3 ms·

Say I agree (at least I get your point ;) - I'm curious how does one test (in the empirical science sense) a systems perspective with regards to predictions / s

by wjnc 9y ago

Say I agree (at least I get your point ;) - I'm curious how does one test (in the empirical science sense) a systems perspective with regards to predictions / scenario's? Because at least in popular opinion the Club of Rome is known for a pretty grim worldview and political advocacy. Perhaps I'm too much schooled in traditional economics but Robert Nobel wasn't too kind [1] and even in my environmental economics classes we spent quite some time deconstructing the Club of Rome. [1] https://en.wikipedia.org/wiki/Club_of_Rome#Critics https://en.wikipedia.org/wiki/Club_of_Rome#Critics

Certhas 9y ago
It's important to understand that the aim of systems isn't to predict things but to understand potential mechanisms. The fact that certain scenarios that the Club of Rome anticipated seemed to not come true also unveiled new mechanisms to us. Predictions can't be the benchmark, we are simply not advanced enough in the science for that. Certainly the social sciences, like economics, should understand that best. Economics has hardly been able to predict anything.
smallnamespace 9y ago
At some point, there are limits to our ability to empirically test large-scale predictions because we have no ability to run history more than once. This certainly applies to macroeconomics -- how does one empirically 'test' any prediction from a macroeconomic model based on real-life inputs, other than to see what happens? But this shouldn't preclude us from asking the really difficult questions that we care about, rather than restricting ourselves to the realm of physics, where we can get 5 sigmas in classical statistical tests. With regards to Robert Nobel's criticism, the whole point of the Club of Rome's approach is to model a simplified, 'low resolution' view of the world economy in order to embed it in a larger network of information, goods, and energy flows constrained by the environment -- because the boundary conditions matter a lot to how the macroeconomy functions. Sometimes you can get away with a much simpler modeling approach if you actually get all the inputs right. The biggest criticism the Club of Rome has on most macroeconomic models is that it doesn't model the environmental inputs in sufficient detail, and these things really matter a lot in the long run. It is not simply sufficient to take current 'wage levels' or 'capital stock' or even 'level of technology' as a direct input to your DSGE RBC model -- our macroeconomies perform a quasi-ecological function of converting flows of matter and energy, largely taken from Nature and the environment, and converting them further for human use. Understanding the Nature-human interface, then explicitly modeling the stock quantities of the inputs into the macroeconomy must on some level be important, because they are key to how our entire technological society functions. I you're inclined to learn more about their modeling approach at a high level, I highly recommend Thinking in Systems [1]. Read it a week ago and it blew my mind, as its intuitions match up with a lot of patterns that we see over and over again in dealing with any complex network, but presented in a systematic and convincing manner. [1] http://wtf.tw/ref/meadows.pdf http://wtf.tw/ref/meadows.pdf

by wjnc 9y ago

Certhas 9y ago

It's important to understand that the aim of systems isn't to predict things but to understand potential mechanisms. The fact that certain scenarios that the Club of Rome anticipated seemed to not come true also unveiled new mechanisms to us. Predictions can't be the benchmark, we are simply not advanced enough in the science for that. Certainly the social sciences, like economics, should understand that best. Economics has hardly been able to predict anything.

smallnamespace 9y ago

At some point, there are limits to our ability to empirically test large-scale predictions because we have no ability to run history more than once. This certainly applies to macroeconomics -- how does one empirically 'test' any prediction from a macroeconomic model based on real-life inputs, other than to see what happens? But this shouldn't preclude us from asking the really difficult questions that we care about, rather than restricting ourselves to the realm of physics, where we can get 5 sigmas in classical statistical tests. With regards to Robert Nobel's criticism, the whole point of the Club of Rome's approach is to model a simplified, 'low resolution' view of the world economy in order to embed it in a larger network of information, goods, and energy flows constrained by the environment -- because the boundary conditions matter a lot to how the macroeconomy functions. Sometimes you can get away with a much simpler modeling approach if you actually get all the inputs right. The biggest criticism the Club of Rome has on most macroeconomic models is that it doesn't model the environmental inputs in sufficient detail, and these things really matter a lot in the long run. It is not simply sufficient to take current 'wage levels' or 'capital stock' or even 'level of technology' as a direct input to your DSGE RBC model -- our macroeconomies perform a quasi-ecological function of converting flows of matter and energy, largely taken from Nature and the environment, and converting them further for human use. Understanding the Nature-human interface, then explicitly modeling the stock quantities of the inputs into the macroeconomy must on some level be important, because they are key to how our entire technological society functions. I you're inclined to learn more about their modeling approach at a high level, I highly recommend Thinking in Systems [1]. Read it a week ago and it blew my mind, as its intuitions match up with a lot of patterns that we see over and over again in dealing with any complex network, but presented in a systematic and convincing manner. [1] http://wtf.tw/ref/meadows.pdf http://wtf.tw/ref/meadows.pdf