Final Report Summary - ALLEGRO (Consumer Behavior and Energy Taxation: Exploiting Psychological Biases in Designing a Green Tax Reform)
1.1 Main Goals of the Research
The main goal of this research is to design a carbon tax taking account of insights from behavioral economics. The research should identify how the predicted acceptability of a Green Tax Reform and predicted impacts on behavior depend on the assumptions one makes about the psychology of economic agents. The project has been developed in two parts.
The first has consisted in the design of experiments aimed at providing experimental evidence on psychological effects biasing acceptability of energy taxes.
The second has involved the estimation of a demand system for US consumers specifying how individuals or households allocate their disposable income to available goods and services, using micro data from the Consumer Expenditure Survey (CES), and the calculation of direct and cross-price elasticities for the basket of goods and services on which households allocate their current expenditure.
Influencing consumers’ behavior with regard to environmentally relevant actions by fiscal signals is a new and important policy area. This project aim is to provide important additional knowledge to this area and to contributing to improving policy making in a key area of European Economic Policy.
1.2 Summary of progress towards objectives
The outgoing phase of the project has been carried out by the researcher at Carnegie Mellon University, Department of Social and Decision Sciences, which has provided the ideal environment where the experimental part of this project has been developed. Controlled laboratory experiments were designed to test the hypothesis that the delay of the benefits of taxation can be one of the main determinants of low public support for taxation. In particular, controlled laboratory experiments were conducted to examine whether, in a dynamic market experiment with negative externalities, people’s willingness to accept Pigouvian taxes aimed at restoring efficiency is affected by whether individuals bear the negative consequences of the externality now or later.
The results achieved in the first phase are described in the first periodic report submitted at the end of the outgoing phase (attached Document 1) and the main deliverable of this phase is the essay “Time Delay and Support for Taxation” (attached Document 2) co-authored by Silvia Tiezzi (the Researcher) and Erte Xiao. In this paper, the authors provide strong experimental evidence of a negative relationship between Time Delay and Support for Taxation. More specifically, we show that when the negative external effects of consumption are delayed, people are less willing to accept the introduction of Pigouvian Taxes as incentives to change consumption behavior. Such lower public support for taxation is not explained by a conventional (exponential) model of discounting. One attractive unifying explanation of these results is that greater tax aversion (i.e. lower public support for taxation) results from both a preference for the present and bounded rationality.
The second phase of the project has involved the estimation of a demand system for US consumers specifying how individuals or households allocate their disposable income to available goods and services, using micro data from the Consumer Expenditure Survey (CES), and the calculation of direct and cross-price elasticities for the basket of goods and services on which households allocate their current expenditure. We have produced new results on the different consumers’ reaction to tax or price changes. We have separately estimated the compensated gasoline retail price elasticity and the gasoline tax elasticity and have shown that consumers overreact to taxes as compared to price variations. A novel element in this analysis is that we compare reactions to tax-inclusive retail prices to reactions to information on excise taxes that is made available to consumers. A complete system of demand for the U.S. population of households has been estimated using quarterly data from the Consumer Expenditure Survey from 2007 to 2009. Relying on a complete system of demands rather than on single equations avoids imposing an implausible separability restriction, thus allowing estimation of accurate elasticities that take behavioral responses into account, i.e. that account for the way in which consumers reallocate their expenditure on a bundle of goods after a price/tax change in one of the goods. This analysis shows that the reaction to a gasoline tax change is, on average, about 20% stronger than the reaction to a corresponding price change. The implications of these findings for the design of energy policies are also discussed in the main deliverable of this phase of the project, the essay “Overreaction to Excise Taxes: the case of Gasoline” (attached Document 3) co-authored with Stefano F. Verde and recently published as a working paper of the Climate Policy Unit of the Robert Schuman Centre for Advanced Studies at the European University Institute
(http://cadmus.eui.eu/handle/1814/31365). Since we get a differential response to price and tax changes, a development of this research, currently in progress, is to estimate the differential welfare and distributional effects, on U.S. households, of introducing a carbon tax that is bundled with the price and thus fully hidden or a carbon tax that is fully visible to the consumers (i.e. separate information on the carbon tax is made available to the consumers).
1.3 Potential impact of results
Our findings have relevant implications for the design of efficient taxes, because they confirm that salience manipulation can influence consumers responsiveness to price incentives. When policy makers choose to keep some taxes hidden, or not prominent, this will contribute to keep consumers reaction low. This has obvious implications for the carbon tax debate not only in the U.S. where public support for fiscal efficiency-enhancing policies remains fragile, but also in Europe. Our results suggests that the carbon tax rate that would reduce greenhouse gases emissions to any targeted level could be set lower than predicted by the current literature, if consumers’ reactions to carbon taxes can be increased by increasing their visibility. A lower carbon tax rate would, in turn, probably be perceived as more acceptable than a correspondingly higher tax rate. Because policies that are effective in changing behavior, such as price incentives, also tend to have greater impacts on consumers, they also usually receive less public support, and therefore are less politically palatable, producing a negative relationship between effectiveness and acceptability of such policy measures as in Figure 1. Manipulating tax salience, by manipulating information availability, can change this tradeoff. Making excise tax information visible can increase consumers’ reaction to them and shift the effectiveness-acceptability tradeoff upwards so that a higher level of effectiveness is associated with a given level of acceptability.
Figure 1
The main goal of this research is to design a carbon tax taking account of insights from behavioral economics. The research should identify how the predicted acceptability of a Green Tax Reform and predicted impacts on behavior depend on the assumptions one makes about the psychology of economic agents. The project has been developed in two parts.
The first has consisted in the design of experiments aimed at providing experimental evidence on psychological effects biasing acceptability of energy taxes.
The second has involved the estimation of a demand system for US consumers specifying how individuals or households allocate their disposable income to available goods and services, using micro data from the Consumer Expenditure Survey (CES), and the calculation of direct and cross-price elasticities for the basket of goods and services on which households allocate their current expenditure.
Influencing consumers’ behavior with regard to environmentally relevant actions by fiscal signals is a new and important policy area. This project aim is to provide important additional knowledge to this area and to contributing to improving policy making in a key area of European Economic Policy.
1.2 Summary of progress towards objectives
The outgoing phase of the project has been carried out by the researcher at Carnegie Mellon University, Department of Social and Decision Sciences, which has provided the ideal environment where the experimental part of this project has been developed. Controlled laboratory experiments were designed to test the hypothesis that the delay of the benefits of taxation can be one of the main determinants of low public support for taxation. In particular, controlled laboratory experiments were conducted to examine whether, in a dynamic market experiment with negative externalities, people’s willingness to accept Pigouvian taxes aimed at restoring efficiency is affected by whether individuals bear the negative consequences of the externality now or later.
The results achieved in the first phase are described in the first periodic report submitted at the end of the outgoing phase (attached Document 1) and the main deliverable of this phase is the essay “Time Delay and Support for Taxation” (attached Document 2) co-authored by Silvia Tiezzi (the Researcher) and Erte Xiao. In this paper, the authors provide strong experimental evidence of a negative relationship between Time Delay and Support for Taxation. More specifically, we show that when the negative external effects of consumption are delayed, people are less willing to accept the introduction of Pigouvian Taxes as incentives to change consumption behavior. Such lower public support for taxation is not explained by a conventional (exponential) model of discounting. One attractive unifying explanation of these results is that greater tax aversion (i.e. lower public support for taxation) results from both a preference for the present and bounded rationality.
The second phase of the project has involved the estimation of a demand system for US consumers specifying how individuals or households allocate their disposable income to available goods and services, using micro data from the Consumer Expenditure Survey (CES), and the calculation of direct and cross-price elasticities for the basket of goods and services on which households allocate their current expenditure. We have produced new results on the different consumers’ reaction to tax or price changes. We have separately estimated the compensated gasoline retail price elasticity and the gasoline tax elasticity and have shown that consumers overreact to taxes as compared to price variations. A novel element in this analysis is that we compare reactions to tax-inclusive retail prices to reactions to information on excise taxes that is made available to consumers. A complete system of demand for the U.S. population of households has been estimated using quarterly data from the Consumer Expenditure Survey from 2007 to 2009. Relying on a complete system of demands rather than on single equations avoids imposing an implausible separability restriction, thus allowing estimation of accurate elasticities that take behavioral responses into account, i.e. that account for the way in which consumers reallocate their expenditure on a bundle of goods after a price/tax change in one of the goods. This analysis shows that the reaction to a gasoline tax change is, on average, about 20% stronger than the reaction to a corresponding price change. The implications of these findings for the design of energy policies are also discussed in the main deliverable of this phase of the project, the essay “Overreaction to Excise Taxes: the case of Gasoline” (attached Document 3) co-authored with Stefano F. Verde and recently published as a working paper of the Climate Policy Unit of the Robert Schuman Centre for Advanced Studies at the European University Institute
(http://cadmus.eui.eu/handle/1814/31365). Since we get a differential response to price and tax changes, a development of this research, currently in progress, is to estimate the differential welfare and distributional effects, on U.S. households, of introducing a carbon tax that is bundled with the price and thus fully hidden or a carbon tax that is fully visible to the consumers (i.e. separate information on the carbon tax is made available to the consumers).
1.3 Potential impact of results
Our findings have relevant implications for the design of efficient taxes, because they confirm that salience manipulation can influence consumers responsiveness to price incentives. When policy makers choose to keep some taxes hidden, or not prominent, this will contribute to keep consumers reaction low. This has obvious implications for the carbon tax debate not only in the U.S. where public support for fiscal efficiency-enhancing policies remains fragile, but also in Europe. Our results suggests that the carbon tax rate that would reduce greenhouse gases emissions to any targeted level could be set lower than predicted by the current literature, if consumers’ reactions to carbon taxes can be increased by increasing their visibility. A lower carbon tax rate would, in turn, probably be perceived as more acceptable than a correspondingly higher tax rate. Because policies that are effective in changing behavior, such as price incentives, also tend to have greater impacts on consumers, they also usually receive less public support, and therefore are less politically palatable, producing a negative relationship between effectiveness and acceptability of such policy measures as in Figure 1. Manipulating tax salience, by manipulating information availability, can change this tradeoff. Making excise tax information visible can increase consumers’ reaction to them and shift the effectiveness-acceptability tradeoff upwards so that a higher level of effectiveness is associated with a given level of acceptability.
Figure 1
