The archive.

Assembled from roughly three hundred pre-war price series across England, Germany, the United States and beyond, this statistical study argues that the trade cycle cannot be read off any single index number. Published in Vienna by Springer with a foreword by Oskar Morgenstern and backing from the Austrian Institute for Trade Cycle Research and the London School of Economics, Tintner applies Anderson's Variate Difference Method and moving averages to decompose each series into trend, cyclical and seasonal components. His finding is that prices move unevenly — metals and interest rates on their own rhythm, textiles and foodstuffs on another — so that the notion of a general price level dissolves. He offers the results not as proof of causes but as ordered material for the theorist, cautiously favouring the monetary cycle theories of Wicksell, Mises and Hayek.

We consider time, on the contrary, only as a kind of auxiliary variable, which we must eliminate in order to bring out the economic relations.

For Gerhard Tintner, reliable confirmation can matter more than striking novelty. His 1936 review of Allen and Bowley’s Family Expenditure praises their use of household budgets to connect demand theory with statistical evidence, while questioning one simplifying assumption: a linear preference scale maintained throughout the investigation. Where that assumption appears to fail, he wants statistical tests, not merely a workable approximation. This brief review offers a concrete view of Tintner’s standards for econometric research: close knowledge of the data, explicit testing of theoretical assumptions, and economic interpretation of numerical results. His praise turns on what the calculations establish about household spending—not on mathematical sophistication alone.

Discarding observations can make a statistical test more defensible. In this 1939 mathematical note, Gerhard Tintner confronts a difficulty in time-series analysis: successive differencing may remove a smooth trend, but it also creates correlations even when the original errors are independent. His response is to select differences built from disjoint observations, allowing their variances to be compared using familiar significance tests. He applies the same selection principle to lagged products in deriving a serial-covariance distribution. The note offers a precise encounter with the trade-off between retaining information and securing a tractable sampling distribution. Readers can see how the observations chosen determine which tests become available—and why the assumptions of normality, independence and a sufficiently smooth trend matter.

When observed variables contain measurement error, strong correlations alone cannot establish how many independent relationships underlie them. Gerhard Tintner’s 1945 note links that question to a second: how should those relationships be estimated once their number is determined? His distinctive move is to make one generalized eigenvalue problem serve both purposes, using an independently estimated disturbance covariance matrix to separate systematic dependence from noise. The emphasis is structural estimation rather than prediction: the aim is to recover coefficients of underlying relations, not simply improve a forecast. Readers can follow how error-weighted least squares leads back to the roots used in rank testing—and see why the resulting unity depends on assumptions about disturbance estimation and large samples.

What can an estimated demand curve tell policymakers—and what can it never decide for them? In this 1949 article, Gerhard Tintner connects the methodological claims of econometrics to the difficulties of measuring American agricultural demand and supply. Prices and quantities do not identify those relationships unaided: economic assumptions must first make them distinguishable. His empirical results sharpen the point. Demand estimates yield interpretable elasticities, while insignificant supply coefficients expose missing influences, including weather and possibly lagged prices. Tintner treats such failures as grounds for revising models, not concealing uncertainty. The article offers a concrete account of how theory becomes a testable numerical claim, while keeping estimates of policy consequences distinct from judgements about which social objectives deserve priority.

Different statistical aims can lead to closely related estimating equations without making the methods interchangeable. In this mathematical article, Gerhard Tintner connects canonical correlation, principal components, weighted regression, and discriminant analysis through constrained optimization. His distinctive move is to identify the matrices and normalization conditions that give a shared algebraic structure its different statistical meanings: a covariance matrix paired with the identity in principal components, for example, becomes sample covariance paired with error covariance in weighted regression. Readers can discover how apparently separate procedures meet through reductions to common stationary equations—and why their objectives and assumptions still matter. Tintner confines the comparison to estimation, leaving sampling distributions outside the argument.

Writing in tribute to Abraham Wald, Gerhard Tintner asks what mathematical rigor can secure for economics—and where its assumptions limit practical use. This 1952 memorial survey distinguishes the formulation of equilibrium equations from proofs that economically admissible solutions exist. It also shows how cost-of-living comparisons depend on information about preferences that observed prices and purchases alone cannot supply. Tintner’s appreciation is not uncritical: he questions minimax decision rules that treat an indifferent Nature as an adversary, and the feasibility of measuring social losses for policy decisions. The article offers a compact encounter with Wald’s achievements through an economist’s discriminating perspective, revealing both the power of explicit assumptions and the empirical work still needed to make formal results useful.

A fitted relationship need not be the economic relationship an investigator seeks. In Econometrics, first published in 1952, Gerhard Tintner makes this gap between statistical calculation and economic interpretation a central teaching problem. Prices and quantities are jointly determined, so reversing a simple regression cannot by itself recover supply or demand; smoothing a time series can introduce dependence rather than merely uncover it. Tintner’s emphasis is statistical, but his tests of method remain tied to economic questions and worked applications. Readers can discover why assumptions about structure, measurement and temporal dependence change what an estimate warrants—and why describing past observations is not enough to justify a forecast. This reprint preserves a textbook concerned as much with the limits of quantitative inference as with its procedures.

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.

When household budgets grow, does consumption expand in quantity, shift towards dearer varieties, or move into different goods altogether? Gerhard Tintner examines these alternatives through Austria’s 1954/55 urban consumption survey, distinguishing expenditure responses from changes in quantities and average prices paid. His estimates show why a simple contrast between necessities and luxuries is insufficient: rent protection and social insurance can weaken the connection between spending and household resources, while a food’s classification as “inferior” may depend on the social group examined. The report’s distinctive interest lies in its scrutiny of what such estimates warrant. Statistical uncertainty qualifies apparent differences, and forecasts depend on assumptions linking comparisons between households to changes over time. Readers can discover both concrete patterns of Austrian consumption and the limits of using household budgets to anticipate demand.

An investment allocation can maximize projected income yet leave productive capacity idle, consumption sacrificed, or investment exposed to greater uncertainty. In this article, Jati K. Sengupta and Gerhard Tintner examine that tension through Dutch long-term planning and India’s Mahalanobis model. Their concern is not simply to calculate an optimum, but to ask which assumptions make it feasible and desirable: whether saving keeps pace with investment, whether production techniques can change, and whether planners value output or consumption. A numerical exercise using Indian Third Five-Year Plan constraints makes the stakes concrete, showing how a preference for lower investment risk can alter the income-maximizing allocation. Readers can discover how seemingly technical choices about coefficients and objectives shape the economic priorities embedded in a development plan.

How much economic detail can a planning model responsibly promise when national statistics are scarce and unreliable? Gerhard Tintner and Oswaldo Dávila address this problem through a deliberately compact, five-equation Keynesian model of Ecuador, estimated from 1950–1961 data. They defend aggregation as a safeguard against false precision while arguing that coherent development policy requires explicit econometric relationships. The article’s concrete interest lies in its comparison of policy effects: within the model, investment and government consumption plus net exports can produce similar gains in output yet opposite movements in employment and wages. Readers can examine how limited evidence becomes a tool for distinguishing policy choices—and where that tool needs caution, particularly when reported numerical interpretations do not consistently match the equations.