Limits To Growth is kind of a test case for whether people got the memo about the academic research crisis or not (which incorporates the replication crisis, but is wider than it).
The World3 model came out of social studies and isn't even really science, given that the authors never tried to obtain evidence for their hypothesis before publishing. World3 is a toy with no predictive value, which is why reimplementing it in HTML is possible: it's just a handful of equations. Such a thing never could make accurate predictions, which was extremely obvious at the time and pointed out by many (e.g. Nordhaus, Smil), but this didn't stop its authors being feted as great scientists regardless.
There's a lot of similarities here with COVID modelling in the sense that these sorts of academic modelling exercises all seem to suffer the same kinds of problems and social failure modes:
• Over-simplified to the point that laypeople can easily spot what's missing.
• Guesses pulled out of thin air are treated as having equal value to measured data.
• An implicit assumption that the population is purely static and doesn't react to changing circumstances in any way. When later the predictions turn out to be wrong it's argued that they would have been right if not for the modelling causing behavioral changes, making the entire exercise unfalsifiable pseudo-science by its own premise.
• Either never validated against reality or actually falsified, but academics pretend otherwise.
• Predictions phrased vaguely enough that what most people would consider to be falsification is later argued to be within the bounds of one of the many possible scenarios.
The uselessness of such modelling exercises is evident in the fact that despite at least 50+ years of investment by academia and government into them, there are no academic computerized models widely being used to predict the course of the society or economy.
The World3 model came out of social studies and isn't even really science, given that the authors never tried to obtain evidence for their hypothesis before publishing. World3 is a toy with no predictive value, which is why reimplementing it in HTML is possible: it's just a handful of equations. Such a thing never could make accurate predictions, which was extremely obvious at the time and pointed out by many (e.g. Nordhaus, Smil), but this didn't stop its authors being feted as great scientists regardless.
There's a lot of similarities here with COVID modelling in the sense that these sorts of academic modelling exercises all seem to suffer the same kinds of problems and social failure modes:
• Over-simplified to the point that laypeople can easily spot what's missing.
• Guesses pulled out of thin air are treated as having equal value to measured data.
• An implicit assumption that the population is purely static and doesn't react to changing circumstances in any way. When later the predictions turn out to be wrong it's argued that they would have been right if not for the modelling causing behavioral changes, making the entire exercise unfalsifiable pseudo-science by its own premise.
• Either never validated against reality or actually falsified, but academics pretend otherwise.
• Predictions phrased vaguely enough that what most people would consider to be falsification is later argued to be within the bounds of one of the many possible scenarios.
The uselessness of such modelling exercises is evident in the fact that despite at least 50+ years of investment by academia and government into them, there are no academic computerized models widely being used to predict the course of the society or economy.