1. Integrated Assessment Models (IAMs) often assume the same carbon prices in each region (left, orange dots): this is efficient but leads to large inequities (right).
More equitable distributions of carbon prices (left, blue dots) is less efficient.
doi.org/10.1038/s41586…
More equitable distributions of carbon prices (left, blue dots) is less efficient.
doi.org/10.1038/s41586…
2. @NB_pik addresses this problem in a new paper: "The core finding of this research is the strongly nonlinear trade-off between cost-efficiency and sovereignty in achieving the long-term PA climate target in an equitable way."
Small changes to efficiency have big equity gains.
Small changes to efficiency have big equity gains.
3. The gaps between uniform & differentiated carbon prices (first tweet) was modified to create the trade-off curve (previous tweet). This was done by applying an exponential function to adjust pairs of regional prices.
Is the "core finding" dependent on the "exponential"?
Is the "core finding" dependent on the "exponential"?
4. A sensitivity is done in the Annex, compared to a linear. A linear scaling does not give the trade-off gains as does the exponential, so there is quite some sensitivity to how the scaling is done.
The trade-off frontier lies to the left of the exp, but is hard to compute.
The trade-off frontier lies to the left of the exp, but is hard to compute.
5. What does this mean?
First, there are many near optimal solutions that may have better characteristics. Important to explore this "near optimal" solution space, as it could hold some gems.
Another study here:
sciencedirect.com/science/articl…
First, there are many near optimal solutions that may have better characteristics. Important to explore this "near optimal" solution space, as it could hold some gems.
Another study here:
sciencedirect.com/science/articl…
6. Second, in this particular study, to get the rapid gains would require a rather top-down specification of regional carbon prices, that relate quite specifically to this exponential scaling (?). These will also vary by IAM.
The better solution is even harder to achieve?
The better solution is even harder to achieve?
7. @NB_pik gets some bonus points for mentioning "corner solutions" & "trade-off curves" in a Nature paper as this is what it is all about in a model. The challenge is how to map the model world to the real world to make better policy.
8. While a model operates in a model world, the implication here is that we need regional differentiation in carbon prices (or stringency of climate policy). I don't think this needs to follow the scaling suggested in the paper, as we will never get optimal policies.
9. This leads to the discussion of broader policy mixes, which is perhaps more relevant and interesting, as these are the sorts of policies most likely to be implemented. How to sort out the equity vs efficiency issues here...
10. Thoughts welcome.
@NB_pik, my understanding correct? It struck me most that the "core finding" is quite sensitive to scaling assumptions, but I presume the exponential is closer to the theoretical optimal than the linear. Then it is really an interpretation issue.
/end
@NB_pik, my understanding correct? It struck me most that the "core finding" is quite sensitive to scaling assumptions, but I presume the exponential is closer to the theoretical optimal than the linear. Then it is really an interpretation issue.
/end
• • •
Missing some Tweet in this thread? You can try to
force a refresh
