Pre on Nostr: Read "Making Sense Of Chaos" by J Doyne Farmer, in which he describes some of the ...
Read "Making Sense Of Chaos" by J Doyne Farmer, in which he describes some of the problems with standard static economic models and his successes with instead using "complexity economics", where he's building giant models of economies out of interactions between software simulated people called agents.
The thing, he reckons, about the mathematical equation driven models is that they are static, and so only really reflect changes when exogenous shocks hit the system. So you can see a change in the model if there's some supply shock caused by a pandemic or whatever, but the economic model won't evolve over time to reflect changes which come from within the economy itself. It always reaches an equilibrium until the next shock.
Whereas his agent-based models are chaotic, and can model the chaotic behavior of a real economy. Dynamic changes within the model can lead to boom and bust cycles, or to market crashes based on nothing but the endogenous behavior of the model without requiring some shock from outside it like a war or a pandemic.
He describes some success stories over his career of models getting increasingly complex as computational power increases, and paints a convincing picture of how better models can help decision makers determine the best courses of action.
Maybe it's because I'm a programmer from the games business, and not really much of a mathematician, but this is the method I'd have used if you asked me to model an economy. Hard to see how it could possibly work with just an equation of demand and supply.
But of course economics as a discipline is older than computers so they couldn't do it that way in the first place.
It would be good to have the field of economics catch up to trying to model the real world instead of just justify their pre-conceived ideas, to validate concepts that allow the rich getting to rule the world and the politicians to do what they were planning anyway.
#reading #books #economics
The thing, he reckons, about the mathematical equation driven models is that they are static, and so only really reflect changes when exogenous shocks hit the system. So you can see a change in the model if there's some supply shock caused by a pandemic or whatever, but the economic model won't evolve over time to reflect changes which come from within the economy itself. It always reaches an equilibrium until the next shock.
Whereas his agent-based models are chaotic, and can model the chaotic behavior of a real economy. Dynamic changes within the model can lead to boom and bust cycles, or to market crashes based on nothing but the endogenous behavior of the model without requiring some shock from outside it like a war or a pandemic.
He describes some success stories over his career of models getting increasingly complex as computational power increases, and paints a convincing picture of how better models can help decision makers determine the best courses of action.
Maybe it's because I'm a programmer from the games business, and not really much of a mathematician, but this is the method I'd have used if you asked me to model an economy. Hard to see how it could possibly work with just an equation of demand and supply.
But of course economics as a discipline is older than computers so they couldn't do it that way in the first place.
It would be good to have the field of economics catch up to trying to model the real world instead of just justify their pre-conceived ideas, to validate concepts that allow the rich getting to rule the world and the politicians to do what they were planning anyway.
#reading #books #economics