research

This paper examines the equilibrium impacts of emissions regulations on the time path of CO₂ emissions for the maritime shipping industry. Notably, it explores the interactions between travel speed, price, and capital turnover: Fleet fuel efficiency improves when larger ships replace existing ones, but the long lifespan of ships makes turnover slow. Regulations that reduce travel speeds lower emissions quickly, but also limit the supply and increase the price of shipping, thereby impacting shipbuilding and scrapping incentives. To quantitatively assess these mechanisms and their time horizons, I construct a dynamic model of the dry bulk shipping industry with endogenous entry, exit, and travel speed, as well as fleet heterogeneity across age and size. Using a rich dataset on the global fleet and its operation, I structurally estimate the model and use it to simulate the dynamic effects of a fuel tax, an efficiency standard that limits speeds, and an entry subsidy. I find that a fuel tax has a persistent impact, while the effect of a speed limit diminishes considerably over time due to induced ship building. Counterintuitively, rather than hastening the exit of older ships, both policies initially suppress exits, even while reducing emissions. An entry subsidy is more effective at removing old ships from service.

Machine learning predicts outcomes well, but predictive accuracy does not ensure reliable counterfactual responses. We examine how to combine machine learning and theory for measurement and counterfactual analysis, using maritime CO₂ emissions where physics provides a benchmark speed response. Matching hourly tracking data for dry bulk and container ships to annual fuel consumption reported under EU regulations, we compare engineering calculations, structural regressions, hybrid models, and machine learning. Out of sample, all estimated models predict within a few percent of reported totals, outperforming standard engineering calculations. Yet pure machine learning and unrestricted structural regression imply attenuated speed responses. Hybrids preserve the structural component’s speed response by excluding speed-related inputs from the machine learning component. A cost-benefit analysis of speed reductions illustrates the policy stakes: an attenuated speed response can flip the sign of net benefits. Accurate aggregate predictions therefore cannot substitute for scrutiny of the restrictions determining counterfactual responses.

Time preferences are omnipresent, but they are difficult to measure in the contexts in which they are applied. In agriculture, farmers’ time preferences drive choices that impact food security, industry sustainability, and the environment. I structurally estimate the discount rate of farm operators in Alberta, Canada using a dynamic discrete choice model of crop rotation decisions. My estimation strategy leverages the finite temporal dependence of expected yields on crop history and builds on a recent identification result for dynamic discrete choice models. My estimates suggest a strong present bias, somewhat in line with experimental estimates and in contrast to common modelling assumptions.

How does climate change affect investment in peaking power plants that ensure electricity grid reliability? Flexible generation is essential when supply and demand fluctuate rapidly, but is not profitable under normal conditions, which has resulted in an ongoing concern over underinvestment in deregulated markets. However, climate change is expected to increase volatility through (a) increasing extreme weather events that disrupt generation and distribution and fuel demand, and (b) spurring expansion of intermittent renewable generation that in turn depends on weather. Using data from deregulated regions in the US, I provide evidence that these phenomena are indeed associated with increased peaking plant investment. I construct a dynamic model of peaking plant investment to explore the interaction of these phenomena and their joint impact on the evolution of the electricity supply.

While shipping emissions are clearly linked with trade volumes, the quantitative relationship is unknown. We first quantify monthly, fleet-level CO₂ emissions from worldwide maritime shipping activity before and during the COVID pandemic. We then examine the change in these emissions during the COVID pandemic in terms of changes in bilateral trade volumes and provide a decomposition analysis. Finally, we model the heterogeneous elasticities of CO₂ emissions from maritime shipping with respect to international trade and use them to conduct a counterfactual analysis of potential emissions policies.