TL;DR
This paper tackles the significant uncertainty in climate change damage estimates by reformulating a standard economic model (DICE) to incorporate a wide range of expert opinions. The findings suggest that accounting for these uncertainties, particularly the possibility of extreme "catastrophic" events, leads to significantly higher optimal carbon taxes than previously estimated, highlighting the importance of strategic hedging in climate policy.
What Problem Does This Paper Address?
Climate change policy decisions, particularly the optimal level of carbon taxes, are heavily influenced by estimates of the economic damages caused by rising global temperatures. However, these damage estimates are fraught with uncertainty, stemming from complex climate science, economic modeling, and differing expert opinions. Traditional approaches often rely on "point estimates" or "best guesses," which can mask the full spectrum of potential outcomes, including low-probability but high-consequence "catastrophic" events. This paper addresses the critical need to quantify and incorporate these uncertainties into policy analysis, arguing that a more comprehensive understanding of the potential damage range is essential for making robust and strategic climate policy decisions.
Methods
The authors utilize the Dynamic Integrated Climate-Economy (DICE) model, a well-established framework for analyzing climate policy. Their primary methodological contribution is the reformulation of the DICE model to incorporate a wider range of climate damage functions. Instead of relying on a single, "best guess" damage estimate, they:
- Incorporate diverse damage functions: They derive and use several alternative damage functions based on published IPCC estimates and expert surveys, moving beyond Nordhaus's original formulation.
- Utilize expert opinion distributions: Rather than just point estimates, they leverage subjective probability distributions provided by experts, capturing the range and skewness of their opinions.
- Perform Monte Carlo simulations: To analyze the impact of these uncertainties, they conduct Monte Carlo simulations, running the reformulated DICE model thousands of times with randomly sampled damage parameters. This allows for a probabilistic assessment of optimal carbon tax rates.
- Focus on damage function sensitivity: The core of their analysis is to demonstrate how variations in damage function assumptions, particularly the inclusion of outliers and extreme scenarios, affect policy recommendations.
Datasets
The paper primarily uses data from two sources:
- Published IPCC damage estimates: Data from Cline, Fankhauser, Titus, and Tol, as compiled by the IPCC [Bruce et al., 1996].
- Expert survey data: Data from a survey conducted by Nordhaus [Nordhaus, 1994a], which collected damage estimates (10th, 50th, and 90th percentiles) from 18 experts across different disciplines.
The paper does not introduce new datasets but rather re-analyzes existing ones within a more comprehensive uncertainty framework.
Key Insights and Ideas
- "Best guess" estimates can be misleading: Relying on single point estimates for climate damages can obscure the significant risks associated with extreme events, leading to potentially inadequate policy responses.
- Expert opinions are diverse and often skewed: The survey data reveals a wide range of opinions among experts, with a notable tendency towards right-skewness, meaning experts believe there's a higher probability of a "nasty surprise" (larger damages) than a "pleasant surprise."
- Uncertainty drives higher carbon taxes: Incorporating a broader range of damage estimates, especially those reflecting pessimistic outcomes and potential catastrophic events, significantly increases the calculated optimal carbon tax.
- Strategic hedging is crucial: The possibility of extreme damage outcomes suggests that policy makers should consider "strategic hedging" – similar to purchasing insurance – to mitigate the risks of low-probability, high-consequence events.
- Non-linearities matter: The non-linear nature of the DICE model amplifies the impact of higher damage estimates, further pushing up optimal carbon taxes.
Key Results
- Optimal carbon taxes are significantly higher: When accounting for a wide range of expert opinions and potential extreme damages, the optimal carbon tax rates are substantially higher (three to eight times higher on average, and over twenty times higher in "surprise" scenarios) than those suggested by the original DICE model.
- Pessimistic scenarios lead to drastic tax increases: For instance, in the most pessimistic damage runs, optimal carbon taxes can reach nearly $200 per ton in 1995 and climb to nearly $500 per ton by the end of the 21st century.
- Original DICE model is optimistic: The original DICE model's damage function is found to be more optimistic, leading to lower recommended carbon taxes compared to the bulk of expert opinion.
- Damage function choice is critical: The choice of damage function has a profound impact on policy recommendations, underscoring the need for careful consideration of uncertainty.
Limitations and Future Work
- Model Simplification: The authors acknowledge that the DICE model is a simplification of complex reality and that its quantitative results should be interpreted qualitatively. They do not reformulate the Ω(t) term, which links economic and climatic sides, to focus on damage function sensitivity.
- Expert Opinion Credibility: While using aggregate expert opinion, the authors note that it should not be considered a definitive "consensus" and that new research may reshape these opinions.
- Future Work: The paper suggests that a more comprehensive analysis incorporating uncertainties in all major components of climate models, explicitly categorized and displayed, would lead to better-informed policy choices, including options for strategic hedging. They also point to the need for explicit and transparent communication of uncertainty ranges to policymakers.
Summary updated