Gerhard Tintner on Demand Analysis: A Study in Econometrics (1953)

Gerhard Tintner’s review of Herman Wold’s Demand Analysis, written in association with Lars Juréen, presents the book as a major contribution to econometrics whose importance rests on the conjunction of economic theory, statistical technique, and empirical prediction. Tintner’s assessment is emphatically favorable, but it identifies specific grounds for that judgment: an ordinal theory of consumer choice, regression methods informed by time-series analysis, and Swedish consumption estimates that yield unusually accurate forecasts.

This is without doubt one of the most important contributions to the young science of econometrics ever published.

The review first maps the book’s five-part architecture. Part I surveys its questions, methods, and findings in accessible terms. Part II develops the theoretical assumptions of demand analysis, restating and elaborating Wold’s articles of 1943–44. Part III treats stationary random processes, drawing on Wold’s earlier thesis and P. Whittle’s work on hypothesis testing in time-series analysis. Part IV addresses regression theory and methods. Part V, attributed to Wold and Juréen together, presents results from Swedish family-budget surveys of 1913, 1923, and 1933 and market statistics for 1921–1939. This progression gives the book both a technical foundation and an empirical destination.

Tintner distinguishes the expertise required to appreciate its original contributions from the accessibility of its introductory and empirical sections. Full engagement requires substantial mathematics, statistics, and mathematical economics; Parts I and V nevertheless offer economists and specialists in agriculture, marketing, and consumption a comparatively direct route to the findings. His praise thus concerns both research achievement and the book’s capacity to communicate beyond a narrowly mathematical readership.

The conceptual foundation is an updated Pareto theory of consumer choice. Tintner singles out the distinction between ordinal and cardinal utility:

The author makes an excellent case for ordinal utility functions (indifference systems) and rejects cardinal utility.

Here demand analysis rests on the ordering of preferences through indifference systems rather than on cardinal utility measurement. Tintner pairs this theoretical commitment with Wold’s statistical choice: classical least squares is retained, while newer, more sophisticated methods are rejected. The review reports that choice approvingly without explaining the arguments against the alternatives. It does specify that time-series corrections modify the standard errors, so the statistical framework combines a familiar estimation method with adjustments informed by the temporal character of the data.

The empirical contribution consists chiefly of income and price elasticity estimates for selected foods in Sweden, differentiated by social class and family composition, including children’s number and age. Tintner places particular weight on forecasting:

One particular successful feature of the book is a prediction of per capita consumption of various food items for 1949-50. This prediction is surprisingly accurate and shows the stability of the consumption pattern in Sweden.

Predictive success supplies the review’s strongest practical evidence for the analysis. Tintner contrasts it favorably with other econometricians’ forecasting efforts, although he provides neither comparative figures nor named examples. He also notes forecasts for 1960 and 1970, without claiming that these had been verified.

The review closes by praising the extensive international bibliography and recommending the book for a somewhat advanced course in demand analysis. Its governing judgment is that theoretical coherence, statistical competence, and successful empirical prediction together make Wold and Juréen’s study essential reading for those concerned with econometrics’ development.