Conference Agenda

Decarbonization and digitalization have triggered a demand boom for critical minerals, yet resource-rich countries have historically struggled to turn finite natural resources into lasting, broad-based industrialization. This paper provides the first global analysis of contemporaneous export restrictions on raw materials and critical minerals and their role in promoting downstream industrialization in green value chains worldwide. I combine data on over 3,200 export restrictions across 60 countries with detailed trade and production data, and classify more than 700 HS codes and 41 minerals across 12 green value chains into production stages. Using newly constructed product-level input–output linkages, the value chain stage classification, and an event study design, I estimate both direct upstream and indirect downstream effects. The theoretical framework highlights how export restrictions can reallocate rents from upstream to downstream sectors via domestic price wedges—particularly in settings with strong input–output linkages. The empirical estimates strongly reflect this mechanism: while exports decrease, domestic mining production increases in the short-term driven by expanded downstream activity, which, eventually, leads to a net rise in upstream exports in the medium term. The results suggest that, under the right conditions, resource-rich countries can leverage export restrictions to support broader green industrialization.

In this paper, we investigate the claim that rich countries use development aid to ensure access to natural resources. We provide a theoretical model that suggests that even an altruistic donor may be inclined to allocate a higher share of their aid expenditure on infrastructure and other trade promoting measures if they rely on the recipient’s resource exports for their own production. We use a panel dataset from 2001 to 2011. Our results suggest that bilateral resource trade on average positively affects the number of infrastructure projects and the average size of projects. The effect seems to be driven mostly by fuels and road infrastructure projects. While the effect of resources is weaker for landlocked countries, we find that the transport capacity of the recipient’s fleet of bulk carriers —used in the maritime transport of many resources— reinforces the effect of resources on infrastructure aid. Finally, we find a decreasing influence of resources over time.

Economic sanctions increasingly target large, complex firms. But targeted firms can adapt intertemporally across extensive asset portfolios, often resulting in delayed or unintended sanction impacts. Using oil field-level panel data and a natural experiment created by Russia’s 2014 invasion

of Ukraine, I show that sanctioned oil companies increase production in the short-term but sharply reduce exploration and abandon undeveloped fields – accelerating reserve depletion. To rationalize this sanction-induced short-termism, I develop a dynamic model of oil company decision-making

that decomposes sanction impacts by policy type: technology bans, financing bans, and restrictions on multinational participation in joint ventures. Estimates indicate that sanctions imposed on Russian oil companies in 2014 will accelerate average field abandonment by 23% (11 years) and reduce government oil revenues by 6.5% (US$39 billion) over 20 years.

The rapid expansion of artificial intelligence and data center infrastructure has generated substantial electricity demand, but its implications for the global transition to net zero emissions remains uncertain. Despite growing concern that data center development may delay the phase-out of coal power, this relationship has not been empirically assessed. Here, using comprehensive global datasets on data centers and coal-fired power plants, we apply survival analysis to quantify the link between data center exposure and the risk of coal plant retirement. Using Cox proportional hazard model with continuous proximity-based measures of data center exposure, we find that coal plants exposed to post-2020 data centers exhibit significantly lower retirement risk. This effect is more pronounced for cleaner plants. Further estimation using accelerated failure time model shows that each standard deviation increase in post-2020 data center exposure extends plant lifespan by approximately 0.73%. Finally, we identify 585 coal-fired power generation units that are exposed to data center infrastructure, projected to emit an additional 1.952 Gt CO₂ from 2025 onward considering extended operational lifespans. This study provides the first quantification of the unintended consequences of large-scale data center construction in recent years on fossil fuel infrastructure, highlighting the necessity of integrating digital infrastructure considerations into future energy policy design and net-zero emission projections.

Energy use in road transport and buildings are key contributors to climate change, air pollution, and the corresponding health burden. Given technological progress in recent years, replacing fossil fuels by electricity generated from clean sources emerges as a potential strategy to mitigation harmful environmental and health consequences. Yet, a global assessment of broader benefits of end-use electrification is not available in the academic literature. Here, we line up a multi-disciplinary modelling toolbox to quantify the global air quality and health implications of electrification in a broad scenario space encompassing variation in climate policy, clean air controls, technological progress, and enabling conditions. Under current climate policies, we find that electrification reduces premature deaths related to air pollution by 73,000 (28,000-184,000) globally in 2050. More ambitious climate policies, in line with the 2°C target of the Paris Agreement, raise these benefits by 50% to 110,000 (66,000-270,000) avoided premature deaths per year by mid-century. Around 10% (16%) of the global gains stem from the simultaneous evolution of the electricity generation mix under current (2°C-compatible) climate policies, highlighting that further grid decarbonisation amplifies the health benefits of electrification. Comparing buildings with transport, the results indicate that the electrification of energy use in buildings contributes more strongly to reductions of fine particulate matter, particularly in Africa, India, and China, while road transport electrification provides stronger clean air benefits for ground-level ozone due to nitrogen oxides emission reductions. Overall, the results indicate strong, local, and near-term benefits of policies that increase the pace of end-use electrification through heat pumps, clean cooking, and electric vehicles, jointly contributing to progress on sustainable development goals related to climate, air quality, and health.

This paper investigates the aggregate and dynamic effects of commodity price fluctuations and exchange rate movements on deforestation in the Brazilian Amazon. We identify exogenous shocks to the prices of two key commodities, grains and cattle, as well as to the exchange rate, and use these shocks as external instruments in a Structural Vector Autoregression (SVAR) framework. We find that, while grain price shocks have little effect on forest loss, cattle price shocks generate a persistent increase in deforestation. In addition, we identify a new channel by which exchange rate depreciations lead to higher deforestation, operating through their impact on commodity prices in local currency terms, which underscores the importance of macroeconomic conditions, beyond global commodity markets alone, in shaping deforestation dynamics.