Options
Essays on the economics of carbon taxation
Maison d'édition
Neuchâtel
Date de parution
2022
Mots-clés
- taxe carbone
- taxe CO<sub>2</sub> suisse
- réforme fiscale environnementale
- évaluation de politique
- comportement des ménages
- demande d’énergie
- combustible fossile
- énergie renouvelable
- opinion publique
- hétérogénéité
- référendum. KEYWORDS: Carbon tax
- Swiss CO<sub>2</sub> levy
- environmental tax reform
- policy evaluation
- household behavior
- energy demand
- fossil fuel
- renewable energy
- public opinion
- heterogeneity
- referendum.
taxe carbone
taxe CO<sub>2</sub> s...
réforme fiscale envir...
évaluation de politiq...
comportement des ména...
demande d’énergie
combustible fossile
énergie renouvelable
opinion publique
hétérogénéité
référendum. KEYWORDS...
Swiss CO<sub>2</sub> ...
environmental tax ref...
policy evaluation
household behavior
energy demand
fossil fuel
renewable energy
public opinion
heterogeneity
referendum.
Résumé
Dans cette thèse, trois différentes problématiques liées à l’imposition des émissions de CO<sub>2</sub> sont traitées. La première, qui est analysée dans le Chapitre 1, est le fait que les taxes carbones ne constituent pas des sources de revenus fiscaux viables dans le long terme car elles ont pour finalité de réduire leur assiette fiscale, et même de finalement atteindre la neutralité carbone avec l’aide de mesures complémentaires. Ce problème de finance publique survient lorsqu’une juridiction ayant implémenté une taxe carbone, dans le cadre d’un projet de réforme fiscale environnementale, utilise les revenus que celle-ci génère afin de réduire les impôts distortifs préexistants qui touchent les facteurs de production – l’hypothèse du double dividende. La solution proposée dans ce chapitre est de progressivement imposer les sources d’énergie renouvelables à mesure que le recours aux énergies fossiles perd de l’ampleur, ce afin que l’utilisation totale d’énergie constitue l’assiette fiscale au lieu des seules énergies fossiles. Une telle mesure encouragerait également une plus faible consommation d’énergie, ce qui serait bénéfique pour la société, car même les énergies renouvelables causent des externalités négatives. Un modèle théorique est construit sur la base d’une fonction de production fondée sur l’énergie, qui traite cette dernière comme un intrant nécessaire au capital. L’énergie est modélisée comme étant générée à partir de sources fossiles et renouvelables, avec des fonctions de coût unitaire en forme de dôme inversé distinctes. Des impôts sont levés sur les deux sources d’énergies, ainsi que sur la consommation. Des simulations numériques sont utilisées comme illustrations, et celles-ci montrent qu’augmenter les impôts sur les deux sources d’énergie à des taux différents peut soutenir une sortie des énergies fossiles tout en allégeant les impôts sur la consommation. Cependant, des améliorations de l’efficacité énergétique sont requises afin d’éviter un impact négatif sur la production, car l’utilisation d’énergie est liée au niveau du capital utilisé, de sorte qu’imposer l’énergie équivaut à imposer indirectement un facteur de production. <br>La deuxième problématique, qui est traitée dans le Chapitre 2, est l’efficacité de la taxe CO<sub>2</sub> suisse. Ce chapitre analyse la réaction de court terme des ménages suisses aux augmentations du taux de ladite taxe qui ont eu lieu en 2016 et en 2018 quant à leur consommation d’énergies fossiles pour le chauffage et l’eau chaude. Des données issues de l’Enquête suisse sur la demande d’énergie des ménages collectées entre 2016 et 2020 sont utilisées pour comparer les trajectoires de deux groupes : le groupe de traitement, qui consiste en l’ensemble des ménages dont le système de chauffage est basé sur les énergies fossiles (mazout ou gaz naturel), et le groupe de contrôle, qui contient tous les ménages dont les systèmes de chauffage n’utilisent pas d’énergies fossiles, comme les pompes à chaleur ou le bois. Un modèle de régression de type « différence dans la différence » est utilisé pour estimer l’impact des deux hausses du taux de la taxe CO<sub>2</sub>. Une technique de pondération égale à l’inverse de la probabilité de recevoir le traitement est appliquée au groupe de contrôle afin d’améliorer l’équilibre de la distribution des covariables observables entre les deux groupes, dont la composition n’est pas entièrement comparable. Bien que la consommation globale de chauffage et d’eau chaude a diminué depuis 2016, les résultats ne permettent pas d’identifier un quelconque effet touchant spécifiquement les utilisateurs d’énergies fossiles par rapport aux utilisateurs d’énergies non-fossiles. Cette absence de résultat suggère qu’imposer les émissions de CO<sub>2</sub> des ménages n’est pas efficace pour les pousser à réduire leur consommation de chauffage et d’eau chaude, du moins dans le court terme. Une explication potentielle est le manque de perceptibilité du signal de prix de la taxe CO<sub>2</sub>. De plus, de nombreux ménages manquent de pouvoir décisionnel sur leur consommation de chauffage et leur choix de système de chauffage parce qu’ils sont locataires, ou parce qu’ils logent dans des appartements, ou les deux. Le design de la taxe CO<sub>2</sub> n’est donc pas adapté à leur situation. <br>La dernière problématique étudiée dans cette thèse est celle de l’opinion publique à propos des taxes carbones, qui est le sujet du Chapitre 3. Pour obtenir assez de soutien pour être mises en œuvre, ainsi que pour survivre et se développer à travers le temps, les taxes carbones doivent être acceptées par la population. Les différents éléments de leur design ainsi que les caractéristiques personnelles des individus forment les préférences relatives aux propositions de taxe CO<sub>2</sub> et leur hétérogénéité. Pour analyser ces préférences, des données issues d’une expérience de choix sont utilisées. Cette expérience a été menée en 2019 sur la base de la révision de la taxe CO<sub>2</sub> qui était alors débattue en Suisse. Les résultats de l’analyse économétrique de ces données, qui a été effectuée en suivant une approche fondée sur le concept de classe latente, montrent que deux profils de préférences distincts émergent : les pro-révisions, qui soutiennent une révision de la taxe CO<sub>2</sub> avec un taux plus élevé, et les anti-révisions, qui préfèrent s’en tenir au statu quo, c’est-à-dire à la taxe déjà existante. Le premier groupe est composé en moyenne de personnes politiquement plutôt de gauche et préoccupées par l’environnement, tandis que le second est composé de personnes plus de droite avec un point de vue plus neutre sur les problématiques environnementales. Le résultat d’un référendum sur une réforme de la taxe CO<sub>2</sub> est difficile à prédire à partir de ces données, étant donné la partition relativement équitable de la population entre ces deux groupes. Le véritable référendum sur la question, qui a eu lieu le 13 juin 2021, a finalement résulté en un rejet de la révision par une courte majorité, ce qui est cohérent par rapport aux résultats de l’expérience de choix.
ABSTRACT:
In this thesis, three different issues related to the taxation of CO<sub>2</sub> emissions are investigated. The first one, analyzed in Chapter 1, is the fact that carbon taxes are not sustainable sources of fiscal revenue in the long term because such taxes are meant to shrink their base, and even to eventually reach carbon neutrality with the help complementary measures. The public finance issue arises when a jurisdiction that implemented a carbon tax uses the revenue it generates to lower preexisting distortionary taxes on factors of production as part of an environmental tax reform project, with the final goal of both acting against the causes of climate change and benefiting from improved economic efficiency—the double dividend hypothesis. The solution proposed in this chapter is to progressively tax renewable energy sources as the use of fossil fuels decreases, so that overall energy use constitutes the tax base instead of only fossil energy. Such a measure would also encourage lower energy use through improvements in energy efficiency, which is beneficial to society, as even renewables cause negative externalities. A theoretical model is constructed on the basis of an energy-based production function, which treats energy as a necessary input to capital. Energy is modeled as being generated from fossil and renewable sources, with distinctive U-shaped unit cost functions. Taxes are levied on both energy sources, as well as on consumption. Numerical simulations are run as illustrations, and they show that increasing taxes on both sources of energy at different rates can push to a transition away from fossil fuels, while at the same time reducing the consumption tax rate. However, improvements in energy efficiency might be required not to face a detrimental impact on production, because energy use is tied to the level of capital used, so that taxing energy amounts to indirectly taxing a factor of production. <br>The second issue, which is investigated in Chapter 2, is the effectiveness of the Swiss CO<sub>2</sub> levy. This chapter analyzes the short-term response of Swiss households to the 2016 and 2018 increases in the CO<sub>2</sub> levy rate regarding their fossil fuel consumption for heating and warm water. Data from the Swiss Household Energy Demand Survey collected between 2016 and 2020 are used to compare the trajectories of two groups: the treatment group, which consists in households whose heating systems rely on fossil fuels (heating oil or natural gas), and the control group, which contains all households whose heating systems use non-fossil technologies such as heat pumps or wood. A difference-in-difference regression model is run to estimate the impact of the two CO<sub>2</sub> levy rate increases. Inverse probability of treatment weights are included in the estimation procedure to improve the balance of the distribution of observable covariates between the treatment and control groups, as their compositions are not fully comparable. Although overall heating and warm water consumption seems to have declined since 2016, results fail to identify any specific effect on fossil fuel users in comparison to non-fossil fuel users. This lack of results suggests that taxing CO<sub>2</sub> emissions from households is not effective to push them to lower their heating and warm water consumption, at least in the short term. An explanation could be that the levy’s price signal is not salient enough. In addition, many households lack decision power over their heating consumption and the choice of their heating system because they are tenants, or they live in apartments, or both. The design of the CO<sub>2</sub> levy is thus not well adapted to their situation. <br>The last issue studied in this thesis is public opinion on carbon taxes, which is the topic of Chapter 3. To gather enough support to be implemented and to survive and develop through time, carbon taxes need to be accepted by the population. Elements of design as well as personal characteristics of individuals shape preferences regarding CO<sub>2</sub> tax proposals and their heterogeneity. To analyze these preferences, data from a discrete choice experiment are used. This experiment has been conducted in 2019 on the basis of the then under discussion reform of the Swiss CO<sub>2</sub> levy. Results from the econometric analysis of these data with a latent class approach show that two distinct preference profiles emerge: the Pro-tax reform, who support a reform of the CO2 levy with an increased rate, and the Anti-tax reform, who stick to the status quo of the already existing tax. The first group is generally more left-leaning and environmentally concerned, whereas the second is more neutral regarding environmental matters and self-identify more with the right of the political spectrum. The outcome of a referendum on a reform of the CO<sub>2</sub> levy is hard to predict from these data given the relatively even split of the population between those two groups. The actual referendum that took place on June 13, 2021, ended up in a rejection of the reform by a short margin, which is consistent with the results of the choice experiment.
ABSTRACT:
In this thesis, three different issues related to the taxation of CO<sub>2</sub> emissions are investigated. The first one, analyzed in Chapter 1, is the fact that carbon taxes are not sustainable sources of fiscal revenue in the long term because such taxes are meant to shrink their base, and even to eventually reach carbon neutrality with the help complementary measures. The public finance issue arises when a jurisdiction that implemented a carbon tax uses the revenue it generates to lower preexisting distortionary taxes on factors of production as part of an environmental tax reform project, with the final goal of both acting against the causes of climate change and benefiting from improved economic efficiency—the double dividend hypothesis. The solution proposed in this chapter is to progressively tax renewable energy sources as the use of fossil fuels decreases, so that overall energy use constitutes the tax base instead of only fossil energy. Such a measure would also encourage lower energy use through improvements in energy efficiency, which is beneficial to society, as even renewables cause negative externalities. A theoretical model is constructed on the basis of an energy-based production function, which treats energy as a necessary input to capital. Energy is modeled as being generated from fossil and renewable sources, with distinctive U-shaped unit cost functions. Taxes are levied on both energy sources, as well as on consumption. Numerical simulations are run as illustrations, and they show that increasing taxes on both sources of energy at different rates can push to a transition away from fossil fuels, while at the same time reducing the consumption tax rate. However, improvements in energy efficiency might be required not to face a detrimental impact on production, because energy use is tied to the level of capital used, so that taxing energy amounts to indirectly taxing a factor of production. <br>The second issue, which is investigated in Chapter 2, is the effectiveness of the Swiss CO<sub>2</sub> levy. This chapter analyzes the short-term response of Swiss households to the 2016 and 2018 increases in the CO<sub>2</sub> levy rate regarding their fossil fuel consumption for heating and warm water. Data from the Swiss Household Energy Demand Survey collected between 2016 and 2020 are used to compare the trajectories of two groups: the treatment group, which consists in households whose heating systems rely on fossil fuels (heating oil or natural gas), and the control group, which contains all households whose heating systems use non-fossil technologies such as heat pumps or wood. A difference-in-difference regression model is run to estimate the impact of the two CO<sub>2</sub> levy rate increases. Inverse probability of treatment weights are included in the estimation procedure to improve the balance of the distribution of observable covariates between the treatment and control groups, as their compositions are not fully comparable. Although overall heating and warm water consumption seems to have declined since 2016, results fail to identify any specific effect on fossil fuel users in comparison to non-fossil fuel users. This lack of results suggests that taxing CO<sub>2</sub> emissions from households is not effective to push them to lower their heating and warm water consumption, at least in the short term. An explanation could be that the levy’s price signal is not salient enough. In addition, many households lack decision power over their heating consumption and the choice of their heating system because they are tenants, or they live in apartments, or both. The design of the CO<sub>2</sub> levy is thus not well adapted to their situation. <br>The last issue studied in this thesis is public opinion on carbon taxes, which is the topic of Chapter 3. To gather enough support to be implemented and to survive and develop through time, carbon taxes need to be accepted by the population. Elements of design as well as personal characteristics of individuals shape preferences regarding CO<sub>2</sub> tax proposals and their heterogeneity. To analyze these preferences, data from a discrete choice experiment are used. This experiment has been conducted in 2019 on the basis of the then under discussion reform of the Swiss CO<sub>2</sub> levy. Results from the econometric analysis of these data with a latent class approach show that two distinct preference profiles emerge: the Pro-tax reform, who support a reform of the CO2 levy with an increased rate, and the Anti-tax reform, who stick to the status quo of the already existing tax. The first group is generally more left-leaning and environmentally concerned, whereas the second is more neutral regarding environmental matters and self-identify more with the right of the political spectrum. The outcome of a referendum on a reform of the CO<sub>2</sub> levy is hard to predict from these data given the relatively even split of the population between those two groups. The actual referendum that took place on June 13, 2021, ended up in a rejection of the reform by a short margin, which is consistent with the results of the choice experiment.
Notes
Doctorat, Université de Neuchâtel, Institut de recherches économiques
Identifiants
Type de publication
doctoral thesis
Dossier(s) à télécharger