Modeling emissions markets experimentally: The impact of price uncertainty

Combating climate change requires deep cuts of greenhouse gas emissions. It is crucial that policies intended to implement these cuts be efficient and drive investment in carbon reducing technologies. Experiments are being increasingly used as test beds for government policy to efficiently reduce emissions. Emissions trading schemes increasingly being implemented all over the world, including planned schemes in Australia. The schemes should lead to efficient emissions reductions (Montgomery, 1972), but this result rests on strong assumptions, such as no transaction costs, perfect information, and most importantly to the current study, risk neutrality of participants. These assumptions do not always hold in the field. For example, firms, even those owned by well-diversified stockholders, still exhibit risk-averse strategies. Do emissions trading schemes still lead to efficient outcomes when Montgomery's stringent assumptions are being relaxed? We simulate a carbon emissions market experimentally. All firms in the system are allocated emissions permits for free. The total number of permits corresponds to policy makers' overall emissions target. Since the adoption of emissions reduction technologies often involve high initial investments for the long term, such decisions are modelled as irreversible for the time of the experiment. In a well-functioning permit market the efficiency gains are reached since firms able to invest in emissions reduction at relatively low cost ("net sellers") do so. They then sell the emission permits originally allocated to them. Buyers are firms with relatively high emissions reduction costs ("net buyers"), for whom it is less costly to acquire additional permits than to invest in emissions reduction technologies. The study examines how uncertainty about the price of emissions permits affects market volume and firms' decisions to invest in long-term emissions reduction technologies. Results support recent theoretical results by Baldursson & von der Fehr (2004) and Ben-David et al. (2000) which suggest that firms' uncertainty aversion depresses the volume and efficiency of emissions markets. If the market price of permits is uncertain, "net sellers" tend to under-invest in abatement technologies and place fewer than expected permits on the market, while net buyers over invest in excessively costly emissions reductions technologies. Results also indicate that a market for future permits (a kind of forward market) which could reduce uncertainty would not alleviate the problem. Results are discussed in light of the challenges that the measurement of risk attitudes generally presents.