestic light seems to be correlated with more whitewashed walls and fewer insects. Allows for more efficient public administration. Increase working time and improves quality of service. Improved quality of life (light, TV, radio). Raises productivity: increased profit and employment e.g. light extends work time; electricity allows applications such as water pumping (irrigation), soldering, motive applications (drilling, sawing, mills), cold chain (e.g. for small shops and restaurants, milk processing, beef storage), fish ponds, electric fences, video, cinemas, etc, permits use of ICT. Light: children and women gain additional time at night (reading, homework); improved light quality (200 times brighter) and cost per lumen; reduced cooking times and easier cleaning due to illuminated room; increases productivity for self-consumption. Safety: street lighting allows children and women to socialize at night; facilitates community activities (light, TV, radio, discotheques); potential effect on birth rates. Source: Motta and Reiche (2001). Electricity generation in Brazil is highly based on renewable energy sources. In 2014, 77.2% of total electricity supply was provided by renewables sources. This contrasts with only 28.2% in isolated areas, where fossil fuels are responsible for 71.8% of electricity generation. To supply the county’s electric system in 2014, 78.30 MtCO2 were emitted, from which almost 10% came from isolated systems where electricity consumption is only 0.8% of total demand in Brazil. In that sense, the choice of supply source for isolated systems is critical for improving energy access without increasing total greenhouse gas emissions (EPE 2015). reached over 3.3 million households, equivalent to more than 15 million people (MME 2017). More than enabling access to electricity, an important benefit of the program was recognizing electricity supply as a way to promote social and economic development in less developed regions of the country. The program was a key component of the national strategy for poverty reduction, sustainable development and reduction of social inequality (Gómez and Silveira 2010). Therefore, the results of electrification projects should not be measured just by the number of new households connected, but also by the social and economic benefits promoted by electricity access. Identifying social, environmental and infrastructure evolution caused by the implementation of the LpT policy is critical to understanding the welfare improvement and evaluating the return of the capital invested in the program (Gómez and Silveira 2010). Table 2 identifies potential improvements to welfare associated with electricity service in rural areas (Motta and Reiche 2001). Historically, thermal-power plants fuelled by diesel were the main supply choice for isolated systems, but renewable energy systems are being increasingly regarded as a favourable option for providing power to isolated communities. Despite the higher capital cost, generation from renewable sources can have lower operational costs. When considering local realities of isolated communities, the use of renewable energy options can be a preferable solution to providing electricity access (Di Lascio and Barreto 2009, Gómez and Silveira 2015, Sanchez et al 2015). Table 2 shows that electricity uses are associated with many dimensions of development. Not only can the population have the choice of consuming electrical appliances, but also education and health improve- ment can be achieved. Moreover, electrification can change the local reality in terms of social, economic and environmental aspects. The use of government incentives in the form of laws, technological research and institutional frameworks is important to change the current fossil-fuel-based generation in isolated communities (Pereira et al 2010). The LpT program can be considered as a mean to foster the use of renewable energy sources. In November 2008, the MME promoted activities to assist local utilities in develop- ing and implementing small projects for electricity supply using renewable energy sources. These activi- ties were executed with the support of Inter-American Development Bank and focused on training profes- sionals and utilities to find solutions based on local capacities for using alternative energy sources (Barreto 2008).3.1. Environmental aspects Access to electricity can change in many forms the way of living in a community. In addition to social and economic impacts related to electrification, there are also some environmental impacts. One of the main choices in the electrification process is which energy sources to use in isolated areas, where grid connection is not possible. nan Environ. Res. Lett. 12 (2017) 095004 Also, after 2009, the LpT program launched special project guidelines with the main objective of developing the use of renewable energies in areas with difficult access, by preferably funding projects in isolated regions with the use of renewable energy sources considering the regions’ potentials. There is a significant potential for increasing electricity access in isolated systems through the use of PV, biomass, and small hydro. In addition to being appropriate to local reality, these projects also guarantee electricity supply with lower environmental impacts, and energy independence for the communities (Di Lascio and Barreto 2009). Sanchez et al (2015) evaluated the most significant rural electrification projects using renewable sources in isolated areas of the Brazilian Amazon region during the first stages of the LpT program (2003–2011). These projects showed the convenience of substituting, totally or partially, the use of diesel, which had to be shipped in. More importantly, they showed that electricity generation from local renewable sources is a way of empowering disadvantaged communities, giving them energy independence along with the benefits of electricity access. kerosene lamps and candles as the main sources of lighting. After accessing electricity, family expenses with kerosene, diesel, oil, gas and batteries dropped to half the initial values, indicating a substitution of fossil fuel sources by electricity (MME 2013). Concerning land use impacts, electrification can have two opposing effects as a result of changes in agricultural productivity. Electricity access can lead to an improvement in agricultural productivity, as it allows a more efficient irrigation with the use of water pumps, as shown in Assunção et al (2015). The study suggests that a 10% increase in electrification could lead to a 0.66% increase in the proportion of farms with irrigation and a 9.8% increase in agricultural production per hectare. However the same study found that such improvements can lead to two opposing effects on the protection of forests and native vegetation: (i) an expansion of farm size and/or frontier land con- versions, and (ii) a shift away from cattle ranching, which is more environmentally destructive, and into crop cultivation, allowing farmers to retain more native vegetation within rural settlements. Even though the authors estimated that electrification caused a small net decrease in deforestation in a specific region in Brazil, decreases in deforestation cannot be correlated to higher electricity access given that it depends on many other key variables, including the type of agricultural crops involved. Yet, electricity can add value to local traditional production of extracted products from native forests reinforcing subsistence agriculture, which can account for a high share of family income. Therefore, extensive agricul- ture is not used as a substitute for improving family income and local vegetation can be preserved (Di Lascio and Barreto 2009). Pereira et al (2010) compared the energy consump- tion mix of an average household before and after getting access to electricity (figure 3). Before access to electricity, LPG, firewood and diesel combined represented90%oftotalenergy demand,theremaining 10% was due to the use of charcoal, gasoline, kerosene and others, from a total of 5.16 GJ yr−1 per capita. After getting access to electricity, the total consumption increased 28%, and the share of those energy sources dropped to 65%. Also, it is worth highlighting the fast penetration of electricity, reaching 34% of the average energy consumption basket. The control group confirmed that electricity access was responsible for this change in the composition of the average house- hold’s energy basket.3.2. Socio-economic development To measure socio-economic impacts, a survey devel- oped by MME (2013) evaluated the profile of beneficiaries and the impacts of the program in the communities. Results show that 89.8% of the beneficiary families had a total monthly income equal Changes in the energy basket used by households were also observed in a national survey made in 2013 with beneficiaries of the LpT program. This survey showed a transition in a family’s energy basket from nan Environ. Res. Lett. 12 (2017) 095004 Table 3. Brazilian situation in 2000 and 2010: electrification rate and income per capita by state. Brazilian states Country region Households with electricity Per capita income 2000 (%) 2010 (%) 2000–2010 Growth rate (%) 2000 (US$) 2010 (US$) 2000–2010 Growth rate (%) Brazil 93.5 98.6 5.5 182.91 245.10 34.0 Acre (AC) North 75.8 91.1 20.2 111.34 161.21 44.8 Alagoas (AL) Northeast 89.8 99.0 10.2 88.08 133.55 51.6 Amazonas (AM) North 82.2 92.2 12.2 108.56 166.66 53.5 Amapá (AP) North 95.1 98.3 3.3 131.08 184.93 41.1 Bahia (BA) Northeast 80.9 96.5 19.2 99.43 153.36 54.2 Ceará (CE) Northeast 88.2 99.1 12.3 95.77 142.21 48.5 Distrito Federal (DF) Centre-west 99.7 99.9 0.2 370.31 529.52 43.0 Espírito Santo (ES) Southeast 98.7 99.8 1.2 177.27 251.75 42.0 Goiás (GO) Centre-west 97.3 99.4 2.2 176.44 250.38 41.9 Maranhão (MA) Northeast 78.7 96.1 22.2 67.39 111.25 65.1 Minas Gerais (MG) Southeast 95.7 99.4 3.9 169.46 231.46 36.6 Mato Grosso do Sul (MS) Centre-west 95.6 98.6 3.2 177.93 246.79 38.7 Mato Grosso (MT) Centre-west 89.5 98.0 9.5 179.88 235.42 30.9 Pará (PA) North 76.7 91.9 19.8 103.66 137.93 33.1 Paraíba (PB) Northeast 94.5 99.4 5.3 92.34 146.63 58.8 Pernambuco (PE) Northeast 95.5 99.5 4.2 113.40 162.28 43.1 Piauí (PI) Northeast 74.5 93.0 24.9 78.66 128.72 63.6 Paraná (PR) South 97.7 99.6 2.0 197.06 275.05 39.6 Rio de Janeiro (RJ) Southeast 99.5 99.9 0.4 255.03 320.87 25.8 Rio Grande do Norte (RN) Northeast 94.1 99.4 5.6 108.37 168.39 55.4 Rondônia (RO) North 83.9 97.3 15.9 144.23 207.11 43.6 Roraima (RR) North 86.0 90.7 5.5 142.69 186.97 31.0 Rio Grande do Sul (RS) South 97.8 99.7 1.9 218.62 296.15 35.5 Santa Catarina (SC) South 98.6 99.8 1.2 214.21 303.77 41.8 Sergipe (SE) Northeast 91.8 99.2 8.1 100.86 161.63 60.3 São Paulo (SP) Southeast 99.6 99.9 0.3 272.43 334.81 22.9 Tocantins (TO) North 77.2 94.7 22.7 106.33 181.11 70.3 Source: IPEA (2013).Table 3. Brazilian situation in 2000 and 2010: electrification rate and income per capita by state. to or below two times the minimum wage6 and 18.8% only received half the minimum wage. Nearly half of the targetedfamilieswereruralworkers.Amongthefamilies interviewed by the program’s survey, 41.2% considered that theprogram brought incomeriseand 40.5%saw an increaseintheamountofjobopportunities.Thisaddsto the evidence of the positive social and economic co- benefits of the program. regions with the highest electrification rates also had a higher income increase in the same period. Although a causal relationship between the electrification process and income cannot be inferred, a correlation between them can be noticed (IPEA 2013). It is important to mention that after 2003 other governmental social programs were established with the objective of reducing poverty in all dimensions. The main program was Bolsa Família, a cash transfer social program. By August 2016, the program had benefited 13.8 million families, with an average cash transfer of US$ 56.00 per month per family (MDS 2016). This will be further explored in section 4. The income per capita in each state between 2000 and 2010 improved significantly. Regions, such as the Northeast and Midwest, showed a higher monthly income in 2010 than in 2000. The Southeast and South regions had the highest electrification rates and income per capita in 2000, while the lowest values were in the North and Northeast regions, for both cases. In 2010, an improvement could be observed in the latter regions in both dimensions. Bolsa Familia was integrated with many other programs, such as LpT. The government understood that a monthly stipend was by itself not enough to lift most of these individuals and their families out of extreme poverty. In conjunction with LpT, however, Bolsa Família’s benefits made it possible for families to make use of electricity benefits, investing in appliances, for studying, or for small family busi- nesses (IPEA 2013)7. Table 3 compares electrification rates and per capita income in each Brazilian state in 2000 and 2010. The greatest increases in electrification rates were in poorest and largely rural states (mostly in the North and Northeast regions). Generally speaking, the 7 Even though no data is available on the electricity consumption rates of households included in the Bolsa Familia, it is reasonable to assume that they would likely fall in the ‘up to 30 kWh’ bracket, which is entitled to receive a 65% discount (ANEEL 2017). 6 The monthly minimum wage in 2003 was equivalent to US$ 74.10, currently it is US$ 271.70. Environ. Res. Lett. 12 (2017) 095004 program at that time, with the goals of reducing poverty, promoting food security, and increasing access to public services, especially health, education, and social assistance. Since it was launched, 5 million people left extreme poverty living conditions, reducing inequalities in Brazil (Fultz and Francis 2013). In the regions included in the LpT program (most of them rural areas), the rise in income levels can be associated with more productive rural activity, as well as the diversification of economic activities. Electrifi- cation allows the creation of small businesses, such as bakeries, local markets and drugstores. After LpT, for instance, the presence of local markets, bars and bakeries increased 24%, 22% and 7%, respectively (MME 2013). As mentioned by Soares (2012), the strategy of the Brazilian government has been based on the complementarity of programs. These include adult education, opportunities for youth, job training, labour intermediation, subsidized electricity, rural electricity grid expansion (LpT), rural extension of microcredit to those who either are or may soon be Bolsa Família beneficiaries. The integration of complementary programs and actions contributes to families’ socio-economic inclusion and their emanci- pation from the program in a long term perspective (Quinhões and Fava 2010). Bolsa Família can be considered a driver of the social achievements observed, and electrification process is one of the important keys used to give possibilities to many families to alleviate poverty. Therefore, electricity has a role to make the development possible, not by itself, but integrated to other social efforts. Also, according to MME (2013), 462 000 new direct and indirect jobs have been created as a result of the program implementation, and around 244 000 women started in a productive activity (MME 2013). In addition, in another survey made in the state of Tocantins, in the North region, Guimarães (2011) reports the economic improvements triggered by the LpT program. The author presented two case studies on how electrification increased both productivity and family income in rural areas. Guimarães (2011) also reveals that after electrification, communities were able to increase their income and expand their economic activities. For instance, farmers were able to use electrical machinery in farming and processing activities, which increased their productivity consid- erably. In some cases, households increased their income by 250%. Despite the several evaluations of the results of the LpT program, there is a lack of formal empirical assessments that attempted to quantitatively measure the socio-economic improvements associated with the LpT program. The empirical assessment performed in this paper is an attempt to complement some knowledge gaps on the effects of the LpT program by performing a statistical analysis at the municipality level in Brazil. y According to MME (2013), almost all beneficiaries reached by the program have reported improvements in their quality of life, mainly due to comfort and home needs. According to Pereira et al (2010), what distinguishes a poor household from a better-off one is also the wide range of choices in terms of which fuels to use (more efficient, more convenient, less polluting, etc) and which equipment and appliances to buy. The government appraises that US$ 2.0 billion were injected in the household appliances market due to the LpT program, through electrical appliances bought by beneficiaries of the program. It is estimated that 81% of families purchased new TV sets, 71% refrigerators and 62% cell phones. Considering all the appliances, a total of 14 million new pieces of equipment were bought.4. Empirical assessment of the results of the LpT program When the program was launched, nearly 90% of the target families that did not have access to electricity had low income—up to three times the minimum wage—and lived in areas with low HDI8 (Eletrobras 2016). It is expected that electricity supply has a large impact on well-being in regions with low HDI by improving health, education and communication services (Gómez and Silveira 2010). Actually, the role of advances in energy services in improving the HDI of a country at early stages of development is demon- strated by some studies (Pasternak 2000, Martinez and MME (2013) also measured social impacts associated with the electrification process. A survey with program beneficiaries showed an improvement in public services (e.g. education) and welfare. Most of LpT program beneficiaries believe that morning and night shift educational activities were improved. In addition, according to the survey, 309 000 women were enrolled in primary and secondary schools. The survey also evaluated the population opinion about health services. Nearly half of the beneficiaries believed that health care improved given the better access and quality of health centres. 8 HDI is formed by three basic components: (i) longevity, measured by life expectancy at birth, (ii) educational level, measured by two variables, average literacy rate of people aged 25 or older and the expectancy of years of study, and (iii) income, measured as GDP per capita in purchasing parity power. The three components have the same weight (IPEA 2013). The indices are relative and normalized, such that each component is calculated with respect to the minimum value in the sample, and then normalized to the maximum difference found in the sample (Steckel et al 2013). A country potentially having the highest score across all three components would have an HDI value of 1.0. Despite the results, poverty is a complex and multidimensional phenomenon; as so, it cannot be reduced to a single component, as electricity. It is also important to understand the role of other govern- ment programs in Brazil. Bolsa Família was the main Environ. Res. Lett. 12 (2017) 095004 nan Ebenhack 2008, Jackson 2009, Steinberger and Roberts 2009 and 2010). In this context, the HDI can be one way to analyse the success of a policy. economic development are closely linked. The study also found that electrification efforts made by the LpT program seems to have achieved more success in municipalities that had a low electricity access rate but a relatively high HDI, implying that the drive to bring electricity to the countryside brought the most benefits to municipalities that were already doing relatively well in other development-relevant measures. In contrast, municipalities that previously had both low electrifi- cation rates and a low level of socio-economic development appear to have fallen further behind in relative, if not in absolute terms. The Brazilian government uses the HDI as a tool for planning and monitoring development policies, including the LpT program (Gómez and Silveira 2010). Comparing 2000 data on development with observed electrification growth shows that low HDI levels were a reality in areas with the lowest electricity attendance. According to IPEA (2013), in 2000, North and Northeast regions presented the lowest HDI in Brazil and also the lowest electrification rate at the time, just 87.7% in the Northeast and 81.6% in the North. On the other hand, more developed states, like the Federal District and São Paulo, had high HDI and presented high electrification rates (respectively, 99.7% and 99.6%). Regarding the evolution of the HDI between 2000 and 2010 in each Brazilian state, four states had improvements in electrical coverage higher than 20%: Acre, Maranhão, Piauí and Tocantins. All of them had progress in the HDI levels of around 30%, Maranhão being the state with the highest improvement, 34.2%, with an increase in the HDI from 0.476 to 0.639. According to 2010 data, all Brazilian states left the group of lowest human development regions, and were considered to be medium development regions, with HDI levels higher than 0.600. At the time, the lowest HDI was in the state of Alagoas (0.631). It is worth mentioning, however, that HDI in Alagoas in 2000 was 0.471. In that way, to Slough et al (2015) the strong correlations found cannot tell us whether electrification drives development or development drives electrifica- tion. The study concluded that the LpT program targeting poor communities is important for reducing inequality of electricity access, but not sufficient to drive transformational development effects, since the latter depend on the government’s ability to promote economic growth and social development. Comple- mentary interventions are necessary to allow local communities to exploit rural electrification for produc- tive uses, not limiting electricity access for the provision of basic household services. In fact, it is equally possible that LpT actually targeted the most advanced munici- palities and did not contribute much to development itself. Despite the correlation observed in the studies, HDI evolution cannot be inferred as a result of the electrification process.However, the latter isunarguably a pre-condition for high HDI levels. The social benefits regarding electrification access can only be achieved if otheractionsare executed jointly withtheelectrification process. Cunha (2007) shows that correlation between HDI levels and total per capita electricity consump- tion for 177 countries and 27 Brazilian states are similar to most countries with medium development levels. Also, statistically, there is a significant correlation between residential electricity consump- tion and HDI, as found by Pasternak (2000), Kanagawa and Nakata (2008), Mazur (2011), Martinez and Ebenhack (2008), Steinberger and Roberts (2009) and Oliveira (2013). An empirical quantitative assessment of the program’s results based on a panel data regression model is proposed to assess the relationship between HDI and its components and electrification rate and, thereby, provide further insight into the socio- economic impacts of rural electrification in the country. In the Amazon region, Gómez and Silveira (2010) finds evidence about the relationship between per capita residential electricity consumption and HDI. The author concludes that, if electricity access is provided to those with low HDI, a significant improvement in HDI can be achieved. Strong benefits can apparently be achieved in the Amazon region, as electricity helps break isolation and increases oppor- tunity for the socio-economic inclusion of many communities.4.1. Database The database used in this work was constructed from the concatenation of Brazilian population data from Brazilian Institute of Geography and Statistics (IBGE 2016) and the Atlas of Human Development in Brazil (Atlas Brasil 2016), which provides the Municipal Human Development Index (MHDI) and other 200 indicators for demography, education, income, labour, housing and vulnerability of Brazilian municipalities. Slough et al (2015) examined the correlation between HDI in Brazilian municipalities in 2000 and the improvements in electrification rates between 2000 and 2010, during the LpT program. The study reveals that the improvement in electricity access goes along with an increased HDI and increased per capita income. In each case, the association was strong and suggested that rural electrification and socio- The database was constructed at the municipal level, including observations for the 5565 municipali- ties from all of the 27 Brazilian states for the years 2000 and 2010. The period was selected according to the availability of information. Descriptive statistics of all variables, as well as the correlation matrix, were calculated. The results are nan Environ. Res. Lett. 12 (2017) 0950044.2. Methodological approach M i i li l i4.2. Methodological approach 4.2.1. Municipality selection found in the Appendix. The correlation matrix seeks to contribute to the verification of correlation between the explanatory variables. Variables used in the estimations, as well as their theoretical and empirical references are described below (Atlas Brasil 2016).4.2.1. Municipality selection There are no official data about the actual municipali- ties that took part in the LpT program. Therefore, it was necessary to identify and filter the municipalities that were served by the program based on the variation of the rate of electrification: all municipalities that had an increase above 40% in the period were considered in the analysis. By applying this selection criteria, 805 municipalities were selected, comprising 12 million people in 2010, the approximate number of people served by the program according to MME (2017). The dependent variables used by this study are the human development index and its three basic dimensions—income, education and health—as de- scribed below. a. MHDI: Municipal Human Development Index. Geometric mean of the indices for the Income, Education and Longevity dimensions, described below.4.2.2. Panel data regression model 4.2.2. Panel data regression model A panel data regression model was used to assess the relationship between the HDI and its components and electrification rate, in particular, the estimates assuming random and fixed effects will be presented, as well as the robustness tests to choose the best econometric model. b. Municipal Human Development Index—Educa- tion Dimension (MHDI_E): is obtained by the geometric mean of the frequency of children and young people at school, with weight of 2/3, and the education of the adult population, weighing 1/3. The regression models with panel data combine time series and cross-sectional observations. There- fore, there are more observations and additional degrees of freedom compared to the specific use of cross-sectional or time series analysis (Baltagi 2001, Hsiao 2003). c. Municipal Human Development Index—Longevity Dimension (MHDI_L): is obtained from the indicator of life expectancy at birth, using the formula: For modelling the unobserved effects there are two possibilities, both of which were tested: the fixed effects and the random effects. The fixed effects model considers that the specific intercept of each individual can be correlated with one or more regressors. As for the random effects model, it assumes that the (random) intercept of an individual unit is not correlated to the explanatory variables (Wooldridge 2002). In this case, when considering that the variables are not correlated, the random effects method is more appropriate. On the other hand, if the unobserved effects are correlated to some explanatory variable, the estimation by fixed effects would be more appropriate. For the selecting the method—fixed or random effects —the Hausman test will be performed (Wooldridge 2002). Where: O is the observed value of the indica- tor; Min is the minimum value; Max is the maximum value and the minimum and maxi- mum values are 25 and 85 years, respectively. d. Municipal Human Development Index—Income Dimension (MHDI_Y): is obtained from the per capita income indicator, using the formula: For the selecting the method—fixed or random effects —the Hausman test will be performed (Wooldridge 2002). Where: O is the observed value of the indicator; Min is the minimum value; Max is the maxi- mum value and the minimum and maximum values are R$ 8.00 and R$ 4033.00 (at August 2010 prices). The econometric model adopted is represented by the equation (3): The explanatory variables used in the study were as follows: Where: Y i;t represents the dependent variable for municipality i in period t (MHDI, MHDI_E, MHDI_L, MHDI_Y); a is the intercept; b1 and b2 are the parameters to be estimated; ILIGHT,i,t and V BFi;t are the explanatory variables; and ei;t represents the error term. a. Share of the population living in households with electric power (I_LIGHT): the ratio of the population living in permanent private house- holds with electricity access to the total popula- tion living in permanent private households, multiplied by 100.4.3. Results The results for the random effects and the fixed effects regression models were sequentially estimated using equation (3). The Hausman test rejected the null hypothesis that the random effects are consistent, pointing out that the best selection is the fixed effects b. Bolsa Família control variable (V_BF): financial amount passed on to municipalities for the management of the Bolsa Família family grant program (in Brazilian reais). Environ. Res. Lett. 12 (2017) 095004 nan Table 4. Panel regression model results. Coefficientsa Random effect Fixed effects Dependent Variable: MHDI T_LIGHT 0.2286 0.2054 V_BF 0.0245 0.0258 R2 0.90 0.95 Hausman test 0.008956 Dependent Variable: MHDI_E T_LIGHT 0.2286 0.5210 V_BF 0.0245 0.0543 R2 0.90 0.94 Hausman test 2.2e-16 Dependent Variable: MHDI_L T_LIGHT 0.2286 0.0425 V_BF 0.02456 0.0112 R2 0.90 0.94 Hausman test 2.2e-16 Dependent Variable: MHDI_Y T_LIGHT 0.0819 0.0528 V_BF 0.0105 0.0121 R2 0.64 0.78 Hausman test 5.06e-05 a : Significant at 1%.Table 4. Panel regression model results. being education perhaps the transmission channel for that. This means that labour productivity rises, due to education, to then cause income growth in the Brazilian poorest municipalities. The Bolsa Família value variable coefficient (V_BF) is also positive for all models and it is statistically significant at the 1% level of significance. The results show that the program, which transfers income to families living in poverty and extreme poverty, does not have a large influence on the HDI. When analysing the monthly values transferred per capita, the results can be better understood. On average each family served by the program received around R$ 26 per month9. Thus, the program is more associated with the relief of hunger than with later stages of human development. It helps the extreme poor but has a small influence on HDI, since other factors need to be developed to increase the Municipal Human Development Index (MHDI), especially in its health and education components.5. Final remarks In 2003, Brazil launched the LpT program aiming to universalize access to electricity. The program focused on rural and isolated areas, also targeting to bring development to the region along with electrification. With an initial target of reaching 10 million rural people until 2008, nowadays and after four phases, the program has reached almost 15.8 million people. The program is expected to continue until 2018. modelling. The estimation results and the test performed are shown in table 4. Results show that the MHDI is positively related to both explanatory variables, which is expected. Namely, the higher the level of electrification, the higher the MHDI is expected to be. The coefficient for electrification rate (T_LIGHT) is positive for all models and it is statistically significant at the 1% level. The results show that the electricity access sector is relevant for human development. LpT is considered the first electrification govern- mental policy that focused not only in guaranteeing electricity access to communities, but also in reducing social inequality in rural communities. The LpT program created a priority level based on social welfare parameters, such as HDI and electricity access inequality. Also, the program’s execution along with other initiatives allowed electrification actions to be integrated to other governmental programs like Brazil Without Misery 10 (Brasil sem Miséria, in Portuguese), Water for All (Água para Todos, in Portuguese), National Program for the Strengthening of Family Farming (Programa Nacional de Fortalecimento da Agricultura Familiar—PRONAF, in Portuguese), National Technical Assistance and Rural Extension When assessing each HDI component separately, results show that the education component is the most affected by electrification, indicating that electricity plays a fundamental role in the indexes related to schooling. In other words, electricity access in the Brazilian rural area was closely related to the increase in the population’s access to the education system. Although parallel educational policies are needed to increase MHDI_E, and it is not safe to say that electrification is the cause for this, electricity access is a major requirement to improve education. The assessment conducted by Kanagawa and Nakata (2008) confirm this influence. According to the study, which aimed to reveal quantitative relations between access to electricity and advancements in socioeco- nomic condition in rural Assam state, India, it is estimated that the literacy rate could rise to 74% from 63% with the electrification in the area. 9 Around US$ 8 at current rates. 10 One of LpT’s priorities in its later phases was to reach Brasil sem Miséria beneficiaries (MME 2011). Brasil sem Miséria program was launched in 2011 with the objective of reducing extreme poverty through a cash transfer program, access to basic public services and families’ productiveness inclusion. Infrastructure programs, as LpT and Water for All (Água para Todos, in Portuguese) were key to rural productive inclusion, given that they provide the necessary infrastructure for farming families in Brazil’s semi-arid region, thus enhancing the productive structure needed to strengthen families’ autonomy. Up to June 2014 a total of 369 000 families were connected to electric power from the outset of Brasil sem Miséria plan, with 267 000 of them being Bolsa Família beneficiaries. Of these, 262 000 were in a situation of extreme poverty before the plan (MDS 2015). 9 Around US$ 8 at current rates. 10 The other two components of HDI—health and income—are statistically explained but not strongly influenced by the increase in the municipal electrifi- cation rate. Other explanatory variables may be more relevant in influencing these factors and should be tested. Or even, in the case of income, there should be a delay between electrification and income growth, nan Environ. Res. Lett. 12 (2017) 095004 Program (Programa Nacional de Assistência Técnica e Extensão Rural—PRONATER, in Portuguese), Na- tional Rural Housing Program (Programa Nacional de Habitacao Rural, in Portuguese), My House My Life (Minha Casa Minha Vida, in Portuguese) and University for All (Universidade para Todos, in Portuguese). In that way, the program could reach communities that were not covered by previous programs and foster sustainable development in those regions. Despite the results achieved by the program, Brazil still has people with no access to electricity. Brazil is a continental country with areas that are hard to access due to the presence of large rivers and dense forests. Part of the population living in those areas are sparse, therefore, supplying electricity to these isolated communities is a challenge for the program. Another challenge is maintaining the affordability of electricity for low-income house- holds benefitted by LpTafter the end of the Program, which depend on cross-subsidies provided by the Brazilian interconnected electricity system, guaran- teed only until 2018. Regarding the achievements of LpT, it is important to evaluate the role of electrification in development goals. Electrification is expected to provide the means through which new jobs and income can be generated and welfare can be improved. The presence of electricity can be correlated with HDI, income improvement, educational and health access and withAppendix household’s electrical appliances use. But, these benefits can only be reached if other complementary actions are executed alongside the electrification process. Electricity is key to development, but is not in itself a sufficient condition for achieving social development. The empirical results of this study showed that the education component of HDI was the one most influenced by electrification. Chances are that labour productivity growth (hopefully caused by education) will later generate income. But the analysis using the existing database is not able to indicate that yet. Also, development is a complex and multi- dimensional phenomenon, as such it requires a concerted, holistic approach based on complementary programs. These findings are very much in line with those from Slough et al (2015), who also found that electrification efforts made by the LpT program were apparently more successful in higher HDI regions, implying that electricity access is more effective when accompanied by, or in addition to, other development- relevant policies and measures. Acknowledgments This paper isa product of theCD-Links project, andhas received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No. 642147. The authors would also like to acknowledge the support of the Brazilian Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq) and the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES). ORCID iDS Roberto Schaeffer https://orcid.org/0000-0002- 3709-7323 References ANEEL 2005 Atlas de Energia Elétrica 2nd edn (Brasilia: Agência Nacional de Energia Elétrica) Table A1. Descriptive statistics of all the variables inserted in the model. Variables Min. 1st Qu. Median Mean 3rd Qu. Max. MHDI 0.20 0.39 0.49 0.48 0.57 0.71 MHDI_E 0.04 0.19 0.31 0.32 0.45 0.61 MHDI_L 0.55 0.65 0.71 0.71 0.76 0.85 MHDI_Y 0.33 0.46 0.51 0.51 0.55 0.70 T_LGITH 10.30 54.50 68.68 70.87 92.74 100.00 V_BF 2864 23 960 39 940 51 390 65 680 316 900 Table A2. Correlation matrix of the model variables. — IDHM IDHM_E IDHM_L IDHM_Y T_LIGTH V_BF IDHM 1 IDHM_E 0.9764982 1 IDHM_L 0.8731050 0.8052615 1 IDHM_Y 0.7972515 0.6733530 0.7399156 1 T_LIGTH 0.8568461 0.8720774 0.7178726 0.5756068 1 V_BF 0.5473096 0.5758726 0.4751958 0.3130929 0.5838101 1Table A1. Descriptive statistics of all the variables inserted in the model. h h ld’ l i l li h A k l d Table A1. Descriptive statistics of all the variables inserted in the model. Variables Min. 1st Qu. Median Mean 3rd Qu. Max. MHDI 0.20 0.39 0.49 0.48 0.57 0.71 MHDI_E 0.04 0.19 0.31 0.32 0.45 0.61 MHDI_L 0.55 0.65 0.71 0.71 0.76 0.85 MHDI_Y 0.33 0.46 0.51 0.51 0.55 0.70 T_LGITH 10.30 54.50 68.68 70.87 92.74 100.00 V_BF 2864 23 960 39 940 51 390 65 680 316 900 Table A2. Correlation matrix of the model variables. — IDHM IDHM_E IDHM_L IDHM_Y T_LIGTH V_BF IDHM 1 IDHM_E 0.9764982 1 IDHM_L 0.8731050 0.8052615 1 IDHM_Y 0.7972515 0.6733530 0.7399156 1 T_LIGTH 0.8568461 0.8720774 0.7178726 0.5756068 1 V_BF 0.5473096 0.5758726 0.4751958 0.3130929 0.5838101 1Table A2. Correlation matrix of the model variables.Acknowledgments household’s electrical appliances use. But, these benefits can only be reached if other complementary actions are executed alongside the electrification process. Electricity is key to development, but is not in itself a sufficient condition for achieving social development. The empirical results of this study showed that the education component of HDI was the one most influenced by electrification. Chances are that labour productivity growth (hopefully caused by education) will later generate income. But the analysis using the existing database is not able to indicate that yet. Also, development is a complex and multi- dimensional phenomenon, as such it requires a concerted, holistic approach based on complementary programs. These findings are very much in line with those from Slough et al (2015), who also found that electrification efforts made by the LpT program were apparently more successful in higher HDI regions, implying that electricity access is more effective when accompanied by, or in addition to, other development- relevant policies and measures. This paper isa product of theCD-Links project, andhas received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No. 642147. The authors would also like to acknowledge the support of the Brazilian Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq) and the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES).ORCID iDS Roberto Schaeffer https://orcid.org/0000-0002- 3709-7323References ANEEL 2005 Atlas de Energia Elétrica 2nd edn (Brasilia: Agência Nacional de Energia Elétrica) Environ. R nan Environ. Res. Lett. 12 (2017) 095004 ANEEL 2012 Resolucao Normativa No. 488, de 15 de Maio de 2012 (Brasilia: Agência Nacional de Energia Elétrica) Fultz E and Francis J 2013 Cash Transfer Programmes, Poverty Reduction and Empowerment of Women: A Comparative Analysis: Experiences from Brazil, Chile, India, Mexico and South Africa GED Working paper 4/2013 (Geneva: International Labour Office) (www.ilo.org/wcmsp5/groups/ public/—dgreports/—ender/documents/publication/ wcms 233599.pdf) (Accessed: 20 August 2016) ANEEL 2017 Tarifa Social de Energia Elétrica—TSEE (Brasilia: Agência Nacional de Energia Elétrica) (www.aneel.gov.br/ tarifas-consumidores) (Accessed: 30 May 2017) Assunção J, Molly L and Mushfiq Mobarak A 2015 Infrastructure development can benefit the environment: electrification, agricultural productivity and deforestation in Brazil LACEA 2015 Papers (http://hdl.handle.net/123456789/53092) (Accessed: 30 January 2017) GNESD 2007 Reaching the Millennium Development Goals and Beyond—Access to Modern Forms of Energy as a Prerequisite (Roskilde: Global Network on Energy for Sustainable Development) Goldemberg J, La Rovere E L and Coelho S T 2004 Expanding access to electricity in Brazil Energy Sustain. Dev. 4 86–94 Atlas Brasil 2016 Human Development Atlas in Brazil [Atlas de Desenvolvimento Humano no Brasil] [in Portuguese] (http://atlasbrasil.org.br/2013/) (Accessed: 25 January 2017) Goldemberg J, Reddy A K N, Smith K R and Williams R H 2000 Rural energy in developing countries World Energy Assessment: Energy and the Challenge of Sustainability ed J Goldemberg (New York: United Nations Development Program) Baltagi B H 2001 Econometrics Analysis of Panel Data 2nd edn (Chichester: Wiley) Barreto E J F 2008 Renewable Energy Technologies: Energy Solutions for the Amazon Region [Tecnologias de energias renováveis: Soluções energéticas para a Amazônia] [in Portuguese] 1st edn (Brasília: Ministry of Mines and Energy, Government of Brazil) Gómez M F and Silveira S 2010 Rural electrification of the Brazilian Amazon—achievements and lessons Energy Policy 38 6251–60 Gómez M F and Silveira S 2012 Delivering off-grid electricity systems in the Brazilian Amazon Energy Sustain. Dev. 16 155–67 Bittencourt E B 2010 Evaluation of the Light for All Program in Ceará State [Avaliação do processo de implementação do programa Luz para Todos no estado do Ceará] [in Portuguese] (Ceará: Universidade Federal do Ceará) Gómez M F and Silveira S 2015 The last mile in the Brazilian Amazon—a potential pathway for universal electricity access Energy Policy 82 23–7 Borges P, Portugal-Pereira J, Koberle A, Lucena A F P, Szklo A and Schaeffer R 2016 Policy case study: evaluation of the Brazilian electrification program: Luz para Todos (Light for All Program) CD-Links Project Guimarães L N M R 2011 Effectiveness of the National Programme of Universal Access to Electricity—Light for All [A efetividade do programa nacional de universalização do acesso e uso da energia elétrica—Luz para Todos] [in Portuguese] vol 1 (UniCEUB Revista Brasileira de Políticas Publicas) pp 207–23 Brazil 2003 Decreto 4 873 of 11th November 2003 Brazil 2016a Indicators Management Platform [Plataforma de Gestão de Indicadores] [in Portuguese] (https://i3gov. planejamento.gov.br/?p=catalogo) (Accessed: 8 June 2016) Hsiao C 2003 Analysis of Panel Data 2nd edn (New York: Cambridge University Press) p 359 Brazil 2016b Debate about uncertainties of the Light for All Program future [Debate expõe duvidas sobre futuro do Luz para Todos] [in Portuguese] (Agência do Senado) (www12. senado.leg.br/noticias/materias/2016/10/27/debate-expoe- duvidas-sobre-futuro-do-luz-para-todos) (Accessed: 30 January 2017) IBGE 2000 Censoo Demográfico 2000 [Demographic Census 2000] [in Portuguese] (Rio de Janeiro: Brazilian Institute of Geography and Statistics) IBGE 2016 Socio-economic database, Brazilian Institute of Geographical Statistics (https://sidra.ibge.gov.br/pesquisa/ censo-demografico/demografico-2010/inicial) (Accessed: 25 January 2017) Coelho S and Goldemberg J 2013 Energy access: lessons learned in Brazil and perspectives for replication in other developing countries Energy Policy 61 1088–96 IPEA 2013 Bolsa Família Program: A Decade of Social Inclusion in Brazil (Brasília: Instituto de Pesquisa Econômica Aplicada) Cook P 2011 Infrastructure, rural electrification and development Energy Sustain Dev. 15 304–13 Jackson T 2009 Prosperity Without Growth (London: Earthscan) Cunha K B, Wlater A and Rei F 2007 CDM implementation in Brazil’s rural and isolated regions: the Amazonian case Clim. Change 84 111–29 Kanagawa M and Nakata T 2008 Assessment of access to electricity and the socioeconomic impacts in rural areas of developing countries Energy Policy 36 2016–29 Di Lascio M and Barreto E 2009 Energy and Sustainable Development in the Rural Amazon [Energia e desenvolvimento sutentável para a Amazônia rural Brasileira] [in Portuguese] (Brasilia: Kaco gráfica e Editora Ltda) Martinez D M and Ebenhack B W 2008 Understanding the role of energy consumption in human development through the use of saturation phenomena Energy Policy 36 1430–5 Mazur A 2011 Does increasing energy or electricity consumption improve quality of life in industrial nations? Energy Policy 39 2568–72 Dinkelman T 2011 The effects of rural electrification on employment: new evidence from South Africa Am. Econ. Rev. 101 3078–108 MDS 2015 Brazil without extreme poverty 1st edn (Brasília: Ministry Of Social Development And Fight against Hunger) Eletrobras 2015 Annual and Sustainability Report 2015 (http:// eletrobras.com/pt/SobreaEletrobras/Relatorio_Anual_ Sustentabilidade/2015/Relatorio-Annual-e-de-Sustentabilidade- Eletrobras-2015.pdf) (Accessed: 24 May 2017) MDS 2016 Bolsa família programme [Programa Bolsa Família] [in Portuguese] Ministry of Social Development and Fight Against Hunger (Brasília: Government of Brazil) (http:// mds.gov.br/assuntos/bolsa-Família) (Accessed: 20 August 2016) Eletrobras 2016 Light for All Program [Programa Luz para Todos] [in Portuguese] (www.eletrobras.com/elb/data/Pages/ LUMIS32AB99AAPTBRIE.htm) (Accessed: il )8 April 16) MME 2004 Anexo a Portaria 447 of 31st of December Ministry of Mines and Energy (Brasília: Government of Brazil) EPE 2015 Statistics of Electricity Production and Consumption— Base Year 2014 [Anuário estatístico de energia elétrica 2015 —ano base 2014] [in Portuguese] (Rio de Janeiro: Energy Research Company) MME 2008 Factsheet about the Light for All program [Informativo do programa Luz para Todos Nr. 9.] [in Portuguese] (Brasília: Government of Brazil) (www.mme. gov.br/luzparaTodos/Asp/informativos.asp) (Accssed: 30 June 2016) Fugimoto S K 2005 Universal Service to Electricity: Access and Use [A universalização do serviço de energia elétrica: Acesso e uso contínuo. Universidade de São Paulo] [in Portuguese] (São Paulo) MME 2011 Anexo a Portaria 628 of 3rd of November (Brasília: Government of Brazil) Environ. Res. Lett. 12 (2017) 095004 nan MME 2013 Impacts of the Light for All program [Impacto do Programa Luz para Todos: Pesquisa realizada pela empresa MDA Pesquisas] [in Portuguese] (Brasília: Government of Brazil) (www.mme.gov.br/luzparatodos/downloads/ pesquisa_de_satisfacao_2013.pdf) (Accessed: 25 January 2017) Sanchez A S, Torres E A and Kalid R A 2015 Renewable energy generation for the rural electrification of isolated communities in the Amazon region Renew. Sustain. Energy Rev. 49 278–90 Slough T, Urpelainen J and Yang J 2015 Light for all? Evaluating Brazil’s rural electrification progress, 2000–2010 Energy Policy 86 315–27 Slough T, Urpelainen J and Yang J 2015 Light for all? Evaluating Brazil’s rural electrification progress, 2000–2010 EnergyPolicy 86 315–27 MME 2017 Light for All program [Programa Luz para Todos] [in Portuguese] (Brasília: Government of Brazil) (www. mme.gov.br/luzparatodos/Asp/o_programa.asp) (Accessed: 25 January 2017) Soares S 2012 Bolsa Familia, its design, its impacts and possibilities for the future Working paper number 89 (Brasilia: International Policy Centre for Inclusive Growth) Steckel J, Brecha R J, Strefler J, Jakob M and Luderer G 2013 Development without energy? On the challenge of sustainable development in the context of climate change mitigation Ecol. Econ. 90 53–67 Motta M and Reiche K 2001 Rural Electrification, Micro-finance and Micro and Small Business (MSB) Development: Lessons for the Nicaragua Off-grid Rural Electrification Project (World Bank) (http://siteresources.worldbank.org/ EXTRENENERGYTK/Resources/5138246-1237906527727/ Rural_Electrification,_Micro-finance_and_Micro_and_ Small_Business.pdf) (Accessed: 27 July 2016) Steinberger J K and Roberts J T 2009 Across a moving threshold: energy, carbon and the efficiency of meeting global human development needs IIF Social Ecology working paper number 114 Oliveira P M C P 2013 Universalization of the access to electric power as a driver of economic development [Universalização do acesso à energia elétrica como vector de desenvolvimento econômico (Porto: Faculdade de Economia do Porto) [in Portuguese] Steinberger J K and Roberts J T 2010 From constraint to sufficiency: the decoupling of energy and carbon from human needs 1975–2005 Ecol. Econ. 70 425–33 UNDP 2002 Energy for sustainable development: a policy agenda (New York: United Nations Development Program) Pasternak A 2000 Global energy futures and human development: a framework for analysis US Department of Energy Report UCRL-ID-140773 (Livermore, CA: Lawrence Livermore National Laboratory) UNGA 2015 Transforming our world: the 2030 Agenda for Sustainable Development United Nations General Assembly Res. 70/1 (https://sustainabledevelopment.un.org/post2015/ transformingourworld) (Accessed: 26 January 2017) Pereira M G, Freitas M A V and da Silva N F 2010 Rural electrification and energy poverty: empirical evidences from Brazil Renew. Sustain. Energy Rev. 14 1229–40 Valer R L et al 2017 Issues in PV systems applied to rural electrification in Brazil Renew. Sustain. Energy Rev. 78 1033–43 WHO 2014 Indoor Air Quality Guidelines: Household Fuel Combustion (Geneva: World Health Organisation) Quinhões T A and Fava V M D 2010 Intersectoral and Transversal Programs Associated to Bolsa Família Programme [Intersetorialidade e transversalidade: a estratégia dos programas complementares do Bolsa Família] [in Portuguese] vol 61 (Brasília: Revista do Serviço Publico) pp 67–96 WHO 2016 Household air pollution and health WHO Fact sheet 292 (Geneva: World Health Organization) (www.who.int/ mediacentre/factsheets/fs292/en/) (Accessed: 26 January 2017) mediacentre/factsheets/fs292/en/) (Accessed: 26 January 2017) Wooldridge J 2002 Econometric Analysis of Cross section and Panel Data (Cambridge, MA: MIT Press)table_1nan Research Articletable_2Oil and Gas Companies — Are They Shifting to Renewables? A Study of Policy Mixes for Energy Transition in Brazil Alexandre Noguchi1 , Farley Simon Nobre1 1 Universidade Federal do Paraná, Curitiba, PR, Brazil chi, A., & Nobre, F. S. (2023). Oil and gas companies — Are they shifting to renewables? A study of policy mixes for energy transition in B nistration Review, 20(1), e220087. How to cite: Noguchi, A., & Nobre, F. S. (2023). Oil and gas companies — Are they shifting to renewables? A study of policy mixes fo BAR-Brazilian Administration Review, 20(1), e220087. DOI: https://doi.org/10.1590/1807-7692bar2023220087ABSTRACT Keywords: oil and gas; fossil fuel subsidies; energy transition; policy mix; sustainability. We argue that there is a need to advance further research that strengthens the anal- ysis of policy mixes for the energy transition in major emerging economies. In this context, this article aims to answer the following question: How do Brazil’s policies favor or hinder an energy transition of oil and gas companies (O&G) to renewables? To achieve this purpose, we conducted literature and archival research and interviews with experts to analyze (a) Brazil’s energy policy mixes that address O&G and renew- ables issues; and (b) major O&G companies’ activities and perspectives that influence the energy transition. Results demonstrated that though some of the O&G companies have made significant renewables investments in the last years, they continue focusing on O&G activities. We discuss the main policy mix features that hinder the prioritization of renewables by these O&G companies and that can undermine a sustainable energy transition in Brazil. JEL Code: L520. Received: May 27, 2022. This paper was with the authors for two revisions. Accepted: January 16, 2023. Publication date: February 02, 2023. Funding: The authors have stated that there is no financial support for the research in this article. Conflict of Interests: The authors have stated that there is no conflict of interest. Corresponding author: Farley Simon Nobre Universidade Federal do Paraná, Av. Prefeito Lothário Meissner, n. 632, Jardim Botânico, CEP 80210-170, Curitiba, PR, Brazil f.nobre@ufpr.br Editor-in-Chief: Ivan Lapuente Garrido (Universidade do Vale do Rio dos Sinos, Brazil). Associate Editor: Claudio Zancan (Universidade Federal do Paraná, Brazil). Reviewers: Laura Albuquerque (Universidade Federal do Rio de Janeiro, Brazil), and five anonymous reviewers. Editorial assistants: Kler Godoy and Simone Rafael (ANPAD, Maringá, Brazil). Data Availability: BAR – Brazilian Administration Review encourages data sharing but, in compliance with ethical principles, it does not demand disclosure of any means of identifying research subjects. documents received to the plagiarism check, using specific tools, e.g.: iThenticate Plagiarism Check: BAR maintains the practice of submitting all documents received to the plagiarism check, using specific tools, e.g.: iThenticate. Peer review: is responsible for acknowledging an article’s potential contribution to the frontiers of scholarly knowledge on business or public administration. The authors are the ultimate responsible for the consistency of the theoretical references, the accurate report of empirical data, the personal perspectives, and the use of copyrighted material. This content was evaluated using the double-blind peer review process. The disclosure of the reviewers’ information on the first page is made only after concluding the evaluation process, and with the voluntary consent of the respective reviewers. Copyright: The authors retain the copyright relating to their article and grant the journal BAR – Brazilian Administration Review, the right of first publication, with the work simultaneously licensed under the Creative Commons Attribution 4.0 International license (CC BY 4.0) The authors also retain their moral rights to the article, including the right to be identified as the authors whenever the article is used in any form. BAR-Brazilian Administration Review, 20(1), e220087, 2023. Oil and gas companies — Are they shifting to renewables? A study of policy mixes for energy transition in BrazilINTRODUCTION fuel production, which are less researched than sub- sidies for consumption (Rentschler & Bazilian, 2017). The central research question (RQ) in this article is: How do Brazil’s policies favor or hinder an energy transition of oil and gas companies to renewables? There is continuing interest in how policy mixes can favor energy transition (Haddad et al., 2022; Kern et al., 2019). Facing opportunities and threats in a glob- al energy transition movement, oil and gas (O&G) companies started to diversify their business mod- els to comprise new portfolios driven by renewables (Hartmann et al., 2021). O&G companies such as Shell, Total, bp, and Equinor have all created divisions for renewable energies. Former O&G company Ørsted completely divested its O&G segment in 2017, and it is now a renewable energy organization and one of the most sustainable firms in the world (Corporate Knights, 2022; Pickl, 2019; Stevens, 2016; Timperley, 2021). Public policies involve strategic instruments for an energy transition, as they directly affect firms’ in- vestment decisions (Markard, 2018; Rogge & Reichardt, 2016; Sovacool & Geels, 2016). In a statement about the 2021 IPCC report, UN Secretary-General António Guterres said, “… Countries should also end all new fossil fuel exploration and production, and shift fossil fuel subsidies into renewable energy” (United Nations, 2021). While most publications on this subject are from developed countries (Ghosh et al., 2021; Kern et al., 2019) and especially from European nations (Rogge et al., 2017), we found very few publications about the energy transition of O&G companies and policy mix- es from major emerging economies — subsumed by the BRICS — that have idiosyncratic institutions and complex socioeconomic challenges compared to de- veloped nations. We argue that there is a need to ad- vance further research that strengthens the analysis of policy mixes for energy transitions in major emerging economies. This article aims to understand the fea- tures of Brazil’s energy policy mix that favor or hinder a transition from O&G businesses toward renewables, moving away from fossil fuels. While the world re- newables usage was merely 14% in 2021, Brazil’s re- newable energy sources shared 46% of its national energy matrix, backed by a large production of sug- ar cane derivatives (e.g., ethanol) and electricity from hydropower — which accounted for 19.1% and 12.6%, respectively, of the total energy supply. In addition, Brazilian renewables also comprise a fast-growing so- lar and wind energy systems market, which account- ed for 1.7% and 8.8% in 2021, respectively (Empresa de Pesquisa Energética [EPE], 2022; International Energy Agency [IEA], 2021). However, the country still faces grand societal challenges regarding poverty and in- equality issues, in which context the royalties from O&G can be a valuable resource to tackle them. By g p To answer this inquiry, we organized qualitative research around two processes. First, we relied on the policy mix framework proposed by Rogge and Reichardt (2016) to analyze Brazil’s energy policy mix regarding its elements’ consistency1 with the tran- sition goals toward renewables. Among the three building blocks of the framework, we analyzed ele- ments (i.e., policy strategy and instrument mix) and characteristics (i.e., the consistency of the elements). Our analysis does not include political processes (i.e., the policymaking and implementation) because they are not relevant to our research question. Therefore, we studied the consistency of the elements to un- derstand how aligned the policy strategy and instru- ments are toward the transition of O&G companies to renewables. This first stage contributes to enlighten- ing how the policy mix impacts the O&G companies in Brazil. Second, we conducted archival research and interviews to gain reliability in our findings by draw- ing data from multiple sources. We performed archival research about seven major O&G companies in Brazil to find evidence of renewables and O&G activities. We conducted the interviews with O&G industry experts to capture their perceptions of Brazil’s policy mix for energy transition and O&G companies’ activities. We focused our research on public policies and O&G activ- ities relevant to the exploration and production (E&P) segment. We leave other O&G value chain segments for future research (e.g., refining and distribution). We discuss the main barriers in the Brazilian pub- lic policies that can hinder a transition toward renew- ables, including the fossil fuel subsidies that under- mine the global efforts to shift resources to a cleaner and sustainable energy matrix. Brazil heavily subsi- dizes its O&G production because it stimulates short- term economic growth and creates tax revenue to address social issues. We further discuss if these sub- sidies have effectively accomplished these two (eco- nomic and social) objectives and if the country should still need them. However, at least in the short run, we found that Brazil keeps going in an opposite direction of a needed transition to renewables since it still relies on a fossil fuel exploration regime with plenty of sub- sidies. Finally, we propose directions for the Brazilian policy mix to favor the transition of the O&G compa- nies toward renewables and to reform their fossil fuel subsidies. 2 BAR-Brazilian Administration Review, 20(1), e220087, 2023. A. Noguchi, F. S. NobreTHEORETICAL CONCEPTS AND FRAMEWORK expenditure (CAPEX) on renewables and electric mo- bility during 2022-2025 (Total Energies, 2021a). The transition of O&G companies O&G companies are unlikely to transition to renew- ables as Ørsted did ultimately. It is perilous to move out of their core business, and petroleum products will still be needed for many more decades (Hartmann et al., 2021; Stevens, 2016). As a first step to decarbonizing, O&G companies are likely to reduce their carbon inten- sity, deaccelerate their O&G exploration and production (E&P), and diversify their business portfolio with cleaner technologies (Fattouh et al., 2018; Stevens, 2016). O&G companies have sustained their business-as-usu- al models by continuously searching for new reserves, executing enormous projects, and not worrying too much about their operations’ externalities (like flaring). However, in a world of growing preoccupation with cli- mate issues and commitment to reducing fossil fuels, their old businesses show signals of failure. One of the pillars of this business model is to maximize the com- pany’s proven reserves, which means constantly drill- ing and acquiring new oilfields to increase their expect- ed future revenue. As the access to low-cost oilfields is getting scarce, companies have been exploring places like ultra-deep waters (e.g., the Brazilian pre-salt layer) or shale (typical in the USA). These oilfields increase the costs of adding new reserves and producing O&G, re- duce profitability, and make it more difficult for O&G companies to increase their value (Fattouh et al., 2018; Stevens, 2016). Like Brazil’s Repetro tax exemption pro- gram for O&G, production subsidies are essential to commercially make feasible many of the costly pre-salt layer fields (Centro Brasileiro de Infraestrutura [CBIE], 2019). Nevertheless, the growing number of legislations worldwide that restrict or phase out fossil fuels can fa- vor energy transition policy plans. France and Spain’s long-term decisions to date the end of all O&G produc- tion in their territory (for 2040 and 2042, respectively), and Canada, which has imposed restrictions for new licenses for offshore O&G in the Arctic (London School of Economics [LSE], 2021), are examples of the plans favoring progress in the energy transition. Intriguingly, national O&G companies (NOC), like Petrobras (the state-owned Brazilian O&G compa- ny, founded in 1953), seem to be behind the private companies, like Shell and Equinor (called international O&G companies, or IOCs), in the shift to renewables. According to one of the interviewees in this study — a petroleum politics researcher from an O&G multina- tional in Brazil —, NOCs have different concerns than the IOCs, like ensuring the nation’s oil supply and re- solving social issues. Indeed, NOCs are not driven by stock prices, and they are not pressed for climate ac- tions as the IOCs are. Thus, IOCs are generally pushed to decarbonize faster than NOCs. The interviewee said, “you don’t see protests at CNOOC and Gazprom’s doors like you see at Exxon’s.” Cheon et al. (2015) argue that NOCs are generally oriented by their ‘national pur- pose,’ and that their political and economic goals come before profit. Petrobras, for example, is a NOC, and it has a clear strategy to focus on O&G production for the following years, with very few activities in renewables (Petrobras, 2021). The state should serve as an exam- ple, but these contradictions suggest that private O&G companies are more interested in the energy transition than governments of oil-exporting countries. Aware of the increasing difficulties of operating their oil and gas-based business model, many O&G com- panies diversify their portfolios. One common strate- gy is mergers and acquisitions or joint ventures with renewable energy companies, like bp with Bunge and Shell with Raízen in Brazil for ethanol production. Shell created a ‘New Energies’ division in 2016 to work with hydrogen, renewable energies, and electrical vehicles (Pickl, 2019), and Total plans to spend 20% of its capitalThe Brazilian energy matrix According to the Brazilian Company of Energy Studies (EPE, 2022), Brazil’s energy matrix comprises 46% of re- newables. In contrast, the world average is merely 14% (EPE, 2022; IEA, 2021). Table 1 presents the breakdown of each energy source in the matrix. Table 1. Break down of the Brazilian energy matrix (in 2021).table_3Note. Source: EPE (2022). BAR-Brazilian Administration Review, 20(1), e220087, 2023. Oil and gas companies — Are they shifting to renewables? A study of policy mixes for energy transition in Brazil sugar cane producer globally (Statista, 2023). While the country has had little growth and sometimes a decrease in its hydropower production in the last few years, the production of sugar cane derivatives is still growing (Empresa de Pesquisa Energética [EPE], 2021b). When it comes to the electricity matrix, the share of renewables is 83% in Brazil, and the world average is 27%. This high share of renewables comes from hydro- power (65.2%), biomass (9.1%), wind power (8.8%), and solar power (1.7%) (EPE, 2022). In Brazil, the total hy- dropower production has slightly decreased from 34.6 Mtoe in 2010 to 34.0 Mtoe in 2020, while the petroleum production increased from 106.5 to 152.6 Mtoe in the same period (Empresa de Pesquisa Energética [EPE], 2021a). It is noteworthy that the capacity for hydropow- er electricity production is expected to increase only 4.2% from 2020 to 2030, while petroleum production is expected to grow 62.2% in the same period (Ministério de Minas e Energia, 2021a; 2021b). Therefore, to in- crease its share of renewables, Brazil needs to boost the development of additional renewable sources. Sugar cane derivatives include ethanol and its ba- gasse, which is widely used for heat production in in- dustry and electricity generation. These two renew- able energy sources have been under accelerated development since the 1970s. However, other mod- ern renewables, such as wind and solar power, have only become significant to the energy matrix after the 2000s. Biomass energy production started to become a fast-growing activity in 2000, wind power in 2014, and solar in 2015 (EPE, 2021b). That was mainly due to successful policies created at that time, such as the PROINFA (Programa de Incentivo às Fontes Alternativas de Energia Elétrica, or Incentive Program for Alternative Sources of Electric Energy) renewables incentive pro- gram, the Reserve Energy Auctions (LER) for wind and long-term solar contracts, and regulations for net metering (EPE, 2021b; Lozornio et al., 2017; Silva et al., 2020). The country promotes these alternative re- newable sources to diversify its energy matrix and to reduce its dependency on traditional energy sources. From 2020 to 2030, wind power capacity is expected to grow by 202% (from 15.9 to 32.2 GW), solar power by 270% (from 3.1 to 8.4 GW), biomass by 8.6%, and distributed generation by 583% (from 4.2 to 24.5 GW) (Ministério de Minas e Energia, 2021a; 2021b).Oil and gas in Brazil Until 1997, only Petrobras was allowed to produce O&G in Brazil. When the O&G monopoly ended, the gov- ernment created public policies and subsidies to en- courage foreign companies and new players to join the market. The Repetro program was created around that time, in 1999, to achieve those objectives, and it still is one of the most influential production subsidies for the O&G industry in Brazil. The Repetro program is a special customs regime that exempts specific equipment and components for O&G activities from federal taxes,2 thus increasing the feasibility and profitability of O&G projects (CBIE, 2019; PWC, 2022; Santos & Avellar, 2017). When the pre-salt layer reserves were confirmed, the government be- came even more interested in developing the oil busi- ness and increasing its production. Therefore, new taxes were created, like special participation fees and signature bonuses, and the Social Fund was formed, in 2010, to provide resources for social development (Agência EPBR, 2021; Jesus et al., 2017; Oliveira & Laan, 2010; Pereira & Neto, 2017).Policy mixes Multiple policies that influence an energy transition comprise conflicting goals. Therefore, it is important to understand their interactions and influence on the over- all goal when studying public policies. Policy mixes are essential in studying sustainability transitions because they guide the direction and pace of the transition (Gunningham & Sinclair, 1999; Kern et al., 2019; Rogge & Reichardt, 2016). Policy mix refers to a combination of multiple policy instruments such as a country’s pub- lic policies. The policy instrument is a generic term to describe government programs, public measures, laws, regulations, and other tools used by the government to achieve strategic goals. Policy instruments can reduce taxes, directly provide resources, or indirectly mobilize other actors to spend their resources (Kern et al., 2019; Rogge & Reichardt, 2016). Examples of policy instru- ments are feed-in tariffs, carbon emissions regulations, and decarbonization credits.Renewables in Brazil Brazil has a long history of promoting the development of petroleum, biofuels, and hydropower. However, only in the past few decades the Brazilian government has made significant progress in supporting alternate re- newable sources, like solar and wind power (Lozornio et al., 2017; Oliveira & Laan, 2010; Silva et al., 2020). Brazil’s most traditional and vital renewable energy sources come from sugar cane derivatives and hydro- power. The country benefits from a large hydropow- er capacity ranked only behind China (International Energy Agency [IEA], 2022). Its land is well suited for sugar cane production, standing as the number one Many authors use policy mixes to advance research on sustainability transitions, especially on energy tran- sitions (Kern et al., 2019; Rogge et al., 2017). We use 4 BAR-Brazilian Administration Review, 20(1), e220087, 2023. A. Noguchi, F. S. Nobre Rogge and Reichardt's (2016) policy mixes framework to analyze the policies in this article. Their frame- work organizes terminology in policy mixes and offers sub-elements and categories for public policies, allow- ing a clear scope analysis. Rogge and Reichardt (2016, p. 1622) define “the policy mix as a combination of the three building blocks elements, processes and charac- teristics, which can be specified using different dimen- sions.” Elements comprise two sub-elements: the pol- icy strategy and the instrument mix. Policy strategy is divided into policy objectives and the principal plans. The first refers to long-term targets, such as Brazil’s target to achieve 10% efficiency gains in the electrici- ty sector by 2030 (International Energy Agency [IEA], 2018), while the latter indicates the general path that the government wants to take, such as the objective of Brazil’s National Energy Policy to “increase the use of natural gas” (Lei No. 9478, 1997). The instrument mix is the combination and the result of the interaction of all policy instruments of a policy mix. Policy processes subsume the policymaking and implementation pro- cesses, and the last block refers to the characteristics of the elements and policy processes: (1) the consis- tency of elements, (2) the coherence of processes, (3) credibility, and (4) comprehensiveness of a policy mix. Moreover, their framework conceptualizes dimensions to delineate the policy mix: policy field (or domain) (e.g., transport, education, energy), governance level (e.g., federal laws, state laws), geography, and (4) time. All three building blocks influence social and techno- logical change, but researchers can choose to focus on one block, a combination of two blocks, or some of their minor components. The framework helps define a focus or scope of analysis (e.g., the interaction between political processes and the policy strategy). Using this approach helps clarify the blocks, links, and scope of the policy mixes under study and avoids jeopardizing the research’s findings and validation (Ossenbrink et al., 2019). vant subjects, like political processes, disputes of power, and policy implementation. Rogge and Reichardt's (2016) policy mixes framework to analyze the policies in this article. Their frame- work organizes terminology in policy mixes and offers sub-elements and categories for public policies, allow- ing a clear scope analysis. Rogge and Reichardt (2016, p. 1622) define “the policy mix as a combination of the three building blocks elements, processes and charac- teristics, which can be specified using different dimen- sions.” Elements comprise two sub-elements: the pol- icy strategy and the instrument mix. Policy strategy is divided into policy objectives and the principal plans. Carefully choosing the boundary settings of the study is very important to research. We understood that to answer our research question and ‘draw a picture’ of the status of the transition, we needed not only to ad- dress the policies that support the renewables’ regime but also the encompassing regime (oil and gas) and the policies that support or put pressure on it. The methodology section will show how we started by defining which blocks and components of the framework we used in the research to study our boundary setting and scope of analysis properly. Then, it shows how we captured the policy instruments, plans, and strategies relevant to our scope of analysis and their characteristics (i.e., their consistency). The results, shown in the Results section, are orga- nized by the policy mix framework. We show in each table the data for each component of the framework. For example, Table 3 shows the policy objectives, Table 4 the principal plans, and Table 5 the policy instruments, and all these tables also show the characteristics. This way of presenting the results helps readers familiar with the policy mix literature assimilate the results.METHODOLOGY This article’s central question is: How do Brazil’s policies favor or hinder an energy transition of oil and gas com- panies to renewables? We answer this inquiry by devel- oping knowledge on two interwoven topics: (a) Brazil’s energy policy mixes that address O&G and renewables issues, and (b) Brazilian O&G companies’ activities and perspectives that influence the energy transition. Regarding the first topic, we conducted literature and archival research to comprehensively analyze Brazil’s policy instrument mix and its influence on the energy transition. Our analysis relied on the policy mix concept (Flanagan et al., 2011; Rogge & Reichardt, 2016) to understand not only a single instrument but also the combination and interaction between multiple policy instruments.Connecting the theoretical framework to the methodology and results Sustainability transitions occur in complex political spaces with an extensive and sophisticated network of actors, comprising technological, economic, socio-cul- tural, and institutional changes. Therefore, research- ers must eliminate irrelevant and biased elements to avoid an overly complicated and inefficient analysis. We borrowed terminology and analytical tools from the Rogge and Reichardt's (2016) policy mix framework to achieve better acceptance, validity, and uniformity, allow a more straightforward comparison of findings (Kern et al., 2019), and constitute a consistent set of in- Rogge and Reichardt's (2016) policy mix framework is used in this research because it suits the approach of studying not one but multiple policy instruments and analyzing its effects on the phenomenon of interest (i.e., the transition to renewables). The framework also allows the researchers to choose which of its blocks and components they will use or not in their research. This work allowed us to adhere only to the blocks and components relevant to study the actual status of the transition, leaving out specific elements to non-rele- Oil and gas companies — Are they shifting to renewables? A study of policy mixes for energy transition in Brazil terwoven policy blocks (Ossenbrink et al., 2019; Rogge & Reichardt, 2016). focuses on the policymaking and implementation pro- cesses. Our research question concerns the present state of the policy mix. We did not include the design features of the instruments because we did not intend to make an in-depth analysis of single instruments, but only to study their influence on each other toward the transition’s goals. Table 2 presents the dimensions used in our search for policy instruments in Brazil. We adopt- ed these dimensions because they capture the space in which interactions can occur in the scope of our re- search inquiry, which is related to the present in Brazil and is about renewables and the O&G industry. We chose to focus on the E&P segment because it is the first stage of the O&G value chain, so that it will gener- ate more activity in the following stages. We analyzed federal policies because they have the most relevance to the Brazilian energy production system, and we left other governance levels for future research. We first identified the policy mix framework’s blocks and components of our research interest. Then, we concentrated our attention on the elements — the pol- icy strategy (policy objectives and principal plans) and the policy instrument (goal, type, and purpose) — and their characteristic of consistency. To understand the combined effect of the policy in- struments on the transition of O&G companies, we an- alyzed the nature of their interactions — which can be positive, negative, or neutral. To study the interactions among the instruments and between the instrument mix and the policy strategy, we chose to analyze the characteristic of consistency because it focuses on the elements’ current state and indicates contradictions in the policy mix that make it inefficient in achieving the transition’s goals. We did not include the political processes in our analysis because this building block Table 2. Public policies’ dimensions adopted in this study.table_4the policy components and linkages within the policy mix framework as applied in this article. Guided by the public policies’ dimensions in Table 2, we conducted archival research of policy objectives, principal plans, and instruments in the context of Brazil’s energy system. We captured the policy objectives from the updated Brazilian national determined contribution (IEA, 2018). It is currently the official document that guides the national renewable energy targets, and we brought the principal policy mix plans from the National Energy Policy (Lei No. 9478, 1997). There are no feder- al-level policy objectives related to the development of O&G, only production forecasts. We searched for the relevant policy instruments on the IEA policy database. However, we also investigated government databas- es, executive reports, and strategic plans (e.g., Lei No. 9478, 1997), news, O&G private and public organiza- tions’ websites, and research institutions’ libraries (e.g., CBIE, 2019; INESC, 2020a) to learn more about these instruments and their impacts on the energy system. Our archival research found policy instruments such as the Repetro O&G tax exemption and the PROINFA re- newables programs. We then proceeded to classify all these policy instruments according to their goal, type, and purpose. Built upon the framework of Rogge and Reichardt (2016), Figure 1 presents a research design of We performed two consistency analyses to under- stand how the instrument mix contributes to our re- search question. A first consistency analysis between the Brazil National Energy Policy’s objectives and the O&G companies’ transition goals toward renewables (objective versus goals) is represented by the linkage 2 in Figure 1. A second one, between the energy poli- cy instruments and the goals of the transition of O&G companies (instrument mix versus goals), is represent- ed by linkage 3. These analyses are further examined in subsection “Energy policy mix in Brazil”. Our research question lies in linkage 1, representing the policy mix’s influence on the transition of O&G companies to renewables. A consistent policy mix can have all its objectives achieved without trade-offs. We assume that the O&G companies’ transition goals toward renewables are: (1) to reduce the efforts in O&G exploration and production (E&P), and (2) to increase renewable energy activities. They are defined as goals, not objectives because they are desired effects that contribute to the energy tran- sition’s long-term objective. We limited our research to 6 BAR-Brazilian Administration Review, 20(1), e220087, 2023. A. Noguchi, F. S. Nobre ciency and carbon capture and storage (CCS), are also important proxies for the energy transition. the E&P activities because they represent the first stage in the O&G value chain. Moreover, investments in E&P can favor progress in the subsequent stages, like refin- ing and distribution. We conducted archival research on these firms’ an- nual and strategic reports (e.g., Equinor, 2021; Petrobras, 2021; Shell, 2021; 2022). We also searched for addi- tional information on their website and the news (e.g., Reuters and the Brazilian executive magazine Exame). Furthermore, we read reports from energy and petro- leum organizations that comprise the International Energy Agency (IEA), the U.S. Energy Information Administration (EIA), the Brazilian Energy Research Company (EPE), and the ANP. We gathered relevant information about investment, property, CAPEX, and budget forecasts of renewables and E&P in Brazil, as presented in subsection “O&G companies’ activities in Brazil”. To enrich the discussion of the policy mixes, we pay special attention to one of the most controversial types of policy instruments: fossil fuel subsidies. Section “Discussion on O&G subsidies in Brazil” debates the fol- lowing question: Does Brazil still need to subsidize its O&G production? We also analyze if the tax revenues from O&G have been effective in developing the local community and addressing social issues. Regarding the second topic, we carefully ana- lyzed E&P and renewables activities of seven major O&G companies in Brazil. They are Petrobras, Equinor, Total, Shell, Galp, Repsol Sinopec, and bp. These are all publicly traded firms, and, except for Petrobras and Sinopec (from the joint venture Repsol Sinopec), they are multinational companies that originated in Europe. According to the National Agency for Petroleum, Natural Gas, and Biofuels in Brazil (ANP), this group of firms account for 95% of Brazil’s oil production (Agência Nacional do Petróleo, Gás Natural e Biocombustíveis [ANP], 2021a). It is noteworthy that this selection con- tains the O&G companies with the highest activity in renewables globally (Shell, Total, bp, and Equinor) (Pickl, 2019. We focused on renewable energy activities be- cause it is an important proxy for the energy transition, and most O&G companies that have been acting on climate issues have made some investment in this sec- tor. However, other actions, like improving energy effi- To foster our data gathering and strengthen anal- ysis, we interviewed two senior professionals in the O&G industry to capture their understanding of policies and O&G companies’ activities for an energy transition. The first interviewee is a VP of renewables at an O&G multinational operating in Brazil. The second is a se- nior researcher in petroleum politics who has worked in O&G companies in Brazil. We conducted semi-struc- tured interviews, which lasted from 30 to 60 minutes. We focused our questions to find answers to our main research inquiry: How do Brazil’s policies favor or hin- der an energy transition of oil and gas companies to re- newables? In the interviews, we focused on those two interwoven topics: (a) Brazil’s energy policy mixes that address O&G and renewables issues; and (b) Brazilian Oil and gas companies — Are they shifting to renewables? A study of policy mixes for energy transition in Brazil from interviews and secondarily from archival research, were collected between April 2021 and January 2022. O&G companies’ activities and perspectives that influ- ence the energy transition. Additionally, we built reliability in our study by as- sembling data from multiple sources. We searched for publicly available material involving written and vid- eo-recorded information from Brazilian managers and experts from these major O&G companies. The infor- mation included interviews, webinars, and workshops delivered and recorded for the 2019, 2020, and 2021 Rio Oil and Gas Congresses, and publicized by epbr agen- cy and other O&G related institutions. These multiple data sources added information to our study especially on what those major O&G companies are doing re- garding the energy transition in Brazil. All data, primarilyRESULTS AND ANALYSESRESULTS AND ANALYSES Energy policy mix in Brazil We started our analysis with the two components of the policy strategy: the policy objectives (Table 3) and the principal plans (Table 4) of Brazil’s energy policies. They were retrieved from Brazil’s Nationally Determined Contribution (NDC) (IEA, 2018) and the National Energy Policy (originally described in the Brazilian Law No. 9,478 from 1997) (Lei No. 9478, 1997), respectively. Then, we classified them regarding their consistency with the goals of the assumed transition of O&G companies to renewables (described in the Methodology section). Table 3. Consistency of Brazil’s policy objectives for energy with the transition’s goals.table_5Note. Source: The authors, with data from IEA (2018).Table 4. Consistency of Brazil’s principal plans for energy with the transition’s goals.table_6newables in the Brazilian energy matrix. Therefore, there are no contradictions with the transition’s ob- jective of increasing renewables activity by O&G com- panies. Nevertheless, there is a contradiction in the transition’s objective of reducing E&P activity with the energy policy’s objective of increasing the use of nat- ural gas. As for the objective of ensuring the supply of petroleum products, it is not necessarily a trade-off with the reduction of activity in E&P, as a reduction can occur, and the supply can still be guaranteed, so we defined it as neutral. With this consistency analysis, we conclude that most policy objectives and principal plans of Brazil’s energy policy align with the transition. The policy objectives (Table 3) are all consistent with the transition’s goals, which is no surprise since they were presented in Brazil’s NDC, which is a document that contains plans for climate action. Despite having some policy instruments with goals for O&G develop- ment, Brazil has no quantifiable objectives for this ac- tivity as the country has for renewables. The Program for the Revitalization of Onshore O&G (REATE), for ex- ample, has a goal to achieve 500 mboe/d by 2030, but this is an instrument goal, not a policy objective, so we did not include it in our tables. As for the analysis of the principal plans (Table 4), most of the objectives of the National Energy Policy aim to increase total production and the share of re- 8 BAR-Brazilian Administration Review, 20(1), e220087, 2023. A. Noguchi, F. S. Nobre For the second consistency analysis, between the in- strument mix and the goals of the transition, we present all the policy instruments relevant to renewables and E&P found in our archival research from the IEA policy data- base and other complementary sources. We sorted them according to their consistency to the goals of the transi- tion. Table 5 shows all the instruments consistent with the transition, and Table 6 shows the ones that are not. All the instruments were classified as per Rogge and Reichardt (2016) categorization of primary type and purpose. Table 5. Group of policy instruments promoting renewables (consistent with the transition Table 5. Group of policy instruments promoting renewables (consistent with the transition).table_7Note. Source: The authors. Table 6. Group of policy instruments promoting E&P (inconsistent with the transition).table_8many inconsistencies with the goals of the transition for it has many instruments aiming to promote O&G. One of the interviewees, the oil company VP, said that “the oil business would secure our income while our company shifts to renewables,” supporting that in- struments that promote O&G may indirectly favor in- vestments in renewables by O&G companies. In line with that view, the other interviewee said that O&G many inconsistencies with the goals of the transition for it has many instruments aiming to promote O&G. Analyzing the instrument mix against the transition goals, the group of policy instruments promoting re- newables (Table 5) has synergy with the transition’s goal of increasing renewables activity. In contrast, the group of policy instruments promoting E&P (Table 6) conflicts with the goal of reducing activity in E&P. Differently from the policy strategy, which has a good alignment with the transition, the instrument mix has One of the interviewees, the oil company VP, said that “the oil business would secure our income while our company shifts to renewables,” supporting that in- struments that promote O&G may indirectly favor in- vestments in renewables by O&G companies. In line with that view, the other interviewee said that O&G Oil and gas companies — Are they shifting to renewables? A study of policy mixes for energy transition in Brazil E&P may indirectly allow more investments in renew- ables by these firms. companies have had “waves” of investment in renew- ables in the past, and all these waves happened in peri- ods of high oil prices, and they ended when the oil cri- ses came. This statement supports the idea that higher profitability in O&G activities motivates O&G compa- nies to invest in renewables; thus, policies supporting Table 7 presents some of the key interviewees’ per- spectives about the factors that slow down the most the energy transition from O&G toward renewables in Brazil. Table 7. Interviewees’ perspectives on policy features that hinder the energy transition from O&G to renewables in Brazil.table_9Note. Source: The authors.O&G companies’ activities in Brazil According to the energy policy and Brazil’s Decennial Energy Plan (Ministério de Minas e Energia, 2021a; 2021b), from the Ministry of Energy, the core energy fuels for Brazil will still be hydropower, biofu- els, and petroleum products in the 2021-2030 period. Other renewables and natural gas are considered com- plementary fuels to the core ones, but the government still promotes them. The strategy for Brazil’s energy pol- icy is not to transition from fossil fuels to renewables, as there is no policy instrument to limit or reduce E&P ac- tivity. Still, it is part of the strategy to increase the share of renewables in the energy matrix. O&G companies’ activities in Brazil Figures 2 and 3 present time distributions of the history of acquiring new O&G exploratory blocks in ANP bid- ding rounds for the seven selected major companies. Exploratory blocks are demarcated areas that are po- tentially abundant in O&G resources, so they are sold from the government to O&G companies for explora- tion and production rights. Figure 2 presents Petrobras’ data, and Figure 3 presents the other six O&G compa- nies (Equinor, Total, Shell, Galp, Repsol Sinopec, and bp). Each bar in the graphs shows the sum of partici- pation shares that were acquired by those companies altogether in that year, where 100% is equivalent to an entire block, and 500% mean, for example, a total of shares equivalent to five blocks. Exploratory blocks do not have the same size or the same potential for O&G production, but we do not make distinction regarding these characteristics in this analysis. The period con- sidered is from 2000 to 2021 (bidding rounds started in 1997). These distributions are based on the ANP database (Agência Nacional do Petróleo, Gás Natural e Biocombustíveis [ANP], 2021b) and represent an im- portant indicator of future E&P activity because if com- panies have purchased O&G blocks recently, they will develop them. Therefore, they are likely to produce O&G for decades. Proposition 1. A policy mix with a policy strategy aiming to increase the share of renewables will be inefficient (in achieving this objective) if the policy mix has inconsistent policy instruments favoring progress in O&G. Proposition 2. The lack of infrastructure, proper regulations, legal frameworks, and federal agencies for renewables holds back foreign investment in re- newables in emerging countries, hindering a sus- tainability transition by O&G companies. Proposition 3. If not supported by incentives and a fair taxation system, renewables will remain as sec- ondary energy sources to fossil fuels in emerging countries that subsidize O&G. We analyzed the acquisition rate of blocks per year of each company. Dividing the average of 2015-2021 and 2000-2021, we found the following: Galp (0.46); Petrobras (0.63); Total (0.68); Shell (1.53); bp (1.80); A. Noguchi, F. S. Nobre nies have increased their acquisition rates in the last six years compared to the average of the last twenty-one years (i.e., a result higher than 1.0). Equinor (1.88); and Repsol Sinopec (2.74). This calcula- tion allowed us to compare the recent acquisition rate with the historical average. Four of the seven compa-table_10table_11keep developing their blocks, like Bacalhau and BM-C- 33 for Equinor and Pau Brasil for bp. Although these companies do not show signs of reducing their E&P activity in Brazil in the next few years, most of them are increasing their renewables activity. Petrobras, Shell, Total, Equinor, and bp all have renewable energy as- sets already producing in Brazil, like biofuels, biogas, onshore wind, and solar power. Shell and bp are nota- ble for ethanol production through joint ventures (with Raízen and bp Bunge, respectively). In solar and wind, Petrobras, bp, Total, and Equinor are already produc- ing significant amounts of energy. Repsol Sinopec is the only one of the seven companies considered here that do not have any renewable’s activities in Brazil (BP, 2021a; 2021b; Equinor, 2020; 2021; 2023; Galp, 2021b; Petrobras, 2021; Repsol, 2021; Shell, 2022; Total Energies, 2021b). Table 8 presents a summary of the O&G companies’ renewables activities in Brazil. When asked if O&G companies are transitioning to renewables or if they are trying to maintain the status quo, both the interviewees said that, in their opinion, the current focus of these companies in Brazil right now is O&G. “Honestly, the focus of O&G companies in Brazil to- day is to make money with fossil fuels. They have invested a large amount of money in Brazil buying fields, including Shell, Equinor, and CNOOC. They’ll want a return on their investment.” (Interviewee — oil company VP). Galp has announced a target to increase by 25% its oil production in Brazil by 2025 (compared to 2021) (Siqueira, 2021). Total has a target to reach 150,000 boe/d, 150% more than 2021 (Total Energies, 2021a). Other companies did not inform a target, but they will BAR-Brazilian Administration Review, 20(1), e220087, 2023. Oil and gas companies — Are they shifting to renewables? A study of policy mixes for energy transition in BrazilTable 8. Summary of activities in renewables for the O&G companies in Brazil.table_12It is important to note that O&G companies are rel- evant players in the renewables market. For example, in the ethanol market, Shell’s joint venture with Raízen is the largest producer in Brazil, while bp Bunge is in the top four (UDOP, 2020), and Lightsource bp has a mas- sive capacity of 2.2 GW from solar power. As we see that O&G companies in Brazil continue to make huge investments in the O&G business, even with the exis- tence of various instruments favoring renewables, we propose that: that the government does not efficiently use O&G tax- es, and they failed to significantly change the situation of the poor in Brazil. Fossil fuel subsidies are significant barriers that hin- der the world’s energy transition toward renewables. On one side, typical justification for them comprises poverty alleviation, industrialization growth, and eco- nomic development (Cheon et al., 2015; Rentschler & Bazilian, 2017). On the other side, they generate un- desired effects such as increased carbon emissions, increased energy demand, and unsustainable fiscal burdens for governments (Moghaddam & Wirl, 2018; Oliveira & Laan, 2010; Timperley, 2021). In 2009, G20 countries (including Brazil) committed to phase out fossil fuel subsidies and reform inefficient subsidies. Although these countries still spend hundreds of mil- lions of dollars annually on it, many oil-exporting coun- tries successfully reform and reduce their fossil fuel subsidies, like India, Iran, and Mexico (Mason & Ennis, 2009; Moghaddam & Wirl, 2018; Rentschler & Bazilian, 2017; Timperley, 2021). Proposition 4. Inasmuch as an emerging country (like Brazil) has no plans or policies to phase out O&G production, O&G companies will continue produc- ing O&G for the next decades on a large scale.DISCUSSION ON O&G SUBSIDIES IN BRAZIL The purpose of this section is to discuss if Brazil still needs to subsidize its O&G production and if the tax revenues from O&G have been effective in developing local communities and addressing social issues. There are subsidies for production and for con- sumption of fossil fuels, and Brazil has both. On the one hand, consumption subsidies aim at reducing the final price of fuel for end users and to promote industrializa- tion by supporting energy-intense industries with low- er energy costs (Moghaddam & Wirl, 2018; Rentschler & Bazilian, 2017; Oliveira & Laan, 2010). On the other hand, production subsidies are meant to encourage com- panies to increase their production of fossil fuels, and they usually increase the profit for producers (INESC, 2020a; Timperley, 2021; Zhao et al., 2019). As this ar- ticle is about policies that affect E&P, we are especially interested in discussing production subsidies. There has been little progress in reforming this type of subsidies, and they have received much less attention from re- searchers than consumption subsidies (Rentschler & Bazilian, 2017). Fattouh et al. (2018, p. 5) argue that for oil-export- ing countries “there is no conflict between renewable investment and hydrocarbon business in these coun- tries” because with the increase of renewable energy domestic production, these countries are allowed to export more O&G. This may make sense in an econom- ic view, but in an environmental point of view, expor- tation of O&G still hinders the global efforts for climate action. The interviewed petroleum politics researcher said, “each country will make the energy transition that it can afford.” She argued that Brazil has poverty and inequality issues that developed countries do not have, and the tax revenues from O&G could change that. She said that Brazil’s strategy for an energy transition could be to maintain and subsidize the O&G business. At the same time, the government can focus on reduc- ing emissions in other areas, like energy efficiency and deforestation. As an illustration case, Hogarth (2016) showed that Brazil could decrease its GHG emissions significantly by reducing deforestation. Still, she agrees Most authors and organizations define subsides as the tax and financial policy instruments that directly reduce the price of fossil fuels for consumers or the production cost for producers (Coady et al., 2010; BAR-Brazilian Administration Review, 20(1), e220087, 2023. 2 BAR-Brazilian Administration Review, 20(1), e220087, 2023. A. Noguchi, F. S. Nobre INESC, 2020a; Timperley, 2017; Timperley, 2021), and by this definition, only the Repetro program and the Act No. 13,586/2017 are considered production subsidies among all the listed policy instruments in this study. & Neto, 2017; Poubel & Santos-Junior, 2017). Jesus et al. (2017) have studied the five cities in Brazil that are most dependent on the revenue from O&G and concluded that, in the period of 2005 to 2015, social inequality in- creased in all the five cities, and in some of these cit- ies the educational and violence levels became worse. According to Oliveira and Laan (2010), poor families did benefit from subsidies in Brazil in the last decades, but the large industrial energy consumers were the ones benefited the most, while the common taxpayers are the ones that paid for all that. Even though fossil fuel subsidies are usually justified as a support to the poor, many times most of the subsidies are received by the rich, who tend to consume proportionally more energy than the poor population (Cheon et al., 2015; Rentschler & Bazilian, 2017). According to the Institute of Socio-economic Studies (INESC) (INESC, 2020a; 2020b), the cost of fossil fuel production and consumption subsidies in 2019 for the Brazilian government were R$ 36.27 billion (US$ 7.10 billion) and R$ 63.01 billion (US$ 12.6 billion), respective- ly. The cost of production subsidies came mostly from foregone tax revenues for O&G from Repetro program (77%) and Act No. 13,586/2017 (17%), while 83% of the cost of consumer subsidies came from diesel and gas- oline tax reductions. Does Brazil still need to subsidize its O&G produc- tion? Rentschler and Bazilian (2017) have analyzed sub- sidy reforms in many countries, and they argue that “in practice, the key rationale for implementing subsidy re- form has typically been fiscal rather than environmen- tal” (Rentschler & Bazilian, 2017, p. 2). They still add that “the necessity and urgency of reform can only be fully understood when considering the complete range of adverse environmental, social and economic side ef- fects of fossil fuel subsidies” (Rentschler & Bazilian, 2017, p. 2). If not for environmental reasons, leaders can re- form their subsidies for the benefit of their economies in the long-term. When the Repetro program was created in 1999, it was supposed to develop the market and to expire in 2020. Whether Repetro was responsible or not, its goal to develop the industry and to bring new players to the market was certainly achieved. In 1997, Brazil’s to- tal petroleum production was almost 1 million boe/d,3 and in January 2022, the daily production was around 3.8 million boe/d, being 74% of that production derived from the pre-salt area and coming from many new players in the market other than Petrobras (Agência Nacional do Petróleo, Gás Natural e Biocombustíveis [ANP], 2022b; Agência EPBR, 2021). With such accom- plishments, one could say that O&G would not need production subsidies after 2020 (when Repetro should expire). Still, in 2018, the government extended the scope of the Repetro program and its validity to 2040. The government’s rationale behind this decision was that the subsidy would continue to promote new in- vestment, increase the country’s competitiveness, and bring more players to the market (Brasil, 2017). As an economic justification to reform the sub- sidies, the INESC institute (2020a) claims that tax re- nounce from production subsidies largely reduces state revenues that are essential to the Brazilian popu- lation, like PIS (Social Integration Program) and COFINS (Contribution to Social Security Financing), which are fundamental for state pension and unemployment insurance. Dr. Fernanda Delgado de Jesus, petroleum politics researcher in Brazil, highlights that “all public policies need to be measured,” and even though the production subsidies for O&G are costly, they bring large economic benefits to the population through royalties, special participation fees, and signature bo- nus, and many cities rely upon these taxes, so the pos- itive effects from these subsidies must be considered (Núcleo WIN Brazil UFBA, 2021, 1h12m). In 2021, the O&G business in Brazil distributed R$ 37.6 billion (US$ 7.5 billion) in royalties and R$ 36.8 billion (US$ 7.3 billion) in special participation fees for the government, and part of this revenue is expected to be used in basic services, such as health and education (Agência Nacional do Petróleo, Gás Natural e Biocombustíveis [ANP], 2022a). Studies have shown that the government revenue We conclude that this subsidy’s main goal is not to support a nascent industry, but to continually increase its production and have economic benefits. While Brazil may have been successful in its economic objectives for the pre-salt, we cannot say the same for the social development goals on poverty alleviation and inequal- ity reduction. Indeed, business cases that prioritize short-term economic decisions will be disconnect- ed to the long-term societal and environmental out- comes (Hahn et al., 2018). In section “Energy policy mix in Brazil”, we showed that Brazil’s energy policy mix has a policy strategy highly oriented to the progress of re- newables, but it includes many O&G policy instruments that are inconsistent with it. In section “O&G companies’ activities in Brazil”, we show data from O&G compa- nies supporting that the O&G instrument mix has been successful in its goals of developing the O&G industry. Studies have shown that the government revenue from the O&G has failed to significantly reduce poverty and to improve the educational levels in cities that re- ceive royalties from O&G (Martinez & Reis, 2016; Pereira BAR-Brazilian Administration Review, 20(1), e220087, 2023. Oil and gas companies — Are they shifting to renewables? A study of policy mixes for energy transition in Brazil and removal of subsidies for fossil fuels while creating more subsidies for renewable technologies that are still not competitive, like second-generation ethanol and offshore wind — following Cheon et al. (2015) and Rentschler and Bazilian (2017). In this section, we present a line of though suggest- ing that, although O&G instruments have achieved its short-term economic goals, they have failed to achieve social goals. As they also hold back the transition to renewables, they might not be beneficial to society in the end. The literature shows that subsidizing O&G is not efficient for the economy or social development in the long run. Fossil fuels subsidies may create short- term stimulus for the economy, but they normally cause detrimental effects for sustainability issues in the long-term.4 They incentivize growth in energy consumption and discourage energy efficiency and low-carbon energy sources (Oliveira & Laan, 2010; Rentschler & Bazilian, 2017). Researchers showed that there are more efficient manners for a government to spend money to alleviate poverty than subsidies, such as direct cash transfer programs or investment in basic services for the population, and that is a major sup- porting argument for eliminating them (Cheon et al., 2015; Jain, 2019; Moghaddam & Wirl, 2018; Rentschler & Bazilian, 2017). According to Cheon et al. (2015, p. 376), the subsidies for fuel in Brazil “encouraged ex- cess and inefficiency and benefited industries more than they did low-income households, widening the gap between the wealthy and the poor.” Proposition 5. Policy instruments that artificially lower the production cost of O&G reduce the com- petitiveness of renewables and discourage invest- ment in low carbon technologies, thus hindering a sustainable energy transition.CONTRIBUTIONS AND FINAL REMARKS One of the main expectations of the current global energy transition is to reduce GHG emissions from energy use radically. A clean transition will be done by quitting fossil fuels and replacing them with re- newables and electrification. Brazil is ahead of most countries regarding total use of renewable energy, with a vehicle fleet that can run almost entirely with biofuels, a meager share of coal, and most of its elec- tricity coming from hydropower (EPE, 2022). Unlike most developed countries, Brazil’s GHG emissions do not come from energy use. Instead, they come pri- marily from land-use change and the forestry sector (Timperley, 2018). All these facts must be understood by the popu- lation and the political parties to avoid opposition to subsidies reforms. Politicians must clearly communi- cate the population what is being done to compen- sate the removal of subsidies and what are the long- term benefits. Society must understand that the extra revenues will be used in their benefit in more efficient manners, like cash transfer or social programs. The short-term fiscal benefits are exchanged for a long- term economic development. Not to say the environ- mental reasons. It is also important to create mech- anisms that will protect the most vulnerable citizens from high prices, as the poorest cannot wait for long- term returns (Jain, 2019; Rentschler & Bazilian, 2017). Brazil’s public policies seem to favor O&G more than renewables as the petroleum segment has more political benefits than other energy sources. Petroleum has federal institutions to coordinate the market (e.g., ANP and the Secretary of Petroleum, Natural Gas and Biofuels — SPG), a mature regulatory framework, tax benefits (e.g., Repetro and Act No. 13,586/2017), fi- nancing programs, and R&D mandates. Petroleum does not have infrastructure limitations for distribu- tion as other sources like biogas and electricity. According to Rogge et al. (2017, p. 2), transforma- tive policy mixes for sustainability transitions “need to combine different instruments addressing multiple market and system failures by fulfilling different pur- poses, such as technology push and demand pull.” In order to speed up investments in renewables in Brazil and similar emerging economies, the authors recom- mend that public policies should include: (a) regu- latory frameworks for all renewable energy sources, which will allow developments in new projects, pre- dictability for investment, and a competitive ener- gy market; (b) regulatory agencies for all renewable energy sources; (c) established plans for the energy sector, which will guide the creation of new policies and instruments; (d) policies to speed up the develop- ment of infrastructure for energy; and (e) restrictions As for the main theoretical contributions of this article, we empirically studied how the interplay be- tween building blocks of policy mixes (in this case, the elements and the consistency characteristics) affects the effectiveness of policy mix in directing a change toward sustainability objectives (i.e., the shift of O&G companies toward renewables). We validated the link- age between consistency, policy strategy, and the in- strument mix of Rogge and Reichardt's (2016) frame- work in an empirical study in an emerging economy. Therefore, we expanded the geographical scope of policy mixes and sustainability transitions previously applied to European studies (Ghosh et al., 2021; Rogge et al., 2017) to a significant emerging country. BAR-Brazilian Administration Review, 20(1), e220087, 2023. A. Noguchi, F. S. Nobre We identified some limitations in our work but that open new fruitful opportunities for future research. First, our scope of analysis is limited to national-lev- el policies and the E&P segment only. There are also important policy instruments at the state government level and policies for other oil and gas segments that can largely influence our research question, for ex- ample, tax reduction on fuel price for end consumer policies and some local incentives for E&P and renew- ables. Second, we analyzed the consistency among the policy instruments. Nevertheless, Rogge and Reichardt (2016) show that other characteristics and design features of policies (like comprehensiveness, credibility, coherence, stringency, and depth) are in- fluencers of the policy mix toward its goal. Therefore, these unexplored elements are subject to further re- search to encompass a more holistic perspective of the problem. Third, we focused our analysis on the shift of O&G companies from E&P to renewables, and there is space for a broader analysis of all the other actions that these companies have done regarding the energy transition, like energy efficiency, increase in natural gas use, CCS, and carbon offset measures. Petrobras, for example, has a clear strategy to focus its energy transition actions on these later technologies, and not on renewables. Furthermore, we suggest the need for additional research in emerging countries — and possibly those belonging to the BRICS — to test the validity of our propositions deduced from the pol- icy mixes and O&G activities in Brazil.NOTES 1. Consistency “captures how well the elements of the policy mix are aligned with each over, thereby con- tributing to the achievement of policy objectives” (Rogge & Reichardt, 2016, p. 1626). 2. The Repetro avoids the incidence of II (Imposto de Importação, or Importation Tax), IPI (Imposto sobre Produtos Industrializados, or Tax over Industrialized Products) and PIS (Programa de Integração Social, or Social Integration Program) and COFINS (Contribuição para o Financiamento da Seguridade Social, or Contribution to Social Security Financing) (PWC, 2022). 3. Boe/d means barrels of oil equivalent per day, ac- counting for petroleum and natural gas. 4. “Indeed, global warming and air pollution are sources of mortality, food insecurity, and diseases that main- ly stress vulnerable populations by overshooting the need for medical treatment and hospitalization” (Nobre, 2022, p. 147). 5. REATE stands for Programa de Revitalização da Atividade de Exploração e Produção de Petróleo e Gás Natural em Áreas Terrestres, or Program for the Revitalization of Oil and Natural Gas Exploration and Production in Onshore Areas.REFERENCES Agência EPBR. (2021, November 29). Passado e futuro do marco regulatório do petróleo, por Décio Oddone. EPBR Agency. https://epbr.com.br/passado-e- futuro-do-marco-regulatorio-do-petroleo-por-decio-oddone/ As developed countries pressure their O&G com- panies to decarbonize, O&G companies might seek new projects in countries without restrictions and with subsidies to O&G, like Brazil. We should expect O&G companies in Brazil to maintain oil as their primary asset for the next decade, and renewables will be a complementary and a growing business. However, for now, O&G companies will most likely invest in renew- ables because they want to de-risk their future oper- ations, not relying solely on petroleum products and because they want to keep their ‘license to operate’ to satisfy their stakeholders. Nevertheless, an energy transition toward renewables by the O&G companies will not occur if public policies do not lead that way. Agência Nacional do Petróleo, Gás Natural e Biocombustíveis. (2021a). Boletim da produção de petróleo e gás natural (novembro de 2021). https://www.gov.br/ anp/pt-br/centrais-de-conteudo/publicacoes/boletins-anp/boletins/arquivos- bmppgn/2021/2021_11_boletim.pdf Agência Nacional do Petróleo, Gás Natural e Biocombustíveis. (2021b). Relação dos concessionários. https://www.gov.br/anp/pt-br/centrais-de-conteudo/dados- abertos/relacao-de-concessionarios#:~:text=A%20rela%C3%A7%C3%A3o%20 dos%20concession%C3%A1rios%20consolida,signat%C3%A1rias%20e%20 suas%20respectivas%20participa%C3%A7%C3%B5es Agência Nacional do Petróleo, Gás Natural e Biocombustíveis. (2022a). Arrecadação com royalties e participação especial foi recorde em 2021. https:// www.gov.br/anp/pt-br/canais_atendimento/imprensa/noticias-comunicados/ arrecadacao-com-royalties-e-participacao-especial-foi-recorde-em-2021 Agência Nacional do Petróleo, Gás Natural e Biocombustíveis. (2022b). Boletim da Produção de Petróleo e Gás Natural (janeiro de 2022). https://www.gov.br/ anp/pt-br/centrais-de-conteudo/publicacoes/boletins-anp/boletins/arquivos- bmppgn/2022/boletim-janeiro.pdf BP (2021a). Annual report and form 20-F 2020. https://www.bp.com/content/ dam/bp/business-sites/en/global/corporate/pdfs/investors/bp-annual-report- and-form-20f-2020.pdf We trust this article contributed with significant analysis and foundational propositions that opened novel perspectives to assist managers and policymak- ers’ strategic decisions on the nature, opportunities, and threats of institutional policy mixes and O&G ac- tivities in a relevant emerging country. This research shed new light on advancing knowledge on tackling a grand societal challenge regarding the need to move toward a sustainable energy transition effectively. BP (2021b). Investor pack: July 2021. https://www.bp.com/content/dam/bp/ business-sites/en/global/corporate/pdfs/investors/bp-esg-investor-pack- march-2021.pdf BP (2021c). Biocombustíveis. https://www.bp.com/pt_br/brazil/home/produtos- e-servicos/bp-bunge-bioenergia.html Brasil, C. I. (2017, August 18). Governo prorroga regime aduaneiro especial para setor petroleiro até 2040. Agência Brasil. https://agenciabrasil.ebc.com. br/economia/noticia/2017-08/governo-prorroga-regime-aduaneiro-especial- para-setor-petroleiro-ate-2040 Centro Brasileiro de Infraestrutura (2019, December 27). Quais os benefícios do REPETRO? https://cbie.com.br/artigos/quais-os-beneficios-do-repetro/ Oil and gas companies — Are they shifting to renewables? A study of policy mixes for energy transition in Brazil Cheon, A., Lackner, M., & Urpelainen, J. (2015). Instruments of political control: National oil companies, oil prices, and petroleum subsidies. Comparative Political Studies, 48(3), 370-402. https://doi.org/10.1177/0010414014543440 Jesus, F. D., Silva, T. B., Resende, L. D., Roitman, T., & Silva, J. F. (2017, October). Discussing royalties in Brazil in oil price volatility times: A public management analysis. Proceedings of the Offshore Technology Conference Brazil, Rio de Janeiro, RJ, Brazil. https://doi.org/10.4043/27978-MS Coady, D., Gillingham, R., Ossowski, R., Piotrowski, J., Tareq, S., & Tyson, J. (2010). Petroleum product subsidies: Costly, inequitable, and rising. IMF Staff Position Notes, 10(5). https://doi.org/10.5089/9781455239108.004 Kern, F., Rogge, K. S., & Howlett, M. (2019). Policy mixes for sustainability transitions: New approaches and insights through bridging innovation and policy studies. Research Policy, 48(10), 103832. https://doi.org/10.1016/j.respol.2019.103832 Corporate Knights (2022, January 19). The 100 most sustainable corporations of 2022. https://www.corporateknights.com/rankings/global-100-rankings/2022- global-100-rankings/100-most-sustainable-corporations-of-2022/ Lei Nº 9478, de 6 de agosto de 1997 (1997). Dispõe sobre a política energética nacional, as atividades relativas ao monopólio do petróleo, institui o Conselho Nacional de Política Energética e a Agência Nacional do Petróleo e dá outras providências. Casa Civil. http://www.planalto.gov.br/ccivil_03/leis/l9478.htm Empresa de Pesquisa Energética (2021a). Relatório síntese do balanço energético nacional – BEN 2021. https://www.epe.gov.br/sites-pt/publicacoes- dados-abertos/publicacoes/PublicacoesArquivos/publicacao-601/topico-588/ BEN_S%C3%ADntese_2021_PT.pdf Lozornio, E. J. C., Moro, R. L., Souza, J. F. T., & Simões, A. F. (2017, December). Políticas públicas para o fomento da inserção da energia eólica na composição da matriz brasileira de oferta de energia elétrica. Proceedings of the XIX Encontro Internacional sobre Gestão Empresarial e Meio Ambiente - ENGEMA, São Paulo, SP, Brazil. http://engemausp.submissao.com.br/19/anais/arquivos/467.pdf Empresa de Pesquisa Energética (2021b). BEN 50 anos. https://www.epe.gov.br/ pt/publicacoes-dados-abertos/publicacoes/ben-50-anos Empresa de Pesquisa Energética (2022). Matriz energética e elétrica (ano base de 2020). https://www.epe.gov.br/pt/abcdenergia/matriz-energetica-e-eletrica London School of Economics (2021, October 19). Domestic limits to fossil- fuel production and expansion in the G20. The London School of Economics and Political Science, Grantham Research Institute on Climate Change and the Environment. https://www.lse.ac.uk/granthaminstitute/news/domestic-limits- to-fossil-fuel-production-and-expansion-in-the-g20/ Equinor (2020). Annual report 2019. https://www.equinor.com/content/dam/ statoil/documents/annual-reports/2019/equinor-2019-annual-report-and- form-20f.pdf Equinor (2021). Sustainability report 2020. https://www.equinor.com/news/ archive/20210319-annual-sustainability-reports-2020 Markard, J. (2018). The next phase of the energy transition and its implications for research and policy. Nature Energy, 3, 628-633. https://doi.org/10.1038/s41560-018-0171-7 Equinor (2023). Facts about our renewable assets, October 2022. https://cdn.equinor. com/files/h61q9gi9/global/fe5ac0501c89e4f7add791583edc619abe00c691. Martinez, A. L., & Reis, S. S. (2016). Impacto dos royalties do petróleo no índice de educação básica: Análise do caso dos municípios capixabas. RACE – Revista de Administração, Contabilidade e Economia, 15(2), 505-530. https://doi.org/10.18593/race.v15i2.9590 pdf?ren-asset-facts-october2022-equinor.pdf Fattouh, B., Poudineh, R., & West, R. (2018). The rise of renewables and energy transition: What adaptation strategy for oil companies and oil-exporting countries? [OIES Paper n. MEP 19]. The Oxford Institute for Energy Studies, Oxford United Kingdom https://doi org/10 26889/9781784671099 Mason, J., & Ennis, D. (2009, September 25). G20 agrees on phase-out of fossil-fuel subsidies. Reuters. https://www.reuters.com/article/us-g20-energy-idUSTRE58O18U20090926 Flanagan, K., Uyarra, E., & Laranja, M. (2011). Reconceptualising the ‘policy mix’ for innovation. Research Policy, 40(5), 702-713. https://doi.org/10.1016/j.respol.2011.02.005 Moghaddam, H., & Wirl, F. (2018). Determinants of oil price subsidies in oil and gas exporting countries. Energy Policy, 122, 409-420. https://doi.org/10.1016/j.enpol.2018.07.045 Galp. (2021a). Presença no mundo - Brasil. https://www.galp.com/corp/pt/sobre-nos/presenca-no-mundo/brasil Ministério de Minas e Energia. (2021a). Plano decenal de expansão de energia. https://www.epe.gov.br/sites-pt/publicacoes-dados-abertos/publicacoes/ PublicacoesArquivos/publicacao-50/topico-87/Sumário%20Executivo%20 do%20PDE%202021.pdf Ministério de Minas e Energia. (2021a). Plano decenal de expansão de energia. https://www.epe.gov.br/sites-pt/publicacoes-dados-abertos/publicacoes/ Galp. (2021b). Integrated management report 2020. https://www.galp.com/ corp/Portals/0/Recursos/Investidores/SharedResources/Relatorios/en/2020/ GalpRC20IMR.pdf PublicacoesArquivos/publicacao-50/topico-87/Sumário%20Executivo%20 do%20PDE%202021.pdf Ghosh, B., Ramos-Mejía, M., Machado, R. C., Yuana, S. L., & Schiller, K. (2021). Decolonising transitions in the Global South: Towards more epistemic diversity in transitions research. Environmental Innovation and Societal Transitions, 41, 106-109. https://doi.org/10.1016/j.eist.2021.10.029 Ministério de Minas e Energia. (2021b). REATE 2020. http://antigo.mme.gov. br/web/guest/secretarias/petroleo-gas-natural-e-biocombustiveis/acoes-e- programas/programas/reate-2020 Nobre, F. S. (2022). Cultured meat and the sustainable development goals. Trends in Food Science & Technology, 124, 140-153. https://doi.org/10.1016/j.tifs.2022.04.011 Gunningham, N., & Sinclair, D., (1999). Regulatory pluralism: Designing policy mixes for environmental protection. Law Policy, 21(1), 49-76. https://doi.org/10.1111/1467-9930.00065 Núcleo WIN Brazil UFBA (2021, June 23). Dia 3: Conexão WINNER - Semana da Transição Energética [Video]. YouTube. https://www.youtube.com/watch?v=y8OGiLSlfCM&t=5164s Haddad, C., Nakić, V., Bergek, A., & Hellsmark, H. (2022). Transformative innovation policy: A systematic review. Environmental Innovation and Societal Transitions, 43, 14-40. https://doi.org/10.1016/j.eist.2022.03.002 Oliveira, A., & Laan, T. (2010). Lessons learned from Brazil’s experience with fossil-fuel subsidies and their reform. International Institute for Sustainable Development. https://www.iisd.org/publications/report/lessons-learned- brazils-experience-fossil-fuel-subsidies-and-their-reform Hartmann, J., Inkpen, A. C., & Ramaswamy, K. (2021). Different shades of green: Global oil and gas companies and renewable energy. Journal of International Business Studies, 52(5), 879-903. https://doi.org/10.1057/s41267-020-00326-w Hahn, T., Figge, F., Pinkse, J., & Preuss, L. (2018). A paradox perspective on corporate sustainability: Descriptive, instrumental, and normative aspects. Journal of Business Ethics, 148, 235-248. https://doi.org/10.1007/s10551-017-3587-2 Ossenbrink, J., Finnsson, S., Bening, C. R., & Hoffmann, V. H. (2019). Delineating policy mixes: Contrasting top-down and bottom-up approaches to the case of energy-storage policy in California. Research Policy, 48(10), 103582. https://doi.org/10.1016/j.respol.2018.04.014 Hogarth, J. R. (2016). Evolutionary models of sustainable economic change in Brazil: No-till agriculture, reduced deforestation and ethanol biofuels. Environmental Innovation and Societal Transitions, 24, 130-141. https://doi.org/10.1016/j.eist.2016.08.001 Pereira, D. A. L., & Neto, A. C. (2017). Índices de desenvolvimento municipais e royalties do petróleo: uma abordagem multivariada de comparação de perfis entre municípios que recebem ou não royalties pelo petróleo produzido. GEPROS. Gestão da Produção, Operações e Sistemas, 12(3), 238-264. https://doi.org/10.15675/gepros.v12i3.1712 International Energy Agency (2018). Country reports: Brazil – 2018 update. IEA Bioenergy. https://www.ieabioenergy.com/wp-content/uploads/2018/10/ CountryReport2018_Brazil_final.pdf Petrobras (2021). Annual report 2020. https://www.investidorpetrobras.com.br/ apresentacoes-relatorios-e-eventos/relatorios-anuais/ International Energy Agency (2021). Implementation of bioenergy in Brazil – 2021 update. IEA Bioenergy. https://www.ieabioenergy.com/wp-content/ uploads/2021/11/CountryReport2021_Brazil_final.pdf Poubel, R., & Santos-Junior, P. (2017, October). A bacia de campos, os royalties e os trabalhadores: Uma análise da pendularidade na Região Norte Fluminense. Proceedings of the XVI Seminário de Integração, Campos dos Goytacazes, RJ, Brazil. https://seminariodeintegracao.ucam-campos.br/wp-content/ uploads/2018/02/A-Bacia-de-Campos.pdf Poubel, R., & Santos-Junior, P. (2017, October). A bacia de campos, os royalties e os trabalhadores: Uma análise da pendularidade na Região Norte Fluminense. Proceedings of the XVI Seminário de Integração, Campos dos Goytacazes, International Energy Agency (2022). Hydroelectricity. International Energy Agency. https://www.iea.org/reports/hydroelectricity Proceedings of the XVI Seminário de Integração, Campos dos Goytacazes, RJ, Brazil. https://seminariodeintegracao.ucam-campos.br/wp-content/ uploads/2018/02/A-Bacia-de-Campos.pdf INESC (2020a). Incentivos e subsídios à produção de petróleo e gás no Brasil: Três motivos para reformá-los. INESC. https://www.inesc.org.br/wp-content/ uploads/2020/12/Estudo-de-Caso-PRODUCAO_V06.pdf Pickl, M. J. (2019). The renewable energy strategies of oil majors – From oil to energy? Energy Strategy Reviews, 26, 100370. https://doi.org/10.1016/j.esr.2019.100370 INESC (2020b). Incentivos e subsídios ao consumo de combustíveis fósseis no INESC (2020b). Incentivos e subsídios ao consumo de combustíveis fósseis no Brasil: Entre amplas renúncias e graves impactos climáticos e sociais INESC INESC (2020b). Incentivos e subsídios ao consumo de combustíveis fósseis no Brasil: Entre amplas renúncias e graves impactos climáticos e sociais. INESC. PWC (2022). Repetro. PwC Brasil. https://www.pwc.com.br/pt/consultoria- tributaria-societaria/incentivos-fiscais/repetro.html#:~:text=O%20Repetro%20 %C3%A9%20um%20regime,da%20marinha%20mercante%20(AFRMM) Brasil: Entre amplas renúncias e graves impactos climáticos e sociais. INESC. https://www.inesc.org.br/wp-content/uploads/2020/12/Estudo-de-Caso- CONSUMO_V04.pdf Jain, A. K. (2019). A fine balance: Lessons from India’s experience with petroleum subsidy reforms. Energy Policy, 119, 242-249. https://doi.org/10.1016/j.enpol.2018.04.050 Rentschler, J., & Bazilian, M. (2017). Reforming fossil-fuel subsidies: Drivers, barriers and the state of progress. Climate Policy, 17(7), 891-914. https://doi.org/10.1080/14693062.2016.1169393 BAR-Brazilian Administration Review, 20(1), e220087, 2023. A. Noguchi, F. S. Nobre Timperley, J. (2021). Why fossil-fuel subsidies are so hard to kill. Nature, 598(7881), 403-405. https://doi.org/10.1038/d41586-021-02847-2 Repsol Sinopec (2022). Repsol Sinopec Brasil divulga plano de sustentabilidade 2021. https://repsolsinopec.com.br/noticias/repsol-sinopec-brasil-divulga- plano-de-sustentabilidade-2021/ Total Energies (2021a). Exploração e produção. Total Energies Brasil. https:// totalenergies.com.br/exploracao-e-producao-total-brasil Repsol (2021). 2021-2025 strategic plan. Repsol (2021). 2021-2025 strategic https://www.repsol.com/en/about-us/2025-strategy/index.cshtml Total Energies (2021b). Annual report form 20-F 2020. Total Energies. https:// totalenergies.com/system/files/documents/2021-03/2020-total-form-20-f.pdf Rogge, K. S., & Reichardt, K. (2016). Policy mixes for sustainability transitions: An extended concept and framework for analysis. Research Policy, 45(8), 1620- 1635. https://doi.org/10.1016/j.respol.2016.04.004 UDOP (2020). Quatro dos maiores processadores de cana brasileiros têm ociosidade; confira. https://www.udop.com.br/noticia/2020/5/11/quatro-dos- maiores-processadores-de-cana-brasileiros-tem-ociosidade-confira.html Rogge, K. S., Kern, F., & Howlett, M. (2017). Conceptual and empirical advances in analysing policy mixes for energy transitions. Energy Research & Social Science, 33, 1-10. https://doi.org/10.1016/j.erss.2017.09.025 United Nations (2021, August). Secretary-general’s statement on the IPCC working group 1 report on the physical science basis of the sixth assessment. United Nations, Secretary-General. https://www.un.org/sg/en/content/ secretary-generals-statement-the-ipcc-working-group-1-report-the-physical- science-basis-of-the-sixth-assessment Santos, R. J., & Avellar, A. P. M. (2017). Políticas de apoio à indústria de petróleo e gás no Brasil: Um estudo das ações públicas para o desenvolvimento da cadeia de valor. Economia e Sociedade, 26(3), 721-750. https://doi.org/10.1590/1982-3533.2017v26n3art7 Zhao, X., Luo, D., Lu, K., Wang, X., & Dahl, C. (2019). How the removal of producer subsidies influences oil and gas extraction: A case study in the Gulf of Mexico. Energy, 166, 1000-1012. https://doi.org/10.1016/j.energy.2018.10.139 Siqueira, C. (2021, June 4). Galp mira gás e renováveis no Brasil. Jornal Petróleo Hoje. https://petroleohoje.editorabrasilenergia.com.br/galp-mira-gas-e- renovaveis-no-brasil/ Shell (2021). Shell Annual Report 2020. https://reports.shell.com/annual-report/2020/ Shell (2022). Strategic report - Shell annual account and reports 2021. https://reports.shell.com/annual-report/2021/strategic-report.htmlAuthorsAlexandre Noguchi Silva, A. A. C., Oliveira Filho, O. D. Q., Araújo, A. M., Silva, C. O. G., Ferreira, C. R., Andrade, L. I., & Arruda Filho, P. H. C. (2020). Análise das atuais políticas de incentivo à mini e microgeração distribuída e da certificação de aerogeradores de pequeno porte no Brasil. Brazilian Journal of Development, 6(7), 52217- 52235. https://doi.org/10.34117/bjdv6n7-755 Alexandre Noguchi Universidade Federal do Paraná Av. Prefeito Lothário Meissner, n. 632, Jardim Botânico, CEP 80210-170, Curitiba, PR, Brazil Av. Prefeito Lothário Meissner, n. 632, Jardim Botânico, CEP 80210-170, Curitiba, PR, Brazil alexandrenoguchi@gmail.comFarley Simon Nobre Sovacool, B. K., & Geels, F. W. (2016). Further reflections on the temporality of energy transitions: A response to critics. Energy Research & Social Science, 22, 232-237. https://doi.org/10.1016/j.erss.2016.08.013 Universidade Federal do Paraná Universidade Federal do Paraná Av. Prefeito Lothário Meissner, n. 632, Jardim Botânico, CEP 80210-170, Curitiba, PR, Brazil f.nobre@ufpr.br Statista (2023). Leading sugar cane producers worldwide in 2021, based on production volume. https://www.statista.com/statistics/267865/principal- sugar-cane-producers-worldwide/Authors’ contributions Stevens, P. (2016). International oil companies: The death of the old business model. Chatham House, The Royal Institute of International Affairs. https:// Stevens, P. (2016). International oil companies: The death of the old business model. Chatham House, The Royal Institute of International Affairs. https:// www.biee.org/wpcms/wp-content/uploads/International-oil-companies-Paul- Stevens.pdf 1st author: conceptualization (equal), data curation (equal), formal analysis (equal), investigation (equal), methodology (equal), project administration (equal), validation (equal), visualization (equal), writing – original draft (equal), writing – review & editing (equal). www.biee.org/wpcms/wp-content/uploads/International-oil-companies-Paul- Stevens.pdf Timperley, J. (2017). Explainer: The challenge of defining fossil-fuel subsidies. Carbon Brief. https://www.carbonbrief.org/explainer-the-challenge-of- defining-fossil-fuel-subsidies 2nd author: conceptualization (equal), data curation (equal), formal analysis (equal), investigation (equal), methodology (equal), project administration (equal), supervision (equal), validation (equal), visualization (equal), writing – original draft (equal), writing – review & editing (equal). Timperley, J. (2018). The carbon brief profile: Brazil. Carbon Brief. https://www.carbonbrief.org/the-carbon-brief-profile-brazil. BAR-Brazilian Administration Review, 20(1), e220087, 2023.Copyright of BAR - Brazilian Administration Review is the property of Associacao Nacional de Pos-Graduacao e Pesquisa em Administracao (ANPAD) and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.C H A P T E R 1 5 Equity, Greenhouse Gas Emissions, and Global Common Resources Paul Baer The core of the climate change problem is as simple as it is daunting: We now know that it will be impossible for the whole world to obtain—or even to approach—the emissions levels of the industrialized countries, without gravely endangering our planetary life support systems. In the United States, emissions average over 5 tons of carbon per person per year;1 even in more efficient Euro- pean economies, average emissions exceed 2 tons of carbon yearly. Yet global annual emissions must fall by more than 50 percent—to a third of a ton per per- son or less—if atmospheric greenhouse gas (GHG) levels are to be stabilized in this century. Meeting this target would be a big enough problem if it were just a matter of the industrialized nations reducing their emissions by a factor of 5 or 10, and bargaining among themselves about how to share the atmosphere. But this halv- ing of total emissions must take place in a world where more than a billion peo- ple live on less than a dollar a day and 30 percent of children under 5 are mal- nourished.2 This matters because it is still generally assumed that the solution to poverty is for the poor nations to “develop” along the same path that the rich nations have—for the South to become like the North. But this model depends on increasing energy use and, given current technology, increasing GHG emis- sions. Thus, if the developing nations follow the energy-technology path of the rich countries, the planet faces the risk of catastrophic climate change. Since this risk puts a limit on allowable GHG emissions, one can think of that limit as defining the available “environmental space.” And there is simply not enough environmental space for the South to develop the way the North has. Therefore, the particular environmental space at issue here—the atmos- PART V. DEVELOPMENT AND EQUITY phere—must be brought under common governance; global rules for its use and allocation must be discussed, decided, and enforced. The UNFCCC and the Kyoto Protocol are steps in this direction. The structure of the Kyoto Protocol, which establishes binding emissions caps on the developed countries, has made the distribution of those caps a cen- tral controversy both in the negotiations and in the ratification debate. Devel- oping countries were explicitly exempted from caps because of their lower his- torical and current emissions, and because of their agreed need to devote their resources to sustainable development and poverty alleviation.3 However, U.S. opponents of the Kyoto Protocol have vehemently argued that the protocol is not fair to the United States because developing countries have no caps and bear no costs. The Clinton Administration did not submit the Kyoto Protocol for ratifica- tion and thus avoided this debate over fairness. On taking office, President Bush used the fairness argument as one reason for rejecting the protocol outright. Environmental groups in the U.S. made some effort to counter it, but their strategy seems largely to rely on ratification without the United States and on seeking domestic reductions outside the Kyoto framework. However, the unre- solved debate over developing country commitments will continue to focus attention on the fair distribution of emissions rights. If the Kyoto Protocol enters into force without the United States, caps for developing countries will be crucial to the negotiation of targets for the second commitment period (after 2012). If the protocol doesn’t enter into force, the requirement that developing countries limit their emissions probably will be a key bargaining issue in negotiating an alternative agreement. And if the United States wants the developing countries to accept caps, it will have to propose an allocation formula that addresses the developing countries’ fundamental con- cerns over equity. Everyone in the developing world cannot emit at the high rates of the North, but why should developing countries agree to restrictions that bind them to their current, much lower per capita rates or that restrict their economic growth? What is an equitable solution to this dilemma? There are other important aspects of equity in the climate debate, such as the risks we impose on future generations (intergenerational equity) and liability for the harm that will be caused by climate change we are unable or unwilling to avoid.4 However, because the question of equitable allocations among countries remains a major—and urgent—unresolved obstacle to an effective global treaty, I focus on the allocation issue in this chapter. One can view the question of an equitable allocation of emissions rights as a political science problem or as an ethical problem. Much of what has been writ- ten about equity in the climate negotiations is relatively traditional political sci- Chapter 15. Equity, Greenhouse Gas Emissions, and Global Resources ence; for example, authors attempt to analyze the rational (economic) interests of the parties and their relative power, and try to predict the outcome of the negotiations. In this framework, the analyst is a neutral observer, and equity (or the perception of inequity) is a possible variable to account for the negotiating outcome.5 Sometimes authors will go so far as to suggest possible allocation for- mulas that they believe could be acceptable to all parties; however, the interests and preferences of countries are taken as given. Alternatively, an ethical analysis looks at the justifications that competing parties offer for their negotiating positions and attempts to critically evaluate them.6 In this framework, the analyst is a participant, and equity is something to be defined and argued for in order to influence the world. This is how I approach the problem in this chapter.Framing the Problem: Burden Sharing Versus Resource Sharing in the Global Commons The climate change problem can be posed as a question of burden sharing or as a question of resource sharing.7 In the burden-sharing framework, the costs of protecting the atmosphere by reducing emissions to a safe level are a burden that must be shared globally. The costs come from the need to introduce lower- emitting technologies—presumed to be more expensive—and the requirement for reduced consumption. The issue of equitable allocations is usually framed this way;8 U.S. opposition to the Kyoto Protocol is based on the argument that it imposes an unfair burden. In this framework, it makes sense to say that the burden should be shared equally unless there are compelling reasons why it shouldn’t be. If we accept a principle of equal sacrifice, and we believe that it is a greater sacrifice for a poor person to pay a dollar than it is for a rich person—in economic terms, the declining marginal utility of income—we might define a person’s or country’s fair share based on ability to pay. Like a progressive tax, this would mean that the wealthy pay a higher proportion of their income than the poor do, but the poor still pay something. However, focusing on the burden of reductions obscures the question of who has been responsible for, and benefited from, the overuse of the atmosphere. Assessing responsibility requires us to focus on the atmospheric carbon sink as an economic resource, and to account for both its unequal appropriation in the past and its unequal use today. We need to ask who has used the resource, what benefits they have acquired from its use, and what losses will be suffered by those who cannot use as much as they otherwise would have. If the finite size of the available atmospheric space defines the total benefits that can come from its use,PART V. DEVELOPMENT AND EQUITY it is necessary to ask whether a person or country has received or will receive a fair share of the benefits. In this way we can meaningfully define overuse and underuse and define a party’s obligation on this basis: Parties that have exceeded their share have obligations to parties that will therefore get less. To understand what would be a fair share, it will be necessary to look further at the nature of common resources.The Nature of Common Resources From one perspective, any system in which the use of a resource by one party causes harm to another can be viewed as a commons. Those harmed necessarily have a moral stake in the use or conservation of the resource, even if they don’t have the ability to exploit it in kind and thus to cause a symmetric harm. How- ever, it is when each party can cause harm to the others that we have a classic commons problem. In a commons, individuals typically gain much more from their use of the resource than they suffer from the degradation their use causes; thus one can increase one’s own well-being by overconsuming and harming the other users. Furthermore, restricting one’s own use does not ensure protection against the harms caused by others’ use of the resource. In these ways, a common resource establishes a moral community. To protect the resource and to protect them- selves, the parties must grant each other the right to a fair share, and accept enforcement of a mutually agreed limit. I argue that the fundamental principle of fairness in the governance of a commons is equality in decision-making and use, and in particular equality among people, not countries. This cannot be simply asserted or deduced, but rather that must be established through moral reasoning. By drawing on an extended analogy to a hypothetical common resource—in this case, a shared aquifer—and by rebutting common critiques, I will show how the principle of equal rights to common resources can be credibly justified. Imagine two people—let’s call them Nora and Sam—who share an island. Each of them has a well that pumps water from a shared aquifer. Nora discov- ers how to make a pump that pumps three times as fast, and is able to irrigate more farmland; soon she has a grain surplus, is feeding cattle, and is clearly healthier. Sam meanwhile is able to irrigate a much smaller plot and to feed only a few chickens, and is regularly falling ill. Eventually, however, he discovers how to make his own pump that is as powerful as hers. Just before he installs his pump, they both find out that the level of the aquifer is starting to fall. Each is aware that the other is using the aquifer and at what rate. They get together and figure out how large the aquifer is and what its Chapter 15. Equity, Greenhouse Gas Emissions, and Global Resources annual recharge rate must be. They know how much each of them has already pumped, and how long it will take to exhaust the remaining stock at the rate they will both soon be able to pump. They know also that if they are forced to immediately reduce to the recharge rate, it will seriously limit their food sup- plies. They are now forced to decide, whether individually or collectively, how fast to pump the water. Assuming that they decide that they can and must trust one another, what might we expect them to decide is a fair agreement, and why? It seems likely that they would agree to share the aquifer equally unless one was willing to compensate the other. Nora does not have a good argument to make why Sam should continue to use less and she more; now that they each have a big pump, why shouldn’t he be able to use his? Should he agree to remain permanently poorer? It would not make sense for Sam to agree to forever use a smaller share simply because he was using less at the time when the agreement was made. annual recharge rate must be. They know how much each of them has already pumped, and how long it will take to exhaust the remaining stock at the rate they will both soon be able to pump. They know also that if they are forced to immediately reduce to the recharge rate, it will seriously limit their food sup- plies. They are now forced to decide, whether individually or collectively, how fast to pump the water. Assuming that they decide that they can and must trust one another, what might we expect them to decide is a fair agreement, and why? It seems likely that they would agree to share the aquifer equally unless one was willing to compensate the other. Nora does not have a good argument to make why Sam should continue to use less and she more; now that they each have a big pump, why shouldn’t he be able to use his? Should he agree to remain permanently poorer? It would not make sense for Sam to agree to forever use a smaller share simply because he was using less at the time when the agreement was made. On the contrary, Sam might point to the wealth Nora accumulated while she was living on an unsustainable share of the water and say that it is not fair for them now to use an exactly equal share. There was a fixed amount of water in the aquifer when they started, and it can only produce a finite amount of wealth before they are both required to learn to live off the sustainable flow; to divide only the remaining part of the aquifer equally would leave him perpetually poorer. Yes, Nora did not know that the level of water that she was pumping was unsustainable, but Sam did not agree to let her become wealthy at his expense. And indeed her wealth is at his expense; what she used, he cannot. He can make a good case that it is fair for him now to use more, or for her to compensate him for using less. Some might recognize this situation as a version of the prisoner’s dilemma and note that there is a noncooperative solution that is equally plausible. Nora or Sam might decide to pump as fast as possible, knowing that their use was unsustainable, but the other probably would do the same, leaving them both worse off. What matters here is that we see the situation of interdependence as necessarily creating a moral community: Each party can harm or be harmed by the other, and depends on the other’s cooperation.9 This, then, is the structure of a common resource: Even if we would like to get more than our fair share of the benefits, we know that it is not ethical for us to do so. Furthermore, absent any other compelling justifications, a fair share is a equal share. What might constitute a justification for an unequal division of the aquifer? If it rains more on someone’s part of the island, we might think it fair for him or her to accept a less than equal share of the aquifer. However, it is important to realize that such an argument for inequality in access to a particular resource is based on an appeal to equal opportunity more generally: No one should be PART V. DEVELOPMENT AND EQUITY better or worse off than anyone else simply because of which part of the island he or she happens to live on. In the real world, this principle is not given much weight since (for example) countries with large fossil fuel resources are not expected to give a free share to less fortunate countries. One might argue that fossil fuel reserves should be shared as common global resources, but the point here is that because in our hypothetical example the water lies underneath (and is equally accessible to) all parties, there is no way for one individual to physi- cally exclude another from using it, and thus to charge for its use. In this hypothetical example, I have placed the question of the allocation of the common resource into a very abstract context, as if it were the only resource in question and the only issue of negotiation between the two parties. It is in part through this abstraction that the principle of equality emerges so strongly; there is no possible gain to either party from accepting a less than equal share. However, the real world is much more complicated. For example, it could well be argued that, when there are a large number of unused com- mon resources, each party would accept a principle of “first come, first served.”10 Allowing one party the right to claim a larger share of certain com- mon resources, in exchange for allowing other parties a similar right to other resources, might be agreed to make everyone better off because it encourages innovation and investment in the development of those resources. This is a major justification for allowing homesteading or the establishment of mining claims or water rights. However, this condition clearly does not hold in the case of the atmosphere. There have never been any negotiations between all the countries of the world, to say nothing of all the people, concerning general principles of allocation of global common resources. Not, that is, until today. The underusing countries have not agreed to allow the North’s overconsumption.An Ethical Analysis of Allocation Principles for Emissions Rights This leads us back to the problem with which we started: the need for an inter- national agreement to regulate GHG emissions and the controversy over what an acceptable allocation of rights would be. There is an extensive literature on this part of the equity debate, to which I cannot begin to do justice.11 I focus on a relatively narrow but crucial aspect: whether the ethical arguments for various allocation principles are convincing. I address the fairness of various principles rather than the likelihood of their being accepted or the ease of implementing them, not because I believe that it is better to be morally righteous than to be Chapter 15. Equity, Greenhouse Gas Emissions, and Global Resources practical but because what a government and its citizens believe is fair is one jus- tification of country’s negotiating positions. Especially because of the large role of the United States in the climate negotiations, this is not just an academic mat- ter; the sense of fairness that eventually becomes dominant in the United States may have a significant influence on the future of the negotiations. I assume in the following discussion that tradeable emissions permits are a plausible and desirable scheme for addressing climate change. There are reason- able arguments against tradeable permits, such as the practical difficulties of implementation and the possible negative impacts of market power (either buy- ers’ or sellers’). Nonetheless, many analysts have concluded that the power of such a scheme to separate efficiency (making the most cost-effective reductions) from equity (determining who will pay for those reductions) makes it the best option for an international agreement.12 Furthermore, the Kyoto Protocol itself explicitly includes mechanisms for emissions trading. Thus, in the remainder of this section I will assume that however emissions rights are allocated, they may subsequently be traded.Why Emissions Rights Can’t Be Equal by Country I will begin by examining two principles that are seldom explicitly advocated but underlie many of the arguments against other principles (such as equal per capita rights). The first of these is the principle that every country should have a right to an equal share of the atmosphere. It simply isn’t ethically plausible that the rights to use a common resource would be attributed in equal shares to every country. The benefits of the use of the resource fundamentally accrue to people; the allocation of emissions rights to countries is a pragmatic compromise. No one would argue that Fiji should have the same emissions rights as the United States. I bring up this seemingly obvious point only because opponents of the Kyoto Protocol often argue that a reason for the United States to oppose Kyoto is that because the protocol restricts U.S. emissions but not China’s, China will soon emit more than the United States. It is reasonable for the United States to be concerned that if China never accepts limits, U.S. emissions reductions will not prevent climate change. However, the Kyoto Protocol only addresses the period through 2012. One cannot claim that it’s wrong if China some day emits more than the United States without a real argument about the basis for emissions rights, and the United States has a very weak argument. After all, China has more than four times the U.S. population; it must be acceptable for them to emit some amount more than we do. PART V. DEVELOPMENT AND EQUITYWhy Emissions Rights Can’t Be Grandfathered The second principle of resource allocation I will consider is grandfathering, the principle that a party’s current level of use establishes a firm property right. Under this principle, if we need to establish a limit to use, a country is entitled to the same proportion of the limited resource that it had been using when use was unrestricted. It is rare for anyone to make an ethical argument for pure grandfathering in the case of climate; freezing the relative emission rates of different countries at their current proportions can plainly be seen to be unfair to the low-emitting countries. Imagine being born in a poor country in the year 2050 and finding that you are allocated fewer permits than people in much wealthier countries simply because your country had been poor in 2005. You might very reasonably conclude that this was not a fair situation and might reconsider whether your nation should continue to abide by the agreement. There is a weaker form of the argument for grandfathering that is more plausible, and that implicitly underlies a large number of proposed mixed or transitional allocation schemes: Because the high-emitting countries did not know that they were overusing a commons, it would be unfair to ask them to immediately restrict their use to a fair, sustainable share. However, this argu- ment confuses two different points. The first is whether it would constitute an undue hardship on the high emitters to restrict their emissions sharply and rap- idly (or to pay for their excess consumption). The second is whether the high emitters are entitled to the benefits of their current overconsumption. There are numerous precedents for allowing parties to stretch the repayment of their debts over time. But the legitimacy of the debt isn’t determined by the harm that is caused by repaying it, and it is usually assumed to be up to the party who is owed to determine whether and how much to reschedule or reduce the debt. Thus this is at best an argument for temporary grandfathering as part of a tran- sition.Why Emissions Rights Can’t Be Proportional To GDP Others have argued that permits should be allocated at least in part proportion- ally to gross domestic product (GDP);13 the greater a country’s GDP, the greater its emission rights. This gives additional permits to the wealthier nations (attrac- tive for getting them to buy in but not in itself an ethical argument), and it cre- ates incentives to use one’s allocation as efficiently (in the sense of reducing emissions per GDP) as possible. However, this principle has some unacceptable effects if carried to its conclusion; the wealthiest countries would always have the Chapter 15. Equity, Greenhouse Gas Emissions, and Global Resources largest share of the permits, which, given their value, would have the effect of increasing inequality. Again, imagine a resident of a poor country some years from now, who may not burn as much coal as a resident of a rich country pre- cisely because he is not as rich; it is hard to see how he or she would consider this to be fair.Why Emissions Rights Should Be Per Capita The central argument for equal per capita rights is that the atmosphere is a global commons, whose use and preservation are essential to human well being. Therefore, as I argued using the aquifer example, all people should hold both decision-making rights and use rights equally unless there is a compelling higher principle. We might be able to determine what would count as a higher principle by considering the implications of the reductio ad absurdum of equal per capita rights.14 No single reduction can capture all possible failures of an ethical prin- ciple, and there are several that might be interesting and relevant in this case. One possible reduction is that emissions permits are allocated immediately on a strict per capita basis and are not tradeable; this would clearly cause a harmful economic shock to the countries that had to make sharp reductions. This might well be judged unacceptable on utilitarian grounds if it caused more harm to those who were forced to reduce than it brought benefit to those who were not or if it actually harmed those it was meant to help due to global economic inter- dependence. However, what we actually seem to care about here is outcomes, not princi- ples of allocation; if an unequal allocation could be shown to permanently ben- efit those who receive lower allocations, few would argue that we should insist on strict equality. However, other than suggesting that developing countries might suffer if the North underwent economic contraction, no one has ever argued that poor countries actually would benefit from having lower emissions allocations than rich countries, especially not permanently. Another possible reduction is that a global energy administration would actu- ally issue a GHG emissions permit to every person on the planet and require them all to buy and sell them in a single enormous global market. This boggles the mind because of its impracticality, not its ethical failure. If the reason for an equal right is because each person is truly entitled to an equal share of the ben- efits, there are only practical reasons, not logical or ethical ones, for the permits to be issued to countries. Similarly, the idea that each person on the globe might vote on the total amount of emissions to be allowed seems absurd, but again for practical rather than ethical reasons. PART V. DEVELOPMENT AND EQUITYBeyond Equal Annual Allocations: The Principle of Historical Accountability Many analysts have also extended the principle of equal per capita rights to the principle of historical accountability. At the individual level, historical account- ability could mean that each individual gets the same amount in their lifetime, regardless of when they are born; those who have already used more than their allowable share would have to purchase permits from those who haven’t. Return- ing to the aquifer example, this is precisely the argument Sam has for why he should get a larger than equal share of the remaining water. In practice, this would mean that a country’s current allowed emissions are reduced if it has cumulatively overused the commons. There are many possible formulas for quantifying overuse and using it to modify current allocations;15 however, the essential point is that countries are assumed to have benefited per- manently (as by increased wealth and infrastructure) from that overuse and to have a debt to repay. There are some plausible ethical objections to historical accountability, such as the dubiousness of holding living persons responsible for the activities of their ancestors or the fact it hasn’t been known for long that overuse was causing a problem.16 Also, not everyone in wealthy countries has contributed equally to or benefited equally from their cumulative emissions. However, the correlation between the cumulative emissions of countries and their levels of overall wealth is clear, and the fact that wealth is unequally dis- tributed within countries does not seem to justify ignoring the common bene- fits that have accrued. In a hypothetical case, if we could identify the precise contribution that over- use of the commons had made to an individual’s current wealth, it would be rea- sonable to consider that benefit to be an individual debt to those who will be unable to obtain a similar benefit. It seems reasonable that a country that has cumulatively but unequally overused the commons should be responsible for fairly distributing the debt among its citizens.Practical Versus Ethical Objections to Equal Per Capita Rights This leads to a more general consideration of the relationship between practical and ethical objections to equal per capita allocations. The three most common objections to equal per capita rights are that it provides an incentive to popula- tion growth, that poor people who would have a surplus of allocations would not benefit from their sale, and that the North (and the United States in partic- ular) would never accept the financial burden. I will address each of these in turn and show that although they may indeed have some practical relevance, they do Chapter 15. Equity, Greenhouse Gas Emissions, and Global Resources not in themselves constitute arguments that per capita allocations are not ethi- cally justified. Because governments would get permits to use or sell based on the size of their populations, opponents of equal per capita rights argue that it would give governments an incentive to increase, or at least not to limit, population growth.17 However, this concerns the practical effect of a per capita allocation principle; it is not an ethical argument that people should not intrinsically have equal rights to the commons. This is not to say that practical arguments do not have ethical implications; again, if an equal per capita allocation led to substan- tial additional population growth that caused identifiable harm, we might there- fore reject per capita rights because of the consequences. But because there are a variety of plausible solutions (one simple example being fixing the allocation to a base-year population), this argument does not carry much weight. Another argument that has been made against equal per capita rights is that the resulting financial transfers would not aid the people they are supposed to help. Because the permits would be traded by governments, there is no guaran- tee that the poorest people who should be the owners of surplus permits would see much of the benefits from their sale. However, this again is a pragmatic, not ethical argument; the fact that there is not currently a channel for permit pur- chasers to pay the rightful owners of the resource does not mean they are not ethically obligated to do so. They certainly may not simply keep the money. For better or worse, we generally accept national sovereignty as a basis for determining the internal allocation of resources; we do not judge the democratic nature of the Saudi royal family before we pay for the oil we import. Nor has anyone suggested that the United States did not deserve a large allocation because the benefits of emissions are unequally distributed domestically. More- over, if we think that individuals should receive the benefits of the use (or sale) of their permits, we can help empower them to make that demand effectively by giving international recognition to the principle of equal rights.Cost as an Objection to Per Capita Allocations Finally, it is necessary to discuss what is usually given as the ultimate argument against equal per capita emission rights: that the industrialized countries, and the United States in particular, would never agree because of the high costs they would incur. In a tradeable permit system with a cap at today’s global level, an immediate transition to an equal per capita allocation would result in trading of roughly 2 billion tons of carbon permits each year. Recent economic studies have estimated that for a cap based on the Kyoto framework, permits might trade in a global market at $20–$100 per ton,18 but estimates of up to $200/ton PART V. DEVELOPMENT AND EQUITY or more have been made for more restrictive global caps. The United States alone exceeds its equal per capita share by more than a billion tons and thus could be required to purchase permits worth tens or even hundreds of billions of dollars. Thomas Schelling, among others, has argued that because poor people in developing countries are the most likely to suffer from climate change, money spent by wealthy countries to prevent it is a form of foreign aid.19 He also argues that it is unreasonable to expect the United States to pay so much more in this case than we currently do for other forms of foreign aid. However, if one accepts that there should be equal rights to global common resources, any costs associ- ated with tradeable permits are payment for the use of resources, not foreign aid. The fact that the United States might not like or agree to such costs is not an ethical argument; I may not like the high price of oil, but that doesn’t mean I can steal it.20 It is possible to argue that the harm that would come from paying a fair price for emissions rights is greater than the benefit that would come to those who sell the permits. There is some evidence that people feel that the loss of a given amount of income causes greater harm than the gain of the same amount of income causes benefit. One way to look at this is to consider that people have expectations built around their material lives and that even wealthy people suf- fer significantly when their expectations fail to be met.21 However, in the case of global emissions trading, this is not a very plausible argument. The standard analysis of transfers between rich and poor, based on the declining marginal util- ity of income, is that the gain of $10 to a poor person means more than the loss of $10 to a rich person; the huge disparities of wealth between North and South suggest that the marginal utility of income in the South must be much higher.22 If rights to the global commons should be shared equally and paying for those rights would not cause more overall harm than good, there is little remain- ing justification for the North to refuse to agree to such payments. A country does not have the ethical right to opt out of the governance of a commons—to be a free rider—simply because it doesn’t want to reduce its overconsumption. Because one country’s use affects all the others, the moral community and moral obligations exist whether they are respected or not.Conclusions As I suggested at the beginning of this chapter, I consider myself to be not merely an analyst but also a participant in the process of defining equity in the climate change debate. Because the economic stakes are quite high, many par- ties are actively engaged in this process. It is my central claim that self-interest Chapter 15. Equity, Greenhouse Gas Emissions, and Global Resources and ethical justification are not the same and that one can and must use rea- soned argument to determine what is right. If the arguments for equal rights are justified, it follows that the U.S. government should change its negotiating posi- tion and agree to a treaty that establishes at least an eventual goal of equal per capita allocations. With a commitment to equal per capita allocations, a global emissions cap covering developing and developed countries becomes possible, with enormous associated advantages. Such an agreement would create a large and (hopefully) efficient market for permits and thus bring down the cost of compliance world- wide. It would eliminate the need to establish baselines that dogs the Clean Development Mechanism and other project-based mitigation schemes. Per capita entitlements would eliminate the incentive for developing countries to delay reductions in emissions in order to increase their claim to atmospheric space. In all these ways, a transition to an agreement based on equal per capita rights would help us to stabilize atmospheric GHG concentration at lower lev- els and to limit the risks of dangerous climate change. I do not presume to have addressed all the ethical questions concerning the equitable allocation of emissions rights. At the very least I hope I have made clear what a justification for a principle of equity must look like to be an ethi- cal rather than practical (or selfish) argument. Finally, I hope that I have demon- strated that it is both possible and necessary for us to take part in the creation of new norms of international equity, and that the climate change debate offers us an opportunity to make an important contribution to a more just and sus- tainable world.Acknowledgments I would particularly like to thank Tom Athanasiou, Barbara Haya, Francesca Flynn, Richard Norgaard, and Paul Wilson for their thoughtful discussions and extensive comments, as well as three anonymous reviewers. All remaining philo- sophical and factual errors are my own.Notes 1. In the form of carbon dioxide from fossil fuel burning and industrial activities; mul- tiply by 44/12 to get the equivalent mass in CO2. 2 2. World Bank, 2000: World Development Report 2000/2001: Attacking Poverty (Oxford: Oxford University Press). y 3. This reasoning is embedded in the UN Framework Convention on Climate Change PART V. DEVELOPMENT AND EQUITY itself, in the Berlin Mandate passed at the first Conference of Parties in 1995, and in the Kyoto Protocol. 4. For an excellent review of issues of intergenerational equity, particularly as it is con- nected with the theory of discounting in economics, see Arrow, K. J., W. R. Cline, K.-G. Mäler, M. Munasinghe, R. Squitieri, and J. E. Stiglitz, 1996: “Intertemporal equity, discounting, and economic efficiency,” in J. P. Bruce, H. Lee, and E. F. Haites (eds.), Climate Change 1995: Economic and Social Dimensions of Climate Change (Cambridge, England: Cambridge University Press), 125–144. For a dis- cussion of the question of liability for climate damages and other equity concerns, see Banuri, T., K. Göran-Mäler, M. Grubb, H. K. Jacobson, and F. Yamin, 1996: “Equity and social considerations,” in the same book. See also Grubb, M. J., 1995: “Seeking fair weather: ethics and the international debate on climate change,” Inter- national Affairs, 71: 463–496. ff 5. See Schelling, T. C., 1997: “The cost of combating global warming: facing the tradeoffs,” Foreign Affairs, 76: 8–14; Rayner, S., E. L. Malone, and M. Thompson, 1999: “Equity issues and integrated assessment,” in F. Tóth (ed.), Fair Weather? Equity Concerns in Climate Change (London: Earthscan Publications Ltd.), or Vic- tor, D., 1999: “The Regulation of Greenhouse Gases: Does Fairness Matter?” in the same book. 6. Important articles using this framework include Ghosh, P., 1993: “Structuring the equity issue in climate change,” in A. N. Achanta (ed.), The Climate Change Agenda: An Indian Perspective (New Delhi: Tata Energy Research Institute). Shue, H., 1993: “Subsistence emissions and luxury emissions,” Law and Policy, 15 (1): 39–59; Bhaskar, V., 1995: “Distributive justice and the control of global warming,” in V. Bhaskar and A. Glyn (eds.), The North the South and the Environment: Ecological Constraints and the Global Economy (New York: St. Martin’s Press). Paterson, M., 1996: “International justice and global warming,” in B. Holden (ed.), The Ethical Dimensions of Global Change (New York: St. Martin’s Press). 7. I am indebted to Sunita Narain for clarifying this distinction. 8. More than a few articles have “burden sharing” in their titles: e.g., Grubb, M., J. Sebenius, A. Magalhaes, and S. Subak, 1992: “Sharing the burden,” in I. M. Mintzer (ed.), Confronting Climate Change: Risks, Implications and Responses (Cam- bridge, England: Cambridge University Press); Parson, E. A. and R. J. Zeckhauser, 1995: “Equal measures or fair burdens: negotiating environmental treaties in an unequal world,” in H. Lee (ed.), Shaping National Responses to Climate Change (Washington, DC: Island Press). Many more articles use it as their conceptual framework, and numerous workshops have been held on the subject. 9. The fact that it may be possible to cheat and that there is incentive to do so may lead to the need for an enforcement mechanism; it is precisely because the situation creates a moral community that each party may grant that it is fair for the other side to enforce a penalty on them if they are caught. 10. I thank Dick Norgaard for this insight. 11. Comprehensive bibliographies can be found in the IPCC’s Second and Third Assessment Reports: Banuri et al., 1996; Banuri, T. and J. Weyant, 2001: “Setting the stage: climate change and sustainable development,” Chapter 1 in B. Metz and Chapter 15. Equity, Greenhouse Gas Emissions, and Global Resources O. Davidson (eds.), Climate Change, 2001: Mitigation (Cambridge, England: Cambridge University Press). Important early work that I haven’t cited elsewhere includes Krause, F., W. Bach, and J. Koomey, 1989: Energy Policy in the Greenhouse, Volume I (El Cerrito, CA: International Project for Sustainable Energy Paths); Rose, A., 1990: “Reducing conflict in global warming policy: the potential of equity as a unifying principle,” Energy Policy, December 1990: 927–948; Agarwal, A. and S. Narain, 1991: Global Warming in an Unequal World: A Case of Environmental Colonialism (New Delhi: Centre for Science and Environment); Solomon, B. D. and D. R. Ahuja, 1991: “International reductions in greenhouse-gas emissions: An equitable and efficient approach,” Global Environmental Change, 1 (5): 343–350; and Bertram, G., 1992: “Tradable emission permits and the control of greenhouse gases,” Journal of Development Studies, 28 (3): 423–446. More recent discussions include Meyer, A., 1999: “The Kyoto Protocol and the emergence of ‘contraction and convergence’ as a framework for an international political solution to green- house gas emissions abatement,” in O. Hohmayer and K. Rennings (eds.), Man- Made Climate Change: Economic Aspects and Policy Options (Heidelberg: Phys- ica–Verlag); Baer, P., J. Harte, B. Haya, A. V. Herzog, J. Holdren, N. E. Hultman, D. M. Kammen, R. B. Norgaard, and L. Raymond, 2000: “Equity and greenhouse gas responsibility in climate policy,” Science, 289: 2287; and an entire edited col- lection: Tóth, Fair Weather? 1999. 12. Some authors explicitly argue that tradable permits are more desirable than a car- bon tax scheme, which in economic theory can also separate efficiency from equity. Epstein, J. M. and R. Gupta, 1990: Controlling the Greenhouse Effect: Five Global Regimes Compared (Washington, DC: The Brookings Institute). Many other authors assume the benefits of trading with little analysis. 13. For example, Sagar includes GDP in a hybrid proposal with population and histor- ical emissions. Sagar, A., 2000: “Wealth, responsibility and equity: exploring an allocation framework for global GHG emissions,” Climatic Change, 45: 511–527. 14. Reductio ad absurdum means roughly what it sounds like: “reduction to the point of absurdity.” In ethical reasoning this means applying a moral principle in an extreme form to a hypothetical situation, typically to show that in cases related to the real example, there are higher principles that take precedence. p g p p p 15. See Grübler, A. and Y. Fujii, 1991: “Intergenerational and spatial equity issues of carbon accounts,” Energy, 16 (11/12): 1297–1416; Smith, K., 1991: “Allocating responsibility for global warming: the natural debt index,” Ambio, 20 (2): 95–96; Hayes, P. and K. Smith (eds.), 1993: The Global Greenhouse Regime: Who Pays (Lon- don: Earthscan Publications Ltd.); Smith, K., 1996: “The natural debt: North and South,” in T. W. Giambelluca and A. Henderson-Sellers (eds.), Climate Change: Developing Southern Hemisphere Perspectives (New York: John Wiley and Sons Ltd.). 16. For an articulation of the case against historical accountability, see Beckerman, W. and J. Pasek, 1995: “The equitable international allocation of tradable carbon per- mits,” Global Environmental Change, 5: 405–413; for a response see Neumayer, E., 2000: “In defence of historical accountability for greenhouse gas emissions,” Eco- logical Economics, 33 (2): 185–192. g 17. In fact, this is raised by defenders of per capita rights more often than by opponents 17. In fact, this is raised by defenders of per capita rights more often than by opponents PART V. DEVELOPMENT AND EQUITY because it is so easy to defeat; several solutions have been discussed in all the litera- ture since Michael Grubb’s seminal work. Grubb, M., 1989: The Greenhouse Effect: Negotiating Targets (London: Royal Institute of International Affairs). 18. Weyant, J. P. and J. H. Hill, 1999: “Introduction and overview,” The Energy Jour- nal, Special Kyoto Issue: vii–xiv. 19. Schelling, 1997. 20. David Victor has made a similar argument that a difficulty of adopting a tradable permit scheme is that wealthy countries will be reluctant to allocate assets worth hundreds of billions of dollars. Victor, D. G., 2000: “Controlling emissions of greenhouse gases,” in D. Kennedy and J. A. Riggs (eds.), U.S. Policy and the Global Environment: Memos to the President (Washington, DC: The Aspen Institute). Oth- ers have pointed out that if the assets indeed have this value, the overconsuming countries are free-riding and increasing their unfair appropriation by delaying the transition to equal per capita rights. Parikh, J. and P. Parikh, 1998: “Free ride through delay: risk and accountability for climate change,” Environment and Devel- opment Economics, 3: 384–389. 21. For an extensive discussion of this argument, see Wesley, E. and F. Peterson, 1999: “The ethics of burden sharing in the global greenhouse,” Journal of Agricultural and Environmental Ethics, 11: 167–196. 22. In fact, if one uses a standard assumption about the declining marginal utility of income (i.e., that utility increases as the log of income), it can be shown that in a system of tradable permits, equal per capita rights are utility maximizing compared with grandfathering or any mixture of the two. Baer, P. and P. Templet, 2001: “GLEAM: a simple model for the analysis of equity in policies to regulate green- house gas emissions through tradable permits,” in M. Munasinghe, O. Sunkel, and C. de Miguel (eds.), The Sustainability of Long Term Growth (Cheltanham, England: Edward Elgar Publishing Company).Energy 237 (2021) 121611 Contents lists available at ScienceDirectEnergyjournal homepage: www.elsevier.com/locate/energyAssessing the advancement of new renewable energy sources in Latin American and Caribbean countriesNuno Silva a, Jose Alberto Fuinhas a, *, Matheus Koengkan b, c a CeBER, and Faculty of Economics, University of Coimbra, Coimbra, Portugal b Faculty of Economics, University of Coimbra, Coimbra, Portugal c DEGEIT, University of Aveiro, Aveiro, Portugala b s t r a c ta r t i c l e i n f o Article history: Received 5 February 2021 Received in revised form 13 June 2021 Accepted 23 July 2021 Available online 2 August 2021 The research focuses on the evolution of non-hydroelectrical renewable installed capacity from 2001 to 2017 in nineteen countries from the Latin America and the Caribbean (LAC) region. The deployment of new renewables has several characteristics that point to nonlinear relationships with its drivers. In accordance, the quantile regressions econometric technique of Canay was used. The option to use this econometric technique stem from Canay model estimation can capture the variation between different countries (cross-section). The Canay model estimation results support that renewable energy finance flows, economic globalisation index, and carbon dioxide emissions positively affect non-hydroelectrical renewable installed capacity. The positive impact of carbon dioxide (CO2) emissions on non- hydroelectrical renewable installed capacity could be related to increased environmental, political, and social pressure regarding the increase of environmental degradation caused by the growth in emissions that encouraged the new investments in renewable energy technologies. It also could be related to the consumption of new renewable energy sources that mitigate CO2 emissions in the LAC countries. © 2021 Elsevier Ltd. All rights reserved. Keywords: New renewable energy sources Quantile regression Latin American and Caribbean countries Financial development © 2021 Elsevier Ltd. All rights reserved.1. Introduction arise from the vast transformations underway. Nevertheless, gov- ernments have shown a limited capacity to finance the energy transition. The private sector's participation in the rise of new re- newables put new renewables' competitiveness at the centre of take-off. This study focuses on non-hydroelectrical renewable installed capacity in Latin American and Caribbean (LAC) countries. The take-off of new renewables, wind, solar, and bioenergy, puts chal- lenging questions to handle without further research. Traditional econometric approaches are unfeasible in the presence of large variable values that result from the early stages of a growth take- off. In fact, the growth values seem enormous, but only because the starting point is shallow, even without any expression. Why is the research of LAC countries necessary for the literature and practitioners? The LAC countries are mostly economies in development. Economies in development are the predominant kind of economies on Earth. LAC countries lessons could be valu- able for countries initiating the first stages of the take-off of new renewables. Indeed, the same processes have a high probability of occurring in the same way for new energy sources in the future. Why is it important to research the emergence of new renew- able energy? Renewable energy sources tend to be diversified and stand at different levels of technological maturity. Given the dimension of the energy's transition process from fossils to re- newables requires the involvement of markets as well as policy- maker guidance. Governments have been an active partner and are involved in mitigating several economic and societal problems that The research question of this study is as follows. What is the impact of the factors behind the high growth of new renewables in LAC countries? Given that the research focuses on analysing the new renewables' growth, the investigation excludes hydroelectrical power. Indeed, hydro should be excluded because most of its ca- pacity was installed in the last century before our new renewable energy dataset starts. Why can the LAC countries contribute to literature? LAC coun- tries share several common factors that turn them into a unique place in the World. Nevertheless, the LAC region shares the same * Corresponding author. Faculty of Economics, University of Coimbra, Av. Dias da Silva 165, 3004-512, Coimbra, Portugal. E-mail addresses: nunos@fe.uc.pt (N. Silva), fuinhas@uc.pt (J.A. Fuinhas), matheuskoengkan@ua.pt (M. Koengkan). https://doi.org/10.1016/j.energy.2021.121611 0360-5442/© 2021 Elsevier Ltd. All rights reserved. * Corresponding author. Faculty of Economics, University of Coimbra, Av. Dias da Silva 165, 3004-512, Coimbra, Portugal. E-mail addresses: nunos@fe.uc.pt (N. Silva), fuinhas@uc.pt (J.A. Fuinhas), matheuskoengkan@ua.pt (M. Koengkan). https://doi.org/10.1016/j.energy.2021.121611 0360-5442/© 2021 Elsevier Ltd. All rights reserved. N. Silva, J.A. Fuinhas and M. Koengkan Energy 237 (2021) 121611 problems that are frequent in developing countries. Indeed, LAC countries choice is based on the fact that they have exceptional natural conditions to produce energy from renewable sources. They have been exposed to massive growth in non-hydroelectrical installed capacity since the early years of this century. Fortu- nately, data on the evolution of renewable installed capacity and its drivers are available for 19 countries allowing a representative cover of the LAC region. “nexus” and later in the form of an “extended nexus” [2]. This long saga has covered all imaginable energy links, economic growth, and a panoply of variables. Despite some stylised facts, many aspects still require further empirical research. Indeed, among the less well understood empirical behaviours is promoting an effective increase in renewable energy, specifically wind, solar, and bioenergy. The deployment of renewables and, in particular, the new re- newables in LAC countries were a multidimensional phenomenon that includes several players and the development and imple- mentation of public policies. The deployment of the new renew- ables has been limited to a non-long time span. Nevertheless, it is visible that the evolution of how it was approached and inflexions can be observable. The deployment of renewable energy (RE) is also linked to political decisions to cope with climate change. Curbing carbon dioxide (CO2) emissions becomes part of the international political agenda and a source of concern for a growing number of citizens. Political decisions are not subject to market behaviour but cannot ignore it, adding more complexity to the RE deployment phenomenon. What is the novelty of our research? Did it fill a gap in the literature? Our approach was developed to handle the initial stages of non-hydroelectrical renewable installed capacity take-off in LAC countries. To measure new renewable energy sources' growth was used the yearly change of Gigawatts (GW) of non-hydroelectrical installed renewable capacity, and this measure was used as our explained variable. Indeed, the coherence of evolution among the new renewables is necessary to achieve an in-depth vision of the drivers of the evolution of the yearly change. The option to measure the evolution of installed capacity through the yearly change in GW, instead of the most used in literature, the growth rate, is condi- tional in it may provide more accurate insights about the causal impact of factors that promote renewable energy sources. The use of growth rates has unwanted side effects resulting from high growth rates in the first years of the sample because the initial installed capacity was relatively low in most LAC countries. We also fill the literature gap about non-hydroelectrical renewable installed capacity drivers in its first stages in LAC countries. The econometric analysis of the initial phase of the development of new renewables is essentially an absence in the literature that needs to be fulfilled. The panorama of promoting RE in Latin American (LA) countries, particularly solar panels, and the establishment of mini-grids, re- quires policymakers to pay attention to areas that do not achieve economies of scale that turn them profitable [3]. Banal-Esta~nol et al. [3] pinpoint that LA countries also have adopted universal service policies to allow low-income households to connect to grids. They also stress the value of establishing strong links between stake- holders to care for the launch of new technologies. The promotion of renewables in LA countries also has was based on measures that include tax incentives. These tax incentives workout mostly through exemptions in income tax and through sales tax/value-added taxes or via tariffs [4]. Washburn and Pablo- Romero [4] pinpoint that most LA countries are using active policies to promote RE and that the intensity of promoting RE is related to their RE adoption performance. They also find that they switch from a system based on feed-in tariffs to use the auction system. y What is under study? In short, the research focuses on the evolution of non-hydroelectrical renewable installed capacity from 2001 to 2017 in 19 countries from the LAC region. The deployment of new renewables has several characteristics that point to nonlinear relationships with its drivers. Indeed, by one hand, it is expected that any process of take-off will behave in a nonlinear way with its determinants. On the other hand, new renewables were under fast technological evolution and achieving maturity or opening a new phase, for example, as energy sources like photo- voltaic become market competitive with fossil sources. Techno- logical innovations have been a significant source of change in the process toward achieving energy efficiency. Technological innova- tion has also led to economic structural changes, mainly in the production's process that influences the renewable energy market competitiveness. As the political option was more often than not to proceed with renewable energy deployment in a market-oriented environment rather than through a public incumbent, renewable energy sources' competitiveness plays a significant role. This role evolves with the deployment of new renewables themselves in a possible nonlinear way. In accordance, the study makes use of the quantile regressions econometric technique. This technique is prepared to handle the heterogeneity that is present at different quantiles of the distribution. One of the advantages of quantile estimation is that it does not require that the variables follow a normal distribution. Kruckenberg [5] pinpoint that international programmes directed to facilitate the adoption of RE technologies among less developed countries have achieved a mixed level of success. The author identifies partnerships as the dominant RE technologies programme. These programmes involve local, national, and inter- national organisations. Nevertheless, Kruckenberg [5] stress that the attention should be put on key actors' relationships more than on barriers and drivers to develop RE. Kruckenberg [6] stress that North-South partnerships for sus- tainable energy face numerous knowledge challenges in imple- menting off-grid RE technologies. The author concludes that the exchange of knowledgeepower in relationships involving sustain- able energy is vital in establishing off-grid RE technologies. Kim and Park [7] analysed the Clean Development Mechanism, created by the Kyoto Protocol to promote low-carbon development, in a panel of 64 host countries from 2001 to 2014. They found that the adoption of RE is conditional on the level of financial devel- opment of countries. More precisely, they found that the Clean Development Mechanism's effect in implementing RE is much more forceful in countries with undeveloped financial markets. The authors also conclude that promoting that kind of RE projects is beneficial to countries with underdeveloped financial markets. Renewable energy sources in developing countries, with unso- phisticated financial markets, face, as a rule, scarcity of financing through debt and equity [7]. Moreover, Zeng et al. [8] complement that in developing countries such as Brazil, Russia, India, China, and South Africa, the industry funds play an important role in renew- able energy projects, where in the early stage of renewable energy development, their financing models still need to mature. In developing countries, most renewable energy projects raise funds The article is organised as follows. Section 2 analyses the liter- ature. Section 3 reveal the data and method used in the research. Section 4 show empirical results. Section 5 discusses the empirical results. Section 6 concludes the research and present policy recommendations.2. Literature review Since the note On the Relationship Between Energy and GNP [1], the search for the link between energy consumption and economic growth has been an active topic in the field of energy economics. This research has evolved first mostly under the designation of N. Silva, J.A. Fuinhas and M. Koengkan Energy 237 (2021) 121611 from debt; these funds are provided by banks and related in- stitutions, which offer developers very low-interest loans. The use of RE technologies is severely constrained in countries with limited access to external financing [9]. of wind and solar sources of energy. They conclude for higher variability for wind than solar power generation. In this context, Brazil is well-positioned to play a significant role in renewable energy integration in LA countries. CO2 emissions growth has been connected consistently with evolution through the time of several variables. Among these de- terminants of CO2 emissions, the literature identifies some drivers, such as energy consumption, globalisation, urbanisation, trade openness (e.g. Refs. [10e13], or even more unexpected like obesity and overweight [14] or renewable energy consumption and out- door air pollution death rate [15].3. Data and methodology This section's main objective is to evidence, clearly and briefly, the data/variables, the group of countries, and the methodological approach that will be used in our experimental study.3.1. Data The necessity of an initial phase of mitigating and in a second one that achieves carbon neutrality had put pressure on govern- ments to act and citizens to behave in ways that limit CO2 emis- sions. For example, Bersalli et al. [16] support that global decarbonisation strategies are based on electricity generation development built on renewable energy technologies. This study focuses on the evolution of non-hydroelectrical renewable installed capacity, between 2001 and 2017, in 19 coun- tries from the Latin America and the Caribbean (LAC) region, such as Argentina, Bolivia, Brazil, Chile, Colombia, Costa Rica, Dominican Republic, Ecuador, El Salvador, Guatemala, Haiti, Honduras, Ja- maica, Mexico, Nicaragua, Panama, Paraguay, Peru, and Uruguay. Indeed, the choice of this group of countries is motivated by several reasons: (a) they possess exceptional natural conditions to produce energy from renewable sources; (b) they experienced rapid and steady growth in non-hydroelectrical installed capacity since the beginning of the new millennium; (c) data on the evolution of renewable installed capacity and its drivers are available for all these countries. One of the most used vehicles to stimulate investment in renewable energies have been public policies. Bersalli et al. [16] indicate that public policies to boost renewable energies sources began in 1980 in developed economies and spread to emerging countries in the 2000s. For example, these authors found that policy's particularities act to turn auction the elected instrument to promote de diffusion of renewable energies in LA countries. Bradshaw [17] analysed market-oriented regulations in Brazil and concluded that reforms in the power sector have begun in the 1990s. The author underlines that Brazil's political and institutional situation act to design a regulatory reform that went beyond the traditional approach centred on efficiency and competition. That include social and environmental concerns in the planning of the energy sector. Indeed, Brazilian regulators followed creative man- dates in promoting renewable energy use. Bradshaw also shows that the evolution of wind and solar shares in energy supply was conditioned by the dominance of hydropower infrastructure in Brazil. This paper aims to assess the evolution of new renewable en- ergy sources in the LAC countries. However, our analysis excluded hydroelectrical power because most of its capacity was installed in the last century before our dataset starts. We measure the growth of new renewable energy sources through the yearly change in Gigawatts (GW) of non-hydroelectrical installed renewable capac- ity (DRen) and use it as our main dependent variable. We also assess, alternatively, the determinants of the evolution of the yearly change in Gigawatts (GW) of the major components of non- hydroelectrical installed capacity, namely: wind (DWind), solar (DSolar), and bioenergy (DBio). All this data was retrieved from Ref. [22]. Our choice of measuring the evolution of installed ca- pacity through the yearly change in GW instead of using the growth rate may provide more accurate insights into the causal impact of factors that promote renewable energy sources. The latter option would produce unreasonably high growth rates in the first years of the sample because the initial installed capacity was fairly low in most countries. Indeed, when computing growth rates, we have very high values in the initial stages, but these values are only very high because they are relative (rate) values. For example, if the initial value is zero, any addition, however insignificant, produces an infinite rate of change. Our approach is, in these cases, more suitable than traditional approaches. Furthermore, several coun- tries have no non-hydroelectrical installed capacity at the begin- ning of the sample, rendering the computation of growth rates impossible. Fig. 1 below shows that installed capacity from non- hydroelectrical renewable sources increased more than tenfold between 2000 and 2017. Alvarado et al. [18] found that output causes renewable energy in LAC countries. This causality was independent of countries that were high-income or medium-income. Renewable energy growth was also seen as a current energy policy trend (e.g., Ref. [19]. Indeed, renewable energy analysis entails economic, social, politi- cal, and technical dimensions. Bersalli et al. [16] analyse the public policies promoting renewable energy technologies for electricity generation in a panel of LA countries and a European country panel. They conclude that LA countries have been proactive in their tentative to implement renewable energy. These authors emphasise promoting public policies to increase renewable energy investment. Another finding of these authors' research is that tax incentives are insufficient to deploy renewable energy technologies. LAC countries are mostly rich in conventional sources of energy. The transition of fossil sources to renewable ones are exposed to prices of exportation of fossil energy. The technology of extraction of non-renewable energy has evolved to allow the exploitation of these reserves abundantly. Cordano and Zellou [20] dissect the possible consequences of super-cycles in natural gas prices on LA countries' environment. They focus on the use of hydraulic frac- turing on underground water. They conclude that the mix of con- ditions prevalent in LA countries, such as political instability, weak institutions, governmental interventionism, and Brazilian pre-salt endowment, turn the impact of gas-price super-cycle uncertain. Indeed, to realize the following investigation, we consider several explanatory variables that are expected to have an impact on renewable energy investments, such as:  Renewable Energy Finance Flows: IRENA [22] provides a database of renewable energy projects that received interna- tional financial institutions' support. This database includes in- formation about the projects' recipient country, supporting institution, technology (hydropower, geothermal energy, bio- energy, solar energy, wind energy, marine energy, multiple re- newables, and other renewable energy), and instrument type Viviescas et al. [21] focused their analyses on the impact of renewable energy on energy security in LA countries. They look to the intermittency of renewable energy sources and inquire how to mitigate it by exploring energy sources' complementarities. Their study also probes the impact of climate changes on the interaction N. Silva, J.A. Fuinhas and M. Koengkan Energy 237 (2021) 121611  Carbon dioxide emissions: This variable was retrieved from the World Bank Open Data [27] and provided a database of the amount of CO2 emissions in Kilotons per capita. Indeed, CO2 is correlated with other pollutants' emissions, and it may be used as a proxy for air pollution. Thus, higher values of this variable are expected to raise public awareness about the environment and increase the political pressure to promote the deployment of renewable energy sources. Moreover, in this investigation, we renamed the variable as (CO2).table_1 Gross Domestic Product (GDP) (cyclical component): The gross domestic product in constant local currency units were retrieved from the World Bank Open Data [27]. Then we applied the Hamilton [28] filter, with four lags and the differentiation parameter equal to 2, following the author's recommendation, and computed the cyclical component of GDP. An above-trend growth may generate a positive investor sentiment that raises the investment in renewable sources. On the other hand, it could increase the expectation of a reversal in the growth trend that causes renewable energy investors to postpone their projects. Thus, the expected effect of this variable on renewable energy installed capacity is ambiguous. Moreover, in this empirical investigation, we renamed the variable as (GDP_Cyc).table_3table_2(grant, loan, insurance, credit line, private development finance, equity investment, concessional loan, guarantee, and other official flows). Indeed, we define, as our explanatory variable, in the main regression whose purpose is to assess the evolution of all the non-hydroelectrical installed capacity, the total amount of finance flows, in constant 2010 United States Dollar (USD), directed at the promotion of renewable energy, except hydro- electrical, regardless of the instrument type. In this study, we named this variable as (REFF). We also build three other vari- ables that include the finance flows directed, specifically, at the promotion of wind energy (WFF), solar energy (SFF), and bio- energy (BFF). International finance flows render renewable in- vestment more affordable. They may foster their development, particularly in LAC countries, most of which do not have a suf- ficiently robust economy to support these investments on their own. In this investigation, we converted the variables DRen, DWind, DSolar, DBio, REFF, WFF, SFF, and BFF into per capita values through their division by the population of each country. Indeed, per capita values allow us to reduce the disparities between the variables caused by population growth over time. Certainly, the descriptive statistics of all variables that will be used in this empirical investigation are shown in Table 1 below. Indeed, before the execution of the main regression model, we applied the Shapiro-Wilk test to verify the presence of normality in the model. The null hypothesis of this test is that the variables follow a normal distribution. Table 2 below displays the results from the Shapiro-Wilk test for normality test. As can be observed in the table above, the p-values of this test lead to a clear rejection of the null hypothesis that the variables follow a normal distribution in all cases. Thus, traditional panel regression methods which rely on the normality of the variables are inadequate and may generate misleading conclusions. In this sub- section, we approached the group of countries and the variables that will be used in our empirical study. In the next section, we briefly show the methodological approach that this investigation will use.  Financial Development Index: This variable was retrieved from the International Monetary Fund (IMF) Financial Development Index [23] (2020) and provides a development index of financial institutions and financial services, and efficiency (the ability of institutions to provide financial services at low cost and with sustainable revenues and the level of activity of capital markets. Indeed, as we already know, electricity production from renewable energy sources, such as wind and solar, requires a large upfront investment but has lower running costs compared to fossil fuels. Thus, an accessible and efficient financial system is fundamental to support these projects and make them prof- itable for the investor. We measure the LAC countries' financial systems' soundness through the IMF Financial Development Index [24] and it is expected a positive relationship between this indicator and the dependent variable. Moreover, in this study, we renamed the variable as (FDI).3.2. Methodology As mentioned before, this subsection will approach theTable 1Table 1 Descriptive statistics of variables. Table 1 Descriptive statistics of variables. Variables Descriptive statistics Obs Mean Std. dev Min Max DRen 323 4.8966 15.5919 2.0529 131.6383 DWind 323 2.7228 12.4824 0.2345 124.0724 DSolar 323 1.0456 4.8523 0.8106 44.9566 DBio 323 1.3422 5.2247 25.5991 51.9247 REFF 323 3.5667 10.2401 0.000 70.6254 FDI 323 24.6110 12.2706 6.2964 63.3827 CO2 323 1.9575 1.2129 0.1584 4.6916 EGI 323 50.6871 12.8314 22.0174 83.4947 GDP_Cyc 323 1.0456 4.4273 19.6317 13.8450 Notes: Obs denotes the number of observations; Std. dev denotes standard devia- tion; Min denotes minimum, and Max denotes maximum. This table was built using the Stata function summarize.  Economic Globalisation Index De facto: This variable was retrieved from KOF Swiss Economics Institute [25] and was developed by Gygli et al. [26]. The index contemplates two sub- dimensions: (i) trade globalisation, which assesses the impor- tance of a country's trade in goods and services relative to GDP, and the diversity of trading partners; (ii) financial globalisation, which is measured by capital flows and stocks of foreign assets and liabilities. As already know, an economically globalised country may find it easier to import state-of-the-art technology designed to produce energy from new renewable sources. Thus, we expect economic globalisation to have a positive effect on the dependent variable. Furthermore, in this analysis, we renamed the variable as (EG). N. Silva, J.A. Fuinhas and M. Koengkan Energy 237 (2021) 121611 (2) Estimate the covariates' coefficients through the solution of the following problem, Table 2 Shapiro-Wilk test for normality. Variables Obs W V Z Prob > z DRen 323 0.3327 151.794 11.831 0.00000 DWind 323 0.3090 157.192 11.914 0.00000 DSolar 323 0.2717 165.685 12.038 0.00000 DBio 323 0.4148 133.142 11.523 0.00000 REFF 323 0.4215 131.600 11.495 0.00000 FDI 323 0.9175 18.770 6.907 0.00000 CO2 323 0.9059 21.404 7.217 0.00000 EGI 323 0.9752 5.649 4.079 0.00002 GDP_Cyc 323 0.9782 4.960 3.772 0.00008 Notes: This table was built using the Stata function swilk.Table 2 where d DReni;t≡Reni;t  bai, T and n denote the number of years and countries included in the estimation, respectively, and rt is the check function for quantile t. Canay [31] shows that the resulting estimates are consistent and asymptotically normal. We also evaluate the determinants of the evolution of wind, solar, and bioenergy installed capacities, with the dual purpose of testing the robustness of our results regarding the main dependent variable (DRen) and assessing the possible differential impact of the covariates on the individual components of renewable energy installed capacity. Specifically, we model the relation between the covariates and the various sources of renewable through the following equations. methodology that will be used in this empirical investigation. Indeed, to assess the effect of the explanatory variables on the change of non-hydroelectrical renewable installed capacity, this investigation will use the panel quantile regression model. This method presents several advantages relative to traditional least- square methods. It provides a complete picture of the indepen- dent variables' impact over all the dependent variable distribution, whereas least-square methods focus only on its conditional mean. Furthermore, quantile regression is broadly insensitive to outliers and, unlike least-square methods, does not require data to be normally distributed [29,30]. This characteristic is especially important for this study, given that our data is strongly non-normal (see Table 2 above). variable (DRen) and assessing the possible differential impact of the covariates on the individual components of renewable energy installed capacity. Specifically, we model the relation between the covariates and the various sources of renewable through the following equations. We model the relation between the change in renewable energy installed capacity and the explanatory variables through the following Equation, where i ¼ 1, …, 19 identifies the country, t ¼ 2001, …2017, corre- sponds to the observation year, XW0 i;t1 ¼ ð1; WFFi;t1; FDIi;t1; CO2i;t1; EGIi;t1; GDP Cyci;t1; TREND Þ, XS0 i;t1 ¼ ð1; SFFi;t1; FDIi;t1; CO2i;t1; EGIi;t1; GDP Cyci;t1; TREND Þ, XB0 i;t1 ¼ ð1; BFFi;t1; FDIi;t1; CO2i;t1; EGIi;t1; GDP Cyci;t1; TREND Þ, are the explanatory vari- ables vectors for the wind, solar and bioenergy equations, bWðUW i;t Þ ¼ ðbW 0 ; …; bW 6 Þ, bSðUS i;tÞ ¼ ðbS 0; …; bS 6Þ, bBðUB i;tÞ ¼ ðbB 0; …; bB 6Þ are the corresponding coefficients vectors, aW i , aS i , aB i are the unobserved fixed effects, and UW i;t  Uð0; 1Þ, US i;t  Uð0; 1Þ, UB i;t  Uð0; 1Þ are uniformly distributed random var- iables. Equations (4)e(6) are estimated through the same proced- ure described above for the main dependent variable, DRen. where i ¼ 1, …, 19 identifies the country, t ¼ 2001, …2017, corre- sponds to the observation year, X0 i;t1 ¼ ð1; REFFi;t1; FDIi;t1; CO2i;t1; EGIi;t1; GDP Cyci;t1; TREND Þ is the explanatory variables vector, bðUi;tÞ ¼ ðb0; …; b6Þ is the coefficients vector, ai is an unobserved fixed effect, Ui;t  Uð0; 1Þ, and TREND represents a time trend that may be seen as a proxy for the reduction in the levelized cost of renewable en- ergy over time. Indeed, in this investigation, we choose to lag the explanatory variables by one year because it takes a substantial amount of time for renewable energy investments to materialize in new installed capacity. where i ¼ 1, …, 19 identifies the country, t ¼ 2001, …2017, corre- sponds to the observation year, X0 i;t1 ¼ ð1; REFFi;t1; FDIi;t1; CO2i;t1; EGIi;t1; GDP Cyci;t1; TREND Þ is the explanatory variables vector, bðUi;tÞ ¼ ðb0; …; b6Þ is the coefficients vector, ai is an unobserved fixed effect, Ui;t  Uð0; 1Þ, and TREND represents a time trend that may be seen f h d i i h l li d f bl All the estimations were executed in the software R, using the QRPanel.R code provided by Canay. Fig. 2 shows the methodological steps we follow in this research. First, we present descriptive statistics for all the variables and perform the Shapiro-Wilk and Pesaran tests to check their normality and stationarity, respectively. Then, we assess the impact of the covariates on the dependent variables through the estima- tion of panel quantile regressions, using Canay's method. Finally, The fixed effect's unobservability renders the traditional quan- tile regression unable to identify the covariates' coefficients because ai can be arbitrarily related to the remaining variables. To circumvent this problem, we follow Canay [31]; who proposed a simple two-stage method that achieves identification, assuming that the fixed effect is a pure location shift, and ai and Ui;t are in- dependent. Let ui;t≡X0i;t½bðUi;tÞ  bm, and bm ¼ ½b0;m… b6;m , the conditional mean of bðUi;tÞ. Then, using Equation (1) we have, The estimator proposed by Canay [31]; for quantile t, involves two steps: (1) Obtain a consistent estimator1 of bm from Equation (2) and define bai≡½Reni;t  X0i;tbbm 1 We use the fixed effects method. N. Silva, J.A. Fuinhas and M. Koengkan Energy 237 (2021) 121611 Table 3 Panel unit root test (CIPS test). Variables Panel unit root test (CIPS) (Zt-bar) Without trend With trend Lags Zt-bar Zt-bar DRen 0 8.907 *** 8.131 *** DWind 0 6.237 *** 5.671 *** DSolar 0 2.815 *** 1.973 ** DBio 0 4.909 *** 4.460 *** REFF 0 4.480 *** 2.426 *** FDI 0 3.245 *** 2.964 *** CO2 0 0.842 ** 0.633 EGI 0 0.637 1.257 GDP_Cyc 0 4.521 *** 2.871 *** Notes: This table was built using the Stata function multipurt. The null hypothesis for CIPS stipulates that the series have a unit root. ***, **, and * denote a statistically significant rejection of the null hypothesis at the 1 %, 5 %, and 10 % levels, respectively. crucial role in promoting renewable energy in LAC countries, particularly in large projects. The estimates for CO2 emissions are positive, sizable, and broadly similar for all the quantiles. This ev- idence is consistent with the hypothesis that CO2 emissions generate political pressure that fosters renewable energy devel- opment. The dependent variable also bears a significant positive relation with economic globalisation for all the quantiles. Thus, countries with an economically open economy are more prone to develop new renewable energy sources. The time trend, which may be interpreted as a proxy for reducing the levelized cost of renewable energy over time, also positively affects the dependent variable in the lowest and highest quantiles. Finally, neither the financial development index nor cyclical GDP seems to be associ- ated with renewable energy development.Notes: This table was built using the Stata function multipurt. The null hypothesis for CIPS stipulates that the series have a unit root. ***, **, and * denote a statistically significant rejection of the null hypothesis at the 1 %, 5 %, and 10 % levels, respectively. The results for the wind installed capacity determinants' (Table 5) is strongly consistent with the main estimation (DRen) about the impact of finance flows, CO2 emissions, and the economic globalisation index: all these covariates exert a positive effect on installed capacity and the impact of financial flows is particularly prevalent at the highest quantiles. However, unlike in the (DRen) estimation, the financial development index negatively affects wind installed capacity.we check the robustness of our results to the presence of outliers using dummy variables. Table 6, which displays the estimated coefficients for the solar energy equation, corroborates our previous findings regarding the importance of finance flows, CO2 emissions, and time for the pro- motion of renewable energy sources, but it shows that the financial development index and the economic globalisation index may hinder the increase in solar installed capacity.4. Empirical results As mentioned previously in the introduction of this investiga- tion, this section will present the quantile regressions' estimation results, using the Canay [31] method. In this investigation, we choose to estimate the effect of the covariates on the change in non- hydroelectrical renewable installed capacity for the 25 %, 50 %, and 75 % quantiles and the 90 % quantile because we want to find the main drivers of the substantial increases in renewable installed capacity in the recent years. The estimation for bioenergy (Table 7) reveals that CO2 emis- sions, the economic globalisation index, the financial development index, and finance flows (only in the 75th quantile) foster the in- crease in bioenergy installed capacity. However, unlike in the pre- vious estimations for other renewable sources, time appears to have a detrimental effect on bioenergy. Before estimating the quantile regressions, we perform the Pesaran [32] panel unit root test to assess the series stationarity. This test, whose null hypothesis assumes that the variable is non- stationary, follows a non-standard distribution, even when the number of members in the sample is large. It is based on the averaging of panel-member specific Dickey-Fuller type regressions' coefficients and is robust in the presence of cross-section depen- dence. Indeed, Table 3 below shows that the null non-stationarity hypothesis is strongly rejected for all the dependent variables in both specifications. This hypothesis is also rejected for most explanatory variables, except CO2 emissions, in the test with a time trend and the economic globalisation Index in both tests. It is well known that in a relatively short panel, such as ours, the estimation results may be driven by a few extreme observations, which distort the causal link between the covariates and the dependent variable and leads the researcher to draw wrong con- clusions. To check our results' robustness to the presence of out- liers, we identified as outliers those observations whose residuals are more than three standard deviations away from the mean. Then, we created a dummy variable for each one of these obser- vations and re-estimated all the models. These exercise results are presented in Table 8, 9, 10, and 11 below. The comparison of Tables 4 and 8 reveals that the results remain broadly unchanged when dummies are included in the estimation of the evolution of non-hydroelectrical installed capacity: finance flows, CO2 emissions, and the economic globalisation index remain its main drivers, while the time trend impacts it only in the 25th Table 4 below shows the estimated coefficients for the change in the non-hydroelectrical installed capacity equation. The effect of Finance Flows on renewable energy development is positive and statistically significant for all the quantiles, except the fiftieth. Its impact is increasing across the quantiles, suggesting that it plays a Table 4 Estimation results - DRen. Independent variables Dependent variable (DRen) Quantiles 25th 50th 75th 90th REFF 0.1557 *** 0.3520 1.3034 ** 1.8932 *** FDI 0.0330 0.0221 0.0235 0.0612 CO2 2.6207 *** 2.3734 *** 2.3609 *** 2.8066 *** EGI 0.1422 *** 0.1514 *** 0.1839 *** 0.2173 *** GDP_Cyc 0.0255 0.0557 0.0863 0.1689 TREND 0.0885 ** 0.0913 0.0789 0.2654 * Obs 323 323 323 323 Notes: The estimations were performed using the R code QRPanel.R; ***, **, and * denote a statistically significant rejection of the null hypothesis at the 1 %, 5 %, and 10 % levels, respectively.Table 4 using the R code QRPanel.R; ***, **, and * denote a statistically significant rejection of the null hypothesis at the 1 %, 5 %, and 10 % Notes: The estimations were performed using the R code QRPanel.R; ***, **, and * denote a statistically significant rejection of the evels, respectively. tions were performed using the R code QRPanel.R; ***, **, and * denote a statistically significant rejection of the null hypothesis aNotes: The estimations were performed using the R code QRPanel.R; ***, **, and * denote a statistically significant rejection of the null hypothesis at the 1 %, 5 %, and 10 % levels, respectively. Energy 237 (2021) 121611N. Silva, J.A. Fuinhas and M. Koengkan Table 5 Estimation results - DWind. Independent variables Dependent variable (DWind) Quantiles 25th 50th 75th 90th WFF 0.7788 *** 1.5451 *** 2.3467 *** 3.6648 *** FDI 0.0709 *** 0.0889 *** 0.0487 *** 0.0479 *** CO2 2.0195 *** 2.4668 *** 2.4109 *** 2.4716 *** EGI 0.0835 *** 0.0942 *** 0.1136 *** 0.1288 *** GDP_Cyc 0.0183 0.0055 0.0089 0.0349 * TREND 0.0052 0.0105 0.0259 0.2457 *** Obs 323 323 323 323 Notes: The estimations were performed using the R code QRPanel.R; *** and * denote a statistically significant rejection of the null hypothesis at the 1 % and 10 % levels, respectively.Table 5 25th 50th 75th 90th WFF 0.7788 *** 1.5451 *** 2.3467 *** 3.6648 *** FDI 0.0709 *** 0.0889 *** 0.0487 *** 0.0479 *** CO2 2.0195 *** 2.4668 *** 2.4109 *** 2.4716 *** EGI 0.0835 *** 0.0942 *** 0.1136 *** 0.1288 *** GDP_Cyc 0.0183 0.0055 0.0089 0.0349 * TREND 0.0052 0.0105 0.0259 0.2457 *** Obs 323 323 323 323 Notes: The estimations were performed using the R code QRPanel.R; *** and * denote a statistically significant rejection of the null hypothesis at the 1 % and 10 % levels, respectively. Table 6 Estimation results - DSolar. Independent variables Dependent variable (DSolar) Quantiles 25th 50th 75th 90th SFF 0.0354 *** 0.2111 *** 0.2727 1.4467 *** FDI 0.2598 *** 0.2477 *** 0.2453 *** 0.2339 *** CO2 0.7301 *** 0.7918 *** 0.8283 *** 0.8706 *** EGI 0.0320 *** 0.0207 *** 0.0132 0.0041 GDP_Cyc 0.0336 0.0404 * 0.0228 0.0432 TREND 0.1677 *** 0.1720 0.1662 *** 0.1554 *** Obs 323 323 323 323 Notes: The estimations were performed using the R code QRPanel.R; *** and * denote a statistically significant rejection of the null hypothesis at the 1 % and 10 % levels, respectively. Table 7 Estimation results - DBio. Independent variables Dependent variable (DBio) Quantiles 25th 50th 75th 90th BFF 0.0983 0.3685 0.8479 *** 0.9930 FDI 0.1421 *** 0.1473 *** 0.1406 *** 0.1406 *** CO2 1.9514 *** 1.9026 *** 1.8436 *** 1.4957 *** EGI 0.0697 *** 0.0795 *** 0.1057 *** 0.1114 *** GDP_Cyc 0.0027 0.0338 0.0487 * 0.0007 TREND 0.0844 *** 0.1030 *** 0.0858 0.0048 Obs 323 323 323 323 Notes: The estimations were performed using the R code QRPanel.R; *** and * denote a statistically significant rejection of the null hypothesis at the 1 % and 10 % levels, respectively.Table 6 Table 6 Estimation results - DSolar. Independent variables Dependent variable (DSolar) Quantiles 25th 50th 75th 90th SFF 0.0354 *** 0.2111 *** 0.2727 1.4467 *** FDI 0.2598 *** 0.2477 *** 0.2453 *** 0.2339 *** CO2 0.7301 *** 0.7918 *** 0.8283 *** 0.8706 *** EGI 0.0320 *** 0.0207 *** 0.0132 0.0041 GDP_Cyc 0.0336 0.0404 * 0.0228 0.0432 TREND 0.1677 *** 0.1720 0.1662 *** 0.1554 *** Obs 323 323 323 323 Notes: The estimations were performed using the R code QRPanel.R; *** and * denote a statistically significant rejection of the null hypothesis at the 1 % and 10 % levels, respectively. Table 7 Estimation results - DBio. Independent variables Dependent variable (DBio) Quantiles 25th 50th 75th 90th BFF 0.0983 0.3685 0.8479 *** 0.9930 FDI 0.1421 *** 0.1473 *** 0.1406 *** 0.1406 *** CO2 1.9514 *** 1.9026 *** 1.8436 *** 1.4957 *** EGI 0.0697 *** 0.0795 *** 0.1057 *** 0.1114 *** GDP_Cyc 0.0027 0.0338 0.0487 * 0.0007 TREND 0.0844 *** 0.1030 *** 0.0858 0.0048 Obs 323 323 323 323 Notes: The estimations were performed using the R code QRPanel.R; *** and * denote a statistically significant rejection of the null hypothesis at the 1 % and 10 % levels, respectively.Table 7 inclusion of dummies weakens the positive effect of the economic globalisation index and strengthens the time trend's impact. and 90th quantiles, in the baseline estimation, and the 75th quantile in the estimation with dummies. g g p Solar energy results seem to be the most sensitive to outliers (Tables 6 and 10). Even though the effect of finance flows remains Tables 5 and 9 confirm the positive influence of finance flows and CO2 emissions on wind energy installed capacity. However, theTable 8 E i iTable 8 Estimation results (with dummies) - DRen. Table 8 Estimation results (with dummies) - DRen. Independent variables Dependent variable (DRen) Quantiles 25th 50th 75th 90th REFF 0.1269 *** 0.3155 *** 1.4292 *** 2.3875 *** FDI 0.0107 0.0045 0.1329 *** 0.1383 *** CO2 3.3662 *** 3.1722 *** 2.7964 *** 4.0406 *** EGI 0.0416 *** 0.0527 *** 0.1251 *** 0.2379 *** GDP_Cyc 0.0006 0.0013 0.0802 0.0125 TREND 0.0565 0.0912 0.1783 *** 0.1692 Obs 323 323 323 323 Notes: The estimations were performed using the R code QRPanel.R; *** denotes a statistically significant rejection of the null hypothesis at the 1 % level. Energy 237 (2021) 121611N. Silva, J.A. Fuinhas and M. KoengkanTable 9 Estimation results (with dummies) - DWind.Table 9 Table 9 Estimation results (with dummies) - DWind. Independent variables Dependent variable (DWind) Quantiles 25th 50th 75th 90th WFF 0.3001 *** 0.8894 *** 2.6437 *** 4.0447 *** FDI 0.0927 *** 0.0781 *** 0.1436 *** 0.0479 *** CO2 2.0901 *** 1.9496 *** 0.6409 *** 0.9926 *** EGI 0.0045 0.0068 0.0316 *** 0.0997 *** GDP_Cyc 0.0241 0.0062 0.0271 0.0103 TREND 0.0450 *** 0.0291 ** 0.0861 *** 0.2313 ** Obs 323 323 323 323 Notes: The estimations were performed using the R code QRPanel.R; *** and ** denote a statistically significant rejection of the null hypothesis at the 1 % and 5 %, levels, respectively. Table 10 Estimation results (with dummies) - DSolar. Independent variables Dependent variable (DSolar) Quantiles 25th 50th 75th 90th SFF 0.0297 *** 0.2912 *** 0.4642 *** 2.5146 *** FDI 0.0521 *** 0.0707 *** 0.0570 *** 0.2140 *** CO2 0.3701 *** 0.2463 *** 0.3227 *** 0.9518 *** EGI 0.0085 * 0.0076 ** 0.0035 0.0090 GDP_Cyc 0.0101 0.0010 0.0090 0.0186 TREND 0.0573 *** 0.0618 *** 0.0566 *** 0.0055 *** Obs 323 323 323 323 Notes: The estimations were performed using the R code QRPanel.R; ***, **, and * denote a statistically significant rejection of the null hypothesis at the 1 %, 5 %, and 10 % levels, respectively.Table 10 Table 11 Estimation results (with dummies) - DBio. Independent variables Dependent variable (DBio) Quantiles 25th 50th 75th 90th BFF 0.0194 0.3471 ** 0.8316 *** 0.6750 *** FDI 0.0913 *** 0.1030 *** 0.1068 *** 0.0812 *** CO2 0.4595 *** 0.3726 *** 0.2794 *** 0.5431 ** EGI 0.0203 0.0066 0.0001 0.0078 GDP_Cyc 0.0290 0.0338 0.0074 0.0272 TREND 0.0473 *** 0.0322 ** 0.0196 0.0048 Obs 323 323 323 323 Notes: The estimations were performed using the R code QRPanel.R; *** and ** denote a statistically significant rejection of the null hypothesis at the 1 % and 5 % levels, respectively.Table 11Table 11 Estimation results (with dummies) - DBio.Estimation results (with dummies) - DBio. broadly unchanged, dummies lead to a total reversion in the co- efficients related to CO2 emissions. This instability is hardly sur- prising, given that solar energy was the newest energy source to be widely deployed in LAC countries. Thus, in the early years of our sample, most countries did not install new solar capacity, which effectively reduces our sample's information content for this en- ergy source. variables and are on the borderline between the I(0) and I(1) orders of integration (see Table 3, above). Indeed, both two results had been already confirmed by Koengkan et al. [33,34], that studied the LAC region and found the same variable characteristics in their investigations. This is pointing that the results from the preliminary tests that were found in this empirical study are in consonance with the literature review. gy Finally, Table 11 confirms the positive impact of finance flows, time, and CO2 emissions (except in the 90th quantile) on bioenergy found in Table 7. However, the inclusion of dummies leads to the disappearance of the statistical significance for the variable eco- nomic globalisation index. Moreover, the model estimations' results indicate that the in- dependent variables of renewable energy finance flows, economic globalisation index, and carbon dioxide emissions positively affect our dependent variable named change in non-hydroelectrical renewable installed capacity. Besides, Fig. 3 below summarises the effect of independent variables on the dependent one. This figure was based on the results of the Canay model estimation.5. Discussion Then, based on these empirical founds we elaborate on the following question - What are the explanations for the positive impact of these variables in our model? e Well, the positive effect of renewable energy finance flows could be related to the increase of capital inflow for finance the new renewable energy projects As mentioned before, this section will present the possible ex- planations which were found in this empirical investigation. Well, the results from the preliminary tests show the non-presence of normal distribution (see Table 2, above), the indication that the N. Silva, J.A. Fuinhas and M. Koengkan Energy 237 (2021) 121611 energy demand. However, to attend to the energy demand will be necessary more investments in alternative energy sources. More- over, this economic liberalisation process will boost the imports of renewable energy technologies to increase energy efficiency and reduce the consumption of non-renewable energy sources. Finally, the positive impact of CO2 emissions could be related to the increase of environmental, political, and social pressure about the increase of environmental degradation caused by these emis- sions' growth. Indeed, this pressure will encourage the develop- ment of renewable energy policies that will facilitate investments, development, and the consumption of renewable energy sources in the LAC countries. Another explanation is related to the capacity to consume new renewable energy sources to mitigate these CO2 emissions in the LAC region. This section showed the results from the main model and a brief explanation for this study's results. The next section will show the conclusions of this experimental investigation. caused by the restructure plans, for example, ‘Washington consensus' and ‘Brady Plan’, which were adopted to promote the restructuring of external debts and the macroeconomic adjust- ments and mentioned by Koengkan [35]; and Santiago et al. [36]. Indeed, this period of adjustments occurred between 1989 and 1992, in several LAC countries (e.g., Argentina, Brazil, Costa Rica, Mexico, Venezuela (RB), and Uruguay) adopted these strategies. The restructuring plans and macroeconomic adjustments allowed a deep financial, trade, and foreign investment liberalisation, as well as the privatisation of significant portions of the public sector and the reduction of import barriers in most countries from the LAC region. As a result, the capital inflows reappearance in the 1990s, at the middle of the decade (1994), the gross fixed capital stock per capita in the LAC region (Argentina, Brazil, Chile, Colombia, Ecuador, Mexico, and Venezuela) reached a mean value of 11,260 US$ (in 1980 international dollars) that compares with 54,089 US$ for the USA [37]. Indeed, the LAC region's capital inflows acceler- ated again between 2004 and 2014, caused by the “commodities boom” where the capital inflows registered a growth rate of 25 % during this period and reached a value of 1743.936 US$ in 2014 [35]. That is, the renewable energy finance flows followed the same trend, where the inflows of investments in new renewable energy sources (e.g., marine, wind, solar, geothermal, solid biofuels and waste and liquid biofuels) were a value of 1.6 US$ billion in 2014 and 2016 reached 8.7 US$ billion [35].6. Conclusion This analysis assessed the advancement of new renewable en- ergy sources in a group of nineteen countries from the LAC region between 2001 and 2017. This investigation is in the initial stages of maturation, where it will supply a solid foundation for second- generation researchers regarding this topic. This study is a kick- off about the advancement of new renewable energy sources in the LAC region. This empirical research has been based on eco- nomic principles to construct a model that provides an accurate explanation of the results that were found in this study. This empirical investigation utilised as a method the Canay model estimation to realize the following analysis. The results from the preliminary tests pointed to the non-presence of normal dis- tribution and that the variables are on the borderline between the I (0) and I(1) orders of integration. Indeed, both two results agree with the literature that approached this group of countries. The Canay model estimation results showed that the indepen- dent variables renewable energy finance flows, economic globali- sation index, and carbon dioxide emissions positively affect our dependent variable named change in non-hydroelectrical renew- able installed capacity. It is worth remembering that this empirical investigation opted to choose from the Canay model estimation because it can capture the variations between countries (i.e., cross- sections). That is, the results found by this empirical investigation can answer the research question that arose. Therefore, the increase of capital inflows caused by financial liberalisation will reduce the financing costs and consequently will encourage the development and investment in new renewable energy technologies. This explanation is confirmed by Narayan and Smyth [38]; and Koengkan et al. [33]; where the increase of capital inflows to invest in renewable energy projects will reduce their costs of financing, making the credit cheaper and makes the new energy sources more feasible. Moreover, another factor that could be related to this positive impact is the existence of efficient renewable energy financial policies that encourage public and private financing in new renewable energy projects. As mentioned before, the possible explanation for the positive impact of renewable energy finance flows on change in non- hydroelectrical renewable installed capacity could be related to the increase of capital inflow caused by financial liberalisation and that consequently reduced the financing costs and encouraged the development and investments in new renewable energy technol- ogies (e.g., marine, wind, solar, geothermal, solid biofuels and waste and liquid biofuels). That is, the reduction of financing cost made the new renewable energy technologies more attractive than the hydropower plants in the LAC region. Moreover, another factor could be related to this positive impact is the existence of efficient renewable energy financial policies that encouraged public and private financing in new renewable energy projects. Regarding the positive impact of economic globalisation index on change in non-hydroelectrical renewable installed capacity, also could be related to the process of deep financial, trade, and foreign investment liberalisation, as well as the privatisation of significant portions of the public sector, and the reduction of import barriers as was mentioned before. All these changes in the economic structure consequently affect the index of globalisation, where financial and trade liberalisation are the main components of this index. Indeed, as we already know, the process of financial and trade liberalisation will boost the capital inflows to realize a new investment in the productive sectors or speculative, the economic activity, and so the The positive impact of the economic globalisation index could be related to the process of deep financial, trade, and foreign in- vestment liberalisation, as well as the privatisation of significant portions of the public sector, and the reduction of import barriers caused by the restructuring plans that were adopted by some LAC countries that encouraged the economic activity and so the energy demand. However, to attend to the energy demand will be neces- sary to carry out new investments in alternative energy sources. N. Silva, J.A. Fuinhas and M. Koengkan Energy 237 (2021) 121611 developing countries. Ecol Indicat 2016;67:543e55. https://doi.org/10.1016/ j.econmod.2014.10.022. Moreover, this liberalisation process in Latin American economies also encouraged the import of new renewable energy technologies to increase productivity and reduce the consumption of non- renewable energy sources. [13] Kasman A, Duman YS. CO2 emissions, economic growth, energy consumption, trade and urbanisation in new EU member and candidate countries: a panel data analysis. Econ Modell 2015;44:97e103. https://doi.org/10.1016/ j.econmod.2014.10.022. Moreover, the positive impact of CO2 emissions on change in non-hydroelectrical renewable installed capacity could be related to the increase of environmental, political, and social pressure regarding the increase of environmental degradation caused by the growth in these emissions that encouraged the new investments in renewable energy technologies, as well as could be related to the own capacity of consumption of new renewable energy sources to mitigate these CO2 emissions in the LAC region. [14] Koengkan M, Fuinhas JA. Does the overweight epidemic cause energy con- sumption? A piece of empirical evidence from the European region. Energy 2021;216:119297. https://doi.org/10.1016/j.energy.2020.119297. [15] Koengkan M, Fuinhas JA, Silva N. Exploring the capacity of renewable energy consumption to reduce outdoor air pollution death rate in Latin America and the Caribbean region. Environ Sci Pollut Control Ser 2021;28:1656e74. https://doi.org/10.1007/s11356-020-10503-x. p // g/ / [16] Bersalli G, Menanteau P, El-Methni J. Renewable energy policy effectiveness: a panel data analysis across Europe and Latin America. Renew Sustain Energy Rev 2020;133:110351. https://doi.org/10.1016/j.rser.2020.110351. [17] Bradshaw A. Regulatory change and innovation in Latin America: the case of renewable energy in Brazil. Util Pol 2017;49:156e64. https://doi.org/10.1016/ j.jup.2017.01.006.Authors contributions [18] Alvarado R, Ponce P, Alvarado R, Ponce K, Huachizaca V, Toledo E. Sustainable and non-sustainable energy and output in Latin America: a cointegration and causality approach with panel data. Energy Strat Rev 2019;26:100369. https:// doi.org/10.1016/j.esr.2019.100369. Nuno Silva: Investigation, Methodology, Validation, Writing- Original draft preparation. Jose Alberto Fuinhas: Conceptualiza- tion, Supervision, Visualization, Writing-Reviewing and Editing. Matheus Koengkan: Data curation, Formal analysis, Investigation. [19] Xu X, Wei Z, Ji Q, Wang C, Gao G. Global renewable energy development: influencing factors, trend predictions and countermeasures. Resour Pol 2019;63:101470. https://doi.org/10.1016/j.resourpol.2019.101470.Declaration of competing interest [20] Vasquez Cordano AL, Zellou AM. Super cycles in natural gas prices and their impact on Latin American energy and environmental policies. Resour Pol 2020;65:101513. https://doi.org/10.1016/j.resourpol.2019.101513. The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. [21] Viviescas C, Lima L, Diuana FA, Vasquez E, Ludovique C, Silva GN, Huback V, Magalar L, Szklo A, Lucena AFP, Schaeffer R, Paredes JR. Contribution of Var- iable Renewable Energy to increase energy security in Latin America: complementarity and climate change impacts on wind and solar resources. Renew Sustain Energy Rev 2019;113:109232. https://doi.org/10.1016/ j.rser.2019.06.039.Acknowledgements [22] IRENA. International Renewable Energy Agency; 2020. [23] International Monetary Fund (IMF). Financial development index database. URL: https://data.imf.org/?sk¼f8032e80-b36c-43b1-ac26-493c5b1cd33b; 2020. CeBER R&D unit funded by national funds through FCT e Fundaç~ao para a Ci^encia e a Tecnologia, I.P., project UIDB/05037/ 2020. [24] Svirydzenka K. Introducing a new broad-based index of financial develop- ment. IMF Working Papers 2016;16(5):1. https://doi.org/10.5089/ 9781513583709.001.References [25] KOF Swiss Economics Institute. Globalization index. URL: https://kof.ethz.ch/ en/; 2020. / [26] Gygli S, Haelg F, Potrafke N, Sturm JE. The KOF globalisation index e revisited. Rev Int Organ 2019;14(3):543e74. https://doi.org/10.1007/s11558-019- 09344-2. [1] Kraft J, Kraft A. On the relationship between energy and GNP. J Energy Dev 1978;3(2):401e3. [2] Fuinhas JA, Marques AC. The extended energy-growth nexus: theory and empirical applications. Academic Press; 2019, ISBN 9780128157190. https:// doi.org/10.1016/C2017-0-02521-1. [27] World Bank Open Data. URL: http://www.worldbank.org/; 2020. [28] Hamilton JD. Why you should never use the Hodrick-Prescott filter. Rev Econ Stat 2018;100(5):831e43. https://doi.org/10.1162/rest_a_00706. [3] Banal-Esta~nol A, Calzada J, Jordana J. How to achieve full electrification: les- sons from Latin America. Energy Pol 2017;108:55e69. https://doi.org/ 10.1016/j.enpol.2017.05.036. [29] Koenker R. Quantile regression (econometric society monographs). Cam- bridge: Cambridge University Press; 2005. https://doi.org/10.1017/ CBO9780511754098. /j p [4] Washburn C, Pablo-Romero M. Measures to promote renewable energies for electricity generation in Latin American countries. Energy Pol 2019;128: 212e22. https://doi.org/10.1016/j.enpol.2018.12.059. [30] Onyedikachi OJ. Robustness of quantile regression to outliers. Am J Appl Math Stat 2015;3(2):86e8. https://doi.org/10.12691/ajams-3-2-8. [31] Canay IA. A simple approach to quantile regression for panel data. Econom J 2011;14(3):368e86. https://doi.org/10.1111/j.1368-423X.2011.00349.x. [ ] l l h f p // g/ /j p [5] Kruckenberg LJ. Renewable energy partnerships in development cooperation: towards a relational understanding of technical assistance. Energy Pol 2015a;77:11e20. https://doi.org/10.1016/j.enpol.2014.11.004. [32] Pesaran MH. A simple panel unit root test in the presence of cross-section dependence. J Appl Econ 2007;22:265e312. [6] Kruckenberg LJ. NortheSouth partnerships for sustainable energy: knowledgeepower relations in development assistance for renewable energy. Energy Sustain Dev 2015b;29:91e9. https://doi.org/10.1016/ j.esd.2015.10.003. [33] Koengkan M, Fuinhas JA, Vieira I. Effects of financial openness on renewable energy investments expansion in Latin American countries. J Sustain Finan Invest 2020;10(1):65e82. https://doi.org/10.1080/20430795.2019.1665379. [34] Fuinhas JA, Marques AC, Koengkan M. Are renewable energy policies upset- ting carbon dioxide emissions? The case of Latin America countries. Environ Sci Pollut Control Ser 2017;24:15044e54. https://doi.org/10.1007/s11356- 017-9109-z. [7] Kim J, Park K. Effect of the Clean Development Mechanism on the deployment of renewable energy: less developed vs. well-developed financial markets. Energy Econ 2018;75:1e13. https://doi.org/10.1016/j.eneco.2018.07.034. [8] Zeng S, Liu Y, Liu C, Nan X. A review of renewable energy investment in the BRICS countries: history, models, problems and solutions. Renew Sustain Energy Rev 2017;74:860e72. https://doi.org/10.1016/j.rser.2017.03.016. [35] Koengkan M. Capital stock development in Latin America and the Caribbean region and their effect on investment expansion in renewable energy. J Sustain Finan Invest 2020:1e18. https://doi.org/10.1080/ 20430795.2020.1796100. [9] Kim J, Park K. Financial development and deployment of renewable energy technologies. Energy Econ 2016;59:238e50. https://doi.org/10.1016/ j.eneco.2016.08.012. [36] Santiago R, Koengkan M, Fuinhas JA, Marques AC. The relationship between public capital stock, private capital stock and economic growth in the Latin American and Caribbean countries. Int Rev Econ 2019:1e25. j [10] Zeng S, Jia J, Su B, Jiang C, Zeng G. The volatility spillover effect of the Euro- pean Union (EU) carbon financial market. J Clean Prod 2020:124394. https:// doi.org/10.1016/j.jclepro.2020.124394. [37] Hofman AA. Standardised capital stock estimates in Latin America: a 1950-94 update. Camb J Econ 2000;24(1):45e86. https://doi.org/10.1093/cje/24.1.45. [11] Guan D, Meng J, Reiner DM, Zhang N, Shan Y, Mi Z, Davis SJ. Structural decline in China's CO2 emissions through transitions in industry and energy systems. Nat Geosci 2018;11(8):551e5. https://doi.org/10.1038/s41561-018-0161-1. [38] Narayan PK, Smyth R. Energy consumption and real GDP in G7 countries: new evidence from panel cointegration with structural breaks. Energy Econ 2008;30(5):2331e41. https://doi.org/10.1016/j.eneco.2007.10.006. ( ) p // g/ / [12] Ertugrul HM, Cetin M, Seker F, Dogan E. The impact of trade openness on global carbon dioxide emissions: evidence from the top ten emitters amongEcological Applications, 18(4), 2008, pp. 885–898  2008 by the Ecological Society of AmericaEXPANSION OF SUGARCANE ETHANOL PRODUCTION IN BRAZIL: ENVIRONMENTAL AND SOCIAL CHALLENGES LUIZ A. MARTINELLI1,3 AND SOLANGE FILOSO2 1CENA-USP, Av. Centena´rio 303, 13416-000, Piracicaba-SP, Brazil 2Chesapeake Biological Laboratory, University of Maryland Center for Environmental Science, P.O. Box 38, Solomons, Maryland 20688 USAEXPANSION OF SUGARCANE ETHANOL PRODUCTION IN BRAZIL: ENVIRONMENTAL AND SOCIAL CHALLENGES LUIZ A. MARTINELLI1,3 AND SOLANGE FILOSO2 1CENA-USP, Av. Centena´rio 303, 13416-000, Piracicaba-SP, Brazil 2Chesapeake Biological Laboratory, University of Maryland Center for Environmental Science, P.O. Box 38, Solomons, Maryland 20688 USA Abstract. Several geopolitical factors, aggravated by worries of global warming, have been fueling the search for and production of renewable energy worldwide for the past few years. Such demand for renewable energy is likely to benefit the sugarcane ethanol industry in Brazil, not only because sugarcane ethanol has a positive energetic balance and relatively low production costs, but also because Brazilian ethanol has been successfully produced and used as biofuel in the country since the 1970s. However, environmental and social impacts associated with ethanol production in Brazil can become important obstacles to sustainable biofuel production worldwide. Atmospheric pollution from burning of sugarcane for harvesting, degradation of soils and aquatic systems, and the exploitation of cane cutters are among the issues that deserve immediate attention from the Brazilian government and international societies. The expansion of sugarcane crops to the areas presently cultivated for soybeans also represent an environmental threat, because it may increase deforestation pressure from soybean crops in the Amazon region. In this paper, we discuss environmental and social issues linked to the expansion of sugarcane in Brazil for ethanol production, and we provide recommendations to help policy makers and the Brazilian government establish new initiatives to produce a code for ethanol production that is environmentally sustainable and economically fair. Recommendations include proper planning and environmental risk assessments for the expansion of sugarcane to new regions such as Central Brazil, improvement of land use practices to reduce soil erosion and nitrogen pollution, proper protection of streams and riparian ecosystems, banning of sugarcane burning practices, and fair working conditions for sugarcane cutters. We also support the creation of a more constructive approach for international stakeholders and trade organizations to promote sustainable development for biofuel production in developing countries such as Brazil. Finally, we support the inclusion of environmental values in the price of biofuels in order to discourage excessive replacement of natural ecosystems such as forests, wetlands, and pasture by bioenergy crops. Key words: biofuel; deforestation; environmental degradation; ethanol; human social cost; riparian orest; Saccharum; Sa˜o Paulo, Brazil; sugarcane. Key words: biofuel; deforestation; environmental degradation; ethanol; human social cost; riparian forest; Saccharum; Sa˜o Paulo, Brazil; sugarcane.INTRODUCTION obtained. The most important limits and conditions are that the new forms of energy be renewable, environ- mentally friendly, and not contribute to the increase in atmospheric CO2 concentration in the atmosphere. In recent years, energy consumption and global carbon intensity (the ratio between carbon emissions and energy supplied) have increased worldwide, reinvig- orating worries about potential depletion of fossil fuel reserves. Such increase, accompanied by growing political instability in oil-producing regions, has insti- gated many countries to search for alternative forms of energy. However, concerns about rising atmospheric CO2 concentration in the atmosphere due to fossil fuel burning and other anthropogenic activities, aggravated by compelling evidence of consequent dangerous chang- es in the climatic system of the Earth (IPCC 2007), have imposed some limits to the types of alternative energy that can be used, and conditions on how this energy is Biofuel is a promising source of energy because it is generated by the process of photosynthesis, where energy from the sun is captured and transformed into biomass that can be combusted to produce energy. In most cases, this alternative source is renewable, since the CO2 emitted into the atmosphere is recaptured by the growing crop in the next growth cycle. Ethanol from sugarcane is one of the most promising biofuels because its energetic balance is generally positive, meaning that the growing sugarcane absorbs more carbon than is emitted when the ethanol is burned as fuel (Oliveira et al. 2005). Moreover, the price of production is relatively low. Brazil has several advantages in this new scenario of biofuel production due to its expansive territory, geographical position, abundant water resources, and Manuscript received 30 October 2007; accepted 19 December 2007. Corresponding Editor: A. R. Townsend. 3 E-mail: martinelli@cena.usp.br LUIZ A. MARTINELLI AND SOLANGE FILOSO Ecological Applications Vol. 18, No. 4 Brazilian Government, particularly in the 1970s and 1980s.nan According to the most recent land use data for Brazil, ;264 million ha of Brazil’s land mass in 2005 was agricultural. Therefore, the area covered by sugarcane represented only ;2.5% of the total. Compared with the area planted with soybeans (23 million ha), sugarcane land cover is relatively small, and mostly confined to the southeast (64%), and along the coastline in the northeast (19%). In the southeast region, the state of Sa˜ o Paulo has .50% of the country’s sugarcane land cover. In the past 15 years this is where most of the expansion of sugarcane plantations occurred by replacing pasture by sugarcane, mostly in the western region of the state (Sa˜ o Paulo Sugarcane Agroindustry Union 2003, Rudorff et al. 2004), as shown in Fig. 2. Since 1990, the expansion of sugarcane in Sa˜ o Paulo averaged ;85 000 ha/yr. The Turvo River basin (area 11 000 km2) is a typical case, where the land cover of sugarcane increased from 7% to 26% between 1997 and 2007, while the area of pasture decreased from 53% to 32% (Silva et al. 2007). solar radiation. Moreover, for .30 years, the country has invested in improving the production of ethanol from sugarcane, reaching an estimated 19 billion liters of ethanol in 2007. This production is similar to that of corn ethanol in the USA. Despite such advantages, the environmental sustain- ability and economic fairness of ethanol production in Brazil are issues that still need to be carefully debated in the scientific, political, and civic communities before sugarcane ethanol can be considered a ‘‘clean’’ fuel. For one thing, unrestrained use of natural resources and consequent excessive environmental degradation related to the expansion of sugarcane in Brazil may jeopardize important services provided by natural ecosystems, which are already experiencing a large degree of degradation worldwide (Millennium Ecosystem Assess- ment 2005); curbing such environmental degradation (e.g., deforestation) will also help to prevent further accumulation of CO2 in the atmosphere. In addition, the exploitation of cane workers for the benefit of the ethanol industry, without any significant return to Brazilian society in terms of investments in education, health, and infrastructure, is also an issue. Presently, only ;8% and 9% of the sugarcane land cover in Brazil is located in the South and Center-West regions, respectively (Fig. 3). In the North, which is mostly Amazon rain forest, the area planted with sugarcane is still only ;0.4% (;21 000 ha) of the total. Expansion to the North will be restricted by physiolog- ical characteristics of the varieties of sugarcane planted in Brazil. Sugarcane typically needs a period of drought during its growing phase to concentrate sugars, and such a drought period does not occur in the Amazon region (Fig. 2). Another restriction has to do with a new Brazilian law being developed that prohibits sugarcane planting in ecologically sensitive areas, like the Amazon and Pantanal regions. In contrast, the Center-West region of the country (Fig. 2), where half of the soybean production in Brazil occurs, provides ideal climatic and topographic conditions for sugarcane. Depending on land and market prices, soybeans in the Center-West can be replaced by sugarcane. This pressures soybean farmers to move farther north toward the Amazon. Indirectly, sugarcane expansion could lead to additional deforestation in that region. Yet the area planted with soybeans in the Center-West region (10 million ha) is ;20 times larger than that of sugarcane; this scenario, therefore, is unlikely to occur in the near future. The main objectives of this study are (1) to discuss environmental and social issues linked to the expansion of sugarcane in Brazil for ethanol production, and (2) to provide recommendations to help policy makers and the Brazilian government establish new initiatives to pro- duce a code for ethanol production that is environmen- tally sustainable and economically fair.PATTERNS OF SUGARCANE EXPANSION IN BRAZIL Despite the fact that sugarcane expansion is not causing any significant decline in the land cover of natural vegetation, there are other important environmental problems associated with the cultivation of sugarcane in Brazil that need to be addressed to avoid the perpetuation of these problems as the ethanol industry expands. From 1960 to 2007, the area planted with sugarcane in Brazil increased from ;1.4 million to 7 million ha. The increase rate averaged almost 120 000 ha/yr (FAOSTAT 2007) (Fig. 1), except for the period between 1985 and 1990, when the rate was slower. Accompanied by the expansion of sugarcane land cover, the productivity of sugarcane also increased dramatically from 45 to 75 Mg/ha. This increase in productivity, ;600 kgha1yr1, was due to better agricultural techniques and an important genetic breeding program promoted by theENVIRONMENTAL ISSUES Soil degradation Among the major problems linked to sugarcane cultivation is soil degradation caused by erosion and BRAZIL’S ETHANOL PRODUCTION CHALLENGES June 2008 compaction. Soil erosion tends to be high in sugarcane fields (Fiorio et al. 2000, Politano and Pissarra 2005) in comparison to pastures and forests because of extensive areas of bare soil that are associated with the management practices used. Bare soils are exposed to intense rain and winds, both during the initial process of land use conversion when grasses are killed to prepare for the planting of sugarcane, then again in the period between crop harvesting and regrowth. When sugarcane stalks are replaced with new ones every 5–6 years, soils remain bare for several months. Soil compaction results from the constant traffic of heavy agricultural machinery associated with cultivation and harvesting operations in sugarcane fields. The compaction destroys soil physical properties such as porosity and density, which in turn decreases water infiltration and further enhances soil erosion (Cerri et al. 1991, Oliveira et al. 1995, Silva and Ribeiro 1997, Silva Ecological Applications Vol. 18, No. 4 LUIZ A. MARTINELLI AND SOLANGE FILOSO et al. 1998, Ceddia et al. 1999, Fiorio et al. 2000, Prado and Centurion 2001). waste are generated in the mills. The two most important are wastewater generated from the washing of sugarcane stems before they go through the mill, and the vinasse produced during the distillation process. Both of these waste products are rich in organic matter, and therefore increase the BOD (biochemical oxygen demand) of waters receiving these effluents. Elevated BOD promotes depletion of dissolved oxygen in the water and often causes anoxia (Ballester et al. 1999). High nutrient concentrations in these effluents also contribute to the problem by enhancing algal blooms and promoting eutrophication of surface waters (Mat- sumura-Tundisi and Tundisi 2005). Sparovek and Schnug (2001) estimated erosion rates up to 30 Mg of soilha1yr1 for sugarcane fields in Sa˜ o Paulo state (Fig. 2), while rates in forests and pastures did not exceed 2 Mgha1yr1. In one particular region of Sa˜ o Paulo, which has been intensively and extensively cultivated with sugarcane for many decades, Politano and Pissarra (2005) found that erosion varied from severe to extremely severe.Deterioration of aquatic systems The negative effects of accelerated soil erosion include not only soil degradation and subsequent poor crop development, but also deterioration of aquatic systems. As colluvium sediments are transported downhill across the landscape from sugarcane fields, they are deposited onto wetlands, small streams, rivers, and reservoirs. Deposition impacts water quality, ecosystem biodiversity (Politano and Pissarra 2005), and ecosystem functions. For each liter of ethanol that is produced from sugarcane, 12–13 liters of vinasse are generated. With the boom of ethanol production in Brazil in the early 1980s, new legislation was created to ban the direct discharge of vinasse onto surface waters. Since then, nutrient and carbon-rich vinasse is mixed with waste- water from washing sugarcane and is recycled back to sugarcane fields as organic fertilizer (Gunkel et al. 2007). This solution has helped to protect aquatic systems to a certain extent. However, it is not uncommon for small mills to discharge vinasse into streams and rivers because they lack the means of transport and applica- tion. Also, accidents or mishandling during storage and transport of vinasse are not uncommon, even in mills with adequate infrastructure. An example that illustrates the problem of sedimen- tation in aquatic systems linked to sugarcane cultivation was described by Fiorio et al. (2000) for a small watershed in Piracicaba County, state of Sa˜ o Paulo (Fig. 2). This watershed had 25% sugarcane land cover in 1978 when a reservoir was built to supply water for a small town nearby. About 20 years later, almost 70% of the watershed was covered by sugarcane crops, and the reservoir could no longer be used as a water supply because of sedimentation and loss of 50% of its water- holding capacity. In a country like Brazil, where most of the water supply for cities and rural areas and most of the electrical power generated are from dammed rivers and reservoirs, sedimentation of aquatic systems can have serious consequences. In a monitoring study conducted on a small stream adjacent to a sugarcane mill in Piracicaba County, Ometto et al. (2000) reported significant changes in water quality along a 1.8-km reach downstream of a sugarcane mill (Fig. 4). Clear increases in water temperature, electrical conductivity, dissolved organic carbon (DOC), and dissolved inorganic nitrogen (DIN) were observed downstream from the mill. Moreover, concentrations of dissolved oxygen were significantly lower downstream, while dissolved inorganic carbon (DIC) was higher. Gunkel et al. (2007) reported similar changes along a reach of the Ipojuca River in the northeastern region of Brazil. The severity of the problem of sedimentation is aggravated even further by the transport of contami- nants such as pesticides and heavy metals used in sugarcane cultivation to aquatic systems. For instance, organochlorides were found in sediment and fish samples collected in the Piracicaba River basin in 1997 (Silva et al., in press), despite the fact that the use of this product was forbidden in Brazil in 1985. Similarly, organochlorides were found in samples collected in 2003 in streams that drain a sugarcane region in the central portion of Sa˜ o Paulo state (Corbi et al. 2006). This suggests that these compounds are still being used by farmers because the pesticides have a short half-life in the environment. (Silva et al., in press). Other contam- inants such as atrazine, a herbicide used in sugarcane crops, and heavy metals such as copper (Cu), were also found in water samples and stream bed sediments collected in rivers draining regions that have extensive sugarcane cultivation (Carvalho et al. 1999, Azevedo et al. 2004, Corbi et al. 2006). Martinelli et al. (1999a) used stable isotope techniques to determine the origin of organic matter in rivers draining watersheds that were predominantly covered by sugarcane and found that, in the samples collected, carbon originated from sugarcane. Sugarcane is a C4 plant, which can be easily distinguished from C3 plants (e.g., trees) because of a distinctive isotopic signature (expressed as d13C). Therefore, these results suggested that either the discharge of vinasse into streams is substantial in the region, or that accelerated erosion in sugarcane fields transports organic materials to the water (Martinelli et al. 1999a, 2004). A combination of both scenarios is likely.Nitrogen pollution The industrial processing of sugarcane for production of ethanol and sugar is yet another source of pollution for aquatic systems, as large amounts of byproducts and Like most annual crops, sugarcane requires the application of fertilizer to support an economically BRAZIL’S ETHANOL PRODUCTION CHALLENGES June 2008 and ;3.13 million tons of fertilizers to sugarcane fields in Brazil in 2006. Therefore, much of this boost in the use of fertilizer in Brazil is attributed to the expansion and intensification of sugarcane production. viable production. In contrast to developed countries, over-fertilization and consequent losses of excess nutri- ents to aquatic systems have not been a major environmental problem in Brazil. However, the recent expansion of agriculture in the country coincides with a rapid increase in the consumption of fertilizer (ANDA 2006). The single largest annual crop in Brazil is soybeans, covering ;23 million ha of the 64 million ha of arable land in the country. Maize and sugarcane are grown on 13 and 7 million ha, respectively. While soybeans do not require application of N fertilizer, ;3.25 million tons of fertilizers (including nitrogen, phosphorus, and potassium) were applied to cornfields As more fertilizer is used in sugarcane crops, excess nutrients are likely to accumulate in the environment. Because of the high mobility of N, much of the excess is transported to aquatic systems. Excess nitrogen in agricultural fields has increased the export of N in streams and rivers worldwide and is one of the main causes of eutrophication of coastal waters and estuaries (Howarth 2005). In Brazil, eutrophication of dams and reservoirs are also related to increased inputs of reactive Ecological Applications Vol. 18, No. 4 LUIZ A. MARTINELLI AND SOLANGE FILOSO N to landscapes (Matsumura-Tundisi and Tundisi 2005). Large losses of N from sugarcane fields occur not only because of the relatively low use efficiency of N fertilizer by the plant, but also via ammonia volatilization during plant senescence due to the high evapotranspiration rates of sugarcane, which is characteristic of C4 plants (Trivelin et al. 2002, Costa et al. 2003, Gava et al. 2005). Senescence is responsible for the emission of ;80 kg of N/ha (Trivelin et al. 2002), while volatilization from soils fertilized with N amount to 30–40% of the fertilizer applied, or 30–40 kg of N/ha, assuming a fertilizer application rate of 100 kg N/ha. Therefore, total gaseous losses from volatilization in sugarcane crops are an estimated 110 kg (80 þ 30 kg) to 120 kg (80 þ 40 kg) of N/ha annually. Although the soil organic matter reservoir is an additional source of reactive N, biological fixation may supply an important amount of reactive N. In Brazil, several varieties of sugarcane are able to fix N biologically in symbiosis with endophytic bacteria located in roots and foliage tissues (Boddey et al. 2001). Depending on environmental conditions, fixation by Brazilian sugarcane varieties is a maximum of 150 kgha1yr1, with in situ rates averaging between 30 and 50 kgha1yr1 (S. Urquiaga, personal communication). Nitrogen fertilizer is applied to sugarcane crops in Brazil at a rate of 80–100 kgha1yr1. When compared to average amounts applied to annual crops in developed countries like the United States (140–160 kg Nha1yr1) and The Netherlands (300 kg Nha1yr1), and also in comparison to other important crops in Brazil like coffee and citrus, application in sugarcane fields is not considered excessive. Yet significant amounts of N from fertilizer can be lost to the environment in moist tropical and subtropical regions if application is poorly timed (Harmand et al. 2007). In addition, most of the N fertilizer in Brazil is applied in the form of urea (CO(NH2)2) (Cantarella 1998), and can easily be lost through volatilization of ammonia (NH3) or during nitrification. Volatilization from soils occurs when urea is transformed into ammonium carbonate through the action of urease, an enzyme that catalyzes urea hydrolysis. Under specific conditions (high temper- atures and high pH), ammonium carbonate is trans- formed into NH3 and emitted to the atmosphere. Emission of N2O occurs when NH3 is oxidized into NH4 þ, which is then nitrified into NO2  and NO3 . Ammonia emitted during volatilization to the atmo- sphere is likely to be redeposited in the region in dry atmospheric deposition, and ammonium (NH4) depos- ited regionally in wet deposition (Holland et al. 1999). In addition, most of the N from fertilizer absorbed by sugarcane and transported to ethanol production mills is likely to be recycled back into the fields. Therefore, while the addition of fertilizer N to sugarcane crops is relatively low in comparison to that of other crops, very little is likely to be exported from regions where sugarcane is cultivated. Hence, it is possible that sugarcane cultivation in Brazil may lead to an excess of N in soils, and subsequent higher delivery rates to aquatic systems. High rates of N export into rivers draining watersheds heavily cultivated with sugarcane in Brazil, such as the Piracicaba and Mogi river basins, have been reported (Filoso et al. 2003). Several studies have shown that the use efficiency of fertilizer N by sugarcane crops in Brazil is rather low. On average, only ;20–40% of the N applied to sugarcane is assimilated in plant tissues, including roots, stalk, leaves, and tip (Oliveira et al. 2000, Trivelin et al. 2002, Basanta et al. 2003, Gava et al. 2005). Therefore, for every 100 kg N/ha applied to sugarcane fields, between 60 and 80 kg N/ha stays in the soil. Depending on soil and climate conditions, different fractions of this N in the soil are lost via volatilization, denitrification, erosion, and surficial water runoff, or assimilated later by sugarcane ratoons that will sprout in the next crop season (Trivelin et al. 1995, 1996, 2002, Cantarella 1998, Oliveira et al. 2000, Gava et al. 2001, 2005, Basanta et al. 2003). Leaching to groundwater also occurs, and ranges from 5 to 15 kg Nha1yr1, depending on soil characteristics (Oliveira et al. 2002). Besides higher delivery rates to aquatic systems, high inputs of reactive N to the landscape can accelerate the N cycle in sugarcane regions, and consequently, emissions of nitrous oxide (N2O) associated with the processes of nitrification and denitrification. Nitrous oxide is a potent greenhouse gas which is ;300 times stronger than CO2. Therefore, emissions of N2O by sugarcane fields is an important trade-off in the global warming scenario, and must be taken into account in the whole balance of the sugarcane crop as renewable fuel (Crutzen et al. 2007). Tracer experiments used to determine the fate of fertilizer 15N within sugarcane revealed that 10% or 20% of the total fertilizer N applied to crops is assimilated in the stalks, which is the part actually used in the industry, while another 10–20% ends up in other aerial parts of the plant such as the tips and straw, which are either burned or decomposed in the field (Oliveira et al. 2000, Gava et al. 2001, 2005, Basanta et al. 2003). Based on these values, an estimated 10–20 kg N/ha of the fertilizer N applied to crops is transported to the mills in the sugarcane stalks for production of ethanol and sugar, while the remainder stays in the field or is emitted to the atmosphere. Eventually, even the relatively small amount of N transported to the mills is recycled back into the fields, because byproducts from sugarcane processing such as vinasse are used to fertilize and irrigate crops.Destruction of riparian ecosystems The movement of solutes and eroded soils from the uplands to surface waters can be controlled by riparian forests that usually occupy a narrow belt of land along streams and rivers. When riparian forests are removed, the detrimental impacts of sugarcane cultivation and BRAZIL’S ETHANOL PRODUCTION CHALLENGES June 2008 which has been reported as the cause of seven deaths in the region. ethanol production on aquatic systems are exacerbated by degrading water quality, decreasing biodiversity, and increasing sedimentation (Corbi et al. 2006; see Plate 1). According to Brazilian legislation, the riparian vegeta- tion on both margins of a stream or river must be preserved at a width that is relative to the channel width, with a minimum of 10 m for streams, and a maximum of 500 m for rivers. However, assuming a generalized buffer width of 30 m for both streams and rivers, Silva et al. (2007) reported that only 25% (500 km2) of riparian forests remain in the seven major agricultural water- sheds in the state of Sa˜ o Paulo. The other 75% (4500 km2) of the riparian zone was converted to sugarcane and pasture. Environmental consequences of sugarcane burning Sugarcane burning is a common crop management practice in Brazil, as it is used to facilitate manual harvesting by burning most of the straw and leaves. Generally, sugarcane is burned during the night, between April and December. An estimated 2.5 million hectares, or 70% of the sugarcane area, was burned in 2006 in the state of Sao Paulo (Folha de Sa˜ o Paulo, 11 August 2007, page A21 [available online]).4 If we assume this same percentage area for other sugarcane regions in Brazil, we estimate a total of ;4.9 million hectares were burned in Brazil because of sugarcane. In the two major watersheds that had the largest percentage cover of sugarcane in 1997 (Piracicaba and Mogi), only 13–18% of the riparian vegetation was preserved (Silva et al. 2007). The approximate cost for restoration of riparian forest in the region was estimated at US$3500/ha (R. R. Rodrigues, personal communica- tion). Hence, restoring riparian vegetation in these basins would cost between US$200 and 250 million for both basins. Although costly, the price for restoration was equivalent to only 6% and 8% of the gross domestic product (GDP) from agricultural products and indus- tries in the Piracicaba and Mogi basins, respectively. Sugarcane burning increases soil temperature, de- creases soil water content and bulk density and, consequently, leads to soil compaction, higher surface water runoff, and soil erosion (Dourado-Neto et al. 1999, Oliveira et al. 2000, Tominaga et al. 2002). Additionally, Pereira-Netto et al. (2004) detected high concentrations of polycyclic aromatic hydrocarbons (PAH) in soils located near sugarcane burning areas. Soils contaminated with these compounds, which are often carcinogenic, represent a risk for human health when leached to water bodies. In one of the most important sugarcane regions of the state of Rio de Janeiro (Southeast region), PAHs were detected in recently deposited sediments in lakes (Gomes and Azevedo 2003). Because PAHs were also found in the atmosphere in this region (Azevedo et al. 2002, Santos et al. 2002), sugarcane burning was determined as the most likely source of PAHs in soils and sediments. Atmospheric PAHs were found in the region of Araraquara, one of the important sugarcane areas of the state of Sa˜ o Paulo (Zamperlini et al. 1997, Godoi et al. 2004), and one of the compounds, benzo[a]pyrene, which has high carcinogenic properties, was found in higher concentrations than in the atmosphere of large cities. Riparian forests are complex ecological systems localized at the land–water ecotone that perform a disparate number of ecological functions compared to most upland habitats (Naiman et al. 2005). These ecosystems typically maintain high levels of biodiversity (Naiman and De´ camps 1997). Therefore, the ecological consequences of alterations to riparian forests include not only obvious changes in the plant community, but also in the animal community. In sugarcane areas in Brazil, reported effects of degradation of riparian forests on animal communities range from increased numbers of fish species (Gerhard 2005) to decreased small-mammal species richness (Dotta 2005, Gerhard 2005, Gheler-Costa 2006, Barros Ferraz et al. 2007). The widespread increase of generalist species (which have the ability to live in many different places while tolerating a wide range of environmental conditions) that are typical of degraded areas (Gheler- Costa 2006), such as capybaras (Hydrochoerus hydro- chaeris), and small rodents such as Cerdocyon thous and Lepus europaeus (Dotta 2005), is common. Moreover, 18 species of medium- to large-sized mammals typical of more preserved and larger forested fragments of the state of Sa˜ o Paulo could not be found in riparian forest fragments of sugarcane-dominated watersheds (Dotta 2005). Biomass burning has been recognized as one of the main sources of pollution from aerosol particles, especially in tropical regions of the world (Crutzen and Andreae 1990). Aerosol particles are not only important in the radiative budget of the atmosphere (Oglesby et al. 1999, Ramanathan et al. 2001, Streets et al. 2001), but also affect the concentration of cloud condensation nuclei, which affect cloud albedo and rain droplet formation (Roberts et al. 2001). The average concentration of total suspended aerosol particles (particulate matter 10 lm) collected in Piracicaba county during the sugarcane burning season was significantly higher (91 lg/m3) than the average in the nonburning season (34 lg/m3) (Lara et al. 2005). Principal component analysis (Lara et al. 2005) and Besides the alteration in community structure, the widespread occurrence of generalist species of animals in sugarcane areas has been associated with public health problems. For instance, the increase of the capybara population in the Piracicaba River basin has lead to the spread of Brazilian spotted fever (Labruna et al. 2004), 4 hhttp://www1.folha.uol.com.br/folha/ciencia/ ult306u319319.shtmli LUIZ A. MARTINELLI AND SOLANGE FILOSO Ecological Applications Vol. 18, No. 4 stable carbon isotopic composition of aerosol particles (Martinelli et al. 2002) revealed that sugarcane burning was the main source of particles during the burning season, while soil particles were the main source during the nonburning season. The annual average for sus- pended aerosol particles established by Brazilian law is 70 lg/m3, but concentrations found in the Piracicaba region were higher during the burning season. Similar concentrations (70–90 lg/m3) were reported for the city of Sa˜ o Paulo during the winter (Castanho and Artaxo 2001), and for the state of Rondoˆ nia, in the Amazon region, where extensive areas of forest are burned every year (Artaxo et al. 2002). The concentrations found in the Piracicaba region were also similar to those found during the burning season in Araraquara, another sugarcane region of Sa˜ o Paulo. In both regions, peak concentrations reached 240 lg/m3 during the burning season (Allen et al. 2004). Black carbon concentrations were also significantly higher during the burning season, and days with burnings had higher black carbon concentration than days without burnings (Lara et al. 2005). further acidification and impoverishment of this type of soil (Matson et al. 1999, Krusche et al. 2003). The end of this process is the replacement of Hþ by Alþ3, which, in high concentrations, is toxic to plants (Schulze 1989).PUBLIC HEALTH AND SOCIAL ISSUESPUBLIC HEALTH AND SOCIAL ISSUES Sugarcane burning and respiratory diseases In addition to the negative environmental effects, sugarcane burning also affects the health of people living in areas where burning is intense (Arbex et al. 2000, 2007). Epidemiological studies conducted in two coun- ties in the state of Sa˜ o Paulo (Araraquara and Piracicaba), which are surrounded by sugarcane fields, show that respiratory morbidity increased significantly with the concentration of aerosol particles from sugarcane burning (Arbex et al. 2000, 2007, Canc¸ ado et al. 2006). During the sugarcane burning season of 1995 in Araraquara, a study found a significant correlation between the daily number of patients who visited hospitals in the region for inhalation treatment for respiratory diseases, and the mass of particle aerosols (Arbex et al. 2000, 2007). In a second study, conducted in the Piracicaba region, Canc¸ ado et al. (2006) found a significant correlation between PM2.5 (particulate matter 2.5lm), PM10 (particulate matter 10 lm), and black carbon concentrations, and the number of children and elderly patients admitted to hospitals. According to their results, increases of 10 lg/m3 of the PM2.5 concentration lead to an increase of ;20% in the number of hospital admissions. Gas emissions to the atmosphere are another aspect of atmospheric pollution associated with sugarcane burn- ing. In sugarcane areas, concentrations of CO and O3 are commonly high (Kirchoff et al. 1991), while nitrogen oxides (NOx) emissions reached 25 kg N/ha (Oppen- heimer et al. 2004). Part of the N lost to the atmosphere during burning of sugarcane fields returns to the Earth’s surface via wet and dry deposition. Rates of wet atmospheric deposition in the Piracicaba River basin were 5 and 6 kg Nha1yr1 (Lara et al. 2001), which are similar to those found in a region of intense forest fires in the Amazon, and also in areas of the northeastern United States where emissions of nitrogen from the combustion of fossil fuels by vehicles and from power plants are considered high. In comparison to pristine areas of Sa˜ o Paulo state and the Amazon region, N deposition rates in the Piracicaba basin are 3–4 times higher (Almeida 2006). As nitrogen oxides react with water to form nitric acid (HNO3), acid rain becomes another problem associated with burning of sugarcane. In the Piracicaba basin where sulphur emissions are low, the average annual pH of rain varies between 4.5 and 4.8. Therefore, most of the acidity found in the rain comes from the formation of nitric acid (HNO3) (Lara et al. 2001). Arbex (2003) concluded that sugarcane burning is responsible for aggravating the health of people prone to respiratory diseases, which, in turn, increases the demand and expenditure in the public health system. Thus the burning of sugarcane affects several sectors of society, and has negative impacts even for people living outside of sugarcane-ethanol industry areas. Yet sugar- cane burning continues to be a widely used land management practice in Brazil, despite efforts by state and federal governments to eradicate it. For instance, Law 11,241, established by the state of Sa˜ o Paulo in September 2002, requires that by 2006 only 30% of areas with slope lower than 12%, could be burned for sugarcane harvesting. The same law states that by 2011 half of the area in sugarcane farms must be protected from burning, and burning would be entirely outlawed by 2021. The Sa˜ o Paulo state government is trying to convince sugarcane growers to stop sugarcane burnings in 2014. Acidification of surface waters has been one of the most prominent problems associated with emissions of acidifying compounds such as NOx to the atmosphere in the temperate zone. In contrast, high buffer capacity of streams and rivers of Sa˜ o Paulo, especially where untreated domestic sewage discharges are high (Ballester et al. 1999, Martinelli et al. 1999b, Krusche et al. 2003), have protected these systems from acidification. How- ever, as the already basic cation-poor tropical soils in Brazil receive continuous acid rain, Hþ may replace the few remaining cations in the clay surface, causingExploitation of cane cutters Sugarcane brought by the Portuguese from Africa was one of the first crops cultivated in Brazil. In the beginning, the Portuguese tried forcibly to make Native Brazilians work in the sugarcane fields and mills, but after failing, they started bringing African people to work as slaves. Although slavery was abolished in Brazil in 1888, most of the sugarcane in Brazil today is still BRAZIL’S ETHANOL PRODUCTION CHALLENGES June 2008 TABLE 1. Name, age, date, and cause of death for sugar cane cutters in counties within the state of Sa˜ o Paulo, Brazil, since 2004. Name Age Date of death Cause County Galva˜ o, J. E. 38 April 2004 cardiac arrest Macatuba Santos, M. A. 33 April 2004 cardiac arrest Valparaı´so Pina, M. N. 34 May 2004 cardiac arrest Catanduva Pinto, L. R. 27 March 2005 respiratory arrest Terra Roxa Santos, I. V. 33 June 2005 acute pancreatitis Prado´ polis Lima, V. P. 38 July 2005 cerebrovascular accident Ribeira˜ o Preto Sales, J. N. G. 50 August 2005 respiratory arrest Batatais Diniz, D. 55 Sept. 2005 unknown Borborema Souza, V. A. 43 Oct. 2005 unknown Valparaı´so Gomes, J. M. 45 Oct. 2005 unknown Rio das Pedras Lopes, A. R. 55 Nov. 2005 pulmonary edema Guariba Santana, J. 37 June 2006 unknown Jaborandi Borges, M. N. 54 July 2006 unknown Monte Alto Gonc¸ alves, C. 41 July 2006 unknown Taiac¸ u´ Almeida, O. 48 Sept. 2006 unknown Itapira Martins, J. P. 51 March 2007 cardiac arrest Guariba Souza, L. P. 20 April 2007 unknown Colina Souza, J. D. 33 June 2007 unknown Ipaussu Source: Ineˆ s Fascioli and Garcia Peres, 14 May 2007 (see footnote 5).TABLE 1. Name, age, date, and cause of death for sugar cane cutters in counties within the state of Sa˜ o Paulo, Brazil, since 2004. harvested manually, and the conditions for laborers have not improved much from over a century ago (Rodrigues 2006). to transport 15 kg of stalks at a time to a distance varying from 1.5 to 3 m. Although the cane cutters are paid by the mass of sugarcane stalks cut per day, the mass is an estimate based on the linear distance that a cutter covered in one day. Employers convert the linear distance into an estimated area by multiplying the distance by a swath of 6 m. For instance, for every 1000 m covered by the cane cutter, the calculated area is 6000 m2. This area is then converted into sugarcane mass by the industry according to numbers obtained from randomly selected areas where the sugarcane was actually cut and weighed in the field. The workday consists of 8 to 12 hours of cutting and carrying sugarcane stalks, while inhaling dust and smoke from the burned residue (Rodrigues 2006). In addition, working conditions such as clean water, restrooms, and food storage facilities are usually absent in sugarcane fields. These poor working conditions often result in lawsuits against sugarcane employers by the Brazilian Ministry of Labor (Luze de Azevedo, ‘‘Forc¸ a tarefa autua usinas na regia˜ o de Prudente,’’ 19 July 2007; Davi Venturino, ‘‘Juı´za garante condic¸ ˜oes para bo´ ia-fria,’’ 4 August 2007; available online).5 As many cane cutters are short-term migrants from other regions of Brazil, they have no alternative but to reside in inadequate lodging during cane harvest (Costa and das Neves 2005) (Luze Azevedo, ‘‘Alojamentos na mira do Ministe´ rio do Trabalho,’’ 3 August 2007; see footnote 5). The problem with this method is that the mass of sugarcane can vary widely temporally and spatially depending on local conditions. The weighing is also done by the industry, and does not include any oversight by the field workers. The result is that cane cutters are usually underpaid (Alves et al. 2006). To compensate for low wages, cutters often try to maximize their daily earnings by working long hours, even under inappro- priate conditions. Consequently, between 2004 and 2007, several deaths of cane cutters in sugarcane fields were reported in the state of Sa˜ o Paulo (Table 1). All of the deaths were of people younger than 55 years old, and most of them (67%) were younger than 45. Although the reported cause of death was unknown in half of the cases (Table 1), for the other half, cause of death could be linked to high working loads and poor working conditions (Scopinho et al. 1999). g ; ) Manual sugarcane harvesting in Brazil includes first the burning of dry leaves and other plant residues such as the tips to facilitate harvesting, and then the cutting of stalks as close to the ground as possible. Once cut, the stalks are transported manually to a depository, and the laborer is paid by the mass of sugarcane cut per day. According to Alves (2006), a cutter in 1950 cut an average of 3000 kg of sugarcane per day, while 30 years later the average was ;6000 kg/d. More recently, numbers have increased to almost 12 000 kg/d, but in the most intensively cultivated sugarcane regions in Sao Paulo, the average ranges between 7000 and 8000 kg/d (EPTV, ‘‘Pesquisa constata aumento no corte de cana,’’ 18 May 2007 (see footnote 5). Alves (2006) estimated that for every 6000 kg of stalks cut, a cutter swings his machete ;70 000 times and walks a distance of 4.5 km The average price paid for 1000 kg of sugarcane stalks cut is equivalent to ;US$1.2. Therefore, for an average yield of 10 000 kg/d, the cane cutter earns US$12. On a monthly basis, and assuming 24 workdays per month, a cutter makes ;US$300 (Jose´ Maria Tomazela, ‘‘Cana- vieiro e´ o piro servic¸ o do mundo,’’ 31 March 2007; (see footnote 5). Considering that the total cost of produc- 5 hwww.pastoraldomigrante.org.bri Ecological Applications Vol. 18, No. 4 LUIZ A. MARTINELLI AND SOLANGE FILOSO tion of ethanol in Brazil was ;US$1.10 per gallon (1 gallon ¼ 3.78 L) during the 2005 crop year, with variable costs of US$0.89 per gallon and fixed costs of US$0.21 per gallon, and that in early 2006, the wholesale price paid to the mills for anhydrous ethanol was US$2.05 per gallon (Martines-Filho et al. 2006), revenue values were in the order of 50% of the costs. However, it is clear that such high return rates are not being reflected in the salary and working conditions of cane cutters. industries in the sugarcane business, Cosan, signed an agreement under the Brazilian Law called ‘‘adjustment of conduct,’’ promising sugarcane cutters better working and living conditions. In addition, this agreement assures the proper recruiting and hiring of sugarcane cutters to eliminate the work of ‘‘gatos’’ and the problems associated with them. As real employees, sugarcane cutters will also have the same rights and benefits as other categories of workers in the industry. The typical sugarcane worker in Brazil is a migrant from a poor area of the northeastern region of the country who moves to the Southeast to work for 6–8 months as a sugarcane cutter. These migrants are usually hired and transported to sugarcane plantations by people known in Brazil as ‘‘gatos.’’ The cane cutters are hired under false promises of a good salary and decent living conditions (Costa and das Neves 2005). In such situations, Brazilian labor legislation is entirely disregarded.CONCLUDING REMARKS AND RECOMMENDATIONS Poets—and city folks—love to romanticize agricul- ture, portraying it as some sort of idyllic state of harmony between humankind and nature. How far this is from the truth! Since Neolithic man—or most probably woman—domesticated the major crop and animal species some 10–12 millennia ago, agriculture has been a struggle between the forces of natural biodiversity and the need to produce food under increasingly intensive production systems. However, increasing pressure from authorities in Brazil and from the international community is making the ethanol industry recognize that it is not possible to claim that ethanol is a clean or sustainable fuel, when laborers are still working in extremely poor conditions, are severely underpaid, and worst yet, often die of exhaustion in the fields during sugarcane harvesting (Scopinho et al. 1999). Recently, one of the major —Norman Bolaug These words, written by the father of the Green Revolution (Borlaug 2002), eloquently say that agricul- ture is in constant conflict with the environment. Yet why is this conflict usually downplayed when it comes to sugarcane and ethanol production in Brazil, while it is debated for other cultures that also cover extensive areas BRAZIL’S ETHANOL PRODUCTION CHALLENGES June 2008 in Brazil, such as maize (13 million ha) and soybeans (23 million ha)? produces ;500 billion liters of vinasse, along with all the other environmental problems discussed above, and summarized in Fig. 5. One of the reasons is that the sugarcane ethanol program in Brazil has been considered, by most Brazil- ians, a successful solution to the 1970s oil crisis, and for the problem of the country’s past dependence on foreign oil. More recently, ethanol production in Brazil started being seen as a potential solution to global warming problems, which would also benefit the Brazilian economy by creating new international trade opportu- nities. Therefore, generally speaking, the loss of natural resources associated with the expansion of sugarcane and increase of ethanol production in Brazil is commonly seen as necessary and justifiable. Still, the OECD-FAO Agricultural Outlook 2007–2016 (2007) estimated that in 2016 Brazil will produce ;44 billion liters of ethanol, an increase of 145% relative to 2006. Assuming that one hectare of sugarcane is necessary to produce 6000 liters of ethanol, Brazil would have to double the present land cover of sugarcane in the country, from 7 to 14 million hectares. Approximately 60% of this total area would be utilized for ethanol and 40% for sugar. Furthermore, 44 billion liters of ethanol The sugarcane industry in Brazil has enjoyed support and protection from politicians and lobbyists for centuries. Consequently, the industry has created a legacy of disregard for environmental and social laws in the country. Bold examples include the several occasions that enforcement of laws regulating or banning sugar- cane burning practices have been postponed, and civil suits regarding poor living conditions of sugarcane cutters have been ignored by the judicial system in Brazil. However, it is obvious that this template of environmental degradation and social injustice must go through major changes if the sugarcane-ethanol industry in Brazil is to be truly competitive in a market where biofuel production and use are expected to be sustain- able and socially correct. Such changes should include: (1) proper planning and environmental risk assessments for the expansion of sugarcane to new regions such as Central Brazil, (2) improvement of land use practices to reduce soil erosion and nitrogen pollution, (3) protection of streams and riparian ecosystems, (4) banning of Ecological Applications Vol. 18, No. 4 LUIZ A. MARTINELLI AND SOLANGE FILOSO sugarcane burning, and (5) fair working conditions for sugarcane cutters. It is important to recognize, never- theless, that a number of external factors, in addition to internal ones, contribute to the difficulties faced by Brazil and other developing countries in overcoming environmental and social issues associated with agro- business (Almeida et al. 2004). Hence, we support the idea that rather than creating more barriers to the trade of agricultural products from Brazil, a more constructive approach should be taken by international stakeholders and even the World Trade Organization (Almeida et al. 2004) to promote sustainable development in countries where agricultural expansion and biofuel production are likely to grow substantially in the next few years. In addition, we propose that environmental values be included in the price of biofuels from Brazil and elsewhere in order to discourage the excessive replace- ment of natural ecosystems such as forests, wetlands, and pasture by bio-energy crops. Artaxo, P., J. W. Martins, M. A. Yamasoe, A. S. Proco´ pio, T. M. Pauliquevis, M. O. Andrea, P. Guyon, L. V. Gatti, and A. M. C. Leal. 2002. Physical and chemical properties of aerosols in the wet and dry season in Rondoˆ nia, Amazonia. Journal of Geophysical Research 107:8081. Azevedo, D. A., E. Gerchon, and E. O. Reis. 2004. Monitoring of pesticides and polycyclic aromatic hydrocarbons in water from Paraı´ba do Sul River, Brazil. Journal of the Brazilian Chemical Society 15:292–299. Azevedo, D. A., C. Y. M. Santos, and F. R. Aquino Neto. 2002. Identification and seasonal variation of atmospheric organic pollutants in Campos Dos Goytacazes Brazil. Atmospheric Environment 36:2383–2395. Ballester, M. V., L. A. Martinelli, A. V. Krusche, R. L. Victoria, M. Bernardes, and P. B. de Camargo. 1999. Effects of increasing organic matter loading on the dissolved O2, free dissolved CO2 and respiration rates in the Piracicaba River Basin, Southeast Brazil. Water Research 33:2119–2129. Barros Ferraz, K. M. P. M., S. F. Barros Ferraz, J. R. Moreira, H. T. Z. Couto, and L. M. Verdade. 2007. Capybara (Hydrochoerus hydrochaeris) distribution in agroecosystems: a crossscale habitat analysis. Journal of Biogeography 34: 223–230. Basanta, M. V., D. Dourado-Neto, K. Reichardt, O. O. S. Bacchi, J. C. M. Oliveira, P. C. O. Trivelin, L. C. Timm, T. T. Tominaga, V. Correchel, F. A. M. Ca´ ssaro, L. F. Pires, and J. R. de Macedo. 2003. Management effects on nitrogen recovery in a sugarcane crop grown in Brazil. Geoderma 116: 235–248. If the same environmental and social problems linked with the sugarcane industry persist in Brazil into the future, most of the burden caused by these problems will be experienced by the whole society, especially those with lower incomes. On the other hand, the economic profits will be enjoyed by only a few, since Brazil has one of the largest economic inequalities in the world (CEPAL 2007). Boddey, R. M., J. C. Polidoro, A. S. Resende, J. R. Alves, and S. Urquiaga. 2001. Use of the 15N natural abundance technique for the quantification of the contribution of N2 fixation to sugar cane and other grasses. Australian Journal of Plant Physiology 28:889–895.LITERATURE CITED Borlaug, N. E. 2002. Feeding a world of 10 billion people: the miracle ahead. In Vitro Cellular and Development Biology— Plant 38:221–228. Allen, A. G., A. A. Cardoso, and G. O. Rocha. 2004. Influence of sugar cane burning on aerosol soluble ion composition in Southeastern Brazil. Atmospheric Environment 38:5025– 5038. Canc¸ ado, J. E., P. H. N. L. A. A. Saldiva Pereira, L. B. L. S. Lara, P. Artaxo, L. A. Martinelli, M. A. Arbex, A. Zanobetti, and A. L. F. Braga. 2006. The impact of sugar cane-burning emissions on the respiratory system of children and the elderly. Environmental Health Perspectives 114:725– 729. Almeida, L. T., M. F. Presser, and S. L. M. Ansanelli. 2004. Trade, environment and development: the Brazilian experi- ence, Working Group on Development and the Environment in the Americas. hhttp://ase.tufts.edu/gdae/Pubs/rp/ DP01TogeiroJuly04.pdfi Cantarella, H. 1998. Adubac¸ a˜ o nitrogenada em sistema de cana crua. STAB-Ac¸ u´ car, A´ lcool, e Subprodutos 16:21–22. Almeida, V. P. S. 2006. Acidez orgaˆ nica na precipitac¸ a˜ o e uso do solo no Estado de Sa˜ o Paulo: variabilidade espacial e temporal. Tese apresentada para obtenc¸ a˜ o do tı´tulo de Doutor em Cieˆ ncias. A´ rea de concentrac¸ a˜ o: Quı´mica na Agricultura e no Ambiente. Universidade de Sa˜ o Paulo, USP, Brazil. Carvalho, C. E. V., A. R. C. Ovalle, C. E. Rezende, M. M. Molisani, M. B. Saloma˜ o, and L. D. Lacerda. 1999. Seasonal variation of particulate heavy metals in the Lower Paraiba do Sul River, RJ, Brazil. Environmental Geology 37:297–302. Castanho, A. D., and P. Artaxo. 2001. Wintertime and summertime Sa˜ o Paulo aerosol source apportionment study. Atmospheric Environment 35:4889–4902. Alves, F. J. C. 2006. Por que morrem os Cortadores de Cana? Sau´ de e Sociedade 15:90–98. Ceddia, M. B., L. H. C. dos Anjos, E. Lima, A. Ravelli Neto, and L. A. da Silva. 1999. Sistemas de colheita da cana-de- ac¸ u´ car e alterac¸ ˜oes Nas propriedades fı´sicas de um solo podzo´ lico amarelo no Estado do Espı´rito Santo. Pesquisa Agropecua´ ria Brasileira 34:1467–1473. ANDA. 2006. Associacao Nacional para Difusao de Adubos. Statistics. hhttp://www.anda.org.br/estatisticas_2006.pdfi Arbex, M. A. 2003. Avaliac¸ a˜ o dos efeitos do material particulado proveniente da queima da plantac¸ a˜ o de cana- de-ac¸ u´ car sobre a morbidade respirato´ ria na populac¸ a˜ o de Araraquara (SP). Boletim de Pneumologia Paulista O´ rga˜ o Informativo da Sppt, Sa˜ o Paulo-SP 20:06–06. CEPAL (Comisio´ n Econo´ mica para America Latina y el Caribe). 2007. Panorama social da America Latina. CEPAL and ONU, Brasilia, Brazil. Arbex, M. A., G. M. Bo¨ hm, P. H. N. Saldiva, G. M. S. Conceic¸ a˜ o, A. C. Pope, and A. L. F. Braga. 2000. Assessment of the effects of sugar cane plantation burning on daily counts of inhalation therapy. Journal of the Air and Waste Management Association 50:1745–1749. Cerri, C. C., C. Feller, and A. Chauvel. 1991. Evoluc¸ a˜ o das principais propriedades de um Latossolo Vermelho-Escuro apo´ s desmatamento e cultivo por doze e cinqu¨ enta anos com d ´ C hi ORSTOM S ´ i P ´ d l i 26 37 50 cana-de-ac¸ u´ car. Cahiers ORSTOM Se´ rie Pe´ dologie 26:37–50. Corbi, J. J., S. T. Strixino, A. Santos, and M. Del Grande. 2006. Diagno´ stico ambiental de metais e organoclorados em co´ rregos adjacentes a a´ reas de cultivo de cana-de-ac¸ u´ car (Estado de Sa˜ o Paulo, Brasil). Quı´mica Nova 29:61–65. Arbex, M. A., L. C. Martins, R. C. Oliveira, A. Pereira, F. F. Arbex, J. E. Cancado, P. N. Saldiva, and A. L. F. Braga. 2007. Air pollution from biomass burning and asthma hospital admissions in a sugar cane plantation area in Brazil. Journal of Epidemiology and Community Health 61:395– 400. Estado de Sa˜ o Paulo, Brasil). Quı´mica Nova 29:61–65. Costa, C., and C. S. das Neves. 2005. Superexplorac¸ a˜ o do trabalho na lavoura de cana-de-ac¸ u´ car. Pages 81–87 in Relatoria Nacional para o Direito Humano ao Trabalho, BRAZIL’S ETHANOL PRODUCTION CHALLENGES June 2008 S. Solomon, D. Qin, M. Manning, Z. Chen, M. Marquis, K. B. Averyt, M. Tignor, and H. L. Miller, editors. Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge, UK. Comissa˜ o de Trabalho, de Administrac¸ a˜ o e Servic¸ o Pu´ blico, Brasilia, Brazil. Costa, M. C. G., G. C. Vitti, and H. Cantarella. 2003. Volatilizac¸ a˜ o de N-NH3 de fontes nitrogenadas em cana- de-ac¸ u´ car colhida sem despalha e fogo. Revista Brasileira de Cieˆ ncia do Solo, Vic¸ osa 27:631–637. Kirchoff, V. W. J. H., E. V. Marinho, P. L. S. Dias, E. B. Pereira, R. Calheiros, R. Andre´ , and C. Volpe. 1991. Enhancements of CO and O3 from burnings in sugar cane fields. Journal of Atmospheric Chemistry 12:87–102. Crutzen, P. J., A. R. Mosier, K. A. Smith, and W. Winiwarter. 2007. N2O release from agro-biofuel production negates global warming reduction by replacing fossil fuels. Atmo- spheric Chemistry and Physics Discussion 7:11191–11205. Krusche, A., P. B. Camargo, C. E. Cerri, M. V. Ballester, L. Lara, R. L. Victoria, and L. A. Martinelli. 2003. Acid rain and nitrogen deposition in a sub-tropical watershed (Piraci- caba): ecosystems consequences. Environmental Pollution 121:389–399. Dotta, G. 2005. Diversidade de mamı´feros de me´ dio e grande porte em func¸ a˜ o da paisagem na sub-bacia do rio Passa- Cinco, Sa˜ o Paulo. Thesis. Ecology of Agroecosystems, Universidade de Sa˜ o Paulo, USP, Brazil. Labruna, M. B., T. Whitworth, M. C. Horta, D. Bouyer, J. McBride, A. Pinter, V. Popov, S. M. Gennari, and D. H. Walker. 2004. Rickettsia species infecting Amblyomma cooperi ticks from an area in the state of Sao Paulo, Brazil, where Brazilian spotted fever is endemic. Journal of Clinical Microbiology 42(1):90–98. Dourado-Neto, D., C. Timm, J. C. M. Oliveira, K. Reichardt, O. O. S. Bacchi, T. T. Tominaga, and F. A. M. Ca´ ssaro. 1999. State-space approach for the analysis of soil water content and temperature in a sugarcane crop. Scientia Agricola 56:1215–1221. g FAOSTAT. 2007. Food and Agriculture Organization of the United Nations, Statistical Databases. hhttp://apps.fao.org/i Lara, L. B. S. L., P. Artaxo, L. A. Martinelli, R. L. Victoria, P. B. de Camargo, A. Krusche, G. P. Ayers, E. S. B. Ferraz, and M. V. Ballester. 2001. Chemical composition of rain water and anthropogenic influences in the Piracicaba River basin, southeast Brazil. Atmospheric Environment 35:4937– 4945. FAOSTAT. 2007. Food and Agriculture Organization of the United Nations, Statistical Databases. hhttp://apps.fao.org/i Filoso, S., L. A. Martinelli, M. R. Williams, L. B. Lara, A. Krusche, M. V. Ballester, R. L. Victoria, and P. B. Camargo. 2003. Land use and nitrogen export in the Piracicaba River Basin, Southeast Brazil. Biogeochemistry 65:275–294. Filoso, S., L. A. Martinelli, M. R. Williams, L. B. Lara, A. Krusche, M. V. Ballester, R. L. Victoria, and P. B. Camargo. 2003. Land use and nitrogen export in the Piracicaba River Basin, Southeast Brazil. Biogeochemistry 65:275–294. Lara, L. L., P. Artaxo, L. A. Martinelli, P. B. Camargo, R. L. Victoria, and E. S. B. Ferraz. 2005. Properties of aerosols from sugar-cane burning emissions in Southeastern Brazil. Atmospheric Environment 39:4627–4637. Fiorio, P. R., J. A. M. Dematteˆ , and G. Sparovek. 2000. Pesquisa Agropecua´ ria Brasileira 35:671–679. Gava, G. J. C., P. C. O. Trivelin, and C. P. Penatti. 2001. Crescimento e acu´ mulo de nitrogeˆ nio em cana-de-ac¸ u´ car cultivada em solo coberto com a palhada. Pesquisa Agro- pecua´ ria Brasileira 36:1347–1354. Martinelli, L. A., M. V. Ballester, A. Krusche, R. L. Victoria, P. B. de Camargo, M. Bernardes, and J. P. H. B. Ometto. 1999a. Land-cover changes and d13C composition of riverine particulate organic matter in the Piracicaba River Basin (southeast region of Brazil). Limnology and Oceanography 44:1826–1833. Gava, G. J. C., P. C. O. Trivelin, A. C. Vitti, and M. W. de Oliveira. 2005. Urea and sugarcane straw nitrogen balance in a soil-sugarcane crop system. Pesquisa Agropecua´ ria Brasi- leira 40:689–695. Martinelli, L. A., P. B. Camargo, L. B. L. S. Lara, R. L. Victoria, and P. Artaxo. 2002. Stable carbon and nitrogen isotopic composition of bulk aerosol particles in a C4 plant landscape of southeast Brazil. Atmospheric Environment 36: 2427–2432. Gerhard, P. 2005. Comunidades de peixes de riachos em func¸ a˜ o da paisagem da Bacia do Rio Corumbataı´, Estado de Sa˜ o Paulo. Dissertation. Ecology Agroecosystems, Escola Supe- rior de Agricultura Luiz de Queiroz, ESALQ, Brazil. Gheler-Costa, C. 2006. Distribuic¸ a˜ o e abundaˆ ncia de pequenos mamı´feros em relac¸ a˜ o a` estrutura da paisagem: a sub-bacia do Rio Passa-Cinco como Modelo. Dissertation. Ecology Agroecosystems. Escola Superior de Agricultura Luiz de Queiroz, ESALQ, Brazil. Martinelli, L. A., P. B. de Camargo, M. Bernardes, and J. P. Ometto. 2004. Carbon, nitrogen, and stable carbon compo- sition and land use changes in rivers of Brazil. Bulletin du Reseau Erosion 22:399–419. Godoi, A. F. L., K. Ravindra, R. H. M. Godoi, S. J. Andrade, M. Santiago-Silva, L. V. Vaeck, and R. V. Grieken. 2004. Fast chromatographic determination of polycyclic aromatic hydrocarbons in aerosol samples from sugar cane burnings. Journal of Chromatography A 1027:49–53. Martinelli, L. A., A. Krusche, R. L. Victoria, P. B. de Camargo, M. C. Bernardes, E. S. Ferraz, J. M. Moraes, and M. V. Ballester. 1999b. Effects of sewage on the chemical compo- sition of Piracicaba River, Brazil. Water, Air and Soil Pollution 110:67–79. Martines-Filho, J., H. L. Burnquist, and C. E. F. Vian. 2006. Bioenergy and the rise of sugarcane-based ethanol in Brazil. Choices 21:91–96. Gomes, A. O., and D. A. Azevedo. 2003. Aliphatic and aromatic hydrocarbons in tropical recent sediments of Campos dos Goytacazes, RJ, Brazil. Journal of Brazilian Chemistry Society 14:358–368. Matson, P. A., W. H. McDowell, A. Townsend, and P. M. Vitousek. 1999. The globalization of N deposition: ecosystem consequences in tropical environments. Biogeochemistry 46: 67–83. Gunkel, G., J. Kosmol, M. Sobral, H. Rohn, S. Montenegro, and J. Aureliano. 2007. Sugar cane industry as a source of water pollution—case study on the situation in Ipojuca River, Pernambuco, Brazil. Water, Air, and Soil Pollution 180:261–269. Matsumura-Tundisi, T., and J. G. Tundisi. 2005. Plankton richness in a eutrophic reservoir (Barra Bonita Reservoir, SP, Brazil). Hydrobiologia 542:367–378. Harmand, J. M., H. Avila, E. Dambrine, U. Skiba, S. de Miguel, R. V. Renderos, R. Oliver, F. Jimenez, and J. Beer. 2007. Nitrogen dynamics and soil nitrate retention in a Coffea arabica-Eucalyptus deglupta agroforestry system in Southern Costa Rica. Biogeochemistry 85:125–139. Millennium Ecosystem Assessment. 2005. Ecosystems and human well-being: synthesis. Island Press, Washington, D.C., USA. Naiman, R. J., and H. De´ camps. 1997. The ecology of interfaces: riparian zones. Annual Review of Ecology and Systematics 28:621–658. Holland, E. A., F. J. Dentener, B. H. Braswell, and J. M. Sulzman. 1999. Contemporary and pre-industrial global reactive nitrogen budgets. Biogeochemistry 46:7–43. Naiman, R. J., N. De´ camps, and M. E. McClain. 2005. Riparia: ecology, conservation, and management of streamside communities. Elsevier/Academic Press, Burlington, Massa- chusetts, USA. Howarth, R. W. 2005. The development of policy approaches for reducing nitrogen pollution to coastal waters of the USA. Science in China Series C—Life Sciences 48:791–806. IPCC. 2007. Climate Change 2007: the physical science basis. Contribution of Working Group I to the Fourth Assessment. OECD-FAO Agricultural Outlook 2007–2016. 2007. Organisa- tion for Economic Co-operation and Development, Food Ecological Applications Vol. 18, No. 4 LUIZ A. MARTINELLI AND SOLANGE FILOSO and Agriculture Organization of the United Nations, Rome, Italy. Santos, C. Y. M., D. A. Azevedo, and F. R. Aquino Neto. 2002. Selected organic compounds from biomass burning found in the atmospheric particulate matter over sugar cane plantation areas. Atmospheric Environment 36:3009–3019. Oglesby, R. J., S. Marshall, and J. A. Taylor. 1999. The climate effects of biomass burning: investigations with a global climate model. Environmental Modelling and Software 14: 253–259. Sa˜ o Paulo Sugarcane Agroindustry Union. 2003. UNICA, Statistics. hhttp://www.unica.com.br/i_pages/estatisticas. asp#i Oliveira, J. C. M. de, K. Reichardt, O. O. S. Bacchi, L. C. Timm, D. Dourado-Neto, P. C. O. Trivelin, T. T. Tominaga, R. C. Navarro, M. C. Piccolo, and F. A. M. Ca´ ssaro. 2000. Nitrogen dynamics in a soil-sugar cane system. Scientia Agricola 57:467–472. p i Schulze, E. D. 1989. Air pollution and forest decline in a spruce (Picea abies) forest. Science 244:776–783. Scopinho, R. A., F. Eid, C. E. F. Vian, and P. R. C da Silva. 1999. New technologies and workers’ health: mechanization of sugar cane harvesting. Cadernos de Sau´ de Pu´ bica 15:147– 161. Oliveira, J. C. M. de, C. M. P. Vaz, and K. Reichardt. 1995. Efeito do cultivo contı´nuo da cana-de-ac¸ u´ car em proprie- dades fı´sicas de um latossolo vermelho escuro. Scientia Agricola 52:50–55. Silva, A. J. N., and M. R. Ribeiro. 1997. Caracterizac¸ a˜ o de Latossolo Amarelo sob cultivo contı´nuo de cana-de-ac¸ u´ car no Estado de Alagoas: atributos morfolo´ gicos e fı´sicos. Revista Brasileira de Cieˆ ncia do Solo 21:677–684. Oliveira, M. E. D., B. E. Vaughan, and E. J. J. Rykiel, Jr. 2005. Ethanol as fuel: energy, carbon dioxide balances, and ecological footprint. Bioscience 55:593–602. Silva, A. J. N., M. R. Ribeiro, A. R. Mermut, and M. B. Benke. 1998. Influeˆ ncia do cultivo contı´nuo da cana-de-ac¸ u´ car em latossolo amarelos coesos do Estado de Alagoas: propriedade micromorfolo´ gicas. Revista Brasileira de Cieˆ ncia do Solo 22: 515–525. Oliveira, M. W., P. C. O. Trivelin, A. E. Boaretto, T. Muraoka, and J. Mortatti. 2002. Leaching of nitrogen, potassium, calcium and magnesium in a sandy soil cultivated with sugarcane. Pesquisa Agropecua´ ria Brasileira 37:861–868. Silva, A. M., M. A. Nalon, F. J. do Nascimento Kronka, C. A. Alvares, P. B. de Camargo, and L. A. Martinelli. 2007. Historical land-cover/use in different slope and riparian buffer zones in watersheds of the state of Sa˜ o Paulo, Brazil. Scientia Agricola 64:325–335. Ometto, J. P. H. B., L. A. Martinelli, M. V. Ballester, A. Gessner, A. Krusche, R. L. Victoria, and M. Williams. 2000. Effects of land use on water chemistry and macroinverte- brates in two streams of the Piracicaba River Basin, Southeast Brazil. Freshwater Biology 44:327–337. Silva, D. M. L., P. B. de Camargo, L. A. Martinelli, F. M. Lanc¸ as, J. S. S. Pinto, and W. E. P. Avelar. In press. Organochlorine pesticides in the Piracicaba river basin (Sa˜ o Paulo-Brazil): a survey of sediment, bivalves, and fish. Quı´ mica Nova. hhttp://quimicanova.sbq.org.br/qn/ No%20Prelo/Artigos/AR06436.pdfi Oppenheimer, C., V. I. Tsanev, A. G. Allen, A. J. S. McGonigle, A. A. Cardoso, A. Wiatr, W. Paterlini, and C. de Mello Dias. 2004. NO2 emissions from agricultural burning in Sa˜ o Paulo, Brazil. Environmental Science and Technology 38:4557–4561. Pereira-Netto, A. D., I. F. Cunha, and T. M. Krauss. 2004. Persistence of polycyclic aromatic hydrocarbons in the soil of a burned area for agricultural purposes in Brazil. Bulletin of Environmental Contamination Toxicology 73:1072–1077. Sparovek, G., and E. Schnug. 2001. Temporal erosion-induced soil degradation and yield loss. Soil Science Society of America Journal 65:1479–1486. Streets, D. G., S. Gupta, S. T. Waldhoff, M. Q. Wang, T. C. Bond, and B. Yiyun. 2001. Black carbon emissions in China. Atmospheric Environment 35:4281–4296. Politano, W., and T. C. T. Pissarra. 2005. Avaliac¸ a˜ o por fotointerpretac¸ a˜ o das a´ reas de abrangeˆ ncia dos diferentes estados da erosa˜ o acelerada do solo em canaviais e pomares de citros. Engenharia Agrı´cola 25:242–252. Tominaga, T. T., F. A. M. Cassaro, O. O. S. Bacchi, K. Reichardt, J. C. Oliveira, and L. C. Timm. 2002. Variability of soil water content and bulk density in a sugarcane field. Australian Journal of Soil Research 40:605–614. Prado, R. M., and J. F. Centurion. 2001. Alterac¸ ˜oes na cor e no grau de floculac¸ a˜ o de um Latossolo Vermelho-Escuro sob cultivo contı´nuo de cana-de-ac¸ u´ car. Pesquisa Agropecua´ ria Brasileira 36:197–203. Trivelin, P. C. O., M. W. Oliveira, A. C. Vitti, G. J. C. Gava, and J. A. Bendassolli. 2002. Pesquisa Agropecua´ ria Brasileira 37:193–201. Ramanathan, V., P. J. Crutzen, J. T. Kiel, and D. Rosenfeld. 2001. Aerosols, climate and the hydrological cycle. Science 294:2119–2124. Trivelin, P. C. O., J. C. S. Rodrigues, and R. L. Victoria. 1996. Utilizac¸ a˜ o por soqueira de cana-de-ac¸ u´ car de inı´cio de safra do nitrogeˆ nio da da aquamoˆ nia-15N e ure´ ia-15N aplicado ao solo em complemento a` vinhac¸ a. Pesquisa Agropecua´ ria Brasileira 31:89–99. Roberts, G. C., M. O. Andrea, K. Zhou, and P. Artaxo. 2001. Cloud condensation nuclei in the Amazon Basin: ‘‘Marine’’ conditions over a continent? Geophysical Research Letters 28:2807–2810. Trivelin, P. C. O., R. L. Victoria, and J. C. S. Rodrigues. 1995. Aproveitamenteo por soqueira de cana-de-ac¸ u´ car de final de safra do nitrogeˆ nio da aquamoˆ nia-15N e ure´ ia-15N aplicado ao solo em complemento a` Vinhac¸ a. Pesquisa Agropecua´ ria Brasileira 30:1375–1385. Rodrigues, A. 2006. O corte: como vivem—e morrem—os migrantes nos canaviais de Sa˜ o Paulo. Instituto Superior de Ciencias Aplicadas, Limeira, Brazil. Rudorff, B. F. T., L. M. S. Berka, M. A. Moreira, V. Duarte, and V. G. C. Rosa. 2004. Estimativa de a´ rea plantada com cana-de-ac¸ u´ car em municı´pios do estado de Sa˜ o Paulo por meio de imagens de sate´ lites e te´ cnicas de geoprocessamento: ano safra 2004/2005. INPE-11421-RPQ/762. Sao Paulo, Brazil. Zamperlini, G. C. M., M. R. S. Silva, and W. Vilegas. 1997. Identification of polycyclic aromatic hydrocarbons in sugar cane soot by gas chromatography-mass spectrometry. Chromatographia 46:655–663.https://doi.org/10.1007/s10584-020-02856-6 Climatic Change (2020) 162:1823–1842 nanBrazil’s emission trajectories in a well-below 2 °C world: the role of disruptive technologies versus land-based mitigation in an already low-emission energy system Alexandre C. Köberle1,2 & Pedro R. R. Rochedo1 & André F. P. Lucena1 & Alexandre Szklo1 & Roberto Schaeffer1 Received: 27 October 2017 /Accepted: 30 August 2020/ # The Author(s) 2020 Published online: 2 October 2020Abstract The Nationally Determined Contributions (NDCs) to the Paris Agreement (PA) submitted so far do not put the world on track to meet the targets of the Agreement and by 2020 countries should ratchet up ambition in the new round of NDCs. Brazil’s NDC to the PA received mixed reviews and has been rated as “medium” ambition. We use the Brazil Land Use and Energy System (BLUES) model to explore low-emission scenarios for Brazil for the 2010–2050 period that cost-effectively raise ambition to levels consistent with PA targets. Our results reinforce the fundamental role of the agriculture, forest, and land use (AFOLU) sectors and explore inter-sectoral linkages to power generation and transportation. We identify transportation as a prime candidate for decarbonization, leveraging Brazil’s already low-carbon electricity production and its high bioenergy production. Results indicate the most important mitigation measures are electrification of the light-duty vehicle (LDV) fleet for passenger transportation, biodiesel and biokerosene production via Fischer-Tropsch synthesis from lignocellulosic feedstock, and intensification of agricultural production. The use of carbon capture and storage (CCS) as well as netzero deforestation make significant contributions. We identify opportunities for Brazil, but synergies and trade-offs across sectors should be minded when designing climate policies. Keywords Brazilemissions.Climatemitigation.Bioenergyandbiofuels.Low-carbontransition . Mitigation scenarios . Integrated assessment model This article is part of a Special Issue on “National Low-Carbon Development Pathways” edited by Roberto Schaeffer, Valentina Bosetti, Elmar Kriegler, Keywan Riahi, Detlef van Vuuren, and John Weyant Electronic supplementary material The online version of this article (https://doi.org/10.1007/s10584-020- 02856-6) contains supplementary material, which is available to authorized users. * Alexandre C. Köberle alexkoberle@gmail.com; a.koberle@imperial.ac.uk Extended author information available on the last page of the article Climatic Change (2020) 162:1823–18421 Introduction The Paris Agreement-PA (UNFCCC 2015) shifted emphasis of climate negotiations from a top-down global approach to one in which countries make voluntary pledges called Nationally Determined Contributions (NDCs). Still, the NDCs submitted do not put the world on track to meet the targets of the Agreement (Rogelj et al. 2016) and by 2020 countries should ratchet up ambition in the new round of NDCs and submit their mid-century strategies (MCSs). However, climate action must be framed in the context of other societal objectives such as those embodied in the United Nations’ Agenda 2030 and the Sustainable Development Goals (SDGs) (Krey et al. 2019b; von Stechow et al. 2016). Concurrent attainment of the 17 SDGs requires an integrated approach that accounts for interlinkages between SDGs, maximizing synergies and minimizing trade-offs in policy design and implementation (Nilsson et al. 2018; Rogelj et al. 2018). The IPCC Special Report on 1.5 Degrees (SR15) (Roy et al. 2018) found SDG 13 (Climate Action) to have strong linkages to other SDGs. Brazil’s NDC to the PA (GofB 2015a) received mixed reviews as to its ambition. The absolute targets of 1.3 Gt CO2eq for 2025 and 1.2 Gt CO2eq for 2030 were rated as “medium” ambition, at the “least ambitious level of a fair contribution” (CAT 2017; Pan et al. 2017). The Agriculture, Forests and Land Use (AFOLU) sector is the cornerstone of the Brazilian NDC and includes the Low-carbon Agriculture Plan (ABC Plan) reduction targets of between 133.9 and 162.9 Mt CO2eq, mostly through livestock sustainable intensification measures. The NDC also pledges to end illegal deforestation in Brazil. Targets in other sectors are weak or vague (Escobar 2015; Köberle et al. 2015). Energy targets include higher shares of renewable sources in primary energy mix and in power generation, to reach 45% and 33%, respectively, over 2005 levels by 2030, and also a vaguely defined energy efficiency improvement of 10% by the same date (GofB 2015a). Brazil’s total greenhouse gases (GHG) emissions have been stable around 1.5 Gt CO2eq since 2009 (SEEG 2019). Roughly speaking, agriculture, energy, and land use (deforestation) each account for about one-third of total GHG emissions on average in Brazil (GofB 2015b; SEEG 2019). Agriculture emissions are dominated by methane from enteric fermentation and nitrous oxide from synthetic fertilizers and animal wastes, energy emissions come mostly from the transportation sector and industry, and land use emissions are mostly CO2 from land use change (LUC) (GofB 2015b). Power generation is already low-carbon, relying mostly on hydropower (~ 65%) and combustion of sugarcane bagasse and other biomass (~ 8%), and primary energy consumption (PEC) in the country is currently from roughly 45% low-carbon sources (EPE 2019). See SOM for more details. Potential measures for emission abatement in agriculture include recuperation of degraded pastures (Assad et al. 2015; Strassburg et al. 2014) and integrated crop–livestock–forestry (iCLF) systems (Agroicone 2016; Balbino et al. 2011), which can lead to an increase in biomass and soil organic carbon (SOC), thereby reducing emissions and potentially, in some cases, making the process a GHG-emission sink for a period of 10–20 years (Carvalho et al. 2014, 2009; Cohn et al. 2014) and enable higher sustainable stocking rates, potentially sparing land for food, fiber, and biofuel production (Cardoso et al. 2016; De Oliveira et al. 2013). Tradeoffs include higher demands for energy (mostly diesel) and fertilizers, important sources of GHG emissions. Emissions from land use can be reduced mainly by eliminating defores- tation (Rochedo et al. 2018), the main source of GHG emissions in Brazil until recently (MCTIC 2016; SEEG 2019). nan Climatic Change (2020) 162:1823–1842 Options for abatement of energy emissions include the use of wind and solar as well as bioelectricity with and without CCS for electricity generation, and the use of liquid biofuels in the transportation sector, especially replacing diesel for freight and kerosene for air transpor- tation. The introduction of CCS in ethanol production also has the potential to contribute (Szklo et al. 2017). Alternatively, introduction of electric vehicles (EVs) could take advantage of the low-carbon electricity in the country, but this could reduce the role of ethanol, with effects on electricity generation from sugarcane bagasse (see SOM). Producing biomass feedstock for bioenergy production could raise demand for land, with implications for agriculture and land use (Rathmann et al. 2012). In contrast, electrification of transportation in Brazil and worldwide could reduce demand for biofuels but increase demand for electricity with implications for the power sector. This interplay between AFOLU and energy sectors responds to varying degrees of GHG emission restrictions (see Methods). Therefore, it is critical to examine such trade-offs explicitly. Moreover, impacts on AFOLU sectors have implications for various SDGs, especially those dealing with food security (SDG2), water (SDG6), and biodiversity (SDGs 14 and 15). We use the BLUES1 model (IAMC 2020; Köberle 2018; Rochedo et al. 2018) to explore these linkages and to produce emission scenarios for Brazil for the 2010–2050 period, consistent with global efforts to stay within various temperature targets, comparing them with the Brazilian NDC. The next section introduces the BLUES model and describes the scenarios. Section 3 shows results of key variables, contrasting them across scenarios. Section 4 discusses results focusing on the transportation sector’s role as mediator of inter-sectoral linkages and on implications for the SDGs. Section 5 provides conclusions and key findings. Details on current trends and context in Brazil as well as additional analysis on results are provided in the SOM.2 Methods2.1 The BLUES model The BLUES model is the most recent version of a family of models built on the MESSAGE model platform (Keppo and Strubegger 2010). It was developed for the Brazilian energy system and has been sequentially updated and applied to assess issues relevant to the national reality (see SOM for a detailed description). The model has recently been reconfigured for better detailing of both regional breakdown and endogenous energy efficiency and GHG mitigation options in the end-use sectors (Köberle 2018; Rochedo et al. 2018; Szklo et al. 2018). More recently, a representation of the land-use system (forests, savannas, low- and high-capacity pastures, integrated systems, cropland, double cropping, planted forests, protected areas) was introduced (Köberle 2018; Rochedo et al. 2018), along with a suite of advanced biofuel technologies not present in previous versions. In addition, the cost assump- tions on electric vehicles and photovoltaic (PV) solar power have been updated to reflect recent developments. As a perfect-foresight cost-minimization model, BLUES produces the least-cost pathway to meet emission budgets subject to constraints to reflect socioenvironmental conditions, policies, 1 Brazil Land Use and Energy System model (aka COPPE-MSB_v2.0). For documentation on the BLUES model see https://www.iamcdocumentation.eu/index.php/Model_Documentation_-_BLUES nan Climatic Change (2020) 162:1823–1842 and other limiting factors. The expansion of key technologies was constrained in the model to reflect technoeconomic restrictions to the deployment of new facilities, such as industrial capacity to manufacture equipment, skilled labor to build and operate the plants, and avail- ability of capital to fund them (see SOM Section 1.2). Innovative biofuel production routes with and without CCS were constrained following this rationale. More details about the BLUES model can be found in the SOM.2.2 The scenarios2.2.1 Socioeconomic premises and reference scenario The scenarios explored here are based on Shared Socioeconomic Pathway 2 (SSP2) assump- tions (Fricko et al. 2017; Riahi et al. 2017), meaning that they do not include structural socioeconomic changes from historical trends in Brazil. GDP growth rates for Brazil were updated from SSP2 projections (Dellink et al. 2017) to reflect the near-term 2010–2020 growth rates (Köberle et al. 2018). The Reference scenario (Ref) in this study includes the most relevant public policies in effect in the country today which may impact GHG emissions. These include non-climate policies such as biofuel blending mandates and renewable energy auctions, and climate policies that are already under implementation such as the ABC Plan (MAPA 2012). Most of the mitigation measures included in the Brazilian NDC (GofB 2015a) are already in effect today through existing policies (Köberle et al. 2015). Table S1 shows the policies implemented in the construction of the scenarios described here. The Ref scenario also includes an assumption of full implementation of the 2012 Forest Code via forcing net-zero deforestation post-2030, as compliance with the Code requires afforestation on some 30 Mha (Soares-filho et al. 2014). Finally, EV costs are projected to reach cost parity with conventional vehicles by 2040. Although EV cost parity by 2040 may seem overly conservative given recent developments (BNEF 2017), the lack of structural socioeconomic changes may imply light-duty vehicles (LDVs) in the country will continue to cost more than they do abroad. Thus, since there are currently no policies to incentivize EVs, we assume the country will lag behind global cost developments by about a decade. This also applies to electric trucking, which faces the additional barriers from deficient infrastructure in the country in the form of badly maintained roads and an unreliable electricity grid. For this reason, electric trucks are not included in the options available to the model.2.2.2 Climate policy scenarios Three climate mitigation scenarios explore progressively more stringent emission constraints based on the current policy scenario (Ref) with various emissions constraints, including two early-action scenarios (mitigation from 2020) and one delayed-action scenario (from 2030). Two scenarios are consistent with global efforts to stay below 2 °C average temperature rise above pre-industrial levels: one early-action (2 deg) and one delayed-action scenario in which mitigation begins only after following the Ref trajectory through 2030 (2deg_dly). Finally, the third scenario is a 1.5 °C scenario (1p5deg) that assumes early-action consistent with Brazil’s contribution to a global 1.5 °C target (Kriegler et al. 2019). The climate mitigation scenarios were implemented via emissions budgets, as described in detail in the SOM. We apply resulting mid-century budgets for Brazil’s share of a global 1000 Gt of CO2 (consistent with a > 66% chance of staying within 2 °C) and 400 Gt of CO2 (in nan Climatic Change (2020) 162:1823–1842 line with a > 66% chance of staying within 1.5 °C warming limit by 2100), taken from cost- optimal global runs of the COFFEE model (Rochedo 2016; Rochedo et al. 2018; Roelfsema et al. 2020) for each of the temperature targets. The COFFEE model is a global integrated assessment model (IAM) built on the same platform as BLUES and having a similar structure, in which Brazil is represented as a single region in a world comprised of 18 regions (see SOM). Because COFFEE runs through 2100 and BLUES only through 2050, the COFFEE cumulative emission budgets for 2010–2050 were used. The COFFEE model was chosen for several reasons. First, it is built on the same modeling platform as BLUES and follows a similar logic and structure. Second, it has one of the best sectoral representations of Brazil among all the IAMs since it was built to explicitly represent Brazil’s role in the world and many IAMs do not even have Brazil represented as separate from Latin America (see Section S1.3.2). The 2010–2050 budgets implemented in BLUES were as follows: 23.6 Gt CO2, for the 2 deg and 2deg_dly scenarios and 15.4 Gt CO2 for the 1p5deg scenario. These emission budgets represent about 15 years of the country’s 2010 emissions for the 2 °C scenarios and about 10 years for the 1.5 °C scenarios. The budgets were implemented in BLUES as a cap on cumulative CO2 emissions and, to limit emissions of non-CO2 GHGs, these were priced at GWP-AR4 equivalency2 starting in 2020, as explained in SOM Section S2. The resulting budgets are close to the median from other IAMs for Brazil (see Section S1.3.2). For more details on the scenario protocols used in this paper see Schaeffer et al. (2020, this issue).3 Results Results indicate key linkages between AFOLU, transportation, and energy sectors. They point to the fundamental role of AFOLU in Brazil’s mitigation efforts, with its emissions becoming negative before those of the energy sector. Intensification of livestock contributes significantly to AFOLU CO2 abatement in BLUES, thereby increasing the currently low stocking rate of Brazilian pastures. In fact, this occurs even in the absence of emissions constraints, which is explained by the low or negative abatement cost of this measure, its large potential, and continuing implementation of the ABC Plan3 (MAPA 2012). Introduction of iCLF systems makes important contributions to improve yields. As mitigation targets become more stringent (tighter CO2 budgets), technological shifts further decarbonize sectors with remaining mitigation potential. In the energy sector, mitiga- tion measures include low-carbon sources for all energy carriers, fuel switch in transportation, and energy efficiency in industry and transportation, along with transitions to low-carbon power generation from biomass, wind and solar. The transportation sector stands out as having a significant potential for emission reductions through the use of biofuels, with and without CCS,4 and the electrification of the LDV passenger fleet. Introduction of EVs in budget scenarios leads to higher electricity consump- tion, affecting power generation expansion strategies the mix of biofuel production, also impacting the agricultural sector. 2 The energy module in BLUES was built prior to the release of GWP-AR5 values. 3 The energy module in BLUES was built prior to the release of GWP-AR5 values. 3 The model results provide evidence to the economic advantage of intensifying livestock production in Brazil. However, several market barriers (as identified by Gil et al. (2015) and Köberle et al. (2017)) still hamper large- scale adoption in the country. 4 CCS here is applied in the production phase of the biofuel, not in the transportation sector per se. nan Climatic Change (2020) 162:1823–1842 We explore these dynamics in the following sections and offer deeper analyses in the discussion section.3.1 Emissions Figure 1a shows resulting GHG emission trajectories in each scenario. Emission reductions in budget scenarios result mainly from reductions in CO2 emissions with some contribution from CH4 and N2O reductions in the short-term. Emissions of both CH4 and N2O drop in the short-term but return to 2015 levels by the end of the period due to increased activity of their sources, mainly in the AFOLU sectors. One distinct aspect of Brazil’s emission profile is the high share of non-CO2 gases, particularly CH4 and N2O from AFOLU sectors. Across budget scenarios emissions of non-CO2 gases dominate towards 2050, as net CO2 emissions become negative. The effect of the net-zero deforestation assumption is clear by the sudden drop in AFOLU CO2 emission post-2030, highlighting the impact that full implementation of the Forest Code can have on the country’s climate action. Brazil’s GHG emission targets pledged in the NDC are shown as black dots in Fig. 1a. Some key insights can be drawn from this figure: (i) the Ref scenario shows current policies are only partially in line with NDC targets and that strengthening of ambition may be required for the country to meet its 2030 emissions target; (ii) policies helping Brazil to meet its NDC would put the country within a trajectory consistent with a least-cost, global end-of-century 1.5 °C target in 2025, but the 2030 target would need to be strengthened; and (iii) delayed-action to 2030 would imply drastic measures towards 2040 in order to bring the country closer to a cost-optimal trajectory to a well-below 2 °C global target. These insights should contribute to Brazil’s development of its mid-century strategy for climate action.table_1Fig. 1 GHG emission trajectories 2015–2050 (a) and CO2 sequestration post-2030 (b). FFI fossil fuel and industry. The black dots indicate the NDC pledged targets. The category “Other” includes GHG sources not modeled in BLUES. They are dominated by indirect nitrous oxide emissions but also include emissions of hydrofluorocarbons (HFCs), methane from livestock other than cattle and poultry, and nitrous oxide from organic soils. The value for 2010 from GoB (2015b) and are assumed constant throughout the periodnan Climatic Change (2020) 162:1823–18423.2 Carbon sequestration Figure 1b shows that long-term reductions in CO2 emissions in Brazil can be achieved mainly through a combination of AFOLU efforts and deployment of CCS technologies in the energy sector. Brazil’s large biofuels potential (Smeets and Faaij 2010) and the availability of sequestration sites (Rochedo et al. 2016) mean it can deploy significant levels of bioenergy with CCS (BECCS5) to help further decarbonize its economy. The figure shows total CO2 sequestered and through which measure and is not limited to carbon removal from the atmosphere. AFOLU sequestration is the aggregate result of (i) intensification of livestock production, via conversion of low-capacity pastures to high-capacity pastures and integrated systems, and (ii) afforestation via growth in planted forests and regrowth of natural forests (see Section 3.3). There is more BECCS in the late-accession 2deg_dly than in the 1p5deg scenario due to the shorter time period available to decarbonize. Negative CO2 emissions in Fig. 1a come from contributions of energy-sector CCS and AFOLU sequestration (Fig. 1b). BECCS becomes increasingly important as the mitigation challenges become tougher. Carbon capture occurs in electricity generation and also in the production of ethanol, BTL-kerosene and BTL-diesel (Tagomori 2017; Tagomori et al. 2019), used to decarbonize the transport sector. CO2 capture in the fermentation phase of ethanol production is a result we have reported elsewhere (Herreras-Martínez et al. 2015; Koberle et al. 2015), an important mitigation option for Brazil (see Section 3.6). BECCS deployment in the 1p5deg and 2deg_dly scenarios reaches cumulative sequestration of around 2.2 Gt of CO2 by 2050. It must be said that negative emissions can be costly in energy and economic terms, and very clear governance protocols would need to be in place to account for sequestered CO2 (Gough et al. 2018; Peters and Geden 2017). Although CCS costs are currently high, they are projected to fall in the medium- to long-term (Budinis et al. 2018). In addition, carboduct networks would need to be built to carry captured CO2 to injection sites, adding to the governance issues surrounding CCS (Merschmann et al. 2016; Tagomori et al. 2018). Policies to support and regulate CCS in Brazil are lacking (da Costa 2014).3.3 Land use and land use change As mentioned before, AFOLU sectors play an important part in Brazil’s mitigation effort. Land use change (LUC) from 2015 across scenarios is shown in Fig. 2. Results show that intensification of livestock causes significant reduction in low-capacity pasture area, which is transformed into high-capacity pastures (with average stocking rates twice as high), integrated systems and cropland. This holds across all scenarios, including current policies, where a reversal of the growth trend of low-capacity pastures indicates this as an attractive measure even in the absence of climate policies, in line with increasing interest in such systems given their demonstrated potential for higher profitability (De Oliveira et al. 2013). The steep changes in pasture area from low-capacity to high-capacity in Fig. 2 are equivalent to a conversion of about 50 Mha in 30 years, a little more than 1.5 Mha per year. This is not unprecedented. A transition at this rate and scale has occurred before in Brazilian agricultural 5 BECCS is understood to include both bioelectricity generation and bioliquids production with CCS. nan Climatic Change (2020) 162:1823–1842 production systems. Between 1992 and 2012, no-till agriculture spread across some 30 Mha and is today used in more than 50% of grain area in the country (Inagaki et al. 2016). To meet the demands for food, fiber, and energy, about 12 Mha of low-capacity pastures are converted to high-capacity in the Ref scenario, indicating current trends will almost meet the stated NDC goal of 15 Mha. On the other hand, there is an increase of only 0.2 Mha hectares of integrated systems, well short of the 4 Mha pledged in the NDC. Likewise, the 0.5 Mha of planted forests fall short of the target for additional 4 Mha. Cultivated land increases while natural lands decrease in the short-term and slightly recover post-2035 in mitigation scenarios. Table S1 shows area changes in 2050 relative to 2010 for each land cover type in Fig. 2. Across all scenarios, agricultural land (crops and livestock) peaks between 2025 and 2030 at around 295 Mha. In the Ref scenario, converted lands do not return to their original cover by 2050. However, as the budgets become more stringent, intensification of agriculture drives the conversion of low-capacity pastures to high-capacity pastures and integrated systems, reducing total area devoted to livestock production. Abandoned pastureland is converted to cropland and planted forests or left to regrow with natural vegetation, especially tropical forests (strong carbon sinks). By diminishing demand for land, agricultural intensification allows forests to recover significant shares of lost area. Savannas recover much less because of their lower carbon stock, implying further losses to the biodiversity-rich Cerrado biome. In the 2deg_dly scenario by 2050, roughly the same extent of forest area converted to agricultural use (~ 6 Mha) since 2015 is returned to natural forests (afforestation) (Fig. S8).3.4 Primary energy The continuing growth trend in PEC is clearly reflected in the Ref scenario, with increases across the range of currently-used energy sources (Fig. 3a). In the budget scenarios, (lignocellulosic) biomass and sugarcane increasingly replace oil and coal as emission Climatic Change (2020) 162:1823–1842 constraints become tighter. In the most stringent cases (1p5deg and 2deg_dly), coal virtually disappears and oil consumption drops by more than 80% (completely disappearing in 2deg_dly), forcing the refining sector to adjust. Biomass share grows to scenario-dependent 50–75% of total PEC by 2050, with profound impacts on land demand for bioenergy, which helps explain some of the land-use dynamics described in Section 3.3 (see Section 4). Competition between biomass and sugarcane is largely determined by assumptions for their respective technoeconomic parameters. In the current version of BLUES, biomass production from grasses stands out as the least-cost feedstock source for biofuels production via biomass- to-liquids Fischer-Tropsch (BTL-FT) synthesis of biodiesel and biokerosene. However, BTL- FT synthesis is currently non-existent in Brazil, meaning this novel technology is deployed gradually, ramping up capacity at a rate commensurable with industrial capacity (SOM Section S1.2.2). Figure 3b shows this growth of bioenergy, which reaches between 65 and 80% share of PEC in 2050 across budget scenarios. Additionally, the NDC target of 33% bioenergy in PEC is already met in the cost-optimal Ref scenario.table_2Fig. 3 Primary energy consumption (PEC) (a), share of bioenergy in PEC (b), power generation (c), and share of non-hydro renewable energy in power generation (d) in Brazil. The black dots show the NDC targets for the plotted quantities nan Climatic Change (2020) 162:1823–18423.5 Power generation Hydropower remains the mainstay of the Brazilian power system, but its share drops across all scenarios (Fig. 3c, d). In the absence of climate policies (Ref), coal power generation gains space after 2030 in this cost-optimization model, as it produces the lowest-cost electricity, in line with other baseline scenarios previously reported elsewhere (Köberle et al. 2015, 2018; Lucena et al. 2016; Portugal-Pereira et al. 2016). Nonetheless, wind, solar, and nuclear generation also see significant increases in the Ref scenario. As emission constraints are introduced, bioelectricity increasingly pushes coal out of the energy mix. Coal-fired capacity is replaced in all budget scenarios by bioelectricity production with and without CCS, from both sugarcane bagasse6 and other biomass feedstocks (Fig. 3c). A small amount of legacy coal-fired generation remains through 2050 from lignite mines in the South region of Brazil. Counterintuitively, oil-fired generation without CCS enters the solution of budget scenar- ios, explained by the fact that diesel fuel is gradually decarbonized through blending of BTL- FT biodiesel, surpassing 70% of the diesel pool by 2050 (Section 3.6). In addition, some competition occurs between biomass-fired electricity (mostly from bagasse and residues) and non-biomass renewables (hydro, solar, and wind). This results from the need for negative emissions in the long-term to compensate for higher emissions in the short-term, and so BECCS is used to both meet electricity demand and remove CO2 from the atmosphere. All budget scenarios deploy large amounts of bioelectricity from sugarcane and lignocellulosic sources, with and without CCS. Bagasse is a by-product and has null cost, while biomass provides firm power which gives it the advantage of not requiring back-up capacity as is the case for intermittent sources. The budget scenarios also boast slightly higher electricity demand, spurred by the electrification of the LDV fleet (Section 3.7). Nonetheless, non-biomass renewables grow in all scenarios. Wind power reaches levels comparable to today’s bioelectricity production from bagasse (~ 0.2 EJ/year in 2050), and solar also grows. Brazil has potential for solar and wind exceeding the values deployed in these scenarios by BLUES (CEPEL 2013; Simioni and Schaeffer 2019), but the CCS option makes bioelectricity more attractive. As with any cost-optimization model, different cost and effi- ciency assumptions may yield different results, potentially tipping the scales in favor of wind and solar over bioelectricity. PV costs were updated to reflect recent developments, but the stringent emission constraints favor BECCS deployment. In spite of these increases, non- hydro renewables fail to meet the NDC target share of 33% of power generation.3.6 Biofuels Important changes happen in the biofuel sector as emission budgets tighten (Fig. 4a). The sharp increase in production in the stringent scenarios is driven by the need to decarbonize freight transportation and aviation. By 2050, ethanol and FAME biodiesel (1st Gen) produc- tion have peaked and returned to near 2015 levels. Importantly, most of the ethanol and advanced biofuels are produced with CCS in the production phase of the fuel, leveraging these low-cost options available to the model. Ethanol production decreases following increasing penetration of EVs in the LDV fleet. 6 Sugarcane bagasse is the solid residue after sugarcane is crushed to extract the juice from which sugar and 1st generation ethanol are made. nan Climatic Change (2020) 162:1823–1842table_3Fig. 4 Biofuels production (a) and passenger transportation fuel mix (b) deployed to meet energy services demand in Brazil. Ethanol_A is anhydrous ethanol blended into gasoline at 27.5% by volume BTL-diesel with CCS is used to decarbonize road freight transportation, by far the most important modal for freight in Brazil (EPE 2014). Production of advanced biofuels reaches high-enough levels to completely replace oil products in the 2deg_dly scenario by 2050, meaning both diesel and kerosene are low-carbon biofuels after 2030 since diesel fuel is gradually decarbonized by the blending of BTL-FT biodiesel, surpassing 64%, 69%, and 70% (2 deg, 1p5deg, 2deg_dly, respectively) of the diesel pool by volume in 2050. The high share of BTL-diesel in diesel fuel also enables diesel engine busses to remain an option for passenger transportation throughout the period across scenarios. Interestingly, BTL-biojet fuel enters the mix in the Ref scenario, which is unexpected. The model chooses to deploy BTL technology instead of expanding the jetfuel capacity in refineries, which is currently limited in the country (Carvalho 2017). All scenarios have BTL-kerosene accounting for 100% of jetfuel demand as early as 2040, although in reality some technical limitations may preclude this.3.7 Transportation As global prices of hybrid (HYB), plug-in hybrid (HUYB), and electric vehicles (EV) are projected to drop in the short- to medium-terms, these become important alternatives to decarbonize passenger transportation using the country’s low-carbon electricity. Hybrid and EV costs are implemented in BLUES to reach parity with conventional cars by 2040. Thus, nan Climatic Change (2020) 162:1823–1842 starting in that year, the model switches to EVs in the light commercial category (mostly taxis), that is, the least efficient LDVs in the model delivering transportation services. The model also completely shifts from internal combustion engine (ICE) motorcycles to fully electric 2- wheelers starting in 2030 in all budget scenarios. Figure 4b shows passenger transportation fuel mix across scenarios for modeled years 2015, 2030, and 2050 (for a similar figure for freight fuels, see Fig. S5). Growing demand for mobility services in Brazil is reflected by the growing energy demand of passenger transpor- tation towards 2050. In spite of higher EV penetration, flex fuel vehicles remain the most important private passenger transportation alternative, running mostly on a blend of 70% gasoline (containing 27.5% anhydrous ethanol) and 30% hydrated ethanol.7 In all mitigation scenarios, electrifica- tion of the LDV fleet begins in 2035, with motorcycles and taxis replaced by their electric counterparts. The higher conversion efficiency of EVs causes a significant drop in energy consumption to meet passenger transportation services demand. On an energy basis, EVs account for 25% of passenger private transportation in 2050, increasing electricity demand with implications for power generation (Section 3.5), and reductions in ethanol and gasoline consumption. Implications of these results are addressed in depth in the discussion section (see also SOM Section 1.4.1). Passenger public transport continues to be done by diesel ICE buses, but the increasing share of BTL-diesel in the fuel mix means emissions stabilize. Transportation emissions peak in 2030 and begin to drop as electricity and biofuels gain space in the energy mix of the sector. Total passenger emissions peak in 2030. Stringent scenarios bring higher use of highly specified drop-in BTL biofuels,8 produced with and without CCS (Section 3.6) and blended into the diesel and kerosene pools. This effectively makes diesel and kerosene low-carbon fuels, allowing them to maintain their share in the fuel mix of passenger transportation sector (Fig. 4b). The same occurs in freight transportation, where fuel types remains largely unchanged from the Ref scenario in 2050.4 Discussion Countries are currently preparing to submit revised NDCs in 2020, and the expectation is to ratchet up ambition to close the emissions gap and put the world on track to fulfill PA goals to curb climate change. In addition, Article 4 of the Agreement invited parties to submit, by 2020, long-term GHG emission strategies known as “mid-century strategies.” This paper examined how specific targets of the Brazilian NDC compare to current trends and to more ambitious scenarios to 2050 consistent with global efforts to limit climate change to well-below 2 °C compared with the pre-industrial era. Higher ambition was implemented in the BLUES model as tight emissions budgets equivalent to 15 years of the country’s 2010-level CO2 emissions for a 2 °C global temperature target and to 10 years of 2010 emissions for a 1.5 °C global target. Although it can be argued that these targets go beyond what may be considered a fair 7 Flex vehicles are modeled as operating in two modes: in option 1, the ratio gasoline/ethanol is 30/70, and in option 2 70/30. Option 2 dominates in all scenarios modeled here. Because BLUES models gasoline as blended with anhydrous ethanol at 27.5% by volume, average flex vehicle consumption is 50/50 {0.7*(1– 0.275) = 0.5075}. 8 Drop-in biofuels are highly specified 2nd generation biofuels that can be injected directly into the fuel without any alteration of the engine. nan Climatic Change (2020) 162:1823–1842 contribution from Brazil under other effort-sharing allocation criteria (van den Berg et al. 2019, this issue), this allows exploration of the country’s potential to ratchet up its contribution to the global effort under the PA. Halting deforestation is the main challenge for Brazilian climate action. Full implementa- tion of the Forest Code by 2030 implies afforestation on more than 30 Mha of land which, for the period of analysis, would more than compensate for deforestation pre-2030, effectively delivering on net-zero deforestation. Brazil already boasts a low-emissions energy system with high share of hydroelectricity in power generation and significant penetration of biofuels (ethanol and biodiesel) in liquid fuels. However, rising demand for energy services will drive expansion of energy supply capacity across sectors, which, in the Ref scenario, is met by increasing fossil fuel consumption. In the mitigation scenarios, the stringent constraints on emissions increase the value of negative emissions technologies (NETs), which include afforestation and BECCS in the current version of BLUES. Our results suggest current policies in Brazil are only partially on track to meet the NDC. For AFOLU targets, the Ref scenario approaches the NDC target area for pasture recuperation by 2030 but falls short of the targets for integrated systems and planted forests, suggesting further action may be required. There is also only partial attainment of energy targets, with share of bioenergy well exceeding the 18% of PEC as pledged in the NDC, but the share of non-hydro renewables falling short of the 33% target by 2030. To further reduce the emissions intensity of Brazil’s energy system, some innovative technologies are deployed at scale, especially in the livestock production (integrated systems) and biofuels (FT-BTL) sectors. The agricultural sector can play a central role in Brazil’s future in a low-carbon world. In the mitigation scenarios explored here, a revolution in livestock production leads to a shift towards high productivity livestock systems that break with the old expansionist approach that has driven deforestation in the past. Pasture recuperation and improved management can substantially increase livestock productivity, driving a shift from low- to high-capacity pas- tures as well as an increase in integrated crop-livestock-forestry systems (Figs. 2 and S7). This enables carbon sequestration not only in agricultural soils but also from forest area expansion (Fig. S8) and from the deployment of BECCS (Figs. 3 and 4a) which could make Brazil CO2- negative around 2040, despite non-CO2 gases meaning the country remains a net GHG emitter through 2050 (Fig. 1). However, although agriculture intensification can spare land for regrowth of natural vegetation (afforestation) and bioenergy production, there is evidence that intensification of agriculture may also increase land rents which may lead to further expansion of the agricultural area, increasing deforestation (Nepstad et al. 2009; Rose et al. 2013). In our results, the improved agricultural production techniques enable decarbonization of energy sectors via deployment of various forms of bioenergy, mainly sugarcane and lignocel- lulosic feedstocks. Sugarcane yields ethanol from its juice and bioelectricity from bagasse and both make large contributions in mitigation scenarios. Technical upgrades such as higher efficiency boilers in sugar and ethanol plants raise productivity and produce more electricity with less bagasse. This well-known potential efficiency improvement in Brazil is not realized today due to financial uncertainties but could be spurred with introduction of carbon revenues. First generation ethanol from sugarcane peaks around 2030 and is present through 2050 across scenarios, albeit with CCS in mitigation scenarios. This form of BECCS- liquids begins as early as 2035 but is gradually reduced due to the introduction of 2nd generation ethanol (Fig. 4a) and electrification of the LDV fleet (Fig. 4b). A more substantial contribution from liquid biofuels comes in the form of biodiesel and biokerosene with and without CCS. The lack of mature options in freight and aviation nan Climatic Change (2020) 162:1823–1842 means decarbonization of these sectors relies on bioenergy (Section 3.7), mainly from FT-BTL routes using lignocellulosic feedstock. These results point to important dynamics between (i) electrification of the LDV fleet, (ii) biofuels production and use, and (iii) land use intensification. Each has implications not only for each other but also for the country’s emissions profile and for future development of its energy-infrastructure complex. This has repercussions all the way down to the primary energy level, as seen in Fig. 3. Decarbonizing the transportation sector gives rise to these inter-sector ripple effects in energy and AFOLU sectors. These linkages are mediated by the transportation sector via biofuels, which link the AFOLU, transportation, refining, and power generation sectors. For example, sugarcane ethanol is currently used for passenger vehicles while bagasse from crushed sugarcane is burned to generate CO2-neutral9 electricity. This, coupled with large shares of hydropower, implies low-carbon electricity, which favors introduction of EVs as a mitigation option. However, higher share of EVs reduces demand for light liquid fuels, affecting the refining sector. These dynamics are subject to cost and efficiency assumptions and model results need to be interpreted with care. Nonetheless, it points to specific areas where well-designed policy can have positive impacts across sectors, reducing trade-offs and enhancing synergies. Across budget scenarios, biodiesel supplies between 68 and 73% of diesel demand in 2050, meaning refineries still have to produce some fossil diesel from crude. Due to the nature of the refining process, refineries produce some gasoline when running for diesel production, unless expensive processing units are added. This gasoline byproduct comes at a null cost and is used in spite of stringent carbon budgets. Generally, if one of the oil products remains in high demand (due to low elasticity of substitution), refineries will still produce a basket of products at a minimum level at null cost, reducing potential mitigation from their alternatives. In these results, declining refinery activity also means the model must cover residual demand for maritime bunker fuel by importing heavy fuel oil. Coupled with CCS, both bioelectricity and biofuels have the potential for significant negative emissions, which contribute to meeting the stringent emissions budgets explored in this study. As is the case with most IAMs, the value of BECCS in BLUES is driven by model structure and assumptions and also by what is not represented in the model such as governance structures, water use, or other advanced energy production routes such as power-to-liquids or power-to-hydrogen (Köberle 2019). The levels of BECCS deployment projected here will require proper governance of the whole bioenergy chain, from the production phase to final consumption and sequestration of captured CO2. Monitoring for land use and land use change emissions will need to be robust enough to prevent undesirable consequences from bioenergy feedstock production, especially with regard to prevention of deforestation, sustainable water use, and biodiversity conservation (Fajardy et al. 2019). The dynamics described here have implications for the attainment of the SDGs. As noted in Section 3.3, cultivated area increases in the short-term at the expense of natural lands, especially in the biodiversity-rich Cerrado biome of central Brazil (savannas in BLUES). This implies more deforestation to meet demands for food, fiber, and bioenergy, with potential negative impacts on biodiversity. In mitigation scenarios, deforestation peaks between 2025 and 2030, and forests begin to recover some of the lost area post-2030 (Fig. S8). This indicates that PA targets may be consistent with nature conservation in the long run but pose serious 9 Bioelectricity from sugarcane is the result of burning bagasse, a by-product of sugar or ethanol production. Because bagasse is in fact a residue of the main activity, emissions should be allocated to the other products. nan Climatic Change (2020) 162:1823–1842 threats to the attainment of SDGs by 2030, especially those related to land use, such as biodiversity (SDG 15) and food security (SDG2). Intensification of agriculture and livestock has mixed impacts on SDGs. On the one hand, recuperation of degraded pastures may well have a positive net effect since a small addition of inputs (diesel and fertilizer) may raise productivity, potentially leading to land sparing. On the other, attaining higher yield for crops may have adverse effects through increased use of diesel and fertilizers—impacting emissions (SDG13) and water quality (SDG 14), as well as through the potential need to deploy genetically modified cultivars (through genetically modified organisms—GMOs)—with potentially adverse effects on biodiversity. The net effects of climate mitigation on SDGs are uncertain and beyond the scope of the current analysis. However, for the case of Brazil, these results identify important indicators like forest area and fertilizer use that suggest challenges for the concurrent achievement of the PA and other sustainable development objectives. Such challenges can only be overcome through the implementation of well-designed policy that takes into account the cross-cutting and interrelated repercussions of individual measures. Limitations of this study include uncertainty as to the assumptions in input parameters of the BLUES model (Krey et al. 2019a). However, the complex interactions between bioelec- tricity and other power generation options, biofuels, the transportation sector, land use, and agriculture restrict the benefit of simplistic sensitivity analyses on cost parameters. Indeed, understanding the effects of cost assumptions of the various technologies represented in the BLUES model requires careful design of scenarios to test the sensitivity of the results to the input assumptions of the model. Such a sensitivity analysis is beyond the scope of this study and is reported elsewhere (Köberle et al. 2018). Nonetheless, a discussion on sensitivity tests conducted during the design of the scenarios reported here can be found in SOM Section S1.5.5 Conclusions This paper used the BLUES model to examine potential GHG mitigation options for Brazil in the context of the PA and SDGs. Results indicate that current policies are on track to only partially deliver the pledges in the Brazilian NDC by 2030, but additional policies to meet the NDC would also put the country on a least-cost pathway consistent with global efforts to attain PA objectives. Key measures include net-zero deforestation post-2030, sustainable intensifi- cation of agriculture, electrification of the LDV fleet, and decarbonization of freight transpor- tation and aviation via advanced BTL biofuels production with and without CCS. Results indicate key linkages between AFOLU, transportation, and energy sectors. They point to the fundamental role of AFOLU in Brazil’s mitigation efforts, with its emissions becoming negative before those of the energy sector. Intensification of currently extensive agricultural practices allows for bioenergy feedstock production while curbing deforestation and, in some cases, even allowing for afforestation. Bioenergy production links the land use and agriculture sectors to the energy sectors, especially transportation and power generation. This implies successful mitigation policies relying on bioenergy would need to take a systemic view of the supply chain from the agricultural phase to final consumption and sequestration of combustion CO2. Halting deforestation is the main challenge for Brazilian climate action and full implemen- tation of the Forest Code implies net-zero deforestation through 2050. With no deforestation, the land area required to grow the feedstock comes from sustainable intensification of nan Climatic Change (2020) 162:1823–1842 agriculture, especially of livestock production. Recuperation of degraded pastures and intro- duction of integrated crop–livestock–forestry systems meet demand for agricultural products using less land, with the added benefit of reducing emissions relative to current practices. In the energy sector, mitigation measures include low-carbon sources for all energy carriers, fuel switch in transportation, energy efficiency in industry and transportation, and transitions to low-carbon power generation from biomass, wind, and solar. Brazil’s high bioenergy potential makes negative emissions through BECCS a realistic option. The transportation sector stands out as having a significant potential for emissions reductions through the use of biofuels, with and without CCS,10 and the electrification of the LDV passenger fleet. Power generation is already low-carbon, meaning passenger transportation can be decarbonized through electrifi- cation. However, there are limited options to decarbonize freight transportation and aviation and these stand out as prime candidates for enhanced climate action in the country. To that end, results point to biodiesel and biokerosene made from lignocellulosic feedstock via FT synthesis as an optimal alternative. These dynamics have mixed implications for the attainment of the SDGs. In mitigation scenarios, deforestation peaks between 2025 and 2030, and forests begin to recover some of the lost area post-2030. This indicates climate action in Brazil to be consistent with the SDGs in the long-run despite some short-term challenges up to 2030. The levels of BECCS deployment projected here will require proper governance of the whole bioenergy chain and monitoring for land use and land use change emissions will need to be robust enough to prevent undesirable consequences from bioenergy feedstock production. The interlinked, cross-sectoral measures explored here require a systems approach to policy design to minimize trade-offs and maximize synergies across sectors. Acknowledgements The authors would also like to express their gratitude to the Conselho Nacional de Desenvolvimento Cientifico e Tecnológico (National Scientific and Technological Development Council—CNPq) for the essential support given for this work to be carried out. Funding This work is part of a project that has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No. 642147 (CD-LINKS). Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.References Agroicone (2016) Cattle ranching intensification as a key role on sustainable agriculture expansion in Brazil. Sao Paulo: Agroicone, INPUT Project 10 CCS here is applied in the production phase of the biofuel, not in the transportation sector per se. nan Climatic Change (2020) 162:1823–1842 Assad E, Pavão E, Jesus M de, Martins SC (2015) Invertendo o sinal de carbono da agropecuária brasileira. Uma estimativa do potencial de mitigação de tecnologias do Plano ABC de 2012 a 2023. São Paulo, Brazil: Observtório ABC Balbino LC, Barcellos A d O, Stone LF (2011) Marco referencial: integração lavoura-pecuária-floresta. Embrapa, Brasília BNEF (2017) When Will Electric vehicles be cheaper than conventional vehicles? p Budinis S, Krevor S, Dowell NM, Brandon N, Hawkes A (2018) An assessment of CCS costs, barriers and potential. Energ Strat Rev 22:61–81. https://doi.org/10.1016/j.esr.2018.08.003 Cardoso AS, Berndt A, Leytem A, Alves BJR, de Carvalho, Carvalho I d NO, de Barros Soares LH, Urquiaga S, Boddey RM (2016) Impact of the intensification of beef production in Brazil on greenhouse gas emissions and land use. Agric Syst 143:86–96. https://doi.org/10.1016/j.agsy.2015.12.007 g y p g j g y Carvalho F (2017) Evaluation of the Brazilian potential for producing aviation biofuels through consolidated routes. Federal university of Rio de Janeiro (UFRJ) Carvalho JLN, Cerri CEP, Feigl BJ, Píccolo MC, Godinho VP, Cerri CC (2009) Carbon sequestration in agricultural soils in the Cerrado region of the Brazilian Amazon. Soil Tillage Res 103:342–349. https://doi.org/10.1016/j.still.2008.10.022 Carvalho JLN, Raucci GS, Frazao LA, Cerri CEP, Bernoux M, Cerri CC (2014) Crop-pasture rotation: a strategy to reduce soil greenhouse gas emissions in the Brazilian Cerrado. Agric Ecosyst Environ 183:167–175. https://doi.org/10.1016/j.agee.2013.11.014 p g j g CAT (2017) Climate action tracker: Brazil [WWW document]. Clim. Action Tracker. URL http://climateactiontracker.org/countries/brazil.html (Accessed 6.15.17) p g CEPEL (2013) Atlas do Potencial Eólico Brasileiro. Brasil, Rio de Janeiro Cohn AS, Mosnier A, Havlík P, Valin H, Herrero M, Schmid E, O’Hare M, Obersteiner M (2014) Cattle ranching intensification in Brazil can reduce global greenhouse gas emissions by sparing land from deforestation. Proc Natl Acad Sci U S A 111:7236–7241. https://doi.org/10.1073/pnas.1307163111 Ó Ó Costa IVL da (2014) PROPOSTA DE ESTRUTURA REGULATÓRIA PARA SEQUESTRO GEOLÓGICO DE CO2 NO BRASIL E UMA APLICAÇÃO PARA O ESTADO DO RIO DE JANEIRO. Federal University of Rio de Janeiro Oliveira P De, Freitas RJ, Kluthcouski J, Ribeiro AA, Adriano L, Cordeiro M, Teixeira LP, Augusto R, Castro D, Vilela L, Balbino LC (2013) Evolução de Sistemas de Integração Lavoura-Pecuária-Floresta (iLPF): estudo de caso da Fazenda Santa Brígida, Ipameri, GO Dellink R, Chateau J, Lanzi E, Magné B (2017) Long-term economic growth projections in the shared socioeconomic pathways. Glob Environ Chang 42:200–214. https://doi.org/10.1016/j. gloenvcha.2015.06.004 EPE (2014) Consumo de Energia no Brasil Análises Setoriais. Rio de Janeiro, Brasil: Empresa de Planejame Energético (EPE) g EPE (2019) Balanço Energético Nacional. Empresa de Pesquisa Energetica (EPE) EPE (2019) Balanço Energético Nacional. Empresa de Pesquisa Energetica (EPE) Escobar H (2015) Brazil’s climate targets draw lukewarm reviews. Sci Insid https://doi.org/10.1126/science. aad4637 Fajardy, M., Köberle, AC, Mac Dowell, N., Fantuzzi, A. BECCS deployment: a reality check. Grantham Briefing Paper Number 28. London, UK: Grantham Institute, Imperial College London. Available online at https://www.imperial.ac.uk/media/imperial-college/granthaminstitute/public/publications/briefing- papers/BECCS-deployment—a-reality-check.pdf Fricko O, Havlik P, Rogelj J, Klimont Z, Gusti M, Johnson N, Kolp P, Strubegger M, Valin H, Amann M, Ermolieva T, Forsell N, Herrero M, Heyes C, Kindermann G, Krey V, McCollum DL, Obersteiner M, Pachauri S, Rao S, Schmid E, Schoepp W, Riahi K (2017) The marker quantification of the shared socioeconomic pathway 2: A middle-of-the-road scenario for the 21st century. Glob Environ Chang 42: 251–267. https://doi.org/10.1016/j.gloenvcha.2016.06.004 Gil J, Siebold M, Berger T (2015) Adoption and development of integrated crop-livestock-forestry systems in Mato Grosso, Brazil. Agric Ecosyst Environ 199:394–406. https://doi.org/10.1016/j.agee.2014.10.008 g y p g j g GofB (2015a) Intended nationally determined contribution (INDC) towards achieving the objective of the UNFCCC. Brasilia, Brasil: Presidencia da Republica GofB (2015a) Intended nationally determined contribution UNFCCC. Brasilia, Brasil: Presidencia da Republica GofB (2015b) Terceira Comunicação Nacional do Brasil à Convenção-Quadro das Nações Unidas sobre Mudança do Clima. Brasília, Brasil: Ministério da Ciência, Tecnologia e Inovação (MCTI) Gough C, Garcia-Freites S, Jones C, Mander S, Moore B, Pereira C, Röder M, Vaughan N, Welfle A (2018) Challenges to the use of BECCS as a keystone technology in pursuit of 1.5°C. Glob. Sustain. 1, e5. https://doi.org/10.1017/sus.2018.3 Herreras-Martínez S, Koberle A, Rochedo P, Schaeffer R, Lucena A, Szklo A, Ashina S, van Vuuren DP (2015) Possible energy futures for Brazil and Latin America in conservative and stringent mitigation pathways up to 2050. Tech Forcasting Soc Chang. https://doi.org/10.1016/j.techfore.2015.05.006 nan Climatic Change (2020) 162:1823–1842 IAMC (2020) Model documentation - BLUES [WWW document]. Common Integr. Assess. Model doc. URL https://www.iamcdocumentation.eu/index.php/Model Documentation - BLUES (Accessed 7.1.20) Inagaki T, Sá JC, De Oliveira Ferreira A, Briedis C, Tivet F, Romaniw J (2016) Macroagregados como indicadores de qualidade em sistema plantio direto. Rev plantio direto 151:2–8 Keppo I, Strubegger M (2010) Short term decisions for long term problems – the effect of foresight on model based energy systems analysis. Energy 2033–2042 Köberle A (2018) Implementation of land use in an energy system model to study the long-term impacts of bioenergy in Brazil and its sensitivity to the choice of agricultural greenhouse gas emission factors. Federal University of Rio de Janeiro Köberle, A.C., 2019. The value of BECCS in IAMs: a review. Curr Sustain Energy Rep 6, 107–115. https://doi. org/10.1007/s40518-019-00142-3 Koberle A, Portugal J, Branco DC, Szklo AS, Lucena AFP (2015) Assessing the potential for ethanol - fueled power generation as a flexible dispatchable option for firming intermittent sources in Brazil, in: V Encuentro Latinoamericano de Economía de La Energía. International Association for Energy Economics (IAEE), Medellin, Colombia Köberle A, Rochedo P, Portugal-Pereira J, Szklo AS, de Lucena AFP, Schaeffer R (2015) Brazil Chapter, in: Spencer, T., Pierfedericci, R. (Eds.), Beyond the numbers: understanding the transformation induced by INDCs. A Report of the MILES Project Consortium. IDDRI - MILES Project Consortium, Paris, pp. 34–45 p , , , ( ), y g y INDCs. A Report of the MILES Project Consortium. IDDRI - MILES Project Consortium, Paris, pp. 34–45 Köberle AC, Tagomori IS, Portugal-Pereira J, Schaeffer R (2017) Policy case study: evaluation of the Plano ABC in Brazil. Deliverable for CD-LINKS project WP1 Köberle AC, Tagomori IS, Portugal-Pereira J, Schaeffer R (2017) Policy case study: evaluation of the Plano ABC in Brazil. Deliverable for CD-LINKS project WP1 p j Köberle AC, Garaffa R, Cunha SB, Rochedo PR, Lucena AF, Szklo A, Schaeffer R (2018) Are conventional energy megaprojects competitive? Suboptimal decisions related to cost overruns in Brazil. Energy Policy 122:689–700. https://doi.org/10.1016/j.enpol.2018.08.021 Krey V, Guo F, Kolp P, Zhou W, Schaeffer R, Awasthy A, Bertram C, de Boer HS, Fragkos P, Fujimori S, He C, Iyer G, Keramidas K, Köberle AC, Oshiro K, Reis LA, Shoai-Tehrani B, Vishwanathan S, Capros P, Drouet L, Edmonds JE, Garg A, Gernaat DEHJ, Jiang K, Kannavou M, Kitous A, Kriegler E, Luderer G, Mathur R, Muratori M, Sano F, van Vuuren DP (2019a) Looking under the hood: A comparison of techno-economic assumptions across national and global integrated assessment models. Energy 172:1254–1267. https://doi. org/10.1016/j.energy.2018.12.131 Krey V, Riahi K, Hasegawa T, Luderer G, Poblete-Cazenave M, Parkinson S, Rafaj P, Vrontisi Z, Arvesen A, Fragkiadakis K, Fricko O, Fujimori S, Rogelj J (2019b) Implications of the Paris agreement for achieving the sustainable development goals. Nat Clim Chang submitted p g g Kriegler E, Ueckerdt F, Luderer G, Schaeffer R, Chen W, Gi K, Iyer GC, Kejun J, Krey V, Mathur R, Oshiro K, Roelfsema M, Safonov G, Soest HL van, Vishwanathan SS, Vrontisi Z, Després J, Drouet L, Harmsen M, K b l A V D B t C B tti V C P Ed d J E li J E l l S Kriegler E, Ueckerdt F, Luderer G, Schaeffer R, Chen W, Gi K, Iyer GC, Kejun J, Krey V, Mathur R, Oshiro K, Roelfsema M, Safonov G, Soest HL van, Vishwanathan SS, Vrontisi Z, Després J, Drouet L, Harmsen M, Koberle A, Vuuren D van, Bertram C, Bosetti V, Capros P, Edmonds J, Emmerling J, Evangelopoulou S, Frank S, Fricko O, Fujimori S, Garg A, He C, Huppmann D, Kitous A, Li N, Potashnikov V, Rochedo P, Shekhar S, Stevanović M, Wang H, Baumstark L, Elzen M den, Höhne N, Iacobuta G, Riahi K (2019) Paris Agreement suggests deep collective CO2 emissions reductions in the seven largest greenhouse gas emitting regions by mid century. Nat Clim Chang under revi Lucena AFP, Clarke L, Schaeffer R, Szklo A, Rochedo PRR, Nogueira LPP, Daenzer K, Gurgel A, Kitous A, Kober T (2016) Climate policy scenarios in Brazil: A multi-model comparison for energy. Energy Econ 56: 564–574. https://doi.org/10.1016/j.eneco.2015.02.005 MAPA (2012) Plano Setorial de Mitigação e Adaptação às Mudanças Climáticas para Consolidação da Economia de Baixa Emissão de Carbono na Agricultura – PLANO ABC. Brasilia: Ministério da Agricultura, Pecuária e Abastecimento (MAPA) MCTIC (2016) Estimativas Anuais de Emissões de Gases de Efeito Estufa no Brasil. Brasilia: Ministério da Ciencia, Tecnologia, Inovação e Comunicação (MCTIC) Merschmann PR d C, Szklo AS, Schaeffer R (2016) Technical potential and abatement costs associated with the use of process emissions from sugarcane ethanol distilleries for EOR in offshore fields in Brazil. Int J Greenhouse Gas Control 52:270–292. https://doi.org/10.1016/j.ijggc.2016.07.007 Nepstad D, Soares-Filho BS, Merry F, Lima A, Moutinho P, Carter J, Bowman M, Cattaneo A, Rodrigues H, Schwartzman S, McGrath DG, Stickler CM, Lubowski R, Piris-Cabezas P, Rivero S, Alencar A, Almeida O, Stella O (2009) The end of deforestation in the Brazilian Amazon—supporting online material (SOM). Science 326:1–28. https://doi.org/10.1126/science.1182108 Nilsson M, Chisholm E, Griggs D, Howden-Chapman P, McCollum D, Messerli P, Neumann B, Stevance AS, Visbeck M, Stafford-Smith M (2018) Mapping interactions between the sustainable development goals: lessons learned and ways forward. Sustain Sci 13:1489–1503. https://doi.org/10.1007/s11625-018-0604-z Pan X, den Elzen M, Höhne N, Teng F, Wang L (2017) Exploring fair and ambitious mitigation contributions under the Paris Agreement goals. Environ Sci Pol 74:49–56. https://doi.org/10.1016/j.envsci.2017.04.020 nan Climatic Change (2020) 162:1823–1842 Peters GP, Geden O (2017) Catalysing a political shift from low to negative carbon. Nat Clim Chang 7:619–621. https://doi.org/10.1038/nclimate3369 Portugal-Pereira J, Koberle A, Soria R, Lucena A, Szklo A, Schaeffer R (2016) Overlooked impacts of electricity expansion optimisation modelling: the life cycle side of the story. Energy 1–12. https://doi.org/10.1016/j. energy.2016.03.062energy.2016.03.062 gy Rathmann R, Szklo A, Schaeffer R (2012) Targets and results of the Brazilian biodiesel incentive program— it reached the promised land? Appl Energy 97:91–100. https://doi.org/10.1016/j.apenergy.2011.11.021 ann R, Szklo A, Schaeffer R (2012) Targets and results of the Brazilian biodiesel incentive program—has eached the promised land? Appl Energy 97:91–100. https://doi.org/10.1016/j.apenergy.2011.11.021 Riahi K, van Vuuren DP, Kriegler E, Edmonds J, O’Neill BC, Fujimori S, Bauer N, Calvin K, Dellink R, Fricko O, Lutz W, Popp A, Cuaresma JC, Samir KC, Leimbach M, Jiang L, Kram T, Rao S, Emmerling J, Ebi K, Hasegawa T, Havlik P, Humpenöder F, Da Silva LA, Smith S, Stehfest E, Bosetti V, Eom J, Gernaat D, Masui T, Rogelj J, Strefler J, Drouet L, Krey V, Luderer G, Harmsen M, Takahashi K, Baumstark L, Doelman JC, Kainuma M, Klimont Z, Marangoni G, Lotze-Campen H, Obersteiner M, Tabeau A, Tavoni M (2017) The shared socioeconomic pathways and their energy, land use, and greenhouse gas emissions implications: an overview. Glob Environ Chang 42:153–168. https://doi.org/10.1016/j. gloenvcha.2016.05.009 Rochedo P (2016) Development of A global integrated energy model to evaluate the Brazilian role in climate change mitigation scenarios. Federal University of Rio de Janeiro Rochedo PRR, Costa IVL, Império M, Hoffmann BS, Merschmann PR d C, Oliveira CCN, Szklo A, Schaeffer R (2016) Carbon capture potential and costs in Brazil. J Clean Prod 131:280–295. https://doi.org/10.1016/j. jclepro.2016.05.033 Rochedo PRR, Filho BS, Viola E, Schaeffer R, Szklo A, Lucena AFP, Köberle A, Davis JL, Rajão R, Rathmann R (2018) The threat of political bargaining to climate mitigation in Brazil. Nat Clim Chang Roelfsema M, van Soest HL, Harmsen M, van Vuuren DP, Bertram C, den Elzen M, Höhne N, Iacobuta G, Krey V, Kriegler E, Luderer G, Riahi K, Ueckerdt F, Després J, Drouet L, Emmerling J, Frank S, Fricko O, Gidden M, Humpenöder F, Huppmann D, Fujimori S, Fragkiadakis K, Gi K, Keramidas K, Köberle AC, Aleluia Reis L, Rochedo P, Schaeffer R, Oshiro K, Vrontisi Z, Chen W, Iyer GC, Edmonds J, Kannavou M, Jiang K, Mathur R, Safonov G, Vishwanathan SS (2020) Taking stock of national climate policies to evaluate implementation of the Paris agreement. Nat Commun 11:2096. https://doi.org/10.1038/s41467-020- 15414-6 Rogelj J, den Elzen M, Hohne N, Fransen T, Fekete H, Winkler H, Schaeffer R, Sha F, Riahi K, Meinshausen M (2016) Paris Agreement climate proposals need boost to keep warming well below 2°C. Nature 534:631– 639. https://doi.org/10.1038/nature18307 Rogelj J, Shindell D, Jiang K, Fifita S, Forster P, Ginzburg V, Handa C, Kheshgi H, Kobayashi S, Kriegler E, Mundaca L, Séférian R, Vilariño MV (2018) Mitigation pathways compatible with 1.5°C in the context of sustainable development, in: Flato, G., Fuglestvedt, J., Mrabet, R., Schaeffer, R. (Eds.), Global warming of 1.5 °C: an IPCC special report on the impacts of global warming of 1.5 °C above pre-industrial levels and related global greenhouse gas emission pathways, in the context of strengthening the global response to the threat of climate change. Intergorvernmental Panel on Climate Change (IPCC) g g g ( ) Rose SKR, Glub AA, Sohngen B (2013) Total factor and relative agricultural productivity and deforestation. Am J Agric Econ 95:426–434. https://doi.org/10.1093/ajae/aas113 Rose SKR, Glub AA, Sohngen B (2013) Total factor and relative ag J Agric Econ 95:426–434. https://doi.org/10.1093/ajae/aas113 Roy J, Tschakert P, Waisman H, Halim SA, Antwi-Agyei P, Dasgupta P, Hayward B, Kanninen M, Liverman D, Okereke C, Pinho PF, Riahi K, Rodriguez AGS (2018) Sustainable Development, Poverty Eradication and Reducing Inequalities, in: Masson-Delmotte, V., Zhai, P., Pörtner, H.-O., Roberts, D., Skea, J., Shukla, P.R., Pirani, A., Moufouma-Okia, W., Péan, C., Pidcock, R., Connors, S., Matthews, J.B.R., Chen, Y., Zhou, X., Gomis, M.I., Lonnoy, E., Maycock, T., Tignor, M., Waterfield, T. (Eds.), Global Warming of 1.5°C. An IPCC Special Report on the Impacts of Global Warming of 1.5°C above Pre-Industrial Levels and Related Global Greenhouse Gas Emission Pathways, in the Context of Strengthening the Global Response to the Threat of Climate Change,. In Press Schaeffer R, Köberle A, van Soest HL, Bertram C, Luderer G, Riahi K, Krey V, Vuuren DP van, Kriegler E, Fujimori S, Chen W, He C, Vrontisi Z, Vishwanathan S, Garg A, Mathur R, Shekhar S, Oshiro K, Ueckerdt F, Safonov G, Iyer G, Gi K, Potashnikov V (2020) Comparing transformation pathways across major economies. Clim Change this issue SEEG (2019) Total emissions by sector [WWW document]. Sist. Estim. Emiss. e Remoções Gases Efeito Estufa. URL http://plataforma.seeg.eco.br/total_emission (Accessed 7.1.20) Simioni T, Schaeffer R (2019) Georeferenced operating-efficiency solar potential maps with local weather conditions—an application to Brazil. Sol Energy 184:345–355. https://doi.org/10.1016/J. SOLENER.2019.04.006 Smeets EMW, Faaij APC (2010) The impact of sustainability criteria on the costs and potentials of bioenergy production—applied for case studies in Brazil and Ukraine. Biomass Bioenergy 34:319–333. https://doi. org/10.1016/J.BIOMBIOE.2009.11.003 nan Climatic Change (2020) 162:1823–1842 Soares-filho B, Rajão R, Macedo M, Carneiro A, Costa W, Coe M, Rodrigues H, Alencar A (2014) Cracking Brazil’s forest code. Science 344:363–364 Strassburg BBN, Latawiec AE, Barioni LG, Nobre C a, da Silva VP, Valentim JF, Vianna M, Assad ED (2014) When enough should be enough: improving the use of current agricultural lands could meet production demands and spare natural habitats in Brazil. Glob Environ Chang 28:84–97. https://doi.org/10.1016/j. gloenvcha.2014.06.001 Szklo A, Lucena A, Schaeffer R, Filho BSS, Davis JL, Rajão R, Haddad EA, Domingues EP, Campos A, Koberle A, Ribeiro A, Lima E, Barbosa FA, Império M, Rochedo P, Maia S, Costa WLS (2017) Modelagem integrada e impactos econômicos de opções setoriais de baixo carbono. Brasília: Ministério da Ciência, Tecnologia, Inovações e Comunicações, ONU Meio Ambiente ia, Inovações e Comunicações, ONU Meio Ambiente Szklo A, Lucena AF, Schaeffer R, Soares Filho BS, Davis JL, Rajão R, Haddad EA, Domingues EP, Campos A, Köberle AC, Ribeiro A, Lima E, Barbosa FA, Imperio M, Rochedo PRR, Maia S, Costa WLS (2018) Modelagem integrada e impactos econômicos de opções setoriais de baixo carbono. Brasilia, Brasil: Ministério da Ciência, Tecnologia, Inovações e Comunicações, ONU Meio Ambiente É Ô Tagomori I (2017) POTENCIAL TÉCNICO E ECONÔMICO PARA A PRODUÇÃO de FISCHER- TROPSCH DIESEL A PARTIR DE BIOMASSA (FT-BTL) ASSOCIADA À CAPTURA DE CARBONO NO BRASIL. Federal University of Rio de Janeiro Tagomori IS, Carvalho FM, da Silva FTF, Paulo PR, Rochedo PRR, Szklo A, Schaeffer R (2018) Designing an optimum carbon capture and transportation network by integrating ethanol distilleries with fossil-fuel processing plants in Brazil. Int J Greenhouse Gas Control 68:112–127. https://doi.org/10.1016/j. ijggc.2017.10.013 jgg Tagomori IS, Rochedo PRR, Szklo A (2019) Biomass and bioenergy techno-economic and georeferenced analysis of forestry residues-based Fischer-Tropsch diesel with carbon capture in Brazil. Biomass Bioenergy 123:134–148. https://doi.org/10.1016/j.biombioe.2019.02.018 UNFCCC (2015) Paris agreement, conference of the parties on its twenty-first session. UN Framew Convention on CLimate Chenage (UNFCCC), Paris, France. FCCC/CP/2015/L.9/Rev.1 van den Berg NJ, van Soest HL, Hof AF, den Elzen MGJ, van Vuuren DP, Chen W, Drouet L, Emmerling J, Fujimori S, Höhne N, Kõberle AC, McCollum D, Schaeffer R, Shekhar S, Vishwanathan SS, Vrontisi Z, Blok K (2019) Implications of various effort-sharing approaches for national carbon budgets and emission pathways. Clim Chang. https://doi.org/10.1007/s10584-019-02368-y von Stechow C, Minx JC, Riahi K, Jewell J, McCollum DL, Callaghan MW, Bertram C, Luderer G, Baiocchi G (2016) 2 °C and SDGs: united they stand, divided they fall? Environ Res Lett 11:34022. https://doi. org/10.1088/1748-9326/11/3/034022 Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.Affiliations Alexandre C. Köberle1,2 & Pedro R. R. Rochedo1 & André F. P. Lucena1 & Alexandre Szklo1 & Roberto Schaeffer1 1 Centre for Energy and Environmental Economics-CENERGIA, Energy Planning Program, Graduate Schoo of Engineering, Universidade Federal do Rio de Janeiro, Centro de Tecnologia, Bloco C, Sala 211 Cidad Universitária, Ilha do Fundão, Rio de Janeiro, RJ 21941-972, Brazil 1 Centre for Energy and Environmental Economics-CENERGIA, Energy Planning Program, Graduate School of Engineering, Universidade Federal do Rio de Janeiro, Centro de Tecnologia, Bloco C, Sala 211 Cidade Universitária, Ilha do Fundão, Rio de Janeiro, RJ 21941-972, Brazil 2 Grantham Institute, Imperial College London, London SW7 2AZ, UK nantable_1Engenharia Agrícola ISSN: 1809-4430 (on-line) www.engenhariaagricola.org.brSpecial Issue: Energy in Agriculture Scientific Paper Doi: http://dx.doi.org/10.1590/1809-4430-Eng.Agric.v43nepe20220160/2023 f p oi: http://dx.doi.org/10.1590/1809-4430-Eng.Agric.v43nepe20220160/2023GEOTHERMAL ENERGY: AN ALTERNATIVE TO THE WATER–ENERGY DILEMMA IN NORTHEASTERN BRAZIL Francisco T. G. Lima Verde Neto1, Paulo A. C. Rocha1, Jenyffer da S. G. Santos2,GEOTHERMAL ENERGY: AN ALTERNATIVE TO THE WATER–ENERGY DILEMMA IN NORTHEASTERN BRAZILFrancisco T. G. Lima Verde Neto1, Paulo A. C. Rocha1, Jenyffer da S. G. Santos2, Angel P. Garcia2, Daniel Albiero2* Francisco T. G. Lima Verde Neto1, Paulo A. C. Rocha1, Jenyffer da S. G. Santos2 Angel P. Garcia2, Daniel Albiero2* 2*Corresponding author. Universidade Estadual de Campinas (UNICAMP)/Campinas - SP, Brasil. E-mail: daniel.albiero@gmail.com | ORCID ID: https://orcid.org/0000-0001-6877-8618ABSTRACTKEYWORDS This paper proposes geothermal energy as an alternative solution to the water–energy dilemma in the northeastern region of Brazil (NEB). The main application of this study was to provide a theoretical basis to support a different approach to policies minimizing water scarcity and ensuring sustainability. The analysis developed in this study compares the levelized cost of electricity (LCOE) for many different energy sources. The novelty of this study is the use of geothermal energy in the context of the Brazilian Northeast, focusing on water desalination processes, which are expensive in terms of energy. Therefore, this study is highly important because it offers the potential of addressing the energy/economic barrier related to water desalination in environments with economically viable geothermal energy. This is the case in Northeast Brazil with potential for reuse of abandoned oil wells. In the form of enhanced geothermal systems (EGS), geothermal energy is a competitive energy source compared to other sources in the Brazilian Energy Matrix, especially when considering factors in addition to the economic benefits. In the form of EGS, geothermal energy is a suitable option for addressing water scarcity in the northeast region in a sustainable and low-emission manner. This is a strategic opportunity for NEB in the context of energy production and freshwater production through desalination. renewable energies, water–energy nexus, levelized cost, engineered geothermal systems, desalination.INTRODUCTION Droughts are one of the main contributing factors to the social and economic problems and structural deficiencies of the region. Water scarcity and low rainfall in the NEB are substantial issues. This is predominantly because most of the groundwater in the region is saline, with desalination being a potential solution to the problem (Silva et al., 2018). Renewable energy technologies are a sustainable alternative to fossil fuels. The development of these technologies has been highly responsive to general energy policy guidelines and environmental and social goals, such as diversifying energy carriers, improving access to clean energy, and reducing pollution and dependence on fossil and imported fuels (Turkenburg, 2000; Vogt et al., 2021). Most of the energy used in desalination processes is derived from fossil fuels, which are becoming increasingly depleted and emit carbon dioxide. Incorporating renewable energy sources into the desalination process can increase Among these alternatives, enhanced geothermal systems (EGS) involve injecting fluids at low temperatures, which results in the same flow through high-temperature 1 Universidade Federal do Ceará (UFC)/Fortaleza - CE, Brasil. 2 Universidade Estadual de Campinas (UNICAMP)/Campinas - SP, Brasil. Area Editor: Henrique Vieira de Mendonça Received in: 4-18-2022 Accepted in: 12-22-2022 E 1 Universidade Federal do Ceará (UFC)/Fortaleza - CE, Brasil. 2 Universidade Estadual de Campinas (UNICAMP)/Campinas - SP, Brasil. Area Editor: Henrique Vieira de Mendonça Received in: 4-18-2022 Accepted in: 12-22-2022 Engenharia Agrícola, Jaboticabal, v.43, special issue, e20220160, 2023 Edited by SBEA Francisco T. G. Lima Verde Neto, Paulo A. C. Rocha, Jenyffer da S. G. Santos, et al. regions of the Earth’s crust. This fluid is also produced at another point and used as a heat source for diverse applications, such as binary cycles for energy production. EGS have been developed on at least 18 sites in countries, including Australia, Germany, the USA, China, and the Philippines. Research efforts in different regions have generally focused on government efforts to develop this technology, with an expectation of supplying approximately 70 GWe of power by 2050 (Lu, 2018). Cost-effective desalination can be achieved on resources at 90 °C temperatures and can be improved if technological developments are made (Loutatidou & Arafat, 2015). Brazil. The energy and economic viability of this proposal was also demonstrated.MATERIAL AND METHODS The analysis developed in this study was used to compare the levelized cost of electricity (LCOE) for many different energy sources. The data for EGS were obtained from the Geothermal Electric Technology Evaluation Model (GETEM) free software (DOE, 2018). It consists of three main information categories, namely, the resource temperature, resource depth, and the method of extraction (Hydrothermal or EGS). Abandoned oil and gas wells can be used to implement a desalination system that uses abandoned wells as a heat source to desalinate seawater (Noorollahi et al., 2017). Ali et al. (2018) noted that geothermal resources can be used for different desalination processes because geothermal resources are an uninterrupted source of thermal energy. In some places, geothermal water can be cost- effective, costing as low as 1 Euro/m³ (Tomaszewska et al., 2018), including in regions with water scarcity, such as the Gulf Coast, Sub Saharan, and Middle East, and North Africa (where the cost reaches 1.61–2.0 US$ /m³). It can represent a solution for providing freshwater with low emissions and a competitive cost, and it is cheaper than methods that use PV cells (Chandrasekharam et al., 2018). This model can be used to estimate the leveled cost of electricity (LCOE) for a user-defined geothermal resource type, temperature, and depth. With this information, the GETEM is used to estimate the generation cost using a set of default inputs based on several resource scenarios defined and evaluated by the DOE Geothermal Technologies Office (GTO). The costs, performance, and LCOE based on these default inputs are displayed in the model as default scenarios. A GETEM user can develop an alternative scenario by revising the selected default inputs up to ~109 total for the power plant, well field, exploration, confirmation, operation and maintenance, geothermal pumping, reservoir performance, and economic parameters used. The model then displays the values used in the default scenario. These values can be retained for scenario evaluation or can be revised. As the inputs are revised, the LCOE for the revised scenario (shown at the top of the page) will change. Figure 1 shows the main screen of the GETEM used in this work, as seen on the sheet “Start Here”: Therefore, the purpose of this study was to analyze geothermal energy as a suitable option for NEB. The main contribution of this study was to examine the potential of using abandoned oil wells as geothermal energy generators to enhance water desalination systems in northeastern nanFIGURE 1. Main screen of GETEM. The geothermal data was sourced from Carneiro et al. (2017), and we performed essential statistical characterization, namely, the mean, minimum, and maximum, as shown in Table 1, to infer the influence of resource depth and geothermal gradient on the LCOE. TABLE 1. Geological data for 89 sites across the northeastern region (Carneiro et al., 2017). Engenharia Agrícola, Jaboticabal, v.43, special issue, e20220160, 2023 Geological data for 89 sites Across the Northeastern Region Geothermal Gradient (°C/km) Minimum 7,0 Mean 31,3 Maximum 123,0 Minimum Mean Maximum Geothermal energy: an alternative to the water–energy dilemma in Northeastern BrazilTherefore, data for the geothermal gradient were used for 11 different values, varying from 20 °C/km to 110 °C/km at three different depths from 1500 to 3000 m, assuming a soil temperature of 25 °C. The inputs for the GETEM are listed in Table 2.TABLE 2. Temperature inputs on GETEM for the Gradient x Depth Analysis. TABLE 2. Temperature inputs on GETEM for the Gradient x Depth Analysis. Depth (m) Geothermal Gradient (°C/km) 1500 2000 2500 20 55 65 75 30 70 85 100 40 85 105 125 50 100 125 150 60 115 145 175 70 130 165 200 80 145 185 225 90 160 205 250 100 175 225 275 110 190 245 300 120 205 265 325We compared eight sites that are well known for their high geothermal gradients, as shown in Table 3, and they were then compared at different depths from 1000 to 3000 m. TABLE 3. Geological data for well-known locations across the northeastern Region (Carneiro et al., 2017). Site State Geothermal Gradient (°C/km) Specific Heat Flow (mW/m²) ANT. NAVARRO PB 65,0 195 CAMINDE CE 76,2 229 PARAMOTI CE 79,1 237 QUIXADA CE 82,7 248 CRATEUS CE 86,2 259 FORTALEZA CE 99,8 299 CARIDADE CE 99,9 300 BRE. M. DEUS PE 123,0 370. Geological data for well-known locations across the northeastern Region (Carneiro et al., 2017).E 3. Geological data for well-known locations across the northeastern Region (Carneiro et al., 2017).TABLE 3. Geological data for well-known locations across the northeastern Region (Carneiro et The inputs on GETEM are shown in Table 4: TABLE 4. Temperature inputs on GETEM for depth analysis for specific sites. Depth (m) Site 1000 1500 2000 2500 3000 ANT. NAVARRO 90,0 122,5 155,0 187,5 220,0 CAMINDE 101,2 139,3 177,4 215,5 253,6 PARAMOTI 104,1 143,7 183,2 222,8 262,3 QUIXADA 107,7 149,1 190,4 231,8 273,1 CRATEUS 111,2 154,3 197,4 240,5 283,6 FORTALEZA 124,8 174,7 224,6 274,5 324,4 CARIADE 124,9 174,9 224,8 274,8 324,7 BRE. M. DEUS 148,0 209,5 271,0 332,5 394,0* (*) The geothermal gradient approach for this site would likely reach a temperature greater than the critical temperature for the water (374, 15 °C) Therefore we excluded this data from the calculationsTABLE 4. Temperature inputs on GETEM for depth analysis for specific sites.(*) The geothermal gradient approach for this site would likely reach a temperature greater than the critical temperature for the water (374, 15 °C). Therefore, we excluded this data from the calculations. Engenharia Agrícola, Jaboticabal, v.43, special issue, e20220160, 2023 Francisco T. G. Lima Verde Neto, Paulo A. C. Rocha, Jenyffer da S. G. Santos, et al.RESULTS AND DISCUSSION This technical recommendation is based on the mechanisms of binary power plants, which are mainly dependent on the water phase diagram. At higher temperatures, water emerges on the surface with high enthalpy, and there would be a substantial amount of energy to be extracted from the steam. Therefore, flash power plants would be more suitable for use than binary power plants. We initially worked through the inputs established in Table 2 with the combinations of GETEM simulations, only selecting the binary power plant option. This was despite the technical recommendation that for temperatures above 200 °C, the best option would be flash power plants. The results are presented in Table 5. TABLE 5. Results from the GETEM simulations (US$/MWh), taking the binary cycle as the default. Depth (m) Geothermal Gradient (°C/km) 1500 2000 2500 20 27132,2 3587,9 2258,6 30 2258,3 1219,6 787,4 40 1055,8 581,6 391,4 50 590,1 339,7 252,7 60 369,2 233,5 186,4 70 267,7 188,9 149,2 80 202,9 148,5 120 90 171,3 123,3 112,1 100 140,4 104,6 112,6 110 123,2 98,9 136,7 120 107,2 96,8 173,2TABLE 5. Results from the GETEM simulations (US$/MWh), taking the binary cycle as the default.GETEM simulations (US$/MWh), taking the binary cycle as the default. The need for clean, reliable, and cost-competitive energy is at the core of the Indonesian energy policy challenge. One of the solutions that has been provided by the World Bank is a loan to develop the potential of geothermal energy. LCOE is one of the many criteria that can be used to assess the decision to promote an energy source. However, if we try to monetize all the impacts, which can vary substantially for each energy source, we can have different results. This is because the criteria to monetize such effects and impacts would ignore fundamental value judgments and essential themes, such as wildlife and ecosystems. This may create an illusory method to compare that keeps policies and stakeholders outside the decision process. For some values with a geothermal gradient more significant than 100 °C/km, the LCOE increased with a specific gradient. This observation can be explained because of the efficiency parameters for the binary cycle, and it is in line with the technical recommendation of GETEM if we consider that for geothermal gradients above 90 (in Table 2), the resource temperatures would be above 200 °C. Figure 2 shows the influence of the geothermal gradient on the LCOE for different established resource depths. This information is especially useful because when searching for opportunities to start a geothermal power plant, an objective criterion can be established to start exploratory research based on this criterion, predominantly when focusing on energy/economic efficiency issues. These are essential for evaluating a project with a low-carbon transition perspective (Albiero et al., 2015). The International Energy Association expects that geothermal electricity generation will grow from 87 TWh in 2017 to 277–555 TWh in 2040. Depending on the policy scenario (IEA, 2018), a 218–538% increase will occur. This places geothermal electricity generation with a faster rate than biomass and represents a strategic opportunity for the energy industry. Comparing 2017 to 2016, geothermal electricity was the only renewable energy source that had capacity growth. The importance of geothermal energy as a player in the energy industry has increased worldwide. Countries, such as Costa Rica, El Salvador, Iceland, Kenya, and the Philippines, produce a substantial portion, comprising approximately 20% of their electricity from geothermal energy (Fridleifsson et al., 2008). Geothermal energy does not have the uncertainties associated with renewable energy, and it can be an essential source of carbon-free development in countries, such as Indonesia, where a trilemma for developing countries is present. We performed GETEM simulations using GETEM standards, i.e., using binary power plants for resources below 200 °C and flash power plants for resources with temperatures greater than 200 °C. The results are shown in Table 6 and Figure 2. Engenharia Agrícola, Jaboticabal, v.43, special issue, e20220160, 2023 Geothermal energy: an alternative to the water–energy dilemma in Northeastern Brazil The decrease in LCOE has shown a behavioral change, especially when the conversion technology changes. This has indicated a difference between the costs of operating a flash power plant and a binary power plant. In some cases (2500 m; gradient greater than 80 °C/km), the LCOE values are lower for binary power plants than for flash power plants. in a mixture of steam and liquid. Binary power plants use the geofluid to heat a chosen working fluid in a closed cycle (DiPippo, 2015). In this study, the same type of power plants were analyzed to promote the use of this technology in Brazil, given that for temperatures of geofluid lower than 150 °C, it is “difficult, although not impossible, to build a flash-steam power that can efficiently and economically put such a resource to use” (DiPippo, 2015). y p p p p Flash power plants are a simple way of generating energy because they are used when the geofluid is producedTABLE 6. Results for the GETEM simulations (US$/MWh), with the default software settings. TABLE 6. Results for the GETEM simulations (US$/MWh), with the default software settings. Depth (m) Geothermal Gradient (°C/km) 1500 2000 2500 20 27132,2 3587,9 2258,6 30 2258,3 1219,6 787,4 40 1055,8 581,6 391,4 50 590,1 339,7 252,7 60 369,2 233,5 186,4 70 267,7 188,9 149,2 80 202,9 148,5 144,4 90 171,3 123,3 118,2 100 140,4 104,6 103,3 110 123,2 98,9 92 120 107,2 96,6 82,6 The examples in Table 4 meet the criteria of having a geothermal gradient greater than 60 °C. Taking this into account, we have obtained the results shown in Table 7 and Figure 3. Engenharia Agrícola, Jaboticabal, v.43, special issue, e20220160, 2023 Francisco T. G. Lima Verde Neto, Paulo A. C. Rocha, Jenyffer da S. G. Santos, et al. It is important to clarify the definition of the criterion 60 °C/km. Figure 4 Table 7 show that for every depth simulated, geothermal gradients greater than 60 °C/km provide LCOE lower than the sources that are currently operating in Brazil. Therefore, this “breaking point” highlights an opportunity to start an exploratory campaign to find geothermal resources.TABLE 7. Results for GETEM simulations using the software default settings. TABLE 7. Results for GETEM simulations using the software default settings. LCOE (US$/MWh) Depth (m) ANT. NAVARRO CAMINDE PARAMOTI QUIXADA CRATEUS FORTALEZA CARIDADE BRE. M. DEUS 1000 736,3 488,2 445,0 392,8 354,5 255,8 255,3 169,6 1500 310,0 230,8 206,4 192,5 181,5 150,0 149,8 129,2 2000 207,9 158,6 150,9 141,7 132,6 127,8 127,6 93,6 2500 167,6 155,4 146,7 137,7 130,1 103,6 103,4 79,9 3000 171,0 131,4 125,2 118,1 112,1 92,9 92,8 - The LCOE between geothermal energy and other energy sources was then compared using these data (Figure 4). E between geothermal energy and other energy sources was then compared using these data (Figure 4).The LCOE between geothermal energy and other energy sources was then compared using these data (Figure 4). Engenharia Agrícola, Jaboticabal, v.43, special issue, e20220160, 2023 Geothermal energy: an alternative to the water–energy dilemma in Northeastern Braziltable_2FIGURE 4. LCOE variation In Europe, geothermal energy can be cost- competitive and may win the renewable cost challenge. However, further research is required to reduce the investment risks (Clauser & Ewert, 2018). Geothermal systems can be a source of baseload energy production because they are not weather dependent and can be a renewable option for developing baseload capacity in energy production in countries, such as Turkey (Melikoglu, 2017).TABLE 8. Qualitative comparison between renewable energy sources (Long, 2009). TABLE 8. Qualitative comparison between renewable energy sources (Long, 2009). Energy Source Capacity Factor (%) Reliability Environmental Impact Main Use Geothermal 86 - 95 Reliable and Continuous Minimum Use of Soil Electrical Energy Biomass 83 Reliable Use of Fertile Lands Transportation, Heat, Electrical Energy Hydroelectric 30-35 Weather Related Dam Construction Electrical Energy Wind 25-40 Weather Related Large Occupation Electrical Energy Solar 24-33 Weather Related Large Occupation Electrical Energy Engenharia Agrícola, Jaboticabal, v.43, special issue, e20220160, 2023 Francisco T. G. Lima Verde Neto, Paulo A. C. Rocha, Jenyffer da S. G. Santos, et al. GETEM has presented a more detailed description of costs (Figure 5).table_3technology, which relies heavily on well construction costs and technology. For EGS, well field capital and exploratory costs are nearly one-third of all costs impacting the LCOE. These data are critical because according to the Brazilian National Petroleum Agency, data on the overall production for every oil and gas well onshore in the Northeast Region can be obtained. There are approximately 3800 wells that did not produce either oil or gas from a total of approximately 10500 wells (ANP, 2018). There are extensive opportunities for exchange between the oil and gas industry and an eventual geothermal energy power plant using EGS Geothermal energy can be used to smooth the substantial change in services and rig demand caused by the volatility of oil prices. The maintenance of jobs in the region, especially in places that are highly dependent on the oil and gas industries, should also be considered by policymakers. Some regions in Brazil have the potential use of technologies needed to develop EGS in parallel with other O&G, such as shale gas, especially in the Sergipe– Alagoas basin (Péres et al., 2016) (Figure 6). The installed capacity of the NEB electricity generation should also be considered (Table 9). The installed capacity has 31% reliance on thermal sources, with this number increasing to 79% in the state of Paraíba and to greater than 50% in the states of Ceará, Pernambuco, and Maranhão (Hanbury & Vasquez, 2018). Given that thermal energy generally consumes a substantial amount of freshwater to generate energy, geothermal energy can be considered suitable for regions with water-supply issues, as noted by Chandrasekharam et al. (2018). In the context of freshwater production through desalination, traditional thermal energy is economically attractive. However, geothermal energy has a level of emissions that is three orders of magnitude lower than that of coal (Hanbury & Vasquez, 2018). The emissions for geothermal energy and EGS are generally lower than the emissions from fossil fuel energy sources (Table 10). The high values for emissions for geothermal energy (Bayer et al., 2013) originated from the specifics of different technologies, with higher values associated with non- Engenharia Agrícola, Jaboticabal, v.43, special issue, e20220160, 2023 Geothermal energy: an alternative to the water–energy dilemma in Northeastern Brazil condensable gas in flash power plants. Closed cycles such as EGS tend to have zero emissions associated with the geofluid in the operational phase. Most of the CO2 emissions from EGS are derived from the consumption of diesel in construction and operations (Tomasini- Montenegro et al., 2017). TABLE 9. Installed capacity of electricity generation (GW). Source: (Hanbury & Vasquez (2018). Total Hydro Thermal Wind Solar Nuclear Nordeste 32505 11568 10089 10157 691 0 Maranhão 3388 662 2505 221 0 0 Piauí 1834 119 68 1408 240 0 Ceará 3715 1 1934 1775 5 0 Rio Grande do Norte 4161 0* 511 3533 117 0 Paraíba 775 5 613 157 0* 0 Pernambuco 3500 764 1964 762 10 0 Alagoas 4044 3725 319 0* 0* 0 Sergipe 1707 1581 91 35 0* 0 Bahia 9381 4711 2085 2267 319 0 *negligible valuesLE 9. Installed capacity of electricity generation (GW). Source: (Hanbury & Vasquez (2018).TABLE 10. Emissions for different energy sources. Source: Tomasini-Montenegro et al. (2017).REFERENCES TABLE 10. Emissions for different energy sources. Source: Tomasini-Montenegro et al. (2017). Source Emissions (kgC/MWh) 122 (4-740) 1100 (190-1300) 847(520-1160) 50 (15-800) 500 (250-1234) 31(0-410) 20 (4-100) 7,55-57,5 REFERENCES Albiero D, Cajado DM, Fernandes ILC, Monteiro LA, Esmeraldo GGSL (2015) Tecnologias Agroecológicas para o Semiárido. Available: https://www.repositoriobib.ufc.br/000021/000021cd.pdf. Accessed Dez 17, 2022. Xavier RS, Galvão CB, Rodrigues RL, Garcia AP, Albiero D (2022) Mechanical properties of lettuce (Lactuca sativa L.) for horticultural machinery design. Scientia Agricola 79: 5. DOI: https://doi.org/10.1590/1678-992X-2020-0249 Ali A, Tufa RA, Macedonio F, Curcio E, Drioli E (2018) Membrane technology in renewable-energy-driven desalination. Renewable and Sustainable Energy Reviews 81: 1 - 21. DOI: https://doi.org/10.1016/j.rser.2017.07.047FUTURE PERSPECTIVE AND CHALLENGES ANP (2018) Dados Estatísticos. In: Central Data Centers. Available: https://www.gov.br/anp/pt-br/centrais-de- conteudo/dados-estatisticos. Accessed Mar 03, 2021. The prospects for the use of EGS for water desalination in the Northeast of Brazil are attractive because in addition to it being a clean and available renewable energy source that is relatively easy to use through the reuse of abandoned oil wells, it enables a low-carbon transition policy. Bayer P, Rybach L, Blum P, Brauchler R (2013) Review on life cycle environmental effects of geothermal power generation. Renewable and Sustainable Energy Reviews 26: 446-463. DOI: https://doi.org/10.1016/j.rser.2013.05.039 The challenges in the use of geothermal energy in northeastern Brazil with a view to the desalination of water refer to the development of national technology for the use of EGS, as well as the qualification of personnel to operate and scale the systems. Another challenge is the establishment of a public policy that is effective in providing drinking water for the northeastern population. This should be linked to these resources in the form of credits and counterparts for the development and operation of these systems. Carneiro CDR, Hamza VM, Almeida FFM (2017) Ativação tectônica, fluxo geotérmico e sismicidade no nordeste oriental brasileiro. Revista Brasileira de Geociências 19 (3): 310–322. Cavalcante Júnior R, Vasconcelos Freitas M, da Silva N, de Azevedo Filho F (2019) Sustainable groundwater exploitation aiming at the reduction of water vulnerability in the Brazilian Semi-Arid Region. Energies 12: 904. DOI: https://doi.org/10.3390/en12050904.CONCLUSIONS Chandrasekharam D, Lashin A, Al Arifi N, Al-Bassam AM (2018) Desalination of Seawater Using Geothermal Energy for Food and Water Security: Arab and Sub- Saharan Countries. Renewable Energy Powered Desalination Handbook: Application and Thermodynamics. 177–224. DOI: In the form of EGS, geothermal energy can be a suitable option to address water scarcity in the northeast region in a sustainable and low-emission manner. The large number of oil and gas wells shows that there is an opportunity for further assessment of this potential solution, as demonstrated by this energy and economic feasibility study. y https://doi.org/10.1016/B978-0-12-815244-7.00005-2 Engenharia Agrícola, Jaboticabal, v.43, special issue, e20220160, 2023 Francisco T. G. Lima Verde Neto, Paulo A. C. Rocha, Jenyffer da S. G. Santos, et al. Clauser C, Ewert M (2018) The renewables cost challenge: Levelized cost of geothermal electric energy compared to other sources of primary energy – Review and case study. Renewable and Sustainable Energy Reviews 82: 3683- 3693. DOI: https://doi.org/10.1016/j.rser.2017.10.095 Melikoglu M (2017) Geothermal energy in Turkey and around the World: A review of the literature and an analysis based on Turkey’s Vision 2023 energy targets. Renewable and Sustainable Energy Reviews 76: 485-492. DOI: https://doi.org/10.1016/j.rser.2017.03.082 DiPippo R (2015) Geothermal power plants: Principles, applications, case studies and environmental impact. North Dartmouth, Elsevier. 762p. Noorollahi Y, Taghipoor S, Sajadi B (2017) Geothermal sea water desalination system (GSWDS) using abandoned oil/gas wells. Geothermics 67: 66–75. DOI: https://doi.org/10.1016/j.geothermics.2017.01.008 Geothermal Electricity Technology Evaluation Model (2018) In: Department of Energy. Available: https://www.energy.gov/eere/geothermal/geothermal-electricity- technology-evaluation-model. Accessed Mar 3, 2021. Péres VM, Silva LA da, Jesus TCB de, Barreto TDO (2016) A Tecnologia EGS e sua Aplicação na Exploração de Gás de Xisto no Brasil. Revista Principia 1(31). DOI: https://doi.org/10.18265/1517-03062015v1n31p13-21 Dutra RM, Tolmasquim MT (2002) Estudo de viabilidade econômica para projetos eólicos com base no novo contexto do setor elétrico. Revista Brasileira de Energia 9(1): 1- 16. Silva WF, Santos IFS, Botan MCC, Moni Silva AP, Barros RM (2018) Reverse osmosis desalination plants in Brazil: A cost analysis using three different energy sources. Sustainable Cities and Society 43:134–143. DOI: https://doi.org/10.1016/j.scs.2018.08.030 Fridleifsson IB, Bertani R, Huenges E, Lund JW, Ragnarsson A, Rybach L (2008) The possible role and contribution of geothermal energy to the mitigation of climate change. IPCC scoping Meet. Renewables Energy Sources 59–80. Tomasini-Montenegro C, Santoyo-Castelazo E, Gujba H, Romero RJ, Santoyo E (2017) Life cycle assessment of geothermal power generation technologies: An updated review. Applied Thermal Engineering 114: 1119-1136. DOI: https://doi.org/10.1016/j.applthermaleng.2016.10.074 Hanbury O, Vasquez VR (2018) Life cycle analysis of geothermal energy for power and transportation: A stochastic approach. Renewable Energy 115: 371–381. DOI: https://doi.org/10.1016/j.renene.2017.08.053 Long A (2009) Improving the economics of geothermal development through an oil and gas industry approach. In: Schlumberger Business Consulting. Available: Long A (2009) Improving the economics of geothermal development through an oil and gas industry approach. In: Schlumberger Business Consulting. Available: https://www.smu.edu/- /media/Site/Dedman/Academics/Programs/Geothermal- Lab/Documents/Oil-and-Gas- Publications/Schlumberger Improving the Economics of Tomaszewska B, Pająk L, Bundschuh J, Bujakowski W (2018) Low-enthalpy geothermal energy as a source of energy and integrated freshwater production in inland areas: Technological and economic feasibility. Desalination 435: 35–44. DOI: https://doi.org/10.1016/j.desal.2017.12.032 https://www.smu.edu/- /media/Site/Dedman/Academics/Programs/Geothermal- /media/Site/Dedman/Academics/Programs/Geothermal Lab/Documents/Oil-and-Gas- Publications/Schlumberger_Improving_the_Economics_of _Geothermal_Development.pdf. Accessed Apr 05, 2021. Lab/Documents/Oil-and-Gas- Publications/Schlumberger_Improving_the_Economics_of _Geothermal_Development.pdf. Accessed Apr 05, 2021. Turkenburg WAF (2000) Renewable energy technologies In: World Energy Assess. - Energy Chall. Sustain. Available: http://www.nzdl.org/gsdlmod?e=d-00000-00--- off-0cdl--00-0----0-10-0---0---0direct-10---4-------0-1l--11- en-50---20-about---00-0-1-00-0--4----0-0-11-10-0utfZz-8- 10&a=d&cl=CL2.23&d=HASHd61695cbdbc1bc78f8e16e. 9.3.12. Accessed Mar 2, 2021. Loutatidou S, Arafat HA (2015) Techno-economic analysis of MED and RO desalination powered by low- enthalpy geothermal energy. Desalination 365: 277–292. DOI: https://doi.org/10.1016/j.desal.2015.03.010 Lu SM (2018) A global review of enhanced geothermal system (EGS). Renewable and Sustainable Energy Reviews 81: 2902 - 2921. DOI: Vogt HH, Melo RR, Daher S, Schmuelling B, Antunes FLM, Santos PA, Albiero D (2021) Electric tractor system for family farming: increased autonomy and economic feasibility for an energy transition. Journal of Energy Storage 40: 102744-102760. DOI: https://doi.org/10.1016/j.est.2021.102744 Manju S, Sagar N (2017) Renewable energy integrated desalination: A sustainable solution to overcome future Manju S, Sagar N (2017) Renewable energy integrated desalination: A sustainable solution to overcome future fresh-water scarcity in India. Renewable and Sustainable Energy Reviews 73: 594-609. DOI: h //d i /10 1016/j 201 01 164 fresh-water scarcity in India. Renewable and Sustainable Energy Reviews 73: 594-609. DOI: gy https://doi.org/10.1016/j.rser.2017.01.164 https://doi.org/10.1016/j.rser.2017.01.164 Engenharia Agrícola, Jaboticabal, v.43, special issue, e20220160, 2023Renewable and Sustainable Energy Reviews 133 (2020) 110351Contents lists available at ScienceDirectRenewable and Sustainable Energy Reviewsjournal homepage: http://www.elsevier.com/locate/rserRenewable energy policy effectiveness: A panel data analysis across Europe and Latin America Germ´an Bersalli a,b,*, Philippe Menanteau b, Jonathan El-Methni c a Institute for Advanced Sustainability Studies, Berliner Straße 130, 14467, Potsdam, Germany b Univ. Grenoble Alpes, CNRS, INRA, Grenoble INP, GAEL, 1241 rue des R´esidences, 38400, Saint-Martin-d’H`eres, France c Universit´e de Paris, CNRS, MAP5 UMR 8145, F-75006, Paris, FranceA R T I C L E I N F OA B S T R A C T Keywords: Renewable energy Decarbonisation Electricity Latin America Europe Panel data Auctions Renewable energy (RE) technologies for electricity generation are a central pillar of energy sector decarbon- isation strategies worldwide. Public policies to promote their diffusion have been in place in developed econ- omies since 1980, and, since the 2000s, a growing number of emerging countries began implemented such policies. The Latin American countries have been proactive in RE promotion, but few attempts have been made to evaluate the results. This article proposes an econometric analysis of the effectiveness of RE policies, based on panel data for 20 Latin American and 30 European countries, over 20 years. The results converge for the in- fluence of promotion policies in general: they have a positive and statistically significant effect on RE investment, being the principal determinant in both regions. Nevertheless, on their own, tax incentives are insufficient to assure the deployment of RE technologies. This article also highlights specificities in policy approaches and motivations across both regions and explains why auction became the main instrument in Latin American countries.1. Introduction RE, which were first introduced at the end of the 1980s in Denmark. The numerous policy adjustments and reforms in Europe combined with different country-level approaches, provide valuable data for an eco- nomic analysis that considers differences in socio-economic contexts and should help design and implement similar policies in other world areas. There is a growing body of literature assessing RE policies based on cross-country analyses [4–11], case studies [12–16], and, more recently, econometrics [17–25]. A substantial increase in the share of Renewable Energy (RE) sources in the energy mix is crucial to achieving a level of anthropogenic CO2 emissions compatible with the Paris Climate Agreement. Thus, RE technologies for electricity generation are considered one of the fun- damentals of global energy sector decarbonisation strategies [1]. Public policies to promote their diffusion were implemented in the developed countries from the 1980s, within a niche market strategy, and, since the 2000s, a growing number of emerging and developing countries have also introduced support policies [2]. Following a sharp drop in the cost of several RE technologies, especially wind and solar photovoltaic (PV) [3], these policies aim to promote their large-scale diffusion to achieve a dominant share of the electricity mix. However, there are several open questions regarding the effectiveness, economic efficiency, and equity of selected promotion policies and, especially, in emerging and developing countries where empirical evidence is scarce. Among the emerging and developing economies, those in Latin America have been proactive in promoting RE since the 2000s. How- ever, very little research has aimed at evaluating the results of these policies. There are some valuable cases studies [26–30], but no attempts to perform broader statistical or econometric analysis representative of the Latin American region. Among the few studies that include some countries in Latin America is Pfeiffer and Mulder’s (2013) study [31] of 108 developing economies. Moreover, European and Latin-American countries have implemented the same kind of instruments, which fa- cilitates comparisons between both regions. Europe was a pioneer in the implementation of policies to promote Abbreviations: AUC, Auctions; EU, European Union; FE, Fixed-effects model; FIT, Feed-in Tariffs; PV, Photovoltaic; RE, Renewable Energy; RPS, Renewables Portfolio Standards. * Corresponding author. Institute for Advanced Sustainability Studies, Berliner Straße 130, 14467, Potsdam, Germany. E-mail addresses: german.bersalli@iass-potsdam.de (G. Bersalli), philippe.menanteau@univ-grenoble-alpes.fr (P. Menanteau), jonathan.el-methni@ parisdescartes.fr (J. El-Methni). Available online 19 September 2020 1364-0321/© 2020 Elsevier Ltd. All rights reserved. https://doi.org/10.1016/j.rser.2020.110351 Received 29 August 2019; Received in revised form 5 June 2020; Accepted 3 September 2020 G. Bersalli et al. Renewable and Sustainable Energy Reviews 133 (2020) 110351 incentives are also used to complement price or quantity-based in- struments, but the cases in which they represent the main (the only) support policy are scarce. The objectives of the present article are, firstly, to perform an orig- inal ex-post assessment of the effectiveness of the support policies implemented in Latin America, comparing them with the European experience, and, secondly, to analyse the influence of energy-system and macroeconomic determinants on RE diffusion. We conducted an econometric analysis based on panel data of 50 countries (30 in Europe and 20 in Latin America) during the period 1995–2015. The criterion of effectiveness is understood as the policy’s capacity to trigger new in- vestment in RE which is the primary objective of such measures [7].1 To consider a comparable and homogeneous unit of effectiveness, we selected the annual increase in installed capacity (in MW/inhabitant) of RE technologies as the dependent variable of our econometric model. Although from a theoretical perspective, price and quantity in- struments may be equivalents [39], numerous case studies and cross-country analyses show that, in practice, the effectiveness of policy instruments can differ substantially. Evaluation of European RE policy suggests that effectiveness and efficiency are higher in countries using price instruments than in those that have implemented quantity in- struments [9,10,40,41]. Nevertheless, the effectiveness of these policies depends heavily on each technology’s maturity and the specific “design elements” of each policy instrument [42]. This article contributes to the literature on energy policy assessment by integrating an econometric analysis based on panel data that includes most of the Latin American countries, a region sparsely studied despite its ambitious objectives concerning RE deployment. The model allows us to test novel variables to capture the effects of energy-related and eco- nomic factors on RE diffusion. The results of this econometric analysis are then compared with previous case studies to understand the effects of policies in different contexts. From a political point of view, a com- parison between different decarbonisation pathways in Europe and Latin America should facilitate future cooperation in energy and climate matters, an objective of the European Commission, the Mercosur, and other multilateral organizations. The number of countries that have implemented policies and the hindsight on these actions is sufficient for econometric studies to pro- vide additional analytical elements. These studies differ in their meth- odology and their findings for policy effectiveness. Popp et al. [43] analysed the influence of the ratification (or not) of the Kyoto Protocol. According to the authors, although ratification is not in itself a direct incentive for investment, it can serve to signal a country’s commitment to climate policy and, hence, to future carbon prices. Their model, that included data from the OECD countries between 1991 and 2004, showed a positive and significant relationship between the ratification of the Kyoto Protocol and investment, but no significant influence of direct support policies. Marques and Fuinhas [23] included a broader range of policy vari- ables to evaluate RE effectiveness in a set of 23 European countries from 1990 to 2007. First, the authors constructed a variable for the number of policy measures in each country (the accumulated Number of RE Pol- icies and Measures). They then tested the influence of different types of policies according to the general classification proposed by the IEA Global Renewable Energy Policies and Measures Database: Information and Education Policies, Economic Instruments (including FIT, FIP, etc.), R&D, Regulatory Instruments (including quotas, standards, etc.), Voluntary Approaches and Policy Support. Their results showed a pos- itive and significant impact of public policies in general. However, if considered individually, only subsidies and policy support (e.g., stra- tegic planning and creation of institutional frameworks) showed a major influence. Overall, Marques and Fuinhas found that regulatory in- struments and all other types of policies were not significant for pro- moting RE. The rest of this article is structured as follows. Section 2 reviews the existing literature about RE investment determinants and allows selec- tion of the main explanatory variables included in our model. Section 3 describes the data and the econometric model. Section 4 presents and discusses de results, and finally, Section 5 summarises the main con- clusions and policy implications.2. The multiple drivers of renewable energy investment Analysis of international experiences shows that RE diffusion is a complex process involving several elements [32]. In addition to support policies, other technical, economic, and socio-political variables may have a significant impact on the diffusion rate of RE technologies. A growing corpus of econometric work is investigating the effects of different drivers. However, only a few [18,19,31,33] include in their analyses the developing and emerging countries although these econo- mies account for more than half of total RE investment2 [2]. Aguirre & Ibikunle [44] whose approach was similar to that followed by Marques & Fuinhas (2012), analysed the OECD and BRIC countries from 1990 to 2010. They found a low level of effectiveness of promotion policies and even a negative relationship between tax incentive policies and diffusion of RE. However, the studies by both Marques & Fuinhas (2012) and Aguirre & Ibikunle (2014) included all RE sources, including biomass (in all its forms) and large hydroelectric power plants. These last two traditional RE sources are generally not targeted by support policies and, therefore, their estimation results may have been possibly biased.2.1. Contrasting results on the impacts of support policy Given the cost gap between RE and conventional sources, in- vestments in green energy have relied heavily on the implementation of support policies. These policies, traditionally, fall into two broad cate- gories: price-based and quantity-based [34] Among price-based initia- tives, guaranteed purchase or Feed-in Tariffs (FIT) and Feed-in Premiums (FIP) are intended to ensure return on investment in a stable and predictable framework. The second type of initiative includes quantity-based policies that set targets for the integration of RE into the energy mix and rely on Auctions (AUC) or Renewable Portfolio Standard (RPS) schemes, which include flexibility mechanisms such as green certificates [35–38]. Recently, the “quantity” instruments, especially auctions, tend to replace the “price” instruments in both regions. Tax Pfeiffer & Mulder [31]took again a different approach based on dummy variables to assess a panel of 108 developing countries between 1980 and 2010. Their results indicated that the likelihood of investing in RE was 10% higher after the Kyoto Protocol and even 30% higher in countries with support policies. Regarding policy effectiveness, coun- tries that implemented economic or regulatory instruments showed respectively a 27% and 52% higher likelihood of investing in RE. In terms of amount invested, Pfeiffer and Mulder found a more significant effect where regulatory instruments were used compared to economic instruments, which contrasts with most of the findings for these aspects (see, e.g., Ref. [9]). 1 We also consider the definition in Mitchell et al. (2011): effectiveness is ‘the extent to which intended objectives are met, for instance the actual increase in the output of renewable electricity generated or shares of renewable energy in total energy supplies within a specified time period’. 1 We also consider the definition in Mitchell et al. (2011): effectiveness is ‘the extent to which intended objectives are met, for instance the actual increase in the output of renewable electricity generated or shares of renewable energy in total energy supplies within a specified time period’. 2 The recent development in RE investment has varied by regions, rising in China, Latin America, and the Middle East and Africa while falling in Europe, the United States, Asia-Oceania (excluding China), Japan, and India (REN21, 2018). Cadoret and Padovano [20] measured two other policy variables in a panel that includes 26 European countries from 2004 to 2011. The first i 2 The recent development in RE investment has varied by regions, rising in China, Latin America, and the Middle East and Africa while falling in Europe, the United States, Asia-Oceania (excluding China), Japan, and India (REN21, 2018). Renewable and Sustainable Energy Reviews 133 (2020) 110351 G. Bersalli et al. between energy prices and the contribution of RE to energy supply in regions with high economic growth, although this was not significant for low-growth economies. reflects the country’s level of commitment (as a percentage) to the 2020 European RE targets.3 The second variable concerns environmental taxes, as a percentage of environmental duties in total tax revenues, according to the Eurostat classification. They found no statistically sig- nificant influence from this second variable, which, they believe, might be because environmental tax revenues are not intended for environ- mental protection or the dissemination of new technologies, but are budgetary instruments. It would be optimal to include in our model the costs of the electricity produced by the different sources in each country. However, the only data available are the overall electricity prices by country. These are endogenous since they include the cost of electricity produced by RE. Therefore most previous models included proxy variables such as per capita gas and coal production and, if available, the prices of these sources for each country. Polzin et al. [22]used ordinal variables to represent the different policies in place in the OECD countries between 2000 and 2011. They highlighted the importance of a reliable policy framework with clear medium- and long-term objectives and found that tax and economic incentives (FIT, FIP) are the most relevant measures for investors based on an evident reduced risk associated with such projects. Other indirect instruments, including tradable CO2 emissions permit systems, seem to have a positive impact, but only for the most mature technologies.2.2.3. The energy dependence rate Countries dependent on external energy supply need further to develop local sources of production, including RE. Reducing energy dependence is an energy policy objective and one of the main arguments for increasing RE in both developed and developing economies [44]. As Marques and Fuinhas [23] pointed out, the expected theoretical rela- tionship between these two variables is positive: the higher the country’s dependence on external supply, the greater will be the incentive to deploy RE. The Kilinc-Ata [21] study used data from 1990 to 2008 from Europe and the US. They showed that FITs, tenders and tax incentives are effective for boosting the deployment of RE, while quotas are not. Finally, Kim and Park [33] included a single policy variable: FIT. The results of the various models they tested showed a positive and signifi- cant influence of FIT. 2.2.4. Power sector dynamics: the annual rate of consumption growth y f g The more electricity demand increases or, the higher the prospects for its growth, the more a country will need to invest in new power capacity, including RE. Besides, the need for new investments in the electricity sector is linked to the average age of existing plants. When fossil fuel plants reach the end of their useful lives, they must be replaced by equivalent new capacity (except in the case of an equal drop in de- mand). Although a dynamic power sector should positively influence the spread of RE, the empirical evidence concerning that link is not conclusive. l In summary, econometric studies tend to agree on the effectiveness of policies in general but are less conclusive on the differences between specific instruments. More empirical evidence is required, considering a more recent sample that includes developing countries.2.2. Influences of the structure and dynamics of energy markets Beyond the incentive policies, context-specific variables can have a significant effect on RE diffusion. However, the empirical evidence is inconclusive on several points. In this section, we focus on variables related to the structure, functioning, and evolution of the energy sector (and particularly the electricity subsector). 2.2.5. The level of CO2 emissions from fuel combustion Fighting climate change by reducing energy-related GHG emissions is one of the main reasons for developing RE. Hence, the need for a control variable that captures the emission levels in each country. Most studies include a variable for CO2 emissions (from the energy sector) per capita. Aguirre et al. [44] recall the well-known fact that environmental concerns should stimulate investment in RE. 2.2.1. The energy mix: competition or complementarity between technologies The existing electricity mix may condition the dissemination of RE technologies in several ways. On the one hand, countries with high proportions of hydropower and nuclear-based electricity in their energy mix may be less concerned with developing new low-carbon technolo- gies [43]. Similarly, power generation that is concentrated on a partic- ular source (e.g., nuclear) may reflect a situation of technological lock-in4 and the possible influence of lobbying to maintain their market share [31]. On the other hand, countries with high percentages of hy- droelectricity have the necessary storage capacity to help to balance the intermittency of solar and wind power. From this perspective, new RE and hydroelectric plants are complementary rather than competing technologies. Therefore, the link between these sources of electricity generation is difficult to predict.2.3. Macroeconomic determinants The better the country’s macroeconomic situation, the higher the investment level in RE, ceteris paribus. Economies have specific charac- teristics that make them more or less attractive for investment generally, and in new technologies in particular. These features include, among others, the existence of highly skilled workers, good quality infrastruc- ture, and legal and financial security [24]. Access to project financing sources at reasonable interest rates is also fundamental and, especially, concerning high initial capital-intensive investments.2.3.1. Income per capita2.2.2. Electricity prices and the relative costs of RE and conventional energy sources Several studies [23,44] included GDP per capita in their models. The underlying assumption is that high-income countries are more likely to deploy RE because they can more easily bear the costs of developing these technologies and encourage them through economic incentives. The heterogeneity in our sample of countries is essential for this vari- able, given the large gap between low-income nations, such as Bolivia or Honduras, and countries with significantly higher incomes such as Norway and Luxembourg. RE compete with other sources of electricity generation. Thus, the availability of cheap domestic fossil fuel resources for electricity pro- duction, particularly gas and coal, can affect the attractiveness of RE. The higher the price of any competing sources, ceteris paribus, the more attractive will investments in RE be [43]. Simultaneously, several au- thors [24] have discussed the importance of electricity prices: the higher the price of electricity, the more investors will be encouraged to invest in RE. Chang et al. [45] found a positive and significant relationship2.3.2. Access to funding i One of the most significant barriers to RE deployment is the high initial investment cost. The RE sector’s capital intensity is higher than for fossil fuel energy and requires a proportionately higher initial in- vestment before production can begin. Using panel data for 30 countries 3 According to Directive 2009/28/EC. 4 Learning effects and economies of scale can lead to reductions in the cost of a specific technology in some countries (e.g., nuclear power in France). Renewable and Sustainable Energy Reviews 133 (2020) 110351 G. Bersalli et al. RE capacity (for country x for year t); b) RE generation in kWh per capita [31]; c) share of RE in the electricity mix, as a percentage [44,52]; d) ratio of total RE capacity to total electricity generation [53]; e) newly installed RE capacity per inhabitant (MW/1 million inhabitants) [22, 43]. In our case, investments in new RE are measured not in monetary terms, but as physical quantities (MW). We use indicators of capacity since they are the most accurate proxy for technology deployment. We define net investment per capita in technology j, of country i, in year t as: for the period 2000–2013, Kim & Park [33] examined the relationship between the development of financial markets and RE deployment on a global scale. Their results suggest that RE technologies are diffusing rapidly in countries with well-developed (both equity and credit) financial markets.2.3.3. The openness to international trade l Pfeiffer & Mulder [31] tested the influence of foreign direct invest- ment and the level of openness to international trade. Contrary to their initial hypotheses, they found a negative relationship between both variables and the level of RE generation. From a theoretical point of view, the influence of economic openness on new energy technologies can follow contradictory rules. In principle, trade openness should foster the diffusion of new technologies through lower costs (due to increased competition), removal of barriers, international agreements, etc. [46]. However, greater openness to international competition also requires additional efforts to lower domestic product costs, which, in turn, might lead countries to develop cheaper sources of energy. Therefore, the role of international trade in RE depends on several economic and political factors specific to each country, which are difficult to represent using a single variable. CAP is the total installed capacity at the end of each year; POP is the population. The RE technologies included in the dependent variable are wind, solar, geothermal, and biomass. Large hydropower plants are excluded because hydropower is a more mature technology that is not promoted by the support policies we are evaluating. Small hydropower plants could be included because they tend to be targeted by promotion pol- icies. However, most countries’ data do not distinguish reliably between small and large hydropower plants. Thus, we exclude all of them from our analysis. The data also do not include small generation plants that are not connected to the grid since they are affected by other policies, beyond the scope of this study.2.4. Political and institutional determinants In recent years, some theoretical and empirical studies have focused on the political factors influencing the design and effectiveness of energy and environmental policy. They have highlighted two sets of political determinants: governance quality, including the institutional frame- work in which these policies are implemented, and the ideology (po- litical orientation) of the government in power. Several articles [47–51] suggested that political factors can influence both energy and environ- mental policies. Nevertheless, none of these studies refers specifically to RE diffusion. i Finally, this study includes thirty European and twenty Latin- American countries, two regions using similar policy instruments to promote RE growth. The list of countries is available in Appendix 1. Some countries of these regions were excluded because of the lack of reliable data.3.2. Support policies in Europe and Latin America Cadoret and Padovano [20] is one of the first papers investigating the influence of purely political factors on RE diffusion. The authors studied the determinants of RE’s share in final energy consumption based on panel data for 26 European countries over the period 2004–2011. Their results showed that industry lobbies have a negative influence on the deployment of RE and that left-wing governments tend to favour the deployment of RE compared to their right-wing counterparts. They also showed a positive effect of governance quality on the deployment of RE. The main explanatory variables included in our model are Ex_pol, which denotes the existence or not of support policy, and four variables for the type of instrument. We consider the main support instruments: Feed-in Tariff and Feed-in Premium (FIT), the auction system (AUC), the quota obligations with negotiable green certificates or Renewable Portfolio Standards (RPS), and the fiscal -tax- incentives (FIS). We do not include instruments implemented for short periods, such as voluntary agreements or green pricing. The selected variables reflect the intensity and diversity of the policies implemented in the two regions studied. Other political or geographical factors that can influence RE diffu- sion include population density (related to the space available to install solar and wind parks), wind strength, solar radiation, etc. These vari- ables might have a strong effect on the diffusion of a particular tech- nology but are less relevant when we study general RE policy, independently of technology. Therefore, we exclude them from our model. We exploited several data sources: IRENA/IEA (Joint Policies and Measures database), European RE-Shaping projects [54], and data from Ref. [55,56]Since we are interested in investments in new technologies for electricity generation, we focus on policies specific to the electricity sector. Our database contains 1000 observations indicating the existence or not of a policy in the country i in the year t, and the type of instru- ment. These observations were transformed into binary variables.3. Data and method Fig. 1 depicts the percentage of countries that implemented at least one support policy in the electricity sector. In 1995, 40% of European countries had a policy in place. The number of countries increased significantly after the signing of the Kyoto Protocol (December 1997) and especially after the European Commission published Directive 2001/77/EC. This increase illustrates two central features of RE support policies in the Old Continent. First, they emerged in the context of climate policies and, especially, the need to reduce greenhouse gas emissions from the energy sector. Second, European institutions have played and continue to play a central role in this area through specific regulatory frameworks and common long-term objectives. Thus, the RE development objectives of European Union member states are set at the European level, although the choice of policy instruments to achieve them are chosen nationally. Our dataset covers 50 countries (i = 1, …,i = 50) and 20 years (t = 1995, …,t = 2014) and is balanced.5 In this section, we explain our choices concerning the explained variable (3.1), the explanatory vari- ables, including a brief description of RE policy in Europe and Latin America (3.2 and 3.3), and the econometric model (3.4).3.1. The dependent variable This study seeks to explain the RE diffusion determinants and, accurately, the level of investment in RE technologies over the 20 years to 2014. The literature review showed that several different definitions of the dependent variable had been proposed: a) amount invested in new In Latin America, promotion policies have followed a different logic. These countries were much less concerned about climate policies in the 1990s and 2000s: per capita emissions were significantly lower due to 5 There are no missed observations for any country and any year. Renewable and Sustainable Energy Reviews 133 (2020) 110351 G. Bersalli et al.table_1value 1 for the variable Ex_pol (for a year t) this means it has applied one of the following instruments: FIT, RPS, AUC or FIS. Similar measures for RE policy have been used in previous studies, including [52], and [25]. lower energy consumption and the existence of high levels of hydro- electric production in most countries. Fig. 1 shows that the first policies were put in place only in the late 1990s and early 2000s. They seek to diversify electricity production by attracting private investment needed to meet growing demand. There is also no coordinated planning of the energy sector in this region. The role played by regional institutions, such as Mercosur or Comunidad Andina, is somewhat limited in all but the Central American countries where energy integration is stronger [56].3.3. Other explanatory variables According to our theoretical framework, we included several control variables in our model, accounting for the influence of the structure and dynamics of the energy system and the main macroeconomic de- terminants of RE diffusion. Table 1 presents the name, the definition, the expected influence of each variable to the dependent variable, and also the data sources. Fig. 2 depicts the diversity of policy instruments in Latin America. A specific characteristic of the region is the dominance of quantity in- struments based on auctions, used by five states in 2014: Argentina, Brazil, Uruguay, Peru, and Nicaragua. Four smaller economies - Bolivia, Ecuador, Dominican Republic, and Honduras - chose FIT. A quota sys- tem has been in place in Chile since 2008. Six countries - Colombia, Venezuela, Paraguay, Costa Rica, Jamaica, and Belize - have no eco- nomic or regulatory support policies while Mexico,6 Guatemala, El Salvador, and Panama offer only tax incentives.3.4. The econometric model To assess the effects of RE policies on RE generation, we specified our model as follows: Fig. 3 shows the evolution of support policies in Europe. The price- instruments have tended to dominate, although, in recent years, several countries have reformed their support systems and introduced Feed-in Premiums to replace Feed-in Tariffs (Denmark, Estonia, Ger- many). The quota system is in place in several countries, although Italy and the UK recently decided to withdraw it. It should be noted that auctions, which had almost disappeared, are being used again in Italy, France, the Netherlands or Russia. Also, most countries use the same instrument to promote all types of RE technologies, although with differentiated levels of support. Only a few EU member states apply instruments specific to a technology. For example, in 2014, Italy had auctions for wind and biomass and FIT for solar PV, and Denmark had auctions for offshore wind and FIP for other technologies. l where Yit is the dependent variable (annual new RE capacity) observed for country i at year t, Xit is the time-variant 1 × k regressor matrix, β is the vector of coefficients αi is the unobserved time-invariant country effect and εit is the error term. i Our analysis encompasses several steps. We first analysed the char- acteristic of our data, and we especially confirmed the absence of mul- ticollinearity through the variance inflation factor (VIF) indicator. The next steps consisted of choosing between a fixed or a random-effects model. Using a fixed-effects (FE) model, we assume that specific unob- served characteristics of individuals (in our case countries) can have an impact or a bias on the explanatory variables of our model, and we, therefore, seek to control this. The random-effects assumption is that the individual-specific effects are uncorrelated with the independent vari- ables. The opposite is valid for a FE model, which removes the effect of these time-invariant characteristics, allowing us to assess the explana- tory variables’ net effect on the dependent variable. Given the charac- teristics of the data and the literature review, our panel should correspond to a FE model. We tested this hypothesis through the Hausman test [57]. The null hypothesis implies that the preferred model is the random-effects model, and the alternative hypothesis leads to a FE model. The test rejected the null hypothesis, and in the following, we consider the FE model. In our model, the influence of support policies is estimated in two stages. First, we included the variable Ex_pol, which indicates the exis- tence or not of a support policy (regardless of the type of instrument) in each country at the end of each year. Only direct promotion policies are considered. Thus, regulatory frameworks in the electricity sector, stra- tegic plans and other similar policies that do not include direct diffusion instruments are excluded. Ex_pol is a binary variable, which takes the value 0 if none of these policies are in effect and 1 otherwise. Second, we included four variables for the type of policy tool. If a country i has the 6 Mexico implemented a RPS in 2015. G. Bersalli et al. Renewable and Sustainable Energy Reviews 133 (2020) 110351 nan Then, we conducted several tests to analyses heteroscedasticity and correlation. Firstly, we tested for cross-sectional dependence/contem- poraneous correlation, through the Pesaran cross-sectional test of in- dependence (see Ref. [58]) where the null hypothesis is that residuals across entities are not correlated. The test did not reject the null hy- pothesis, and we conclude that the data have not contemporaneous correlation. Secondly, we tested for serial correlation, applying a Breusch-Godfrey test (see Ref. [59,60]) where the null hypothesis is that there is no serial correlation. The test rejected the null hypothesis, and we conclude that the data have a possible problem of serial correlation. Thirdly, we tested for heteroscedasticity through the Breusch-Pagan test (see Ref. [61]), where the null hypothesis is that there is homoscedas- ticity. The test rejects the null hypothesis, and we conclude that there is heteroscedasticity in the data. Finally, to take into account both heter- oscedasticity and serial correlation for a fixed-effects model, we applied the Arellano’s correction [62]. In summary, after applying several tests, we have chosen a fixed- effects model with Arellano correction. Then we compare the results with those obtained using the random-effect and the PCSE models. We conducted the different steps described here first for the entire sample (50 countries) and then separately for Europe and Latin America.4. Results and discussion This section presents the results of the regressions, first on the whole sample: 1000 observations corresponding to 50 countries over 20 years; and, second, for the European and Latin American countries separately. Also, we discuss the motivation and effects of RE policy in both regions.4.1. General results Table 2 summarises the results of our primary model (fixed-effects with Arellano correction), comparing them with two alternative models (PCSE and random-effects models). The three models are significant, according to the Wald statistic. Several variables are statistically sig- nificant, given the probability attributed to them. Three of our public policy variables appear to be statistically significant: feed-in tariffs, renewable portfolio standards, and auctions. In contrast, the variable Some previous analyses [22,23] applied the Panel Corrected Stan- dards Error (PCSE) method. This technique is well suited to estimating model parameters in the presence of heteroscedasticity and correlation at panel level. However, the estimation of the variance-covariance ma- trix in PCSE depends on large T, which is not the case for our sample (T = 20). Thus, we did not privilege this method. Renewable and Sustainable Energy Reviews 133 (2020) 110351 G. Bersalli et al. Table 1 Explanatory variables. NAME/CODE DEFINITION/INDICATOR EXPECTED INFLUENCE Electricity demand growth (ele_agr) * Growth rate of electricity consumption over the last 5 years. (+) countries facing rapid growth of electricity demand need to invest in new generation capacity, including RE. Share (%) of nuclear in the electricity mix (nuk_shr)* Share of nuclear energy in the total gross electricity production. (−) a high share of nuclear, i.e., low-carbon electricity, provides less incentive to invest in RE. Share of hydropower in the electricity mix (hyd_shr)* Share of hydropower in the total gross electricity production. (−) a high share of hydro, i. e., low-carbon electricity, provides less incentive to invest in RE. Energy dependence rate (ind_shr)* Share of primary energy consumption covered by domestic primary energy production. (−) a low level of energy independence can encourage countries to further invest in RE. CO2 emissions (CO2_pcp)* CO2 emission from fuel combustion per capita (tCO2/capita). (+) high CO2 emission levels can encourage countries to invest in low- carbon energy sources. Coal production per capita (coa_pcp)* Gross annual coal production per capita (Mt/ capita). (−) higher coal production can discourage RE investment. Gas production per capita (gas_pcp)* Gross annual gas production per capita (Mm3/capita). (−) higher gas production can discourage RE investment. Income per capita (gdp_pcp)* GDP US$ at constant price and exchange rate (2005) per capita. (+) it is easier for high- income countries to invest in RE. Access to domestic credit (cdt_shr)** Share of financial resources provided to the private sector by the banking sector and other financial corporations (% of GDP) (+) well developed local financial markets can facilitate RE investment. Source of data: *ENERDATA; **World Bank WDI. countries and years, as in our study, promotion policies appear to be the first determinant for RE investment. These results are in line with those in Ref. [21,22,31], but contrasts with the findings of [43,44] which found no positive effects of RE policies. EXPECTED INFLUENCE (+) countries facing rapid growth of electricity demand need to invest in new generation capacity, including RE. Growth rate of electricity consumption over the last 5 years. Besides, the relationship between growth in electricity consumption and the dependent variable is significant but negative. It indicates that countries with higher demand growth have not invested more in RE, but have relied on other, less expensive electricity sources. For instance, countries like Bolivia and Honduras have had averaged a 6% growth in electricity demand but almost no RE investment. Likewise, countries that have invested the most in RE have simultaneously implemented ambitious energy efficiency policies, which have influenced demand. This is the case of several EU countries like Denmark and Sweden, which have had high RE investment per capita, while the electricity demand stagnated. Popp et al. [43] found similar results in a sample of 26 OECD countries. Our model’s outputs also show that a high share of nuclear power in the electricity mix has a negative relationship with investment in RE. This result can be explained by the fact that countries with high shares of nuclear power (like France) have been less concerned with the rapid development of RE because they already have a decarbonised electricity mix. The coefficient for hydroelectricity is also negative but no signifi- cant. This finding confirms the fact that the complementarity or oppo- sition between RE and hydropower is country-specific. On the one hand, a high share of hydroelectricity facilitates the integration of intermittent technologies (wind and solar) thanks to the storage capacity. On the other hand, countries with a high share of hydropower in the electricity mix (a low-carbon source) may be less concerned about the rapid deployment of other RE sources. l Concerning the influence of coal and gas production, the results are different. Contrary to our hypothesis, a high level of coal production per capita has a positive influence (statistically non-significant, however) on RE diffusion. The explanation for this finding might be that some countries with a high share of carbon in the electricity production have substantially invested in RE as a form to decarbonize the mix or replace nuclear (that is especially true for Germany). The result is negative for gas: a high gas production would be, contrary to coal, a disincentive to develop RE fast. This inverse relationship may be because many coun- tries started to replace coal with natural gas during the 1980s or 1990s to reduce pollution what created a lock-in in these technologies. How- ever, the decarbonisation of the electricity mix is reflected in some countries by the development of RE and gas simultaneously: they phase- out coal and, at the same time, use gas to preserve the grid balance, while intermittent electricity increases. i representing “fiscal incentives” is not statistically significant in any model. We also estimated a model in which the four policy-instrument variables are replaced by the single variable “Ex_pol” which indicates the existence or not of a promotion policy in year t for country x, regardless of the instrument adopted. It allows us to test our primary hypothesis about the effectiveness of RE energy policy when controlling by the macroeconomic and energy-related determinants. The results of this model show a positive and significant influence of Ex_pol on the dependent variable. Therefore, considering a sufficiently large sample of Results in Table 2 show a clear positive and significant relationship Table 2 Summary of results - global model. Variables/model Fixed-Effects PCSE Random-Effects Y = new RE capacity Coeff. SE Coeff. SE Coeff. SE Feed-in Tariffs 7.04 3.45 (+)S ** 7.41 3.15 (+)S ** 9.50 2.00 (+)S *** Portfolio Standards 16.55 4.36 (+)S *** 12.68 5.28 (+)S ** 15.20 2.91 (+)S *** Auctions 7.31 3.12 (+)S ** 5.92 3.64 (+)S * 6.44 2.83 (+)S ** Fiscal incentives 3.25 3.04 (+)NS 3.64 2.98 (+)NS 0.52 2.68 (+)NS Demand growth −0.63 0.27 (−)S ** −0.49 0.22 (−)S ** −0.80 0.20 (−)S *** Share nuclear −0.50 0.25 (−)S ** −0.32 0.11 (−)S *** −0.13 0.06 (−)S ** Share hydropower −0.12 0.10 (−)NS 0.00 0.08 (+)NS −0.08 0.05 (−)S * Energy dependence 0.42 0.23 (+)S * 0.51 0.19 (+)S *** 0.06 0.05 (+)NS CO2 per capita −4.20 1.56 (−)S *** −2.76 1.68 (−)S * −3.48 0.58 (−)S *** Coal production 4.86 3.91 (+)NS 7.12 2.57 (+)S *** 3.66 0.78 (+)S *** Gas production −0.02 0.01 (−)S * −0.02 0.01 (−)S ** −0.00 0.00 (−)NS GDP per capita 1.81 0.62 (+)S *** 1.40 0.41 (+)S *** 1.05 0.13 (+)S *** Credit access 0.04 0.08 (+)NS 0.06 0.07 (+)NS 0.07 0.03 (+)S ** Observations 50 × 20 = 1000 50 × 20 = 1000 50 × 20 = 1000 R2 0.23 0.35 0.24 Notes: Significance levels: *** (1%), ** (5%), * (10%); (S): Statistically significant, (NS): Non-significant.Table 2 Summary of results - global model.Table 2 SSummary of results - global model. * (5%), * (10%); (S): Statistically significant, (NS): Non-significant. Notes: Significance levels: *** (1%), ** (5%), * (10%); (S): Statistically significant, (NS): Non-significant. Renewable and Sustainable Energy Reviews 133 (2020) 110351 G. Bersalli et al. quantity control, and competitive pressure between producers, the financial incentives will have to be gradually allocated by competitive auctions. This evolution of incentive schemes is too recent to appear in the model. However, it is in line with the results observed in Latin America and more broadly throughout the world, with a substantial increase in the auction system as an effective RE supply scheme. This evolution of promotion policies has been accompanied by a very marked dynamic of cost reduction over the last 20 years: while the first feed-in tariffs that enabled the PV sector to take off exceeded €500/MWh [55], the average prices of the 2019 tenders for ground-based PV installations in Germany were below €50/MWh. i between income per capita (GDP per capita) and the dependent variable. That reflects that RE has deployed more rapidly in high-income coun- tries, where the economic and institutional conditions are more favourable to the early diffusion of new technologies. This result is in line with the findings of [43,53]. Finally, the availability of credit does not appear as statistically significant. Further analysis that includes additional refined variables is needed to understand the relationship between the development of equity and credit markets and RE deployment.4.2. About RE policy effectiveness in Latin America and Europe The fixed-effects model for Latin America shows a positive and sig- nificant influence of auctions and RPS, while FIT is not significant. We observe a “sub-performance” of feed-in tariffs in Latin America compared to the European experience, where it was a central instru- ment. In Latin America, only a few countries applied feed-in tariff and feed-in premium, with quite poor results. It was the case in Argentina between 1998 and 2006: public policy based on feed-in premium failed to help RE to take off, affected by an unfavourable macroeconomic and regulatory context [63]. Brazil applied FIT for a few years and then changed for an auctions system. Since 2010 auctions became the main policy instrument with, in general, positive results in countries like Brazil, Uruguay, and, more recently, Argentina. We performed specific analyses for the Europe and Latin America samples to identify possible policy effectiveness differences between both regions. i Firstly, we evaluated the significance of our main explanatory vari- ables (the RE promotion policies) through a t-test. The results in Table 3 show that the existence of a promotion policy (observations from group 1) increased the mean of the dependent variable significantly (p-value <0.05). Furthermore, the policies have a positive and significant impact in both regions but with different relative effectiveness. In Europe, for years and countries with incentive policies (observations from group 1) the additional RE capacity is ten times higher than cases without policy (group 0); the ratio is 5 in Latin America. The standard deviation in group 1 observations is also significant, suggesting differences in policy effectiveness within countries. The auction scheme has several advantages that explain its imple- mentation in Latin American countries: i) It provides a stable and well- known ex-ante revenue stream once projects have been awarded, which facilitating project funding; ii) It stimulates competition among pro- ducers, facilitates the externalization of RE costs and allows control of the total cost of the policy; iii) It is adaptable to promote technologies with different degree of techno-economic maturity; iv) Besides costs, it allows the introduction of other selection criteria (like jobs creation) among the tenders submitted according to the objectives of the public policy. Most importantly, auctions scheme fits well with the institutional design of the electricity system in most Latin-American countries [42]. Nevertheless, the downside of auctions is that it favours “big” existing actors to the detriment of new and smaller producers to enter the mar- kets. Thus, this instrument should be complemented by specific policy targeting small and decentralized RE projects. In recent years, several countries of that region have moved in that direction. Secondly, we estimated the model with a binary “existence of RE policy” variable. The impact of support policy was positive and signifi- cant for both Europe and Latin America. Finally, we introduced the four specific policy instruments in our model. As in Section 4.1, we base our analysed in the fixed-effects model with Arellano correction and compare them with the two alternative models (Tables 4 and 5). The three main policy instruments applied in Europe -FIT, RPS, and auctions-show a positive and significant influence. These results confirm what the review of energy policies for RE in Europe shows. The public policies implemented by the Member States at the European Commis- sion’s initiative in favour of the development of RE have been very effective. This is confirmed by the mid-term progress assessment report,7 which states that the EU Member States are on track to meet the renewable energy targets for 2020, i.e., 20% renewable energy in gross energy consumption.8 To this end, the European Union’s policy has been based on several successive directives setting quantitative objectives for Member States (2010, 2020, 2030) but leaving each of them free to choose the means to be used to achieve them. The very marked oppo- sition during the 2000s between price and quantity instruments has evolved towards the clear domination of the former. Since then, incen- tive mechanisms have been adapted to take into account technological progress and cost trends. Again under the impetus of the European Commission, support schemes have evolved in recent years. Indeed, to allow better integration of renewables into the electricity market, RE policies in Latin America were introduced later and started to be effective when the cost of technology was affordable enough to allow private investments while controlling the cost of the policies. The pri- mary motivation of RE policy was concerns about energy security and the diversification of the electricity mix. Most Latin American countries have an electricity system concentrated in fossil fuels (Chile, Mexico, Argentina) or hydroelectricity (Brazil, Paraguay, Uruguay). A power mix concentrated in a few sources raises risks considerably: a succession of dry years or problems with the supply of imported gas or coal can affect electricity production and its cost substantially. For that reason, coun- tries tried to diversify the electricity mix, and nuclear power and re- newables have represented the main options. However, nuclear power exists only in three countries (Argentina, Brazil, and Mexico), and is a sophisticated technology that requires enormous investment and expertise. Renewables were relatively expensive, but the substantial cost decrease and their modularity made them an increasingly attractive option during the 2000s. Even if some RE technologies are currently cost-competitive in Latin America, public policies still play an essential role in reducing the risk of such investments. Table 3 Effect of policies on RE investment - Student’s t-test by continent. Europe Latin America Group Obs. Mean SE Obs. Mean SE 0: Without policy 129 2.19 6.5 282 1.07 3.94 1: With policy 471 23.32 30.9 118 5.33 13.78 Combined 600 18.7 28.9 400 2.33 8.39Table 35. Conclusion and policy implications This paper’s primary purpose was to evaluate the effectiveness of different RE policy instruments implemented in Europe and, more recently, in Latin America, while controlling by other determinant fac- tors. We first performed a review of the literature on econometric evaluation of RE policy. The determinants explaining the different levels 7 European Commission, 2015, Report from the Commission to the European Parliament, the Council, the European economic and social Committee and the Committee of the regions, Renewable energy progress report. 8 Climate 2020 Energy Package. 8 Climate 2020 Energy Package. Renewable and Sustainable Energy Reviews 133 (2020) 110351 G. Bersalli et al.Table 4 Summary of results - Europe.Table 4 Summary of results - Europe. Variables/model Fixed-Effects PCSE Random-Effects Y = new RE capacity Coeff. SE Coeff. SE Coeff. SE Feed-in Tariffs 12.25 4.38 (+)S *** 11.67 4.42 (+)S *** 14.39 3.12 (+)S *** Portfolio Standards 23.51 5.06 (+)S *** 18.22 6.58 (+)S *** 19.68 4.20 (+)S *** Auctions 10.95 3.32 (+)S *** 8.77 5.71 (+)NS 7.42 5.09 (+)NS Fiscal incentives 8.35 7.37 (+)NS 9.66 6.10 (+)NS 7.83 5.40 (+)NS Demand growth −1.77 0.46 (−)S *** −1.50 0.50 (−)S *** −2.06 0.43 (−)S *** Share nuclear −0.66 0.26 (−)S ** −0.42 0.15 (−)S *** −0.22 0.09 (−)S ** Share hydropower −0.52 0.22 (−)S ** −0.26 0.18 (−)NS ** −0.21 0.10 (−)S ** Energy dependence 0.61 0.32 (+)S * 0.61 0.31 (+)S * 0.10 0.10 (+)NS CO2 per capita −4.02 1.91 (−)S ** −2.45 1.84 (−)NS −3.53 0.86 (−)S *** Coal production 4.53 3.82 (+)NS 6.61 2.78 (+)S ** 3.16 1.18 (+)S *** Gas production −0.02 0.01 (−)S ** −0.02 0.01 (−)S ** −0.00 0.00 (−)NS GDP per capita 1.25 0.66 (+)S * 1.01 0.43 (+)S ** 1.00 0.18 (+)S *** Credit access 0.03 0.09 (+)NS 0.06 0.07 (+)NS 0.05 0.04 (+)NS Observations 30 × 20 = 600 30 × 20 = 600 30 × 20 = 600 R2 0.26 0.34 0.25 Notes: Significance levels: *** (1%), ** (5%), * (10%); (S): Statistically significant, (NS): Non-significant.Notes: Significance levels: *** (1%), ** (5%), * (10%); (S): Statistically significant, (NS): Non-significant. Table 5 Summary of results - Latin America. Variables/model Fixed-Effects PCSE Random-Effects Y = new RE capacity Coeff. SE Coeff. SE Coeff. SE Feed-in Tariffs 0.43 1.87 (+)NS 1.15 1.24 (+)NS 1.72 1.63 (+)NS Portfolio Standards 6.37 2.28 (+)S *** 3.75 2.98 (+)NS 8.52 3.70 (+)S ** Auctions 7.05 3.02 (+)S ** 3.97 2.80 (+)NS 9.52 1.55 (+)S *** Fiscal incentives −0.08 1.27 (−)NS −1.20 1.02 (−)NS 0.21 1.47 (+)NS Demand growth −0.09 0.04 (−)S ** −0.05 0.08 (−)NS −0.11 0.10 (−)S Share nuclear 0.93 0.55 (+)S * −0.11 0.50 (−)NS −0.31 0.33 (−)S Share hydropower 0.09 0.06 (+)NS 0.18 0.07 (+)S ** 0.03 0.02 (−)S Energy dependence 0.25 0.26 (+)NS 0.31 0.17 (+)S * −0.01 0.02 (+)NS CO2 per capita 0.89 1.02 (+)NS −1.27 1.74 (−)NS −1.04 0.62 (−)S * Coal production −0.86 1.03 (+)NS −3.22 1.73 (−)S * −1.13 1.59 (+)S Gas production −0.00 0.00 (−)S 0.00 0.00 (+)NS −0.00 −0.00 (−)NS GDP per capita 2.87 1.69 (+)S 4.45 1.65 (+)S *** 1.23 0.37 (+)S *** Credits acces −0.04 0.10 (+)NS 0.01 0.09 (+)NS −0.01 0.03 (+)S Observations 20 × 20 = 400 20 × 20 = 400 20 × 20 = 400 R2 0.19 0.24 0.19 Notes: Significance levels: *** (1%), ** (5%), * (10%); (S): Statistically significant, (NS): Non-significant.Table 5 1995, 14.6% in 2014, and 19.5% in 20189. In the beginning, the incentive policies, especially subsidies like feed-in tariffs, were para- mount due to the cost gap between fossil fuel and RE projects. In Latin America, RE (excluding hydro) represented 2.5% of total electricity production in 1995, 6.42% in 2014, and 11% in 2018. The countries of this region started to introduce RE policy later than the Europeans one and, thanks to a substantial drop in the costs of RE technologies and favourable natural conditions, most RE projects in Latin America do not need direct subsidies anymore. Currently, that region benefits from one of the greenest electricity mixes in the world thanks to a strong base of hydroelectric power and an increasing share of solar and wind energy. Indeed, Latin America is the first region in the world in terms of the share of RE generation if we include hydroelectricity (the share is around 58% in Latin America, while 36% in Europe, and 25% in the world average). The public policies implemented since the 2000s have had a significant influence on the development of wind, solar, and biomass energy sources. However, the decarbonisation, first, of the electricity sector and then of the whole energy sector, has revealed various problems and requires sustained public policies over time. Several European countries have accumulated vast experience in the implementation of RE policy, including potential adverse effects. or RE diffusion across countries can be classified into four categories: support policy, factors related to the structure and dynamics of energy markets, macroeconomic factors, and political and institutional de- terminants. We identified several conflicting points and lacunas in the econometric literature. We then developed a model which is the first to integrate a significant sample of Latin-American countries. Our results converge for the influence of promotion policies in general: public policies had a positive and statistically significant effect on RE invest- ment. However, the effectiveness of these policies seems stronger in Europe than in Latin America, partially explained by the different temporality in policy implementation. Also, some differences appeared concerning the type of instrument: auctions have consolidated as the main instrument in Latin America, where the institutional conditions felicitate their implementation. We conclude on the effectiveness of the main policy instruments to promote RE, in both Europe and Latin America. Instruments like feed-in tariff or the auction scheme are essential to reduce the risk associated with RE investment, even when some technologies like wind and solar PV are already cost-competitive in several markets. On the opposite, tax incentives, on their own, are not sufficient to incentivize RE deployment, and countries should not base their technology policy only on such instruments. RE generation is increasing in both regions. In Europe, the share of renewables in electricity production (excluding hydropower) was 1% in 9 ENERDATA, 2019. Renewable and Sustainable Energy Reviews 133 (2020) 110351 G. Bersalli et al. Stronger cooperation among the policymakers in both regions would help to boost energy transition. One specific area of plausible collabo- ration is the decentralized RE solutions (like home-based solar and wind technologies) that could enhance energy access in many rural areas of Latin America. [12] del Río P, Mir-Artigues P. Support for solar PV deployment in Spain: some policy lessons. Renew Sustain Energy Rev 2012;16(8):5557–66. ii [13] Bergek A, Jacobsson S. Are tradable green certificates a cost-efficient policy driving technical change or a rent-generating machine? Lessons from Sweden 2003–2008. Energy Pol 2010;38(3):1255–71. [14] Buckman G, Diesendorf M. Design limitations in Australian renewable electricity policies. Energy Pol 2010;38(7):3365–76. Our econometric analysis has some limitations and several possible extensions. First, the available data did not allow us to estimate the influence of some determinants identified in the literature, such as the cost of other sources of power generation, the real potential for different RE sources, and political factors. Similarly, more detailed consideration of the variables for access to financing (e.g., changes in the cost of available financing) would improve our model’s explanatory power. Finally, effectiveness is not the only criterion relevant to an evaluation of energy policies. Other factors, such as static and dynamic efficiency [42,64] and the equity of such policies, are also relevant and should be considered in future research. [15] Verbruggen A. Performance evaluation of renewable energy support policies, applied on Flanders’ tradable certificates system. Energy Pol 2009;37(4):1385–94. i [16] Lipp J. Lessons for effective renewable electricity policy from Denmark, Germany and the United Kingdom. Energy Pol 2007;35(11):5481–95. [17] Liu W, Zhang X, Feng S. Does renewable energy policy work? Evidence from a panel data analysis. Renew Energy 2019;135:635–42. [18] da Silva PP, Cerqueira PA, Ogbe W. Determinants of renewable energy growth in Sub-Saharan Africa: evidence from panel ARDL. Energy 2018;156:45–54. [19] Romano AA, Scandurra G, Carfora A, Fodor M. Renewable investments: the impact of green policies in developing and developed countries. Renew Sustain Energy Rev 2017;68:738–47. [20] Cadoret I, Padovano F. The political drivers of renewable energies policies. Energy Econ 2016;56:261–9. [21] Kilinc-Ata N. The evaluation of renewable energy policies across EU countries and US states: an econometric approach. Energy Sustain Dev 2016;31:83–90. l [22] Polzin F, Migendt M, T¨aube FA, von Flotow P. Public policy influence on renewable energy investments—a panel data study across OECD countries. Energy Pol 2015; 80:98–111.CREdiT author statement Bersalli: Conceptualization, Background, Statistical analysis, Results and Discussion. El-Methni: Statistical analysis, Results and Discussion. Menanteau: Conceptualization, Writing- Reviewing and Editing. [23] Marques AC, Fuinhas JA. Are public policies towards renewables successful? Evidence from European countries. Renew Energy 2012;44:109–18. ´ [24] Del Río P, Taranc´on M-´A. Analysing the determinants of on-shore wind capacity additions in the EU: an econometric study. Appl Energy 2012;95:12–21. [25] Johnstone N, Haˇsˇciˇc I, Popp D. Renewable energy policies and technological innovation: evidence based on patent counts. Environ Resour Econ 2010;45(1): 133–55.Declaration of competing interest [26] Aquila G, de Oliveira Pamplona E, de Queiroz AR, Junior PR, Fonseca MN. An overview of incentive policies for the expansion of renewable energy generation in electricity power systems and the Brazilian experience. Renew Sustain Energy Rev 2017;70:1090–8. The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. [27] Pueyo A. Enabling frameworks for low-carbon technology transfer to small emerging economies: analysis of ten case studies in Chile. Energy Pol 2013;53: 370–80.Acknowledgements & Funding [28] Lampreia J, De Araujo MSM, De Campos CP, Freitas MAV, Rosa LP, Solari R, et al. Analyses and perspectives for Brazilian low carbon technological development in the energy sector. Renew Sustain Energy Rev 2011;15(7):3432–44. The authors acknowledge the support of the GAEL laboratory members for reading the article. The correspondent author gratefully acknowledges his former thesis supervisor Dr. Patrick Criqui and his colleagues at the IASS Potsdam institute. This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. [29] Guzowski C, Recalde M. Latin American electricity markets and renewable energy sources: the Argentinean and Chilean cases. Int J Hydrogen Energy 2010;35(11): 5813–7. [30] Moreno R, Barroso LA, Rudnick H, Mocarquer S, Bezerra B. Auction approaches of long-term contracts to ensure generation investment in electricity markets: lessons from the Brazilian and Chilean experiences. Energy Pol 2010;38(10):5758–69. [31] Pfeiffer B, Mulder P. Explaining the diffusion of renewable energy technology in developing countries. Energy Econ 2013;40:285–96. i [32] Surana K, Anadon LD. Public policy and financial resource mobilization for wind energy in developing countries: a comparison of approaches and outcomes in China and India. Global Environ Change 2015;35:340–59.Appendix A. Supplementary data Supplementary data to this article can be found online at https://doi. org/10.1016/j.rser.2020.110351. [33] Kim J, Park K. Financial development and deployment of renewable energy technologies. Energy Econ 2016;59:238–50. [34] Finon D, Menanteau P, Lamy M-L. Prices versus quantities: environmental policies for promoting the development of renewable energy. Energy Pol 2003;31(8): 799–812.References [35] Butler L, Neuhoff K. Comparison of feed-in tariff, quota and auction mechanisms to support wind power development. Renew Energy 2008;33(8):1854–67. i [ ] , p , q support wind power development. Renew Energy 2008;33(8):1854–67. [36] Del Rio P. The dynamic efficiency of feed-in tariffs: the impact of different design elements. Energy Pol 2012;41:139–51. [1] Plessmann G, Blechinger P. How to meet EU GHG emission reduction targets? A model based decarbonization pathway for Europe’s electricity supply system until 2050. Energy Strat Rev 2017;15:19–32. [36] Del Rio P. The dynamic efficiency of feed-in tariffs: the impact of different design elements. Energy Pol 2012;41:139–51. [37] Dinica V. Support systems for the diffusion of renewable energy technologies—an investor perspective. Energy Pol 2006;34(4):461–80. [2] REN21 R. Global status report, REN21 secretariat. Paris: France. Tech. Rep; 2018. [3] IRENA I. Renewable power generation costs in 2017. Report. Abu Dhabi: International Renewable Energy Agency; 2018. [ ] p , p; [3] IRENA I. Renewable power generation costs in 2017. Report. Abu Dhabi: International Renewable Energy Agency; 2018. [38] Fouquet D, Johansson TB. European renewable energy policy at crossroads—focus on electricity support mechanisms. Energy Pol 2008;36(11):4079–92. [4] Percebois J. Aides publiques aux ´energies ´eolienne et photovoltaïque. Rev Fr Econ 2015;30(4):141–86. [39] Weitzman ML. Prices vs. quantities. Rev Econ Stud 1974;41(4):477 91. [40] Tanaka M, Chen Y. Market power in renewable portfolio standards. Energy Econ 2013;39:187–96. [5] Quirion P. Quels soutiens aux ´energies renouvelables ´electriques? Rev Fr Econ 2015;30(4):105–40. [41] Klessmann C, Held A, Rathmann M, Ragwitz M. Status and perspectives of renewable energy policy and deployment in the European Union—what is needed to reach the 2020 targets? Energy Pol 2011;39(12):7637–57. ´ [6] Sun P, Nie P-y. A comparative study of feed-in tariff and renewable portfolio standard policy in renewable energy industry. Renew Energy 2015;74:255–62. [7] Held A, Ragwitz M, Merkel E, Rathmann M, Klessmann C. Indicators assessing the performance of renewable energy support policies in 27 Member States. available at: www.reshaping-res-policy.eu; 2010. [42] Bersalli G. ´Evaluation et ´evolution des politiques de promotion des ´energies renouvelables: la transition des secteurs ´electriques en Am´erique Latine. Grenoble Alpes; 2017. p g p y [8] Verbruggen A, Lauber V. Assessing the performance of renewable electricity support instruments. Energy Pol 2012;45:635–44. i [43] Popp D, Hascic I, Medhi N. Technology and the diffusion of renewable energy. Energy Econ 2011;33(4):648–62. [9] Haas R, Resch G, Panzer C, Busch S, Ragwitz M, Held A. Efficiency and effectiveness of promotion systems for electricity generation from renewable energy sources–Lessons from EU countries. Energy 2011;36(4):2186–93. [44] Aguirre M, Ibikunle G. Determinants of renewable energy growth: a global sample analysis. Energy Pol 2014;69:374–84. [45] Chang T-H, Huang C-M, Lee M-C. Threshold effect of the economic growth rate on the renewable energy development from a change in energy price: evidence from OECD countries. Energy Pol 2009;37(12):5796–802. [10] Ragwitz M, Held A, Resch G, Faber T, Haas R, Huber C, et al. Assessment and optimization of renewable energy support schemes in the European electricity market: final Report. Germany: Optimization of Renewable Energy Support (OPTRES) Karlsruhe; 2007. gy [46] del Río Gonz´alez P. The empirical analysis of the determinants for environmental h l l h h d l ( ) [46] del Río Gonz´alez P. The empirical analysis of the determinants for environmental technological change: a research agenda. Ecol Econ 2009;68(3):861–78. [46] del Río Gonzalez P. The empirical analysis of the determinants for environmental technological change: a research agenda. Ecol Econ 2009;68(3):861–78. [47] Biresselioglu ME, Karaibrahimoglu YZ. The government orientation and use of renewable energy: case of Europe. Renew Energy 2012;47:29–37. [11] Mitchell C, Bauknecht D, Connor PM. Effectiveness through risk reduction: a comparison of the renewable obligation in England and Wales and the feed-in system in Germany. Energy Pol 2006;34(3):297–305. [47] Biresselioglu ME, Karaibrahimoglu YZ. The government orientation and use of renewable energy: case of Europe. Renew Energy 2012;47:29–37. Renewable and Sustainable Energy Reviews 133 (2020) 110351 G. Bersalli et al. [48] Potrafke N. Does government ideology influence deregulation of product markets? Empirical evidence from OECD countries. Publ Choice 2010;143(1–2):135–55. i Energía Renovable en Centroam´erica y Panam´a. Banco Centroamericano de Integraci´on Econ´omica; 2018. i [49] Fredriksson PG, Vollebergh HR, Dijkgraaf E. Corruption and energy efficiency in OECD countries: theory and evidence. J Environ Econ Manag 2004;47(2):207–31. [57] Hausman JA. Specification tests in econometrics. Econometrica J. Econom Soc 1978:1251–71. y g [50] Fredriksson PG, Svensson J. Political instability, corruption and policy formation: y g [50] Fredriksson PG, Svensson J. Political instability, corruption and policy formation: the case of environmental policy. J Publ Econ 2003;87(7–8):1383–405. [58] Pesaran MH. General diagnostic tests for cross section dependence in panels. 2004. [58] Pesaran MH. General diagnostic tests for cross section dependence in panels. 2004. [59] Breusch TS. Testing for autocorrelation in dynamic linear models. Aust Econ Pap 1978;17(31):334–55. the case of environmental policy. J Publ Econ 2003;87(7–8):1383–405. [51] Neumayer E Are left wing party strength and corporatism good for the [59] Breusch TS. Testing for autocorrelation in dynamic linear models. Aust Econ Pap 1978;17(31):334–55. [51] Neumayer E. Are left-wing party strength and corporatism good for the environment? Evidence from panel analysis of air pollution in OECD countries. Ecol Econ 2003;45(2):203–20. [60] Godfrey LG. Testing against general autoregressive and moving average error models when the regressors include lagged dependent variables. Econometrica J. Econom Soc 1978:1293–301. i [52] Zhao Y, Tang KK, Wang L-l. Do renewable electricity policies promote renewable electricity generation? Evidence from panel data. Energy Pol 2013;62:887–97. [52] Zhao Y, Tang KK, Wang L-l. Do renewable electricity policies promote renewable electricity generation? Evidence from panel data. Energy Pol 2013;62:887–97. [53] Shrimali G, Kniefel J. Are government policies effective in promoting deployment of renewable electricity resources? Energy Pol 2011;39(9):4726–41. [61] Breusch TS, Pagan AR. A simple test for heteroscedasticity and random coefficient variation. Econometrica J. Econom Soc 1979:1287–94. [62] Arellano M. PRACTITIONERS’CORNER: computing robust standard errors for [61] Breusch TS, Pagan AR. A simple test for heteroscedasticity and random coefficient variation. Econometrica J. Econom Soc 1979:1287–94. [53] Shrimali G, Kniefel J. Are government policies effective in promoting deployment of renewable electricity resources? Energy Pol 2011;39(9):4726–41. [62] Arellano M. PRACTITIONERS’CORNER: computing robust standard errors for within-groups estimators. Oxf Bull Econ Stat 1987;49(4):431–4. [54] Winkel T, Rathmann M, Ragwitz M, Steinhilber S, Winkler J, Resch G, et al. Renewable energy policy country profiles. Report prepared within the Intelligent Energy Europe project RE-Shaping. www.reshaping-respolicy.eu; 2011. [63] Bersalli G, Simon J-C. Vers une transition ´energ´etique des pays ´emergents: quelles politiques d’incitation aux ´energies renouvelables dans le secteur ´electrique en Argentine et au Br´esil? D´eveloppement durable et territoires ´Economie, g´eographie, politique, droit, sociologie 2017;8(2). [55] Fulton M, Mellquist N. The German feed-in tariff for PV: managing volume success with price response. DB Climate Change Advisors. Deutsche Bank Group; 2011. g g p p q g [64] Nicholls J, Mawhood R, Gross R, Castillo Castillo R. Evaluating renewable energy policy: a review of criteria and indicators for assessment. 2014. [56] BCIE. An´alisis Comparativo del Marco Regulatorio. Incentivos y Sistema Tarifario de Precios Existentes, para la compra/generaci´on de Electricidad de plantas de\\]^^bbJiQiKq~xxyy~HO!%-Ǩ(0*iq?OWEb(7#9dJSlu>g2< APWkr( 7/00;l?XCVYggK_!(]Վ܎MT~%rz}Lz_:f:=V?E~F]dfȇٝQg nkz,12NNQQQ\^^^fRFICPը- dl_Rh  (h:aXHvgvІ"h''j/{/g7?BJ{QQSS\hnnoz}}T22!!!!"d##S%[%e+Y5?QBKCDGtGGKgKLcQQafijjuloo||V}:tşѡ29bx@L _!+,MMM > *> :> J> Z> j> z> > > > > > > > > ? ? *? :? J? Z? j? z? ? ? ? ? ? ? ? ? @ @ *@ :@ J@ Z@ j@ z@ @ @ @ @ @ @ @ @ A A *A :A JA ZA jA zA A A A A A A A A B B *B :B JB ZB jB zB B B B B B B B B C C *C :C JC ZC jC zC C C C C C C C C D D *D :D JD ZD jD zD D D D D D zS** *( *D *zN zN we check the robustness of our results to the presence of outlifN N N N we check the robustness of our results to the presence of outlifN 2O Baer - Equity, Greenhouse Gas Emissions, and Global CommoJO O de Oliveira Gavira - 2014 - Research and Development Financing jO O A 01cbffe4e4246651P P feb85d3dbf8878d7#P 3P rN zN N N N 2O BO O O O O O O P 3P  section *string8 text *string8 file *string8 " word_count *int648 hash *string8j lance0.17.0z lance0.17%%* 0 *struct  null_count*int648  min_value*string8  max_value*string8 1 *struct  null_count *int648  min_value *string8  max_value *string8 2 *struct  null_count *int648  min_value *string8  max_value *string8 3 *struct  null_count *int648  min_value *int648  max_value *int648 4 *struct  null_count *int648  min_value *string8  max_value *string8 à%IR LANC