Limits to Growth was right about collapse
By Andrew Curry, originally published by thenextwave
May 20, 2025
. . .
Recalibrating the model
Now another group of scientists have gone back to the original World3 model and recalibrated it against the best fit data, which is what has prompted Klement’s article. ‘Recalibration of limits to growth: An update of the World3 model’, by Nebel et al, is published open access in the Journal of Industrial Ecology.
The point of doing this is—as they say in their abstract—is
to better match empirical data on world development.
For assurance, the article goes into detail about their method and their approach, including access to the Python scripts they have used, which have been posted on Github. As they explain:
Since the model was calibrated with the limited capabilities in terms of computing power and data processing in 1972, it seems interesting to what extent a recalibration of the model is possible and what are the effects of such a recalibration. The data situation has improved enormously since then.
Overshoot and collapse
What this means is that the recalibrated model reflects the best available current data.
Klement explains the approach this way:
If the recalibrated model deviates significantly from the forecasts of the 1970s, we have made progress. And given the accuracy of the original model, we can also take some comfort that any progress we made is likely to be real and we have truly extended economic growth further into the future.
And the outcomes of this recalibration? Well, they’re not good. Again from the abstract to the article:
This improved parameter set results in a World3 simulation that shows the same overshoot and collapse mode in the coming decade as the original business as usual scenario of the LtG standard run. [My emphasis]
Industrial production decline
It is worth spelling out the dimensions of this overshoot and collapse. As it happens, there are some handy charts both in Joachim Klement’s article and in the original paper. I’ve reproduced Klement’s versions here because they are easier to read, but the originals are downloadable from the original article.
The first chart is for industrial production, where the recalibrated version tracks the original to the centimetre, pretty much. Does this matter? Klement does the usual wave towards the fact that we live in a much more services-based economy than we did even when the Limits team did the original study.
Peaking food production
I’m not completely convinced of this. As David Mindell observes in The New Lunar Society (I have a review in the works), industrial production has significant multiplier effects on other economic activity, even if this is largely overlooked by economics [p33]. The updated population diagram also follows closely the decline in the mid-2030s that is seen in the Limits base case projection.
On food production, it also seems to be peaking right about now indicating that despite continuous growth in the global population, we are experiencing declining global food production.
Delaying the peak
So it’s possible to imagine that one of the reasons why population starts to decline is because there isn’t enough food to go around. It’s also possible to imagine that the higher peak, and faster overshoot, is a result of the intensification of agriculture and food production, which is about the application of technology. As one of the Limits’ authors, Dennis Meadows, always insisted when asked, technology can delay a peak, but the crash comes harder when it comes.
. . .
Full article:
https://www.resilience.org/stories/2025-05-20/limits-to-growth-was-right-about-collapse/
504 Battery Dr /
npub1ge…wxryp
2025-05-21 14:51:32
Author Public Key
npub1geasa2ewhxp9jcm6lg2c53qk02rcgqx2xw2qm9seyfxysu2svjwsawxrypSeen on
wss://relay.nostr.bandPublished at
2025-05-21 14:51:32Event JSON
{
"id": "237eb52aa05472cf0d8984772c537e7cec5456d69a00d63dd61f779637d34485",
"pubkey": "467b0eab2eb98259637afa158a44167a878400ca33940d9619224c487150649d",
"created_at": 1747839092,
"kind": 1,
"tags": [
[
"proxy",
"https://kolektiva.social/users/504DR/statuses/114546382754984058",
"activitypub"
],
[
"client",
"Mostr",
"31990:6be38f8c63df7dbf84db7ec4a6e6fbbd8d19dca3b980efad18585c46f04b26f9:mostr",
"wss://relay.mostr.pub"
]
],
"content": "Limits to Growth was right about collapse\nBy Andrew Curry, originally published by thenextwave\n\nMay 20, 2025\n\n. . .\nRecalibrating the model\n\nNow another group of scientists have gone back to the original World3 model and recalibrated it against the best fit data, which is what has prompted Klement’s article. ‘Recalibration of limits to growth: An update of the World3 model’, by Nebel et al, is published open access in the Journal of Industrial Ecology.\n\nThe point of doing this is—as they say in their abstract—is\n\nto better match empirical data on world development.\n\nFor assurance, the article goes into detail about their method and their approach, including access to the Python scripts they have used, which have been posted on Github. As they explain:\n\nSince the model was calibrated with the limited capabilities in terms of computing power and data processing in 1972, it seems interesting to what extent a recalibration of the model is possible and what are the effects of such a recalibration. The data situation has improved enormously since then.\n\nOvershoot and collapse\n\nWhat this means is that the recalibrated model reflects the best available current data.\n\nKlement explains the approach this way:\n\nIf the recalibrated model deviates significantly from the forecasts of the 1970s, we have made progress. And given the accuracy of the original model, we can also take some comfort that any progress we made is likely to be real and we have truly extended economic growth further into the future.\n\nAnd the outcomes of this recalibration? Well, they’re not good. Again from the abstract to the article:\n\nThis improved parameter set results in a World3 simulation that shows the same overshoot and collapse mode in the coming decade as the original business as usual scenario of the LtG standard run. [My emphasis]\n\nIndustrial production decline\nIt is worth spelling out the dimensions of this overshoot and collapse. As it happens, there are some handy charts both in Joachim Klement’s article and in the original paper. I’ve reproduced Klement’s versions here because they are easier to read, but the originals are downloadable from the original article.\n\nThe first chart is for industrial production, where the recalibrated version tracks the original to the centimetre, pretty much. Does this matter? Klement does the usual wave towards the fact that we live in a much more services-based economy than we did even when the Limits team did the original study.\n\nPeaking food production\n\nI’m not completely convinced of this. As David Mindell observes in The New Lunar Society (I have a review in the works), industrial production has significant multiplier effects on other economic activity, even if this is largely overlooked by economics [p33]. The updated population diagram also follows closely the decline in the mid-2030s that is seen in the Limits base case projection.\n\nOn food production, it also seems to be peaking right about now indicating that despite continuous growth in the global population, we are experiencing declining global food production.\n\nDelaying the peak\n\nSo it’s possible to imagine that one of the reasons why population starts to decline is because there isn’t enough food to go around. It’s also possible to imagine that the higher peak, and faster overshoot, is a result of the intensification of agriculture and food production, which is about the application of technology. As one of the Limits’ authors, Dennis Meadows, always insisted when asked, technology can delay a peak, but the crash comes harder when it comes.\n. . .\n\nFull article: \nhttps://www.resilience.org/stories/2025-05-20/limits-to-growth-was-right-about-collapse/",
"sig": "16f3610282aed9e6d3b388e923edf5f572c381e26caf158ce3e24c9c784305668fe71c5adec0913fab84090730ddba4a99b6cdde7a5eafc30b02ad279f625cdf"
}