Abstract

Carbon pricing instruments play a crucial role in mitigating CO2 emissions and are widely recognized as effective strategies for addressing climate change. In this context, Chile implemented a carbon tax on large fixed sources in 2017, marking a significant advance in Latin American environmental policy. However, there is a knowledge gap regarding the actual effectiveness of this policy using Bayesian approaches. Therefore, this research aims to evaluate the effect of this tax on CO2 emissions using a novel methodological approach: the Bayesian difference-in-differences method. This method provides an accurate estimate of the carbon tax's effects, taking into account the inherent variability and uncertainty in the data. To carry out this assessment, an annual database at the thermoelectric unit level is used, covering the period from 2014 to 2022. The results of this study reveal that the application of a tax of 5 USD/tCO2 significantly reduces CO2 emissions, with an annual decrease of between 17.7% and 18.1% in the taxed units. However, the magnitude of the observed impact is limited in relation to Chile's decarbonization targets, suggesting that the current design of this climate policy could benefit from improvements to enhance its emission reduction potential.