Gerhard Tintner, Einkommenselastizitäten im österreichischen Konsum: Ökonometrische Untersuchungen zur österreichischen Konsumerhebung 1954/55 (1959)

Tintner’s econometric monograph examines the 1954/55 survey of urban Austrian household consumption through 310 analyses of expenditure, quality, and quantity elasticities. It moves from aggregate spending and differences among social groups to household-size effects, the decomposition of expenditure into quality and quantity, and international comparisons. Its central finding is that additional resources reshape consumption unevenly: necessities generally respond weakly, luxuries strongly, and some staples negatively.

The predictive argument depends on interpreting cross-sectional differences as indications of change over time. Households entering a higher income bracket are assumed to adopt consumption patterns already observed there. Tintner identifies the conditions supporting this inference:

Diese Annahme glaubt man bei gleichbleibenden Preisen in Anbetracht der relativen Konstanz der Verbrauchsgewohnheiten machen zu können.

English translation: One believes that this assumption may be made, so long as prices remain unchanged, in view of the relative constancy of consumption habits.

Stable prices and relatively persistent habits thus connect observed household differences to prospective consumption changes. A further condition governs the interpretation of expenditure elasticities as income elasticities:

Weiters soll mit einer 1%igen Zunahme des Einkommens eine Zunahme der Gesamtausgaben, ceteris paribus, um ungefähr 1% einhergehen.

English translation: Furthermore, a 1 per cent increase in income is to be accompanied, ceteris paribus, by an increase in total expenditure of approximately 1 per cent.

Using 7,019 responding households, Tintner estimates double-logarithmic Engel curves by least squares, weighting grouped observations by their numbers of families. The specification facilitates calculation and comparison with other European studies. Standard errors, significance tests, and confidence limits qualify the coefficients. Numerical detail supports subsequent statistical work, but practical significance depends on the purpose of interpretation:

Eine wirtschaftspolitische Interpretation der Ergebnisse wird mit weit geringerer Genauigkeit das Auslangen finden.

English translation: An interpretation of the results in terms of economic policy will manage with far less precision.

The aggregate estimates suggest that a 1% increase in total expenditure accompanies approximately 0.46% more food spending, compared with roughly 1.51% more non-food spending. Private transport, furnishings, and domestic services are especially responsive. Institutional conditions complicate a simple hierarchy of necessities and luxuries. Tintner attributes rent expenditure’s statistically insignificant relationship with total spending to Austrian tenant protection; social insurance similarly helps explain the weak income dependence of workers’ health expenditure.

The detailed food analysis distinguishes negative coefficients from statistically supported evidence of inferiority. Although several staples have negative estimates, only wheat flour and substitute coffee meet the stated significance thresholds in the aggregate results. Analyses of the self-employed, salaried employees and civil servants, workers, and pensioners reveal socially differentiated patterns: black bread is significantly inferior among salaried employees, while whole milk is so among pensioners. Inferiority emerges as a conditional empirical classification rather than an intrinsic property of a commodity.

Tests across social groups further restrain interpretation. Many broad categories, including aggregate food and non-food spending, exhibit no statistically significant differences in elasticity. Stronger differences appear for personal care, gifts and donations, and membership contributions. Approximate pairwise tests do not always agree with the overall variance analysis. Household-size homogeneity tests assess whether per-capita expenditure adequately represents scale effects. Non-food spending supports the restriction; food spending produces a mixed result, rejected at 5% but not at 1%, which Tintner regards as broadly compatible with the per-capita approach.

The quality analysis treats average prices paid as an indicator of quality. Increased spending may therefore represent purchases of more expensive varieties as well as greater quantities. Tintner estimates quality elasticities for selected foods and obtains quantity elasticities by subtracting them from expenditure elasticities. Most quality estimates are positive, though several are insignificant. Quantity and expenditure elasticities generally remain close relative to their uncertainty, with confectionery providing the principal qualified exception.

Comparisons with Hamburg, Sweden, France, and England broaden the analysis while exposing limits imposed by heterogeneous methods and incomplete measures of uncertainty. Tintner consequently calls for internationally coordinated surveys capable of producing comparable evidence. The monograph joins an empirical ordering of consumption responses to a sustained concern with inference: forecasting requires attention to statistical reliability, institutional conditions, and the assumptions linking household comparisons to changes over time.