The impact of administrative transaction costs in the EU emissions trading system

Abstract

This article empirically investigates the impact of transaction costs for monitoring, reporting, and verification (MRV) of emissions on companies regulated by the EU Emissions Trading System (EU ETS) in Germany. Based on a unique panel dataset, we investigate if MRV costs are dependent on the amount of annual emissions of regulated companies and if there are differences in transaction costs between economic sectors. The results indicate that administrative costs are dependent on the amount of annual emissions for larger companies, which has implications for the economic efficiency of the EU ETS. The most important finding, however, is that there are significant differences in MRV transaction costs dependent on the type and size of companies. This implies the existence of considerable economies of scale. Overall, the EU ETS could benefit from reforms by means of a push towards upstream regulation as this would likely increase administrative efficiency.

Policy relevance statement

Transaction costs are, among other things, an important aspect of market-based climate policy design. A policy instrument with low transaction costs is preferred over instruments with larger transaction costs under equal conditions. This is occasionally referred to as administrative efficiency, and its importance was acknowledged in directive 2009/29/EC of the European Commission. Thoughtful empirical examination of transaction costs is essential in order to inform about the extent and impact of these costs. This article provides an analysis of transaction costs for monitoring, reporting, and verification (MRV) of emissions in the EU ETS. It is shown that administrative costs will likely have negative effects on the cost efficiency of the EU ETS. However, the most relevant finding is that small companies (<250 employees) or firms emitting small amounts of carbon dioxide per year face far higher average transaction costs compared with larger firms or emitters. Thus, there is a tendency for the EU ETS to cause MRV transaction costs that are disadvantageous for small companies. A regulation that is more upstream-oriented could mitigate this negative effect to some extent. The EU ETS could initiate a reform that is targeted on putting a price on the carbon content of fossil fuels instead of directly regulating emissions in a so-called ‘end-of-the-pipe’ way at the installation level.

Keywords:

• cap-and-trade
• EU emissions trading system
• monitoring
• reporting
• and verification
• transaction costs

JEL-Classification:

• D22
• D23
• Q58

Acknowledgements

This research received funding from the Europeans Community's Seventh Framework Programme under Grant Agreement No. 308481.

Disclosure statement

No potential conflict of interest was reported by the author.

Notes

1. Examples are the uncertainty of costs and the benefits of regulation (Hepburn, 2006) or political feasibility. The latter aspect has been of particular importance for policy choice and design in the European Union as coordination between member states and pre-existing regulations in some member states imposed a number of additional restrictions to policy choice and design (Ellerman, Convery, & De Pertuis, 2010).

2. Montgomery (1972) showed that the initial permit distribution is not essential for the final outcome of regulation under cap-and-trade, and that overall control costs will be unaffected by the initial allocation. The allocation of allowances, thus, is an issue related to distributional aspects but does not affect cost efficiency. This is also called the independence property (Hahn & Stavins, 2011).

3. One example is given by Kampas and White (2004). The authors showed that the optimal choice of taxation for the regulation of agricultural non-point source pollutants is dependent on administrative costs. When administrative costs are omitted, optimal policy design will differ as if administrative costs are included. A consistent emissions tax would be preferable in the former case, while a tax on inputs would be cost-minimizing in the latter.

4. It is possible that other additional types of costs are relevant in the MRV process, for instance, costs attributable to such expenditures as office equipment, electronic devices, software and so on. As these types of costs will only be partly attributable to the MRV process, and as the share of the costs attributable to the MRV process is difficult to estimate, we distance ourselves from these type of ‘overhead costs’ and focus on the most important cost components.

5. The average MRV transaction costs presented in this section are derived from a prediction of estimation (1) in Table 3. The values are different from those presented in the data description because the values in the data description are median values covering the whole sample without any differentiation by company type.

6. Efficiency could decrease if low-cost abatement options existed in companies that opt-out or are excluded from regulation.