Chaos and Complexity Economics (with J. Doyne Farmer) 8/26/24

Physicist J. Doyne Farmer wants a new kind of economics that takes account of what we've learned from chaos theory and that builds more accurate models of how humans actually behave. Listen as he makes the case for complexity economics with EconTalk's Russ Roberts. Farmer argues that complexity economics makes better predictions than standard economic theory and does a better job dealing with the biggest problems in today’s society.
For links and more information:
Subscribe to EconTalk on UA-cam: / @econtalkwithruss
Subscribe to the audio episodes:
Apple Podcasts: podcasts.apple...
Stitcher: www.stitcher.c...
Spotify: open.spotify.c...
and wherever you listen to podcasts.

[@robertfarquharson2878]

Has Doyne Farmer catalogued his own inventory of books or is he just pilfering library stock?? 😂

[@wankim1]

I’m either not smart enough or Farmer isn’t explaining it well, but I’m not seeing how his “simulation” is any better than the plethora of economic models out there. 🤷🏻‍♂️

[@alexcipriani6003]

Modeling is relatively static outside of the variable the model considers, imagine a mathematical function where the model simply plays with those variable values. Models follow the divide and conquer engineering paradigm; you disassemble a problem into constituent components you solve each one of them and put the problem back together, this doesn’t account for the dynamic interaction of these components. You usually hear the sum of the parts is greater than the whole. Simulation are more dynamic in their nature and factor in complexities beyond the variables and boundaries of a model.

[@robertfarquharson2878]

They never really do explain how agent-based models work but the basic idea is something like this:
Imagine you’re correlating *speed of car crash* with *probability of dying in the crash*. The classical model just takes the values of those two variables and plots a line of best fit. You can find a relationship between the two and make pretty good predictions based on it.
The “simulations” he’s talking about are like actually putting 3D model cars inside a virtual environment and crashing them into each other and seeing which ones cause the little simulated people inside to die. That is, you’re modelling *the events* that actually cause the values for the two variables you were interested in before.

[@herbwiseman9084]

The variables that Farmer discussed regarding housing ignores the impact of accounting - imposed liabilities, forced transfer of assets and the impact on the relative equity of those setting the prices and those receiving the extra assets. Steve Keen has developed two software programmes that portray the systems features of complexity and dynamic economics. Part of the problem is the misplaced faith in supply and demand dynamics setting prices.