Economic models may be seen as intermediate between theories and the world, and are accordingly differently understood depending on how one sees their relationships to each. Models are specifically economic models when they address forms of social interaction involving prices and other money magnitudes. Daniel Hausman (1992) takes models to be either trivially true or neither true nor false, and thus to constitute conceptual explorations of theories. His emphasis is on theoretical models, not on empirical models, and thus on how models enable us to investigate theoretical concepts. In contrast, Ronald Giere (1988) sees science as a cognitive process, and is concerned with how models help us learn about the world. He sees the relationship between models and the world as one of similarity (though not of isomorphism). This allows models to be both somehow “true” of the world and yet at the same time not be expected to fully reflect the structure of the world.
Models may be defined, then, as instruments of investigation that allow us to concentrate on relationships of special interest, whether in theory or in the world or in combination of the two (see Morgan and Morrison 1999). That they focus attention on aspects of the world implies that they should not be taken as complete explanations, but as particular ways of examining those aspects. Within this framework, several different approaches to Giere’s similarity idea have been suggested. Models have been taken to be idealizations or approximations that provide “realistic” though partial representations of the world. Models can also be seen to be like analogies and metaphors in that they exhibit similarities between different domains or systems. Another view is that models are like stories told to explain how theories may be interpreted or how theories relate to the world.
There are a number of misconceptions about the nature of economic models. While economic models are often seen to be sets of highly formal statements, there is nothing in the concept of an economic model that requires formal representation. Indeed, whether formal models or qualitative ones are preferred depends on the subject matter being modeled. Another widespread belief is that models are essentially instruments for making predictions about the future. But models have many different uses. There are theoretical models, simulation models, policy models, econometric models, measurement models, experiment models, formal models, accounting models, and so forth, all of which are constructed for different, often highly specific, purposes. Further, even good models designed expressly for making predictions can fail to produce reliable predictions in changing environments. Thus, because economic models have both limited and usually quite specific purposes, it is important to avoid uncritical use of the information they generate. Models are instruments for all kinds of investigation, and our investigations of the world are fallible, partial, and subject to revision.
SEE ALSO Macrofoundations; Maximization; Microfoundations; Minimization; Models and Modeling; Multisector Models; Optimizing Behavior; Two-Sector Models
Giere, Ronald. 1988. Explaining Science: A Cognitive Approach. Chicago: University of Chicago Press.
Hausman, Daniel. 1992. The Inexact and Separate Science of Economics. Cambridge, U.K.: Cambridge University Press.
Morgan, Mary, and Margaret Morrison. 1999. Models as Mediators: Perspectives on Natural and Social Science. Cambridge, U.K.: Cambridge University Press.
John B. Davis