If you think I'm advocating models you have read me wrong. The point is I'm advocating _against_ models. I'm advocating for simulations. Simulations are a very simple concept, and I talked about it here before (probably too often). You don't make up a model, you develop a simulation based on individual agents or elements, and the simulation drives the behaviour of the agents. It may require a minimum of 100k elements to begin to get an accurate simulation. You know you've got a working simulation when it begins to reflect the actual behaviour of the world. It's worth using when you're able to provide the basic set of inputs and see it reproduce historical behaviour, say the entire 20th cntury, with some level of fidelity. Then you run it on the current situation and see how well it predicts future behaviour.
When you've got a reasonable set of results, you can start playing around with it, pushing and tweaking things, and seeing what happens. After having done that for a while, you can begin to think about distilling a model to describe what the simulation is telling you about how things actually work. The model will be emergent from the simulation, it doesn't drive the simulation. Only existing real world data drives the simulation, and real world data determines when the simulation is built correctly. -Pete On Sun, 8 Sep 2013, Steve Kurtz wrote: > Pete, > > I humbly suggest that most economists do use models (of their own > design of course) and that has been part of the circular 'proofs' > fallacy of the field. I posted the abstract and .pdf of a debunking > paper on this yesterday. non-equilibrium economics deals with unknown > future wild cards which cannot always be imagined in model design. > Plus, as Korzybski said: "the map is not the territory." > > Steve > On Sep 7, 2013, at 7:47 PM, pete wrote: > > > > > > > On Sat, 7 Sep 2013, Ray Harrell wrote: > > > >> > >> SK: My comments re Crudman relate to neoclassical economics BS > >> > >> > >> > >> I guess Steve that he could be talking about all of us:>)) REH > >> > >> > >> > >> > >> Neo Fights (Slightly Wonkish and Vague) by Paul Krugman Blog nytimes: > >> August 28, 2012, 9:23 am > >> > >> > >> Via <http://economistsview.typepad.com/> Mark Thoma, David Glasner gets > >> upset over claims by Austrians to have rejected or superseded neoclassical > >> economics. > >> <http://uneasymoney.com/2012/08/27/hayek-was-a-neoclassical-yes-neoclassical > >> -economist/> Hayek was a neoclassical economist,Glasner declares. > > > > [...] > > > >> What would truly non-neoclassical economics look like? It would involve > >> rejecting both the simplification of maximizing behavior, going for full > >> behavioral, and rejecting the simplification of equilibrium, going for a > >> dynamic story with no end state. > >> > >> And there is economics like this: > >> <http://www2.econ.iastate.edu/tesfatsi/ace.htm> agent-based economics. It's > >> a project that relies heavily on computing, to keep track of the > >> complexities, and at this point makes simplifying assumptions that are in > >> their own way as unrealistic - but in a different direction! - as those of > >> neoclassical work. Still, it's a good thing to pursue. > > > > Wow, Ray, thanks for that link. It looks like (some small number of > > heretical) economists are finally beginning to look at using real > > simulation software to actually understand economies instead of > > using computers to add pointless decorations to their groundless > > pronouncements. > > > > I like this title which comes up from browsing a ways into the maze of > > links which branch out of that one: > > > > Aki Lehtinen and Jaakko Kuorikoski, "Computing the Perfect Model: Why Do > > Economists Shun Simulation?" (pdf,115KB), Philosophy of Science 74 (July > > 2007) pp. 304.329. > > > > Abstract: "Like other mathematically intensive sciences, economics > > is becoming increasingly computerized. Despite the extent of the > > computation, however, there is very little true simulation. Simple > > computation is a form of theory articulation, whereas true simulation is > > analogous to an experimental procedure. Successful computation is > > faithful to an underlying mathematical model, whereas successful > > simulation directly mimics a process or a system. The computer is seen > > as a legitimate tool in economics only when traditional analytical > > solutions cannot be derived, i.e., only as a purely computational aid. > > We argue that true simulation is seldom practiced because it does not > > fit the conception of understanding inherent in mainstream economics. > > According to this conception, understanding is constituted by analytical > > derivation from a set of fundamental economic axioms. We articulate this > > conception using the concept of economists' perfect model. Since the > > deductive links between the assumptions and the consequences are not > > transparent in .bottom-up' generative microsimulations, microsimulations > > cannot correspond to the perfect model and economists do not therefore > > consider them viable candidates for generating theories that enhance > > economic understanding." > > > > Kinda sounds like my perennial complaint, only a decade after I first > > voiced it here. I believe they're saying it appears economists don't > > like to use simulations because it limits them to speaking about > > actual results, rather than making stuff up. > > > > -Pete > > > > > > _______________________________________________ > > Futurework mailing list > > Futurework@lists.uwaterloo.ca > > https://lists.uwaterloo.ca/mailman/listinfo/futurework > > > _______________________________________________ > Futurework mailing list > Futurework@lists.uwaterloo.ca > https://lists.uwaterloo.ca/mailman/listinfo/futurework > > _______________________________________________ Futurework mailing list Futurework@lists.uwaterloo.ca https://lists.uwaterloo.ca/mailman/listinfo/futurework
