Research
Working papers
The Price of Delay: Infrastructure Lock-in and the Carbon Tax Lock-in Premium in the Oil Sector (Job Market Paper)
Oil projects are typically costly to develop and comparatively cheap to operate, so a carbon tax high enough to have deterred a project can be too low to stop it once it is built. A delay in climate policy therefore locks in emissions that the delayed policy can no longer remove. Existing models of oil supply assign each field a single cost per barrel, which rules this lock-in out. I separate the decision to develop a field from the decision to keep it producing for 25,336 oil fields covering the world’s reserves, and study a 1.5°C-aligned carbon tax delayed from 2016 to 2030. A 2030 tax that ignores lock-in, as these models do, overshoots the remaining carbon budget by 13 GtCO2eq, 10% of it, although it is only 3.5% too low. Almost half of this overshoot would remain even if investors had correctly expected the tax in 2030. Because locked-in fields barely respond to a higher tax, meeting the budget strands other producers instead, thus redistributing the cost of the delay. Lock-in is concentrated in a few hundred capital-intensive projects, mostly offshore, so that targeted bans can remove nearly half of it.
Previously circulated as The long-term cost of delaying carbon taxation in the oil sector, and The Price of Delay: infrastructure lock-in and the carbon tax lock-in premium in the oil sector.
Voice Until the Last Exit: Ownership Structure and the Limits of Voluntary Environmental Commitments
The effectiveness of corporate climate initiatives depends not only on firms’ commitments but also on existing asset ownership structures. We study this question in the context of the World Bank’s Zero Routine Flaring by 2030 initiative, under which oil and gas companies pledge to eliminate routine flaring from their assets. Using asset-level ownership and satellite-based flaring measurements for nearly 20,000 oil and gas assets worldwide (2012–2024), we exploit staggered firm endorsements to study how committed firms navigate the tradeoff between voice, exercising governance influence to reduce emissions, and exit, divesting from assets that are costly to decarbonize. Committed ownership reduces flaring gradually but persistently, reaching approximately 45% below pre-commitment levels after nine years. The persistence of these gains hinges on continued committed presence: when the last committed owner divests, flaring rises by roughly 40% within two years and exceeds 90% within six years, a result we term the “last good owner effect.” Retaining at least one committed voice on the board is both necessary and sufficient to sustain emissions discipline.