Creator of the new and innovative Open Source economic software Minsky, Professor Stephen Keen explains how his program simulates economic conditions for predictive study.
The economic crisis of 2008 has made the public aware that all is not necessarily good with the state of economics – how could economists not foresee such a huge, global event, when they’re supposed to be the experts on the economy?
One answer that normally elicits a “surely you’re joking!” response from the public, is that the vast majority of economic models pretend that banks, debt and even money don’t exist. They also fantasise that the economy is a stable system which will always return to a state of bliss – known as equilibrium – after suffering an unanticipated shock.
These absurd oversights are a product of the 19th century approach to modelling which economics has clung to, despite advances in 20th non-equilibrium modelling – in particular, system dynamics. This is a means to model a complex system using a flowchart metaphor, in which a causal chain is firstly modelled and then numerically simulated by the flowchart. This technology was invented by the MIT management guru Jay Forrester in the 1950s, and now dominates engineering (with programs like Simulink and Vissim) and has many users in management and sociology as well (using programs like Vensim and Stella).
Minsky attempts to introduce this modern approach to modelling into economics by providing a tool tailored for economic issues. Minsky supports the standard “flowchart” paradigm of system dynamics – which you can see in Figure 1 – and it also adds the ability to model monetary flows using interlocking double-entry bookkeeping tables we call “Godley Tables” – as shown in Figure 2.
Minsky is intended to be used by economists to model the economy as a monetary, non-equilibrium system, and by policy makers to test out policies on a virtual economy before they are implemented in the real world. Simulation parameters – which are akin to policy settings – can be varied dynamically while a model is run, and the results of these changes can be seen on the program’s dashboard. This capability is offered by only a few of its rivals in the system dynamics product space, and is one of several features that sets Minsky apart from the competition – even though it is a very new program, with only a few thousand hours of development time behind it.
Other novel features include the ability to type directly onto the program’s “canvas” to develop models, rather than having to “point and click” on a palette, and support for sophisticated text formatting, including subscripts, superscripts and even Greek characters.
As a very new program, the current implementation only touches the surface of what we hope to achieve with Minsky. Our next major ambition is to be able to go from the abstraction of a “single good” economy to modelling a multi-product economy with a single “multisector” command, and to go from modelling a single economy to the global economy with international flows of goods and services and floating exchange rates, with a single “clone” command.
Minsky is Open Source, and if you’d like to try Minsky out for yourself, you can download it from SourceForge at https://sourceforge.net/projects/minsky/. You can also download sample Minsky files from http://www.debtdeflation.com/blogs/2014/01/15/some-sample-minsky-models/. Demonstration videos of using Minsky are available at http://www.debtdeflation.com/blogs/minsky/ and also on my YouTube channel https://www.youtube.com/channel/UCM1ubsbE-tG9ru61mc3zX8A.