Rhetoric of Economics

Deirdre N. McCloskey, The Rhetoric of Economics (2nd ed.). Madison, Wisconsin: University of Wisconsin Press, 1998, 223 pp. $US 17.95 paper.

This is a very combative book. In the right corner is economic modernism - big, bad, positivist and, as a matter of fact, masculine. In some other corner is Deirdre McCloskey who stands for - well, actually, what she stands for is not all that straightforward, as we shall see.

Economic modernism dominates contemporary economics. Its practitioners radically differentiate what they do from literary studies or from softer social sciences. It is behaviourist in the sense that what matters in the final analysis is what people do rather than what they think (p.28). It highly values mathematical formalization, quantification, and statistical tests - particularly tests involving one or another regression procedure. It is positivist in the sense that it treats falsifiability - and the failure to falsify in a test - as the most important gauge of scientific value. It is also, thoroughly unsatisfactory - a form of "voluntary imbecility", in fact (p.99).

To make her case McCloskey shows that the effectiveness of the leading figures in the discipline is enhanced when they stray from the modernist canon. It is not experimental or statistical tests that determine the acceptance of their arguments, it is the rhetoric deployed in making them. In his classic Foundations of Economic Analysis Samuelson sometimes used mathematics not because it is necessary to the argument but to impress. He legitimates his claims through appeals to authority. He uses analogy or metaphor. So does "the Kipling of the economic empire" (p.42), Gary Becker, when he treats children as consumer durables (like refrigerators). Muth's pioneering paper on rational expectations, though badly written, persuaded (after a lag) not because of the evidence it contained - there was little of that - but because of a number of theoretical arguments, including the general case that if reason is assumed in most domains of economics one should not readily abandon it in, in this particular case, agricultural price determination. In identifying and analyzing the rhetorical content in the writings of a number of distinguished economists she is not trying to debunk them. She simply wants to make the case that economic knowledge advances through procedures that diverge wildly from those incorporated in the modernist canon.

She also tries to show that following the precepts of the modernist canon produces inferior work. Quantification is all very well but the resulting quantities do not speak for themselves. An issue that (very properly) exercises economic historians is the extent of the integration of markets in various places, at various times. Integration might be measured by the correlation of prices of, say, wheat in different cities. But how high does the correlation have to be to allow the conclusion that markets are integrated? On its own it tells us more-or-less nothing. Its meaning must be derived from reasoned discussion. McCloskey is arguing that, all too often, quantities are treated as if they speak for themselves, with the accompanying but crucial discussions of importance neglected. The extreme of this is the use of tests of significance where, McCloskey shows us, statistical and substantive significance have been confused, even in the very best journals. Most fundamentally, I think, McCloskey is arguing that economic modernism is intellectually stultifying. No slouch at rhetoric herself, she describes methodology as the "middle manager in a green suit" (p.160), providing neither particularly practical advice (economists usually pay only lip service to its precepts) nor the broader injunction to civilized discourse that underlies, she says (citing Habermas), the very best scholarship. It yields an annual harvest of intellectually mechanical and illiterate young economists like the anonymous young colleague who had the misfortune to sit next to her on a transatlantic flight - and who she cheerfully calls a barbarian (pp.189-190).

For what, then, does McCloskey stand? She is not in favour of the "Santa Monica approach to science ('Hey, man, how do you feel about the law of demand today?')" (p.167). She is not against tests of significance, only their misuse. She is in favour of mathematization, for the precision it brings to thought (p.139). Her concern with the issue of "how large is large?" means that she is for quantification. She is in favour of falsifiable statements. ("Refusing to offer hostages to evidence, though not rare even in modernist circles, is cowardly: so much you can take from the idea of falsification by evidence. Facing facts, we all agree, is good." - p.158). In other words, she is in favour of all the central precepts of economic modernism including the essentials of positivism! She wants, I think, to see more imagination added to this. Intuition, introspection, and simulation are all ways through which imagination can be expressed. They contribute to the debates that are at the heart of science but are ruled out by a strict interpretation of the precepts of economic modernism.

This is a delightful book. It is witty, lucid, and perceptive. I think that, in the final analysis, it is bit less revolutionary than McCloskey seems to think. After all, as she shows at length, undeterred by the tyranny of positivism, the best economists do what she thinks they ought, already. But sociologists should read it with pleasure not least because, for an economist, she is uncharacteristically kind to the discipline!

Michael R. Smith
Department of Sociology
McGill University
SMITH@LEACOCK.LAN.MCGILL.CA