simardcasanova’s avatarsimardcasanova’s Twitter Archive—№ 19,030

  1. …in reply to @undefined
    @MartinSGaynor It’s possible but to me, ABMs are mostly a theoretical tool. They thrive with systems with many degrees of freedom, whereas regular equation-based models quickly become untractable in such context.
    1. …in reply to @simardcasanova
      @MartinSGaynor The best way to understand ABMs is as computational Petri dishes, where you can simulate how an economy would react to policy x/y/z, and so on We need empirical data to calibrate those models, so I’d say ABMs are more complements that substitutes to empirical econ