The demand for earmarking: Results from a focus group study

Abstract

Environmental taxes might be efficient, but plans to impose new taxes are often met with fierce public resistance. In order to design environmental taxes that are both efficient and acceptable to the public — so that they can be politically feasible — it is important to understand public attitudes towards environmental taxes. We conduct a focus group study in Norway to extend the current knowledge on this issue. We find less general resistance to taxation as a policy instrument, and seemingly more trust in government, than what has been reported in similar studies from other countries. The participants are, however, very skeptical and do not see the point of using the revenues from an environmental tax to reduce other taxes, such as the payroll tax. Instead they express a very strong preference for earmarking the revenues for environmental purposes. They also call for more information about environmental taxes, in particular on how the revenues are spent. Providing more information, including what the revenues are spent on — irrespective of whether they are earmarked or not, would seem to provide a relatively cheap and possibly effective way to increase the public acceptability of new or increased environmental taxes.

Introduction

When economists estimate large welfare losses stemming from public policies as if the losses were simple oversights that officials could correct by paying closer attention to what they were doing, it is the economists, not the officials, who are not paying attention.

Winston and Shirley (1998, p 68)

There is almost universal agreement among economist that environmental taxes are the best policy instruments to deal with environmental problems like climate change. Economists also agree on the general principles for how such taxes should be designed (see Cropper and Oates, 1992, for an introduction). Yet in many cases no environmental taxes are imposed, and where they do exist, their design often deviates significantly from the economists’ recommendations. There are several important reasons why this is so. Lobbying pressure from interest groups like industry organizations is one reason. Public attitudes may also explain some of the “simple oversights”, especially when it comes to taxes levied directly on individuals (Kallbekken, 2008).

The general public agrees that it is necessary to do something about many of the environmental problems. For example, a poll of 22,000 respondents from 21 countries found that 83% say it will be necessary to make lifestyle and behavioural changes to reduce emissions of greenhouse gases (Globescan and PIPA, 2007). One poll conducted in the USA found that 68% favour (at least in principle) an international climate agreement that would require the USA to cut its emissions by 90% by 2050. Yet, when it comes to using taxes as a means to achieve such emission cuts, the support vaporizes: 71% oppose higher electricity taxes, and 67% oppose higher fuel taxes (Leiserowitz, 2007).

In some cases there are binding or advisory referenda on the introduction of new environmental taxes, like the congestion charges in Stockholm (Isaksson and Richardson, 2009) and Edinburgh (Gaunt et al., 2007), or the referendum on taxes on fossil fuel in Switzerland (Thalmann, 2004). In other cases it is more difficult to describe exactly what is necessary for a policy proposal to be politically feasible. Gaunt et al. (2007) argue that for road user charging “commentators now acknowledge that the greatest impediment to implementation is public [...] acceptability”. This supports the argument that at least for taxes levied directly on individuals, public support is essential to make a tax feasible. King et al. (2007) does not agree fully, and argues that “the idea that a policy cannot be approved in the absence of popular support is at odds with the way policies are actually advanced.” Although the “political calculus” of environmental taxes might not be an exact science, it is still clear that it is politically risky to propose unpopular policies. While not binding, the immediate political response to the citizen referendum in Edinburgh, where 74% voted against the congestion charge, was to bury the proposal (Isaksson and Richardson, 2009). Several important examples of failures to implement new environmental tax schemes, as well as the abolition of several existing schemes, bear testimony to the significance of public attitudes towards environmental taxes. For instance, in 1999 the UK government had to abandon the fuel tax escalator because of public protests (Dresner et al., 2006b). This paper will address public attitudes towards environmental taxes in Norway.

Section 1 of the paper provides an overview of the literature on attitudes towards environmental taxes, and a quick introduction to environmental taxes in Norway. Based on the review, we define the research question for this paper. In Section 2 we explain why we chose to use focus group discussions, and how they were implemented. Section 3 provides an overview of the results, and we place our results in the context of earlier findings in Section 4. Finally, Section 5 concludes and discusses policy implications.

A large share of the studies conducted on the public acceptability of environmental taxes has been conducted on transport-related taxes. This is perhaps not surprising as most of the environmental taxes paid by individuals are those levied on transport-related goods or activities. What is more surprising is that the more general economic literature on this topic is so scant. The PETRAS study, which is summarized in Dresner et al. (2006a), provides a major exception, and is therefore an important reference point for our study.

Before we turn to the issue of public attitudes it is useful to provide some background on the economics of environmental taxes. There is almost universal agreement among economists on some issues: the tax should be imposed directly on the activity causing the externality (or the closest possible proxy), the rate should equal the marginal damage caused by the externality, and polluters should pay the full tax rate for all emissions (see e.g., Cropper and Oates, 1992, for an introduction). Another text book result is that the government should divorce its spending decisions from its fund-raising or taxation decisions so that spending can be undertaken unconstrained and solely based on benefit–cost considerations (Musgrave and Musgrave, 1984). This means that the revenues should not be earmarked (in a first-best world).

Tullock (1967) introduced the idea that a Pigouvian tax might offer a second benefit (the improvement in environmental quality being the first), if the revenues are used to reduce distortions in the tax system. Pearce (1991) made this idea popular with the term “double dividend”. In the 1990s there was wide debate about how the revenues from Pigouvian taxes should be spent. There seems to be almost general agreement that “revenue recycling” (using the Pigouvian revenues to reduce distortionary taxes) produces a higher welfare gain than returning the revenues lump-sum. Parry and Bento (2000), for instance, find that if production factors are much distorted because of pre-existing taxes, and in certain other circumstances, a double dividend is likely. Ultimately, specific features of the tax system determine the existence and size of the double dividend. While there has been debate over the double dividend, there is a consensus that in a first-best world it is inefficient to earmark the revenues from environmental taxes. The main reason for this is that earmarking leaves governments with less flexibility to shift funding between different budgetary posts, and it is very likely that the activity receiving the earmarked revenue would be either over- or underfunded compared to what is efficient.

For an economist, the primary motivation for introducing an environmental tax is that it gives incentives to reduce polluting activities, which in turn produces an environmental benefit. Dresner et al. (2006a) find that to many people the incentive effect of an environmental tax is far from obvious: the “general public showed a surprising lack of understanding” of the incentive effect. They view taxes “solely as means of raising revenue, rather than in terms of their incentive effects” (Dresner et al., 2006a:901). If people do not understand the incentive effect, it seems logical that if an environmental tax is to provide any environmental benefits, then the revenues must be spent on environmental measures. This is one of the reasons why earmarking the revenues from environmental taxes for environmental measures is found to increase public acceptability in almost all polls and studies on the issue. An international poll of 22,000 people in 21 countries found that 50% supported higher energy taxes, and that the support rose to 77% if the revenues were earmarked for promoting energy efficiency or developing cleaner fuels (Globescan and PIPA, 2007). The finding that earmarking can substantially increase support for an environmental tax is confirmed by for instance Schuitema and Steg, 2008, Hsu et al., 2008, Schade and Schlag, 2003, Thalmann, 2004, Steg et al., 2006.

Earmarking revenues seems to be popular not only because it can produce environmental benefits, but also because of distrust in governments. Dresner et al. (2006a) discover a fundamental lack of trust in government spending of revenue at three levels: distrust in politicians generally, distrust in tax policies, and distrust in intentions of ecological tax reform. There was a strong preference among the respondents for measures that would bring visible environmental benefits, preferably local benefits. Earmarking can be linked to distrust in governments, as earmarking constrains policymaking, and thereby can function as a credible commitment device (Anesi, 2006).

When people are asked about the idea of revenue recycling (or a double dividend), the response is typically negative for two reasons. First, they do not trust that the government will actually do as promised. Dresner et al. (2006b) found that people “did not believe that other taxes would be reduced in reality or that the money would be spent on what was promised.” Second, people find it “nonsensical to link the implementation of an energy tax and the reduction of labour taxes” (Deroubaix and Lévèque, 2006). However, Hsu et al. (2008) find that compared to providing no information about how the revenues will be spent, informing people that the revenues will be used to reduce other taxes, does increase their acceptability.

Several focus group studies find that people would generally like more information about environmental taxes. German participants “complained about not knowing details of [ecological tax reform] and criticised the government for an insufficient information policy” (Beuermann and Santarius, 2006). Danish participants thought that in order to “improve the acceptability of [ecological tax reform], information was crucial” (Klok et al., 2006). While these studies come out in favour of providing information in order to increase acceptability, Winslott-Hiselius et al. (2009) draw a somewhat different conclusion based on the experiences with the Stockholm congestion charge. They find that almost half of the respondents changed their attitudes towards the congestion charge during the trial in Stockholm in 2006. Furthermore, they find that this shift in opinion did not come primarily as a result of improved understanding of the type and magnitude of the effects of the charge, but rather that personal experience through the trial gave a new understanding of what the charge would mean for personal well-being. They suggest that “trials, generally, may be a more useful tool than information in the process of implementing ‘difficult’ policy measures, such as congestion charges” (Winslott-Hiselius et al., 2009).

Another issue that can significantly influence attitudes towards a proposed environmental tax is the perceived distributional fairness of the tax (Dresner et al., 2006a, Eriksson et al., 2006, Fujii et al., 2004). This is important, as pricing of some externalities, for instance emissions associated with energy intensive goods, can be regressive (e.g. Shammin and Bullard, 2009). Again the issue of the use of the revenues (earmarking) is central: Eliasson and Mattsson (2006) find that for the equity effects of the Stockholm congestion charge, the two most important factors are the initial travel patterns and how revenues are used. Shammin and Bullard (2009) show that if the revenues from a US cap-and-trade scheme were used to make equal lump-sum transfers to households, it would be possible to leave the lowest income households with a modest excess revenue, while the richest households would not be fully compensated for the increased energy costs resulting from the scheme.

In addition to distributional concerns, there has been significant attention dedicated to the resistance to coercive policies. Baron and Jurney (1993) found that people would vote against the policy reforms that they expected to produce net benefits, if they perceived the policies to be coercive. They suggest that one reason why there are social norms against voting for coercive policies is that such policies take away choices from people. Steg et al. (2006) argue that the distinction between “push” and “pull” policies is crucial when it comes to their acceptability. Push measures, such as taxes, may be perceived as coercive and not very acceptable because they “restrict people's freedom of choice and force people to change their own behavior”. Pull measures, such as subsidies or information provision, are non-coercive and perceived as more attractive. This is confirmed by for instance Jakobsson et al. (2000), who find that the acceptability of road pricing is “negatively affected by perceived infringement on freedom”. One possibly contradictory result is that Loukopoulos et al. (2005) find more support for prohibiting car traffic in the city centre than for road pricing. It seems reasonable to expect that prohibition would be perceived as the more coercive of the two. However, this does not necessarily contradict that people prefer less coercive policies as such; as other features also differ between the two policy instruments (prohibition for instance involves less redistribution than road pricing).

In addition to being more positive towards pull measures, people also perceive pull measures to be more effective than push measures in changing behaviour (Steg et al., 2006). Furthermore, Geller (2002) suggests that pull measures are likely to produce more positive attitudes, as they are more likely to be perceived as voluntary.

This result links in with the debate on motivation crowding. Coad et al. (2009) study consumers' willingness to support policies to encourage the adoption of energy-efficient cars. They suggest that in the early phases of adoption of a new green product, it might be effective to appeal to the intrinsically motivated consumers by providing information. Coercive measure might risk crowding out their motivation. Subsidies might also be useful at an early stage, as long as “their chief role is to encourage and express support to adopters of the clean technology, rather than to control consumer behavior or fully internalize pollution externalities.” By the time the green consumers have already made the switch to the new and green alternative, financial incentives might be effective to get other consumers to make the switch (and there is no longer any risk of crowding out the motivation of those who are intrinsically motivated).

In addition to these factors, it is also possible to relate attitudes towards environmental taxes to broader socio-economic and political variables. Income, age, sex and pro-environmental attitudes are among the factors analyzed by Eriksson et al., 2006, Fujii et al., 2004, Jakobsson et al., 2000, Loukopoulos et al., 2005. Thalmann (2004) analyzed the votes in Swiss referendum on three proposals for taxes on fossil fuels using a post-referendum survey. He found that people with a leftist political orientation and people with higher education were more favourable towards the tax proposals.

The revenues from environmental taxes were equal to 3% of GDP in Norway in 2006. This can be compared to an EU-27 average of 2.6% (Eurostat, 2009). In terms of revenues, the most important taxes are the fuel taxes, which produced NOK 14 billion in revenues in 2006, followed by the CO₂-tax with revenues of NOK 7.8 billion, and the tax on electricity consumption with 6.1 billion (NOU, 2007).¹ The other environmental taxes are the taxes on fuel oil, waste, disposable packaging, HFC, PFC, sulphur, lubricating oil, glass, plastic, metal, cardboard, and the tax on environment and health damaging chemical products. All revenues from environmental taxes goes into the general budget. Thus there is no earmarking as such. However, there is an “additional” tax on electricity consumption, which is earmarked for an energy efficiency fund, and companies can get out of paying the NO_X-tax if they instead sign up to an industry agreement where they pay the tax (at a lower rate) to a fund where the revenues are used to pay for NO_X mitigation measures.²

To the best of our knowledge, our study is the first academic study of general attitudes towards environmental taxes among the Norwegian public. The media has commissioned several surveys that have addressed people's willingness to pay for specific environmentally friendly goods, and on the acceptability of specific policy proposals. As the media only report main findings, and do not make the data available, it is difficult to evaluate the significance of such results. One exception is Griffin and Lehne (2009), who included questions on the government's responsibility for acting on environmental issues, but focused on consumer willingness to pay for green products. Bogen and Seip (2009) study attitudes towards taxes and the government among Norwegians, but do not focus specifically on environmental taxes. They find that a large share of the population is willing to pay higher taxes if they know that they are earmarked to specific sectors, e.g. 70–80% are willing to pay more taxes if they know that the revenues are spent on welfare services for the elderly, health services and improvements to the education system.

In this paper we extend the existing literature by conducting a focus group study on public attitudes towards environmental taxes in Norway — something which has not been done before. The main objective of this study is to explore and understand perceptions of and attitudes toward environmental taxes. The associated research question is: what are the attitudes towards environmental taxes in Norway which determine public acceptability, and what underlies these attitudes? A secondary research question is: are there any important differences between the results of our study and results from similar studies conducted in other countries?