G. L. S. Shackle, Econometrics (1954)

George Lennox Sharman Shackle’s review of Jan Tinbergen’s Econometrics uses an accessible introduction to quantitative economics to examine the discipline’s ambitions and the organisation of economic research. Its central distinction is between statistical work that clarifies economic experience and the stronger promise of an empirically verified, quantitatively exact science of prediction. Shackle values the former while questioning whether the latter justifies making advanced econometrics the dominant model of postgraduate education. The review moves from the movement’s founding aspirations, through their implications for British universities, to an appreciative account of Tinbergen’s book and its usefulness for economists outside the technical specialism.

Shackle opens with an expedition metaphor: econometric pioneers entered a jungle in search of an “Eldorado” of exact predictive economics, but the reports reaching civilisation suggest diminished expectations. He defines their undertaking as translating economic theory into algebra and statistically estimating the resulting equations’ parameters. Its ambitious promise was to predict both the course of an economy left to itself and the precise consequences of policy interventions. The issue is not simply whether a particular estimation technique succeeds, but whether human affairs admit the kind of predictive mastery associated with astronomy and engineering. Shackle presents that fundamental objection as a view some would advance, then states his own more immediate assessment:

It seems at any rate to be true that even the most complex models which the econometricians have yet constructed have had questionable predictive value and that in one or two dramatic instances econometric prophecies have seriously misled the policy makers who trusted them.

The review supplies no specific instances of those failures. The claim establishes the stakes of Shackle’s argument rather than demonstrating a general impossibility of prediction. If elaborate models remain unreliable, their institutional cost and educational consequences require scrutiny. Should universities invest heavily in econometric centres and recruit research students chiefly for mathematical aptitude, or preserve economics as a philosophically informed, historically illuminated pursuit? His contrast between teams in laboratories and individuals in the study concerns what forms of inquiry should receive support and what kinds of intellectual ability economics should welcome.

Crucially, this defence of qualitative inquiry does not entail indifference to numerical evidence:

To put this question into words is not in the least to be guilty of such folly as to doubt the vital importance of assembling numerical descriptions of the past course of economic history and of the current situation.

Shackle identifies national income accounting, Leontief’s description of the interlocking structure of production, demand studies, and production indexes as indispensable contributions. Their achievement is to make the economic picture precise and detailed, furnishing the economist with something whose meaning must still be explained. He also acknowledges that this work requires equipped centres and collaboration among mathematicians, statisticians, and economists. The distinction therefore runs between supporting specialised quantitative research and treating it as the only worthwhile advanced economics. Establishing such a centre in every British university would, he argues, duplicate effort wastefully and exclude gifted students whose strengths might serve academic, administrative, or business life. His qualified defence of university conservatism rests on preserving intellectual opportunities, not merely resisting technical innovation.

Turning explicitly to Tinbergen, Shackle acknowledges that a book review may seem an ungracious occasion for these policy reflections. Yet he finds the book’s accessibility consistent with his own concern that econometric knowledge should benefit economists who will never become specialist practitioners:

This is not a manual of detailed procedure for the expert practitioner of econometrics, but on the contrary, a book that every student of economics could profitably read.

His account of the book explains that judgment. Tinbergen begins by showing how algebraic formulation improves clarity, precision, and insight; Shackle singles out its explanation of why demand conditions enter the supply equation under imperfect competition. An approachable methodological survey follows, covering time-series methods, multiple correlation, bunch-map procedure, Haavelmo’s approach, multicollinearity, and identification. Shackle praises the author’s grasp of logical essentials: the treatment is introductory, but its brevity reflects mastery rather than superficiality.

The first eighty-eight pages establish these theoretical and statistical foundations. A further seventy survey economic applications and illustrative results. The main text ends with a more demanding discussion of a Frisch “decision model” for choosing policies concerning unemployment and the balance of payments, while an appendix reprints journal articles. This progression gives readers both an understanding of econometric reasoning and a view of its practical uses, without requiring specialist competence throughout.

The review’s conclusion reconciles appreciation of Tinbergen with scepticism about the field’s maximal aspirations. Most people who study or use economics, Shackle argues, will continue to work qualitatively; they nevertheless need to understand what econometricians do. Tinbergen’s achievement is to make that understanding available at a moment when reassessment is especially necessary:

This is a good time for such a popularly accessible survey and guide to be published, for if a fully metric economics has begun to seem a shimmering mirage, it is all the more necessary for economists of both persuasions to take stock and consider where they shall go from here.

The closing “mirage” returns to the opening Eldorado, but the practical conclusion is shared reflection rather than withdrawal from quantitative inquiry. Shackle’s review makes Tinbergen’s book relevant as a bridge between technical specialists and the wider profession, while defending an economics whose evidential resources and educational pathways remain broader than exact prediction.