2021-09-30-未来能源研究所-碳的社会成本_人口_GDP_排放量和贴现率的长期概率预测进展（英）_81页_3mb

Summary of "The Social Cost of Carbon: Advances in Long-Term Probabilistic Projections"

Introduction to Social Cost of Carbon (SCC)

The Social Cost of Carbon (SCC) is an economic metric representing the marginal cost of damages from emitting one additional ton of carbon dioxide (CO₂), used to inform U.S. climate policies. It guides regulations, such as vehicle fuel economy standards and carbon pricing, and is critical for evaluating climate mitigation costs and benefits. Uncertainty characterization and transparency in assumptions are essential due to its influential role in policy decisions.

Key Challenges and Innovations

The paper addresses SCC estimation challenges, such as long-term uncertainties in population, GDP, emissions, and discount rates. It improves upon government methods by incorporating probabilistic projections and stochastic discounting, moving beyond deterministic scenarios and constant discount rates based on NASEM recommendations. This enhances uncertainty characterization and SCC reliability.

Probabilistic Projections

• Economic and Demographic Drivers: Long-term probabilistic projections were developed for population, GDP, and emissions using statistical models and expert elicitation. For example, projections extended to 2300 show global population peaking around 10 billion and then declining, with significant uncertainty.
• Emissions: Future emissions were assessed via expert surveys, incorporating scenarios that reflect economic growth, policy ambition, and technological change. This provides a distribution of emissions pathways, including high and low growth cases.

Climate and Damage Estimation

• Climate Model: The FaIR climate model was used to link emissions to temperature changes under probabilistic projections, accounting for climate uncertainty.
• Damage Estimation: While the paper uses a simplified damage function from the DICE model for illustrative estimates, it highlights the need for updated, empirically derived sectoral damage functions for future SCC updates.

Discounting Approaches

Traditional constant discount rates underestimate long-term uncertainties. The paper introduces stochastic growth discounting, which treats discount rates as uncertain and correlated with economic growth. This approach reduces bias and stabilizes SCC estimates, especially when combined with socioeconomic uncertainty. For instance, under 3 and 2 percent near-term rates, stochastic discounting yields SCC estimates of $56 and $171 per ton CO₂ in 2020 dollars.

Illustrative SCC Estimates

With probabilistic socioeconomic data and stochastic discounting, the illustrative SCC estimate is significantly higher than previous deterministic methods. Key findings:

• Socioeconomic uncertainty drives a 22-53% increase in SCC compared to constant discounting.
• Stochastic discounting reduces SCC volatility and addresses climate-climate correlation, minimizing upward bias in estimates.
• High-emissions scenarios (e.g., SSP5) result in much higher SCC values, underscoring the impact of economic trajectories.

Conclusion and Recommendations

The paper concludes that probabilistic projections and stochastic discounting are crucial for accurate SCC estimation. It calls for continuous updates as scientific and economic understanding advances, including accounting for tipping points, migration, conflict, and equity weighting. Shortcomings include limited integration of sectoral damages, which will be addressed in future research.

**Ethical Considerations**: The discounting approach raises ethical questions about intergenerational equity, as it must balance short-term vs. long-term welfare. Future work should incorporate equity weighting to better reflect distributional impacts.