A Micro-Macro Interactive Simulation Model of the Swedish Economy

Abstract

Work on this micro based macro model of the Swedish Economy began in late 1974 as a joint project between the Federation of Swedish Industries, IBM Sweden and the University of Uppsala. This is an edited version of several instruction manuals I wrote for the coding and later programming of that model. It was published in December 1976 as "A micro - macro Interactive Simulation Model of the Swedish Economy" by the Federation and IBM in two parallel editions. A technical pseudo code to facilitate the final coding of the model in APL, authored jointly by myself, Mats Heiman and Gösta Olavi, was appended to that publication. This republication of the 1976 document, has been slightly trimmed, and lack the Foreword by the President of the Federation, Author’s Remarks, and the Pseudocode. The Pseudocode is however available from the Journal online as (Eliasson et al., 1976). It details the specifications and highlights the modular design of the model and should be helpful for anybody who might want to reproduce a version of the model.

Originally this model, called MOSES, for Model Of the Swedish Economic System, was designed to facilitate understanding and quantifying the role of markets and production firms in economy wide economic development. Of particular interest were the consequences of markets being disturbed by inflation. The modelling approach was bottom up, drawing extensively on empirical research in business economics, not least on my own study on business economic planning practices (Eliasson, 1976a). The model has a general design in that it integrates a standard Keynesian demand driven macro model with a Leontief type input output sector structure. The manufacturing subindustries, furthermore, are populated with boundedly rational (“ignorant”) firms (Simon, 1955; Simon 1959). These firms make up the core market governed supply machinery of the model, and integrate product, labor, and financial market considerations within their business plans. Firms compete in the markets of the model. Their plans and decisions are guided by Stockholm School ex ante price expectations and learning from feed backs of ex post outcome experiences. Schumpeterian entrepreneurs enter markets unexpectedly and compel incumbents to perform. Market competition thus self-coordinates the entire model economy under an upper technology constraint embodied in new investments of firms. The government may intervene in markets, but there are no artificial external equilibrium constraints imposed. I had no ambition to build a micro foundation of existing macro theory, neither has the model been designed to provide forecasts to support policy. The ambition has been to understand what government can do to the economy as one of many monopoly actors with boundedly rational insights.

The ambition was empirical, to begin with to reproduce and understand the long term and cyclical market dynamics of a Sweden like industrial economy. To that end a separate annual Planning Survey to Swedish manufacturing firms, tailored to the needs of the model, was started at the Federation. That survey will eventually provide statistics also to test the model’s capacity to generate distributional characteristics of the Swedish economy.