Why I Don't Believe The Latest Climate Change Economic Damage Estimates

Sure, sure, this is just me but perhaps I can convince

From Tyler Cowen:

You may recall that earlier estimates of climate change costs were more like a five to ten percent welfare loss to the world. I do not however find the main results here plausible. The estimation is extremely complicated, and based on the premise that a higher global temperature does more harm to a region than a higher local temperature. And are extreme events a “productivity shock,” or a one-time resource loss that occasions some Solow catch-up? Is the basic modeling consistent with the fact that, while the number of extreme storms may be rising, the number of deaths from those same storms is falling over time? Lives lost are not the same as economic costs, but still the capacity for adjustment seems considerably underrated. What about the effects to date? The authors themselves write: “According to our counterfactual, world GDP per capita would be more than 20% higher today had no warming occurred between 1960 and 2019.” I absolutely do not believe that claim.

OK, so why shouldn’t we believe that claim? The paper is here and this is the crucial part:

To evaluate the consequences of climate change, we specify a path for global mean temperature. The baseline year t = 0 corresponds to 2024. The world subsequently warms by 3°C above preindustrial levels by 2100, after which temperature asymptotes to 3.3°C. This scenario is broadly consistent with IPCC business-as-usual scenarios that imply 3 to 4°C of warming by 2100 (Hausfather and Peters, 2020; Lee et al., 2023). Given that the world has warmed by approximately 1°C since preindustrial times, this scenario implies 2°C of additional warming since t = 0 (2024) by year t = 76 (2100).

And that’s wrong. Here’s an AI:

IPCC “business-as-usual” scenarios (often represented by high-emission pathways like RCP8.5 or SSP5-8.5) project global mean temperatures to rise by approximately 3oC to over 4oC above pre-industrial levels by 2100. These scenarios assume high fossil fuel consumption, rapid, unmitigated economic growth, and limited climate policies.

And here’s a paper I prepared earlier:

The good news is that the worst predictions of climate change aren’t going to come true. The bad news is that they never were going to and we’ve been misled all these years. The headline prediction was that carbon dioxide emissions were on track for something called Representative Concentration Pathway (RCP) 8.5. This trajectory is what drives the near hysterical predictions of sea-level rise, warming itself, droughts, crop failures and the rest. But we’re not on that path; we’re currently on something closer to RCP4.5. This is where climate change itself is a minor – if chronic – problem, largely solved by standard market processes with, if we should so wish, the addition of a little bit of tinkering. The overturning of consumer capitalism, of civilisation itself, is therefore not warranted.

The central conceit of RCP 8.5 - or of the earlier A1FI that it is based upon - is that the world refuses to use unconventional oil and gas - fracking that is - and also doesn’t build out wind, solar, geothermal and the rest and so reverts back to coal usage to power society. We end up, in this scenario, using not just vastly more coal but more coal as a portion of energy supply than we ever have done.

A1FI - or RCP 8.5 etc - is just about feasible to use as an oh my God this is going to be terrible outcome. Maybe. But it’s near lunatic to use it as something that is going to happen. Or even is possible to happen. Given that solar is now cheap, we have gone fracking, we’re not turning back to coal and so on.

A more useful statement is that RCP 8.5 isn’t going to happen therefore estimates based upon it happening are wrong. But that use of RCP 8.5 - what they call IPCC business as usual scenarios - is the driver of their cost estimates.

Another way to put their claim. If emissions are very high - really, really, high, higher than anyone’s any right to expect - then climate change will be bad, climate change will be economically expensive. Sure, I’ve no problem with that. Well, as a statement, I don;t mean I’d like it to happen. But that expensive is driven by emissions being really really high, not by the likelihood of their being so. This is how they reach their optimal carbon tax of $1,200 a tonne. This is the same assumption used by James Hansen when he said the carbon tax should be $1,000 a tonne (actually “could be up to” but the qualifier was immediately dropped because of course it was) - if emissions are really high, climate is very sensitive to emissions, then and therefore. But, of course, the expense of emissions does depend upon how many of them there are and thus the temperature rise that happens. Fewer emissions would mean a lower cost from less damage.

This new paper on the costs of climate change assumes the very worst of modelled outcomes - RCP 8.5. There’s near no one who believes that RCP 8.5 could even happen let alone will. But it’s that assumption that drives their costs result. Which is why I don’t believe their costs result.

Now, whether I’ve convinced you is another matter but that certainly convinces me.