Generally, my agenda was probably a bit simpler than people might have supposed. This was not intended to be the last word on whether climate change or development interventions are always better. Rather it’s a starting point and “choose your own adventure” model to help prioritizing between a concrete climate and a concrete development charities. Different situations call for the model to be adapted.
Note that there are four parameters that drive the results of this analysis (the SCC, the income adjustment eta, the cost to avert CO2, and the effectiveness of global dev/health vs. cash). For the first two, there really is a lot more uncertainty, but for the latter two, it’s more clear. This makes the model actually valuable and with action guiding potential.
For instance, if you’re a small donor and can’t decide between GiveDirectly and the Coalition for Rainforest Nations, then, if you believe that CfRN really has a cost-effectiveness of $0.02 / tCO2e averted, in many scenarios, especially the realistic one around which there is most consensus, it will often beat unconditional cash-transfers, even if you believe that social cost of carbon is quite low.
However, CfRN does lobbying, not a scalable intervention that one could invest a lot of money in. So, in contrast, if you’re a billionaire and are looking to decide between global development and climate change as a cause area for your foundation, then perhaps global development might be a better bet.
You write that there are very large flaws in my methodology, but because you then adopted the methodology, I think you actually have quarrels with the empirical estimates that I’ve plugged in, correct?
Some comments on the parameter estimates that you use in your model:
Your less $1/tCO2e averted figure for Cool Earth seems fairfor small donors (deforestation prevention can probably absorb a few hundreds of millions—see “Redd+ agreement” and social impact bonds between Norway and Brazil).
However, for large foundations/governments it doesn’t seem quite as scalable in terms of absorbing very large amounts of money as many global development interventions. That’s why I used an intervention such as ocean alkalinity as an example because it might be a way to absorb large amounts of carbon up to 100 billion tonnes / year) for as little as $10 per tonne of CO₂ averted. https://iopscience.iop.org/article/10.1088/1748-9326/aabf9f/pdf
I thought this intervention was representative / similar order of magnitude (10s of $ / ton averted) as some of the bigger ones in the McKinsey report. The next order of magnitude, 100s of $ is for direct air capture, which as far as I understand could absorb most of the CO at scale, but is too expensive. I think this is why getting direct air capture costs down by one more order of magnitude is seen as a climate holy grail, where you can just pump money into and then it solves the whole problem.
But your model seems to be geared towards small donors deciding between to different charities, and then is inconsistent, because you used the median Givewell charity effectiveness (7.95) and not the most effective 17, comparing the best in class low risk offsetting with the median development charity. Even using your model parameters suggests small donors should donate to Deworming over Cool Earth.
Finally, can you say a bit more why you prefer the eta, marginal utility of consumption, to be equal to 1? I felt you you did not provide sufficient empirical justification for this.
See:
“To arrive at estimates of social discount rates consistent with these growth rates, it is necessary to obtain estimates of the elasticity of marginal utility [...] the survey by Groom and Maddison (2019) suggests estimates between 0.5 and 2.0.
The most substantial cross-country analysis available (Evans, 2005; with a focus on advanced economies) arrives at an estimate of 1.4, which we adopt. This estimate is consistent with a review of some 200 experts who have published on social discount rates, which returns a mean value of 1.35 (Drupp et al., 2018). This estimate—drawing on experts who have published on discount rates in highly ranked journals is not necessarily confined to advanced economies, though the authors acknowledge that expertise from developing countries might be underrepresented.”
“We present a quantitative survey of estimates of the elasticity of intertemporal substitution in what we believe is the largest metaanalysis conducted in economics. We collect 2735 estimates from 169 published studies and find that the mean elasticity is 0.5, but that the estimates vary greatly across countries and methods.” https://www.sciencedirect.com/science/article/abs/pii/S002219961500032X
“estimates of η are derived from the so-called Euler-equation although in the macroeconomics literature this information is normally presented in terms of the elasticity of intertemporal substitution (EIS) which is equal to 1/η.” https://link.springer.com/article/10.1007/s10640-018-0242-z
So the last study suggest the eta is equal 2.
This is why I plugged in 1.5 for the realistic case. But I haven’t looked into this in detail and I’d love to have more people look into it.
I hope this did not come across as too critical—I generally really enjoyed reading your treatment and synthesis of the issue.
Generally, my agenda was probably a bit simpler than people might have supposed. This was not intended to be the last word on whether climate change or development interventions are always better. Rather it’s a starting point and “choose your own adventure” model to help prioritizing between a concrete climate and a concrete development charities. Different situations call for the model to be adapted.
That may have been your intention, but the title of your article is “Global development interventions are generally more effective than Climate change interventions” and your summary states “My spreadsheet model below shows that climate change interventions are only more effective than global development interventions, if and only if: [...] under quite pessimistic assumptions about climate change (if the social cost of carbon is higher than $1000 per tonne of carbon)”. I think that these things together would make it very easy for a reader to leave with the simple conclusion that climate change interventions are not cost effective, and I just don’t think the evidence exists to back up that simple conclusion.
I also don’t think that you need to make particularly pessimistic assumptions for the social cost of carbon to be much higher. At the very least, your chosen source of a social cost of carbon used an emission pathway (RCP6.0) which only results in 2.2C of warming by 2100. Based on currently announced national commitments, greenhouse emissions are likely to lead to global temperature increases of 2.3ºC-3.7ºC by 2100 with a 25% chance of exceeding 4°C based on current national policies.
You write that there are very large flaws in my methodology, but because you then adopted the methodology, I think you actually have quarrels with the empirical estimates that I’ve plugged in, correct?
I think the methodology is flawed in the sense that you are combining the low/mid/high estimates of several parameters to produce an estimate which is 10 orders of magnitude wide. That’s so wide as to be almost meaningless. I produced an updated estimate mainly to demonstrate that it’s possible to produce an estimate where climate change is better value than global health with some fairly plausible choices of parameters.
However, for large foundations/governments it doesn’t seem quite as scalable in terms of absorbing very large amounts of money as many global development interventions. That’s why I used an intervention such as ocean alkalinity as an example because it might be a way to absorb large amounts of carbon up to 100 billion tonnes / year) for as little as $10 per tonne of CO₂ averted.
The trouble with these estimates is that I’m not convinced they do a good job of considering how costs change as a technology is scaled. For example, we’ve seen this with solar—http://solarsouthwest.co.uk/solar-panel-cost/. Do you have a recommended source which does somehow take account of these effects? If not, we’re not really comparing costs properly.
Additionally, it’s worth recognizing that the current economic model has huge climate externalities. I really hope we get a climate emissions tax at some point, at which point the fundamental incentive structures change, and I’m not sure how to properly price emissions reductions at that point. At least from a government perspective, the carbon tax could be considered “free”. Can you recommend any papers which have tried to come up with a cost/tonne for a carbon tax?
Finally, can you say a bit more why you prefer the eta, marginal utility of consumption, to be equal to 1? I felt you you did not provide sufficient empirical justification for this.
Honestly, it’s because I haven’t yet had a chance to read up on the marginal utility of consumption and it seemed implausible to me that the value of money would actually be 3 orders of magnitude higher in a cash transfer situation. I’m very much prepared to believe that I’m wrong, and I hope to find the time at some point to read the research you referenced and figure that out for myself. In my defense, I also think the SCC used in my update is perhaps an order of magnitude too small, so I could also have used a higher SCC and the 1260x income adjustment and come to the same conclusion.
I hope this did not come across as too critical—I generally really enjoyed reading your treatment and synthesis of the issue.
Thank you for taking the time to reply! I’ve enjoyed responding to your points :)
Thanks for engaging with this critically!
Generally, my agenda was probably a bit simpler than people might have supposed. This was not intended to be the last word on whether climate change or development interventions are always better. Rather it’s a starting point and “choose your own adventure” model to help prioritizing between a concrete climate and a concrete development charities. Different situations call for the model to be adapted.
Note that there are four parameters that drive the results of this analysis (the SCC, the income adjustment eta, the cost to avert CO2, and the effectiveness of global dev/health vs. cash). For the first two, there really is a lot more uncertainty, but for the latter two, it’s more clear. This makes the model actually valuable and with action guiding potential.
For instance, if you’re a small donor and can’t decide between GiveDirectly and the Coalition for Rainforest Nations, then, if you believe that CfRN really has a cost-effectiveness of $0.02 / tCO2e averted, in many scenarios, especially the realistic one around which there is most consensus, it will often beat unconditional cash-transfers, even if you believe that social cost of carbon is quite low.
However, CfRN does lobbying, not a scalable intervention that one could invest a lot of money in. So, in contrast, if you’re a billionaire and are looking to decide between global development and climate change as a cause area for your foundation, then perhaps global development might be a better bet.
You write that there are very large flaws in my methodology, but because you then adopted the methodology, I think you actually have quarrels with the empirical estimates that I’ve plugged in, correct?
Some comments on the parameter estimates that you use in your model:
Your less $1/tCO2e averted figure for Cool Earth seems fair for small donors (deforestation prevention can probably absorb a few hundreds of millions—see “Redd+ agreement” and social impact bonds between Norway and Brazil).
However, for large foundations/governments it doesn’t seem quite as scalable in terms of absorbing very large amounts of money as many global development interventions. That’s why I used an intervention such as ocean alkalinity as an example because it might be a way to absorb large amounts of carbon up to 100 billion tonnes / year) for as little as $10 per tonne of CO₂ averted. https://iopscience.iop.org/article/10.1088/1748-9326/aabf9f/pdf
I thought this intervention was representative / similar order of magnitude (10s of $ / ton averted) as some of the bigger ones in the McKinsey report. The next order of magnitude, 100s of $ is for direct air capture, which as far as I understand could absorb most of the CO at scale, but is too expensive. I think this is why getting direct air capture costs down by one more order of magnitude is seen as a climate holy grail, where you can just pump money into and then it solves the whole problem.
But your model seems to be geared towards small donors deciding between to different charities, and then is inconsistent, because you used the median Givewell charity effectiveness (7.95) and not the most effective 17, comparing the best in class low risk offsetting with the median development charity. Even using your model parameters suggests small donors should donate to Deworming over Cool Earth.
Finally, can you say a bit more why you prefer the eta, marginal utility of consumption, to be equal to 1? I felt you you did not provide sufficient empirical justification for this.
See:
“To arrive at estimates of social discount rates consistent with these growth rates, it is necessary to obtain estimates of the elasticity of marginal utility [...] the survey by Groom and Maddison (2019) suggests estimates between 0.5 and 2.0.
The most substantial cross-country analysis available (Evans, 2005; with a focus on advanced economies) arrives at an estimate of 1.4, which we adopt. This estimate is consistent with a review of some 200 experts who have published on social discount rates, which returns a mean value of 1.35 (Drupp et al., 2018). This estimate—drawing on experts who have published on discount rates in highly ranked journals is not necessarily confined to advanced economies, though the authors acknowledge that expertise from developing countries might be underrepresented.”
“We present a quantitative survey of estimates of the elasticity of intertemporal substitution in what we believe is the largest metaanalysis conducted in economics. We collect 2735 estimates from 169 published studies and find that the mean elasticity is 0.5, but that the estimates vary greatly across countries and methods.” https://www.sciencedirect.com/science/article/abs/pii/S002219961500032X
“estimates of η are derived from the so-called Euler-equation although in the macroeconomics literature this information is normally presented in terms of the elasticity of intertemporal substitution (EIS) which is equal to 1/η.” https://link.springer.com/article/10.1007/s10640-018-0242-z
So the last study suggest the eta is equal 2.
This is why I plugged in 1.5 for the realistic case. But I haven’t looked into this in detail and I’d love to have more people look into it.
I hope this did not come across as too critical—I generally really enjoyed reading your treatment and synthesis of the issue.
That may have been your intention, but the title of your article is “Global development interventions are generally more effective than Climate change interventions” and your summary states “My spreadsheet model below shows that climate change interventions are only more effective than global development interventions, if and only if: [...] under quite pessimistic assumptions about climate change (if the social cost of carbon is higher than $1000 per tonne of carbon)”. I think that these things together would make it very easy for a reader to leave with the simple conclusion that climate change interventions are not cost effective, and I just don’t think the evidence exists to back up that simple conclusion.
I also don’t think that you need to make particularly pessimistic assumptions for the social cost of carbon to be much higher. At the very least, your chosen source of a social cost of carbon used an emission pathway (RCP6.0) which only results in 2.2C of warming by 2100. Based on currently announced national commitments, greenhouse emissions are likely to lead to global temperature increases of 2.3ºC-3.7ºC by 2100 with a 25% chance of exceeding 4°C based on current national policies.
I think the methodology is flawed in the sense that you are combining the low/mid/high estimates of several parameters to produce an estimate which is 10 orders of magnitude wide. That’s so wide as to be almost meaningless. I produced an updated estimate mainly to demonstrate that it’s possible to produce an estimate where climate change is better value than global health with some fairly plausible choices of parameters.
The trouble with these estimates is that I’m not convinced they do a good job of considering how costs change as a technology is scaled. For example, we’ve seen this with solar—http://solarsouthwest.co.uk/solar-panel-cost/. Do you have a recommended source which does somehow take account of these effects? If not, we’re not really comparing costs properly.
Additionally, it’s worth recognizing that the current economic model has huge climate externalities. I really hope we get a climate emissions tax at some point, at which point the fundamental incentive structures change, and I’m not sure how to properly price emissions reductions at that point. At least from a government perspective, the carbon tax could be considered “free”. Can you recommend any papers which have tried to come up with a cost/tonne for a carbon tax?
Honestly, it’s because I haven’t yet had a chance to read up on the marginal utility of consumption and it seemed implausible to me that the value of money would actually be 3 orders of magnitude higher in a cash transfer situation. I’m very much prepared to believe that I’m wrong, and I hope to find the time at some point to read the research you referenced and figure that out for myself. In my defense, I also think the SCC used in my update is perhaps an order of magnitude too small, so I could also have used a higher SCC and the 1260x income adjustment and come to the same conclusion.
Thank you for taking the time to reply! I’ve enjoyed responding to your points :)