The archive.

A straight-line demand curve makes calculation possible—but what does it leave out? Gerhard Tintner’s textbook trains the prospective econometrician to connect mathematical convenience with economic interpretation and statistical evidence. Economic problems motivate the techniques: income distributions introduce logarithms, marginal costs give derivatives their meaning, and family-income sampling exposes the consequences of selection bias. Represented here by its 1954 second printing, the book joins elementary mathematical instruction to the practical judgment required for estimation and testing. Readers can discover how an economic relationship becomes a calculable model, while learning why an approximation is not an exact description and a significance threshold does not eliminate the possibility of error.

A corrected table can change how farm inputs appear to contribute to production. In this brief notice, Gerhard Tintner and O. H. Brownlee acknowledge computational errors identified by W. O. Jones and replace Table 3 of their 1944 study. Their revised estimates put labor’s marginal productivity per dollar at 1.9555 for hog farms but 0.0361 for dairy farms; the accompanying five-percent fiducial limits distinguish positive estimates from those whose uncertainty spans zero. For readers using the earlier article, this is an essential numerical amendment—not a new model or an explanation of the errors—and a compact source for comparing estimated productivity across farming categories.

Accurate forecasts of Swedish food consumption anchor Gerhard Tintner’s favorable assessment of Herman Wold’s Demand Analysis, written in association with Lars Juréen. In this short review, Tintner weighs the study’s combination of ordinal utility theory, classical least-squares estimation, and time-series corrections against its empirical results. He singles out the predictions for 1949–50 as evidence of stable consumption patterns and contrasts their success with other econometric forecasting efforts. His judgement also distinguishes the mathematical expertise needed to assess the book’s original contributions from the accessibility of its introductory and empirical sections. The review offers a concise view of what Tintner valued in demand analysis: theoretical foundations, statistical methods, and forecasts tested against observed consumption.

How should a theory of choice under uncertainty be judged: by its mathematical consistency, its account of actual conduct, or its rules for rational action? In this 1954 review of the proceedings of the 1952 Paris econometrics colloquium, Gerhard Tintner admires Maurice Allais’s mathematical work without accepting that psychological objections settle the dispute over von Neumann–Morgenstern utility theory. His distinctive concern is methodological: axioms need not be self-evident, descriptions of behavior must be separated from prescriptions, and an idealized model should not be rejected merely because it leaves some motives unexplained. Yet abstraction does not excuse neglect of evidence. The review offers a pointed assessment of where mathematical economics illuminates uncertainty—and where an ostensibly econometric discussion falls short of empirical inquiry.

It is disappointing that a colloquium entitled Econometrics should contain so little empirical material.

Observed prices and quantities do not, by themselves, distinguish demand from supply. This problem gives concrete force to Gerhard Tintner’s account of what mathematics can—and cannot—do for economic knowledge. His 1954 article connects economic modelling with the collection of data and the estimation of relationships, insisting that useful numerical results depend on assumptions that statistical technique cannot supply. Examples from butter demand, investment and industrial production show how models acquire practical meaning, while unidentified equations and short, dependent time series expose the limits of inference. Tintner’s selective methodological pluralism also leaves room for criticism: he questions decision theory where numerical losses cannot be justified. Readers can discover why choosing a mathematical method is inseparable from deciding what an economic model claims and what its evidence can establish.

An optimal cropping plan calculated from average yields leaves a farmer’s exposure to poor outcomes unresolved. M. M. Babbar, Gerhard Tintner, and Earl O. Heady address that gap by attaching probability estimates to a linear-programming solution for an Iowa farm. Their example turns on seasonal labor bottlenecks: corn and flax offer the best return under mean input coefficients, but historical yield variation changes the range of possible production and revenue. The article shows how lower profit bounds might help distinguish equally profitable plans, while making its restrictive assumptions explicit—small, normally distributed coefficient errors and, in the numerical application, fixed prices. Readers can examine a concrete attempt to connect resource optimization with downside risk, rather than treating a calculated optimum as a sufficient basis for decision.

A model chosen for its economic usefulness may still yield misleading measures of statistical precision. In this brief discussion of three agricultural-marketing papers, Gerhard Tintner welcomes modern econometric methods while warning that trying several models before selecting one introduces biases that affect reported standard errors. He also suggests that restrictive linearity assumptions may help explain modest forecasting results. His perspective is constructive rather than merely corrective: he connects agricultural decision models with economic and statistical theory, and values methods that make the choice among alternative technologies part of production analysis. The contribution offers a compact distinction between fitting data, forecasting reliably, and representing the decisions producers actually face.

Joining the ends of a series makes a difficult sampling problem tractable—but what does that simplification buy, and when might it be useful beyond the circular case? Gerhard Tintner studies variance estimates formed from finite differences of normally distributed observations. Overlapping differences remain dependent; his distinctive move is to exploit circular symmetry, factoring their quadratic form by frequency to obtain explicit sampling distributions. Readers can see why an unbiased estimate need not have a familiar finite-sample distribution, and how sample size and the order of differencing affect its uncertainty. Examples with seven and eight observations compare exact densities with normal approximations. Tintner also cautiously proposes using the circular results for non-circular series when endpoint effects are small, making the boundary assumption both the method’s strength and its practical qualification.

Computational progress does not settle the question of how reliable economic estimates are. In this brief discussion, Gerhard Tintner welcomes work by Nerlove, Suits and Koizuma on neglected supply functions and praises Ladd’s demonstration of the value of large-scale digital computers. Yet his approval carries a reservation: viewed against Morgenstern’s concern with the accuracy of economic observations, might Ladd’s model allow too little observational error? Tintner’s compact intervention offers a concrete distinction between advances in technique and adequacy of empirical assumptions, while pointing to cost functions and multicollinearity as subjects still needing attention.

Estimating how often an economic event occurs is not the same as assessing how strongly evidence supports an economic theory. In this short 1957 article, Gerhard Tintner uses Carnap’s distinction between statistical frequency and logical confirmation to clarify these two tasks of econometrics. His interest is practical as well as philosophical: what might confirmation theory contribute to disputes over estimation and hypothesis testing? Tintner identifies promising connections while preserving the obstacles—Carnap’s framework cannot yet handle the continuous variables of classical mathematical economics, and its results depend on choices of language and measure. The article offers a compact account of why numerical inference requires more than observations alone, without presenting Carnap’s developing programme as a finished econometric method.

How much mathematics should an introduction to economic modelling teach, and which tools can it afford to leave out? In this 1958 review of E. F. Beach’s Economic Models—An Exposition, Gerhard Tintner judges accessibility against technical coverage. He values Beach’s use of economic examples to make mathematical and statistical methods intelligible to readers with substantial economic knowledge but limited mathematical training. Yet he finds the explanations compressed and challenges the omission of matrices, determinants, input-output models and linear programming: these, he argues, are no harder than topics already included. This brief assessment offers a concrete view of Tintner’s pedagogical priorities, showing why an example-rich introduction can succeed while still leaving readers without useful analytical tools.

How much can a small sample tell agricultural administrators before a full census is available? In this brief 1958 review of Heinrich Strecker’s Moderne Methoden in der Agrarstatistik, Gerhard Tintner singles out a striking result: a sample covering 2% of West German agricultural enterprises produced preliminary census figures, nine-tenths of whose deviations from complete enumeration were below 10%. Tintner’s interest lies in the practical performance of statistical methods, rather than a separate examination of their theory. His favorable assessment connects this example with ongoing surveys of farm labor and agricultural production, while noting Strecker’s command of international research. The review offers a compact account of the concrete evidence on which Tintner based his confidence in agricultural sampling.