Academic journal article Review of Social Economy

Realism, Regularities, and Prediction

Academic journal article Review of Social Economy

Realism, Regularities, and Prediction

Article excerpt

Tony Lawson's investigations into "economics and reality" represent a transfer and extension into the domain of economics of a set of ideas first developed by Roy Bhaskar (1978) in relation to the natural sciences. This essay starts out from the torsions that arise in the course of the transfer. While Bhaskar's scientific realism offers a plausible (though not uncontroversial) account of the natural sciences, it is readily apparent that his arguments do not carry over directly into the social realm. Lawson (and Bhaskar himself, when he writes of the social) are obliged to modify these arguments substantially. The modified version is, I shall claim, much more problematic than the original. This does not mean that I reject Lawson's substantive conclusions regarding economics; I can agree with many of these. I do want to raise some problems with his way of arguing for those conclusions, and to suggest that his mode of argument leads him to underestimate the importance of both empirical regularities and prediction in economics.

I begin with a brief synopsis of Bhaskar's "realist theory of science" as a basis for comparison. In the most general terms, Bhaskar's realism is essentially the same as materialism in the classic Marxian mold (as in Engels's Anti-Duhring, 1954). In each case, the objects of scientific investigation are taken to be real things or structures that exist independently of human minds (or at any rate, in the case of social science, independently of their scientific concepts). Such objects are endowed with certain causal powers or natural tendencies, and it is the job of science to bring these to light, to form their concepts. Scientific progress is conceived as a progressive deepening of our understanding of the structures and tendencies of this independently existing world.

One of the more distinctive aspects of Bhaskar's realism is his "layered ontology." He distinguishes three nested levels of being: the real, the actual, and the empirical. The domain of the empirical comprises experiences; the domain of the actual comprises both the empirical and those events, in principle observable, that do not, however, feature in anyone's experience; and the domain of the real comprises both the actual and the underlying mechanisms that are responsible (in combination) for the generation of observable events.

Bhaskar's critique of empiricism - upon which Tony Lawson draws heavily in his critique of standard economic theory and econometrics - rests on this distinction of levels of being in conjunction with the distinction between open and closed systems. A closed system, in Bhaskar's terms (not to be confused with those of systems theory - cf. Bertalanffy, 1968), is simply a system characterized by "constant conjunction of events." He identifies three conditions which a system must satisfy in order to achieve such closure (Bhaskar 1978: 69ff). The system must be fenced off from outside influences, or at least any outside influences must remain constant in their effects (extrinsic condition); the individuals composing the system must either be atomic (lacking any internal structure or conditions) or at least their internal conditions must remain unchanged over the period in question (intrinsic condition); and the overall states of the system must be capable of representation as an additive function of the states of individual components of the system, or at least any non-additive principle of organization must remain constant (additivity condition).

These are Bhaskar's concepts. How do they fit together? A key component of his theory as a whole is his transcendental argument from the success of the practice of scientific experimentation. Bhaskar takes it as clear that there are in fact rather few "constant conjunctions of events" in the natural world (outside of astronomy). That is, the world as a whole is an open system. Such constant conjunctions as we see are typically engineered in scientific laboratories, the sites of deliberately produced artificial closures. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.