The evolution of IS-LM models: empirical evidence and theoretical presuppositions

This paper offers an explanation of the resilience of IS-LM models, which are still alive more than sixty years after the birth of the first prototype despite growing criticism within the profession. To this end two significant episodes of their evolution are examined in some detail: the genesis of the prototype of the 'first generation' of IS-LM models (Hicks 1937), and a recent example of what is here called the 'second generation' of IS-LM models (McCallum and Nelson 1997). The resilience of IS-LM models is mainly related to their adaptability to a changing perception of the economic environment within which they are applied. This depends on the evolution of the relevant stylized facts and of the salient policy problems as well as on the evolution of the prevailing theoretical assumptions and methodological approaches. The stock of cumulated empirical evidence and the flow of new additions to it have both played a role in this process, though they have done so mainly indirectly by affecting the perception of the economic environment.