Essays on environmental policy evaluation

• Ano de defesa: 2020
• Autor(a) principal: Christofoletti, Maria Alice Móz
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: eng
• Instituição de defesa: Biblioteca Digitais de Teses e Dissertações da USP
• Palavras-chave em Português: Avaliação de politicas; Carbon taxation; Censored QUAIDS; CO2e emissions; Decentralization; Deforestation; Descentralização; Desmatamento; Emissions; Emissões de CO2; Policy evaluation; QUAIDS censurado; Regression discontinuity design; Taxação de carbono
• Link de acesso: https://www.teses.usp.br/teses/disponiveis/12/12140/tde-15102020-104055/

This thesis examines three topics related to current and potential policies to reduce greenhouse gases emission (GHG) emissions in Brazil. The first chapter, entitled \"Distributional Welfare and Emission Effects of Energy Tax Policies in Brazil\" calculates carbon intensity of 128 products consumed by Brazilian households and utilizes a rich household dataset to investigate the short-run impacts of energy policies in Brazil in the 2000s. Results indicate that 11% of total additional energy emissions between 2010 and 2018 (or 6.5 MtCO2e) could have been avoided in the absence of government tax reductions on diesel, electricity and residential appliances and increments on gasoline taxes. Findings also suggest that taxes on gasoline pump prices are progressive and have a negative impact on total household energy emissions due to substitution effects. Changes in electricity prices are regressive and have large effects on household emissions. More environmentally friendly policies, such as subsidy on ethanol, are the most cost-effective to reduce emissions, despite its small effect on the emissions of the economy. Understanding who benefits from energy taxes and subsidies is key to gaining public support for a greener energy mix, as pledged by the country in its NDC. The second chapter, entitled \"Winners and Losers: The Distributional Impacts of a Carbon Tax in Brazil\" continues to investigate how economy-wide policy alternative instruments, such as a carbon tax, influence the emissions and welfare of Brazilians and the opportunities to implement such instruments in a tax reform context. Estimates suggest that it is possible to observe the first dividend in the Brazilian context, as it could reduce annual GHG emissions by up to 4.2%. However, since low-income households are less price-responsive for the majority of carbon intensive categories, they suffer a larger relative welfare loss due to the carbon tax (0.10% of their total expenditures, vis-à-vis 0.06% for richest households). They are also more likely to suffer from a larger relative indirect effect of \"food and beverages\" and \"housing-related\" consumption, which accounts for a greater budget share of these households. Significant changes in total GHG would require a higher tax rate, which would reinforce the regressiveness of the policy. These results indicate that compensation strategies, such a direct lump-sum transfer, need to be considered by the government to reduce the burden imposed on these households. Given the significant complexities in the Brazilian tax system, the generation of a second dividend effect could be observed only if the country implements carbon pricing mechanism as part of a broader structural tax reform. The third chapter, entitled \"Does Decentralized and Voluntary Commitment Reduce Deforestation? The Effects of Programa Municípios Verdes\" utilizes regression discontinuity and high-resolution spatial dataset (1,781,122 pixels covering 162,242 km2) to examine the effect of a Brazilian state-level programme implemented in 2011 on deforestation rates. The programme was implemented in one of the country\'s state that present the highest deforestation rates. Evidence suggest avoidance of roughly 8.0 MtCO2/year released to the atmosphere, and the extrapolation of estimates to the total area that could be legally deforested in Pará indicates that 41% of deforestation in the Amazon region between 2015 and 2018 could have been prevented using this voluntary initiative. Since Brazil has committed through its NDC to eliminate deforestation in the Amazon by 2030, decentralized programmers with focus on indirect benefits appear to be effective in the long-run, serving as a \"bonus\" to support those regions with higher levels of forest cover.