There are different types of Integrated Assessment Models (IAMs). In the climate world, we usually differentiate two. One type looks at cost-benefit analysis of climate impacts, such as, DICE, FUND, PAGE. Another type looks at cost-effective energy system transformations to stay below a given temperature level, such as, GCAM, IMAGE, MESSAGE, REMIND, and many more.
The models focussing on cost-benefit analysis often look at the social cost of carbon, and because of their relative simplicity, they do rather comprehensive uncertainty analyses.
The models focussing on energy system transformations, the ones that most interest me, focus on multiple scenarios and technological detail. I often see very little sensitivity analysis from these models, or at least I can’t find it!
These days, climate scientists think they are cool when they are working on 1.5°C or geoengineering. Well, Nordhaus had that research done-and-dusted when today’s climate scientists were in nappies, me included. (Ok, I exaggerate, but I was probably still in nappies when Nordhaus started working in this area).
The blogpost gives a nice summary of the different types of models and there uses, but I was immediately drawn to the results.
I am sure people don’t like DICE and don’t like models, but I don’t care! I am interested in processes here, not exact numbers.
Nordhaus compares several scenarios:
- Baseline: No climate-change policies are adopted
- Optimal: Climate-change policies maximize economic welfare, with full participation by all nations starting in 2020
- Temperature-limited: The optimal policies are undertaken subject to a further constraint that global temperature does not exceed 2.5°C above the 1900 average. (The international goal of 2°C is not feasible with current DICE estimates without technologies that allow negative emissions by mid-21st century).
- Stern discounting: These are results associated with an extremely low discount rate as advocated by The Stern Review on the Economics of Climate Change.
Here I want to discuss the three figures, and their key results.
The optimal path represents a balance between costs and benefits. There are many assumptions behind these estimates, but one key one is the discount rate. The discount rate used for this calculation averages 4-5%/yr over the 21st century. This takes the world to 3°C in 2100. Ok, warmer than many would like, but that is not the point here.
The cost-effective pathway takes the world to 2.5°C in 2100. In this calculation, the model is run in a way that ensures that it stays below 2.5°C, irrespective of the cost. The carbon price in this scenario is high, reaching over 200$/tCO2 in 2030, compared to less than 50$/tCO2 in the optimal pathway. In 2100, the carbon price is nearly 500$/tCO2 in the cost-effective pathway, compared to about 100$/tCO2 in the optimal pathway.
The interesting thing is when a cost-benefit calculation is done to get the optimal pathway, but using a low discount rate. The low discount rate causes deep near-term mitigation, similar to that in the cost-effective pathway that forces temperatures to stay below 2.5°C. Of course, carbon prices are high, but temperatures are limited to 2.5°C. The temperature-limited and Stern discounting pathways are probably similar since DICE cannot go any lower without carbon dioxide removal.
Why do I find these results interesting?
The current generation of energy-system integrated assessment models don’t seem to do much sensitivity analysis. I have long been wondering if low discount rates would lead to lower levels of carbon dioxide removal.
Models tend to indicate that too rapid short-term reductions could be expensive compared to the cost of negative emissions in the long term (based on expected technology development rates and assumed discounting rates), and thus favour strategies in which considerable levels of net negative emissions are applied.
[W]hen time preference is zero and no discount is applied to future mitigations costs vis à vis with current ones, we can observe significantly lower deployment of [Direct Air Capture]
I am sure there are other studies here and there, but I am not aware of a comprehensive analysis. Such an analysis is needed, and could be important to understand technology deployment in the current generation of emission scenarios.
Is our love of carbon dioxide removal, and soon solar geoengineering, simply because we discount the future? I would love an answer to that question. Any takers?