I suspect that the effect sizes of impacts of the economy on climate are overestimated from their methodology. The logic is that their econometric model is a fixed-effects distributed lag model. However, over the period where climate impacts were largest (1990 to 2020), the rate of growth of the...
...global economy slowed. This was not largely due to the effect of climate but rather due to changes in population growth rates and population aging, as well as perhaps some effect of declining returns to technological innovation.
We can see this in GDP per capita growth for the United States, which was quite high when the world's temperature was lower, due to mechanization and an increasing labour force. It is now less so when ΔT is high, due to an increasing dependency ratio and population aging.
Without explicitly including population, dependency ratio, technological factors, or other variables into the model as cofactors, the model has no choice but to assign all the recent slowdown in GDP/capita growth to increasing temperatures and other climate variables.
Among other things, the biggest tell for me is that direct modelling studies cited in IPCC reports predict economic impacts that are quadratic in temperature rise (or higher), because the frequency of extremely harmful exceptional events grows quadratically (or more).
This model derives a quasi-linear impact in economic impact. I say quasi-linear because they do recover increasing impacts of climate effects with rising temperature, which will provide some increasing impacts with increasing change in temperature, as other studies show.
Comments