Schumpeter’s seminar report examines how index numbers can describe price movements and measure changes in money’s purchasing power. It moves from a methodological distinction between measurement and causal explanation through a critical comparison of existing tables to increasingly sophisticated weighting procedures. Its organizing principle is that statistical adequacy depends on the question being asked:
Der Wert aller derartigen statistischen Methoden ist ein relativer, das heißt für ihren Aufbau und ihre Beurteilung muß man von dem Gebrauch, den man davon machen will, ausgehen: Der Gebrauch ist das entscheidende Moment, auf absolute Geltung kann keine einzelne der zu erwähnenden Methoden Anspruch machen.
English translation: The value of all such statistical methods is relative, that is, their construction and assessment must start from the use one wishes to make of them: use is the decisive consideration; none of the methods to be mentioned can claim absolute validity.
This principle limits the explanatory ambitions historically attached to index numbers. Prices result from influences on both the monetary and commodity sides; an aggregate movement cannot by itself distinguish their contributions. Schumpeter rejects the assumption that commodity-side influences cancel out, leaving monetary causes visible. Technical progress and other broad tendencies move commodity prices systematically, not merely randomly. Index numbers must therefore first establish changing price relations without presupposing their causes.
The comparison of Soetbeer, Sauerbeck, and the Economist makes this caution concrete. Soetbeer’s Hamburg series combines sources with different and changing errors; Sauerbeck gives important commodities additional entries, but this crude weighting rests on arbitrary judgments; the Economist covers too few commodities. Schumpeter prefers Soetbeer’s consistent principle to Sauerbeck’s makeshift adjustments, without treating either as an exact measure of purchasing power. Compact numerical presentation is useful, but apparent precision in percentages can exceed the reliability of the evidence.
The report next asks what should enter an index. Commodity selection, changing qualities, and the inclusion of successive production stages can radically alter the aggregate. Counting both raw materials and their derivatives multiplies some price influences simply because intermediate products happen to be traded. Wages require separate tables, and an unweighted wage index must not be mistaken for an average wage. Retail prices matter for consumption studies, whereas wholesale prices more clearly register general market tendencies.
Comparison of arithmetic and geometric means, the mode, and the median extends the same purpose-sensitive reasoning to aggregation. Resistance to extreme movements can be useful, but may also suppress economically significant changes. Weighting becomes indispensable when the inquiry moves beyond a descriptive price series:
Unerläßlich aber ist das weighting, wenn man unsere Methode zur Messung der Veränderungen in der Kaufkraft des Geldes verwenden will.
English translation: Weighting is indispensable, however, if we wish to use our method to measure changes in the purchasing power of money.
The deeper obstacle is comparability across changing economies. Purchasing power relative to a particular commodity is observable; purchasing power over the entire stock of goods is not directly known. Moreover, that stock changes in composition. Fixed weights make comparison possible by holding quantities or importance constant, but their reference conditions progressively lose relevance. Marshall’s proposal to compare each year with its predecessor offers a way to incorporate new commodities and changing qualities.
Schumpeter surveys alternative weights and formulas, gives particular praise to Lehr’s procedure based on commodity-specific units obtainable for a monetary unit on average across the comparison years, and then examines fluctuating weights. His criticism of Drobisch exposes the danger of confusing compositional change with price change:
Ein zweiter Einwand ist, daß die Formel Schwankungen angibt, ohne daß sich die Preise geändert haben müssen.
English translation: A second objection is that the formula indicates fluctuations without prices necessarily having changed.
The concluding discussion of Palgrave and the weighted median leads to a defense of mathematical refinement against its dismissal as idle ingenuity:
Diese Methoden führen vielmehr sofort auf die grundlegenden Fragen der Preis- und Geldlehre.
English translation: Rather, these methods lead immediately to the fundamental questions of price and monetary theory.
The report’s relevance lies in this connection between statistical construction and economic concepts. Reliable indexes could support more exact inquiry into distribution and income, and help preserve the real value of debts and claims despite changes in purchasing power. Schumpeter ends with guarded anticipation: such applications require the methods he has examined, but their effective realization remains ahead.
This work was divided into 6 sections when it entered the library's research corpus—an apparatus for search and citation, not necessarily the author's own table of contents. Each title opens its summary.
Put a question to this work; the Librarian answers from its 6 sections and cites the passage.
Ask the Librarian