NetLogo banner

 Contact Us

 Modeling Commons

 User Manuals:


NetLogo User Community Models

(back to the NetLogo User Community Models)

Extinction of Homo Economicus

by Rolf Stenholm (Submitted: 04/24/2014)

[screen shot]

Download Extinction of Homo Economicus
If clicking does not initiate a download, try right clicking or control clicking and choosing "Save" or "Download".

(You can also run this model in your browser, but we don't recommend it; details here.)

A thought experiment to test Homo Economicus ability to survive natural selection.

In most of Economic Theory the use of the representative agent "Homo Economicus" is used to model approximate human behaviour. A decade ago the limitation of this model was well forgotten. Today 2014 after the 2008 financial crisis it has become fashionable to waffle on about the unrealistic assumptions of the representative agent "Homo Economicus". Waffling on about how ricardian equivalence is sheer nonsense and how humans don't play dice when making assumptions. It is at least more constructive to quantify the scale and direction of the shortcomings of economic theory. For instance it is easy to understand that almost all boats need the ability to float in water and therefore logical develop a theoretical model on boats that float in water. Some economists describe this method of testing as passing a model through real world filters.

In the following model the "boat" model is the representative agent and the water is the biological theory of survival of the fittest i.e natual selection. In this model we test to see if "Homo Economicus" has any chance of survival. The economic model assumes rational individuals which may or may not favor survival of the species.

Two kinds of "extinction" events are implemented in this model, the first is a vulcanic eruption event and the second extinction event is famine. Rationaly bounded agents are assumed to not have the ability to predict any of these event which implies that agents have to adopt good survival strategies under uncertainty.
This model incorporates various standard economic measures to relate model actions with economic theory.

To test this theory the model allows the mixing the rational maximising behaviour with a random settlement behaviour. When agents use the random settlement behaviour agents will settle on a randomly selected spot on the map. These agents that use a random settlement behaviour we can could call these agents the counter example of "Homo Economicus" i.e "Homo Quixotics" naming these "random" agents after the famous Don Quixote who lived his life partly in a fantasy world.

The model therefore allows agents to follow either the principals of "Homo Economicus" or those of "Homo Quixotics" or a blend of either. That is, the model implements an implementation of Khanemans two system approach where a human can either use rationality to solve a problem or use intuition to solve the same problem. In this model we have substituted intuition with a simple random function but since survival chances should increase when agents are scattered accross patches this is an adequate simplification in this model.


Agents in this model are assumed to be self sufficent farmers who spend their day farming a single patch. Green patches can produce enough food to keep agents healthy while yellow patches cannot. The amount of food that can be grown on each patch can be changed by exogenous events controlled by the model.

A single patch can only support a single agent. When multiple agents occupy the same patch neither agent will harvest any food from the patch. Agents will try to relocate to other patches to resolve disputed patches.

There is no capital in this model which means that there are no investments in the model.

Agents use a maximizing tree search like method with a point system to determine best target patch. The point system can be altered in the model to allow agents to act as maxmizers and search for an optimal settle patch or settle at a predetermined patch, or any blend of the two.

Agents have offspring after enough time if the amount of food is sufficent. Each agent can reproduce by itself by parthenogenesis. This is not a very realistic model of the human species, however in this model it is adequate and prefered as it makes the study of agent behaviour easier.


* minimum-fertility-level
determine minimum health level for agents to have children
* individualstic,
defines to what degree agents are maximizer or prefer settle into random locations
* lava-power
defines how far lava travels from the vulcano from the center of the world
* inital-popuation
the number of starting agents
* pregnancy-time
the number of ticks needed to for a pregnant agent to have a child
* minimum-food-level
the lowest amount of food needed to keep an agent healthy
* famine-length
define how long a famine period lasts
* innovation-effect
the percentage increase of food that can be grown on every patch
* class-society
set to true if society uses class society as arbitration for disputed patches.


Note that if the model runs long enough the agents will run out of patches that can produce enough food. When overpopulation occurs the civilisation stagnates but does not behave as outlined by Malthus. Children and the young in this model will suffer food shortage and be forced to live out a meager existence on inadequate food producing patches while the older agents will not experience a decline in living standards. Overpopulation effects are strongly affected by the class-society setting in the model.

Using a class society has a large positive effect on most economic indicators. In popular news media class societies and inequality is considered bad but in this model it eliminates much of the cost associated with disputes of overcrowded patches.

Observe that the "Homo Quixotics" agents have a better chance to survive exctinction events than "Homo Economicus".


Try experimenting with creating different events and setting in the model see how the population of "Homo Economicus" compare to a more intuitive population of "Homo Quixotics".


* Hotellingers law is based on assumptions that people are rational utility seekers. Try incorporating hotellingers law in the model.

* Consider introducing capital money and banks in the model to model the "liquidity trap".

* The agents in the model have no forward looking ability. Consider improving agent logic to have deep tree searches to determine best agent behaviour.

* Experiment with taxation and other egalitarian tools to improve the economy.


How bad is a bad model? to answer that we outline below how to refit "Homo Quixotics" into economic theory. The point with this exercise is to illustrate how much of economic theory actually fails when the assumption of rational agents is eliminated.

The alternative model "Homo Quixotics" implemented in this model implies that each agent must consume a minimum amount of goods to stay alive and each agent spend additional income on an exogenous determined set of goods. In other words a reasonable random consumption function. Below we discuss the idea in reference of actual goods in a model economy.

The model assumptions imply that each agent has a consumption vector c which has to start at point g and increases with an increase to agent nominal income y to buy additional goods using price function q. That is to say personal or agent consumption function is
c = g + y * q(p).
On aggregate level using the fact that g is approximately the same for all agents (or person) we get instead that total consumption for an average increase Y that
C = G + Y * sum of ( q(p))
which could be rewritten into a personal consumption function
C = G + YQ(P)* sum(n) where n is n = y(p*) / Y and p* is an exisiting fixed price level.

Below are a couple of examples of classic economic theory evalutated through the lens of "Homo Quixotics"

Marginal Propensity to Consume
It would not be difficult to show that the above theory implies support for marginal propensity to consume as the aggregate consumption function can easily be used to calculate the marginal propensity to consume, Which is not surprising considering that "Homo Quixotics" models a person or agent basing decision on intuition rather than reason or what Kahneman defines as system 1. System 1 is a cheap method of decision making that can produce decisions close to perfectly rational agents with perfect information.

Ricardian Equivalence
It would not be difficult to show that the above theory implies no support for the ricardian equivalence, i.e consumption is invariant to the timing of taxes. The individual consumption model assumes that agents are not forward looking and therefore ricardian equivalence cannot be true (in this model).

Hotellingers law
Will prove false when using this alternative agent as in this case any reasonable application will show that shopping habits of people will at best be based on muddy past shopping experiences. That simply implies that stores are located in adjacent of each other because agents do not generally search for new stores.

Liquidity Trap
Some theory of the "liquidity trap" can be expected to exist as the agent population may decrease and a theoretical interest rate of patches in the model could become negative. However the appearance, effects of this "liquidity trap" can be expected to be radically different from those outlined by Krugman.


* Rand, W. and Wilensky, U. (2007). NetLogo El Farol model. Center for Connected Learning and Computer-Based Modeling, Northwestern University, Evanston, IL.

* Hotelling, Harold. (1929). "Stability in Competition." The Economic Journal 39.153: 41 -57. (Stable URL: ).

* Kahneman, D. (2003). Maps of bounded rationality: Psychology for behavioral economics. American Economic Review, 93 , 1449-1475 (

* Lanir Z., & Kahneman, D. (2006). An experiment in decision analysis in Israel in 1975. Studies in Intelligence, 50(4) , 11-19. (

* Herbert A. Simon, "Rationality as Process and as Product of Thought", Richard T. Ely Lecture, American Economic Review 68(2), May 1978, 1-12. (

* ABC Research Report, Bounded Rationality, Max Planck Institute for Human Development. (

* An Essay on the Principle of Population (1798), Thomas Robert Malthus

* Computer Chess: Algorithms and Heuristics for a Deep Look into the Future (1997), Rainer Feldmann (

* Paul R. Krugman (1998). Its' Baaack: Japan's Slump and the Return of the Liquidity Trap.

* F. A. Hayek 1945). "The Use of Knowledge in Society," American Economic Review, 35(4), pp. 519-530 (


Copyright 2014 Rolf Stenholm

![CC BY-NC-SA 3.0](

This work is licensed under the Creative Commons Attribution-NonCommercial-ShareAlike 3.0 License. To view a copy of this license, visit or send a letter to Creative Commons, 559 Nathan Abbott Way, Stanford, California 94305, USA.

(back to the NetLogo User Community Models)