Modelling the Process of Market Emergence

1 What do we want to model?

The CEE countries developing new market institutions are often characterised as transition economies. The transition is from central planning to some form of market-based exchange arrangements. The exact form is not yet clear ó especially for those countries which are not yet very far along the transitional path. Certainly, an interesting issue is to describe in some useful way the main characteristics of the market arrangements and institutions once they have emerged. It is, however, difficult to see how we can describe the outcome from a process (i.e. the transition) which we have not previously encountered and of which we do not have a clear, shared understanding of the important dynamic forces.

In this paper, we argue that it is more likely to be fruitful if we start from a description of the present position and try to ascertain how relevant agents are in fact reacting to current circumstances in order to develop an understanding of the process we observe. If we model the environment of the emerging market economies and the observed kinds of reactions, this will give us some basis to consider the effects of economic and social policies which are intended to change those reactions in various ways.

Conventional economic modelling is not actually about processes. The behavioural processes are encapsulated in some kind of constrained optimization algorithm involving the maximisation of utility, profit or occasionally some other variable. The point and purpose of the analysis is to identify the states which result from the processes. How the optimization takes place is not usually considered.

This kind of modelling is imperative in the sense that it tells the agents what to do. Imperative modelling is useful if the questions to be asked are of the type: what will be the principal, persistent characteristics of the economy if agents behave in a particular way? Though this question is interesting, it is not the only interesting question.

For policy makers faced with a given environment, a more important question could be: how will agents respond to a particular policy measure and how will the economic system change as a result of those responses. We are not looking here at the states which will eventually emerge but rather the process of change independently of the end point of the transitional process. Because the responses of agents are typically specified as if-then rules, we call such modelling declarative. That is, given a statement of a state, the rules declare (usually by assertion to a database) the consequences of that state.

Declarative modelling is therefore about the responses of the system to a given state whereas imperative modelling is about the states which result from specified modes of behaviour. If those states are persistent, then imperative models describe the long-term consequences of the specified modes of behaviour. Declarative modelling describes sequences of states which are not persistent. The period of the transition is, by definition, a period during which states are transient. Thus, if we want to model the transition itself, declarative modelling is more appropriate than is imperative modelling.

In this paper, we describe the specification of a declarative model and some results obtained from it. We believe the model captures a number of important features of the Russian economy.

Modelling the Process of Market Emergence - 17 MAY 96

[Next] [Previous] [Top] [Contents]

Generated with CERN WebMaker