This dissertation examines the role that technology plays in climate change mitigation. It contains three essays each focusing on different aspects of the process in which advancements in low-carbon energy technologies impact the cost of carbon dioxide (CO2) abatement.

The first essay develops the analytical foundation for understanding how heterogeneous low-carbon energy technologies induce differential impacts on the abatement cost. The analysis derives sets of conditions under which different types of advanced technologies can be evaluated for their respective strengths in reducing abatement costs at different levels of abatement. It emphasizes the weakness of a single point estimation of the impact of a technology and the importance of understanding the pattern of abatement cost reductions throughout the potential levels of abatement.

The second essay focuses on the interactions of the energy technologies in the market. The analysis uses a combinatorial approach in which 768 scenarios are created for all combinations of considered technology groups. Using the dataset, the analysis shows how the reduction in the abatement cost may change significantly depending on the existence of other advanced technologies. The essay shows that many of the fundamental insights from traditional representative scenario analyses are in line with the findings from this comprehensive combinatorial analysis. However, it also provides more clarity regarding insights not easily demonstrated through representative scenario analyses. The analysis emphasizes how understanding the interactions between these technologies and their impacts on the cost of abatement can help better inform energy policy decisions. 

The third essay focuses on the impact technological change has on the cost of abatement, but with special attention paid to the issue of delayed technology development. By combining the probability of advanced technology success estimates from expert elicitations with the abatement cost data estimated with an integrated assessment model, a stochastic dynamic programming model is developed. A multi-period extension of the model allows intertemporal dynamic optimization where the policy-maker can select the technologies to be invested in immediately and the technologies to be invested in later. The analysis emphasizes the benefit of having a wait-and-see option that lets the policy-maker further optimize upon the observation of successes and failures of prior investments.

The three essays collectively serve to demonstrate the importance of clearly understanding the differences among low-carbon technologies. They also provide methodological foundations upon which such technologies can be assessed and compared.  Combining these methods with an enhanced understanding of the technologies will contribute to the body of research aimed at minimizing the cost of mitigating climate change.