Voluntary agreements between an industry or a company and the government to regulate various environmental impacts is a popular policy tool in many OECD countries. Since the adoption of the Kyoto Protocol to the United Nations Framework Convention on Climate Change (UNFCCC) in December 1997 there has been a discussion of choice of policy tools to implement the Protocol in many industrialized countries. Not enough parties to make it enter into force have yet ratified the Kyoto Protocol, but many parties are making preparations for the implementation phase. This study focuses on the suitability of voluntary agreements to curb emissions of greenhouse gases (GHGs). Can such agreements be a more efficient policy tool than taxes, tradable emission permits, or traditional command and control? In this perspective we examine under what circumstances voluntary agreements could be an attractive policy option from the regulator’s (i.e. government’s or its agency’s) perspective.
Voluntary agreements is a very imprecise term since they can range from declarations of intent from a company, to binding contracts with a regulator with penalties specified in case on non-compliance. Furthermore, such agreements are not voluntary in a strict sense since there in most cases is an implicit or explicit threat from the regulator to impose other policy tools if the company is unwilling to negotiate. Also, if the target is not met the company could meet other, stricter regulations, such as emission fees or the introduction of taxes. These are important reasons why other terms like agreements, environmental agreements, negotiated agreements, codes of conduct, industry covenants, eco-contracts, and self-regulation have been used. For the purpose of avoiding misunderstandings on the nature of such agreements we will in this study denote all as Environmental Agreements (EAs).
A general definition of an EA is:
"An agreement between government and industry to facilitate voluntary action with a desirable social outcome, which is encouraged by the government, to be undertaken by the participant based on the participant’s self interest" (Storey, 1996).
In this study we focus on international EAs to curb emissions of GHGs. A general definition of an international EA is:
An agreement between an industry in one country and a regulator in another country, or between an industry in a group of countries and a regional regulator, with the aim to solve a regional or international environmental problem.
The first of these EA types is denoted a bilateral EA, whereas the latter type is denoted a regional EA. Five main conclusions to be drawn from this study are:
Experience with Environmental Agreements in a number of OECD countries suggests that these agreements are most attractive as a supplement to traditional command and control, or to market-based policy tools. This finding is also supported by the academic literature.
The academic literature indicates that skillful design of Environmental Agreements can improve their efficiency, for instance through including a menu of contracts (i.e. agreements) the company can choose between, or through the introduction of subsidies in case of over-fulfillment combined with a tax in case of under-fulfillment of the target.
Bilateral Environmental Agreements is an interesting policy option to regulate pollution from other countries, provided a general agreement between the two governments is established.
Regional Environmental Agreements are rare, even within the European Community, but could have important advantages and be an interesting supplement to other policy tools.
Environmental Agreements can play a role as a ‘soft’ transition stage from traditional command and control to domestic emission trading, and further on to a Kyoto Protocol regime of emission trading and joint implementation.
In terms of the first main conclusion a large variety of EAs is found in a number of OECD countries. Given their undocumented environmental effectiveness, it is odd that EAs have gained such a widespread recognition by governments as an instrument for environmental management. The lack of environmental effectiveness is partly due to a missing specification of baselines and reference scenarios to which the GHG abatement effect is compared and measured, but also due to the fact that regulators do not always participate in setting targets for EAs. Another concern is that the EA negotiation process in many cases is closed for third parties, and consequently more closed than a traditional policy making process in many countries. Such a closed process is undemocratic and can lead to weak EAs. The popularity of EAs may, however, lie in their political feasibility. With the negotiation of EAs, arenas for dialogue, partnership and co-operation between governments and industry have been established. For the industry, EAs are seen as a tool for enhanced predictability in environmental regulation. For governments, EAs may be seen as a tool for a more rapid behavioral change towards environmental accountability within industrial sectors than what is possible in case traditional policy tools are employed. In the long run, however, the societal legitimacy of this instrument will depend upon its demonstrated environmental effectiveness as well as the transparency and openness of the process in which it is developed. Due to these potential advantages of EAs but their undocumented environmental effectiveness, EAs seem most attractive as a supplement to other policy tools.
In terms of the second main conclusion the academic literature on EAs is limited. The literature does not give clear indications of whether EAs should be preferred to other policy tools, but claim that EAs have some potential benefits that can be harvested under some, but likely not the majority of, circumstances. Thus more and broader studies are required to give clear recommendations on the circumstances where EAs should be preferred to traditional command and control and market-based instruments. However, EAs may work well as a supplement to other policy tools under a wider set of circumstances than employing EAs alone.
Studies based on principal-agency theory have shown that menu-based EAs could be an interesting policy tool alternative in a situation of asymmetrical information between regulator and companies. Such EAs can do better than tradable permits and taxes in terms of welfare effects. This conclusion is based on a model where the survival of a specific company (or industry) is a vital constraint. Also of importance is the administrative cost of EAs compared to other policy tools. The administrative cost is likely to be reduced if the EA is negotiated with an industry federation instead of single companies. However, this could mean a heavier administrative burden on the industry federation.
If EAs are analyzed in an incomplete contract framework one tentative conclusion is that there is some potential for designing EAs in a way that reduces the problem of too lax targets for GHG abatement in companies and industries in EA contracts. One solution can be to introduce a subsidy in the case of over-fulfillment of the target, and a tax in the case of under-fulfillment of the target. Another suggestion from such models is that a shared investment in GHG abatement technologies between the regulator and companies is helpful in this context. As a case in point there are likely benefits for the companies in terms of improved energy-efficiency if investment is made in technologies that reduce GHG emissions. Finally, third parties like NGOs or representatives from the local community can play a helpful role in the negotiations and monitoring process of an EA to ascertain targets that are ambitious enough to be welfare-increasing. Through further analyses in an incomplete contract framework more insights can be gained into the efficient design of EAs.
In terms of the third main conclusion a bilateral EA is defined as a case where a regulator in one country negotiates an EA with a company or industry in another country. Bilateral EAs can be part of a national environmental policy or a supplement to an international environmental treaty. One example of bilateral EAs is the agreements from 1991 between the city of Rotterdam in the Netherlands and a number of German chemical firms to reduce pollution to the Rhine River.
In terms of the fourth main conclusion a regional EA is negotiated between a regional regulator and one or more industries in a group of countries. We have surveyed EAs at European Community level, but find that such agreements are rare. To date there are three agreements of this kind, of which one is with the European automobile industry to reduce CO2 emissions from new passenger cars. There may be significant advantages associated with the development of EAs at Community level, particularly in terms of harmonization of policy instruments and the implications this may have for market and competition conditions for European industries. Consequently regional EAs could more or less counteract the decentralization and de-harmonization drive associated with a large number of national EAs in EU member states. In addition there may be a potential for cost reductions attached to regional EAs. It is also clear, however, that a wider use of EAs at Community level would require a change in EU institutions. Given the current ambiguity with regard to the actual environmental effectiveness of this policy instrument, it might be wise to await further documentation of effectiveness in general and conditions for effectiveness in particular, before embarking upon such a process of institutional change.
In terms of the fifth and final main conclusion EAs can play an important role in a transition phase from established policy tools, and in particular command and control, to a Kyoto Protocol regime of emissions trading, joint implementation, and CDM. Thus the emission permits defined by EAs can later be made tradable at a domestic emission trading market, which later could be part of an international emission trading regime. In this manner national EAs could be a ‘stepping stone’ to full-fledged international emission trading. Alternatively, bilateral EAs could be combined with or develop into joint implementation, or possibly CDM projects.