Did markets for emission trading reduce pollution? Empirical evidence from EU emission trading system - phase I (2005-2007)

Did the scarcity of emission certificates deal to emission reduction? In the logic of the market, companies optimize reducing emissions through the scarcity of certificates. In a context of uncertainty companies fail to get enough certificates because they lack a sufficient lobbying power or because they simply do not have adequate information about the future demand.

During the Phase I of the EU Emission Trading System, not every company had more certificates than necessary, but companies were overallocated as a whole. The companies received the first allocation of these certificates for free, then the demand for more certificates means greater cost pressure for these companies. This paper analyses the delegation of emission certificate's responsibility within departments of the company (panel data) in front of increasing cost pressure. We assume that there are two possible strategies of responsibility delegation for emission certificates within company departments: coupling and decoupling. Coupling implies the delegation about responsibility of the emission certificates closer in the process of production departments and decoupling means delegation in the more administrative and financial departments. We assume that the more delegation of emission in the decoupled departments, the less they know about their abatement costs. Abatement costs are a key part of the rational decision to reduce emissions and thus we assume that a greater decoupling means less reduction.

The hypothesis is that the higher the decoupling is, the higher cost pressure increases because companies are in uncertainty and they prefer increasing costs to invest. Therefore the scarcity of certificates do not produce the less reduction.

The empirical support for our research is provided by a yearly survey of companies participating in the EU ETS in four EU countries (Germany, United Kingdom, Netherlands and Denmark / 1040 observations) during the period 2005-2007.