Schumpeter’s short methodological article diagnoses the fragmentation of business-cycle research and proposes a coordinated program of theoretical, statistical, and historical inquiry. Its movement is from a fundamental question about capitalist development, through the interpretation of time series, to the collection of historical evidence and the reconstruction of economic theory. The central claim is that empirical research cannot become explanatory merely by accumulating observations, while theory cannot advance by neglecting the phenomena those observations reveal.
Den schwächsten Punkt im werdenden Gebäude einer befriedigenden Lehre von den Wechsellagen sehe ich in der mangelnden Koordination der Ergebnisse, schon der Fragestellungen, der Materialforschung und der Theorie.
English translation: I see the weakest point in the emerging structure of a satisfactory theory of business cycles in the insufficient coordination of the results—and even of the questions—of empirical research and theory.
The difficulty is institutional and intellectual: researchers turn differences in their instruments into mutually exclusive methodological principles. Schumpeter instead compares their proper relationship to that between theorists and experimentalists in physics. He identifies Spiethoff’s work as exemplary because evidence and theory penetrate one another. Without this cooperation, the field produces meaningless trends, unproductive masses of material, and unusable theories.
The decisive substantive question is whether economic development is fundamentally continuous, with fluctuations imposed by special disturbances, or whether cyclical movement belongs to the development process itself. External events are sufficiently numerous to make particular explanations of almost any crisis or upswing plausible. Yet their availability does not settle whether business cycles constitute the characteristic form of capitalist change. Time-series curves accommodate both interpretations; deciding between them requires historical and theoretical analysis. Schumpeter thus formulates an endogenous-cycle research problem without presenting a developed causal theory of his own here.
The same dependence on conceptual interpretation governs apparently narrower statistical tasks. A series can be decomposed into trends and waves in indefinitely many ways; its shape cannot determine which components represent economically meaningful processes. Likewise, observed sequences among variables do not establish causation. Even the apparently secure association between low short-term interest rates and high stock-market prices becomes uncertain under close inspection, and correlations with preceding and subsequent prices need not identify the direction of influence.
Das Nichtbeachten dieser Tatsache macht unsere Reihen stumm und auf sie gegründete Voraussagen sinnlos.
English translation: Failure to take account of this fact renders our series mute and forecasts based on them meaningless.
This warning concerns the theoretical grounds of decomposition, not the value of statistical evidence. Schumpeter distinguishes theory as an explanatory hypothesis about cycles from theory as the conceptual equipment needed to grasp relationships among economic magnitudes. The latter is already indispensable to connecting symptoms; a fuller explanation of their sequence may also require a hypothesis about the cyclical process as a whole. The alternative to theory is therefore not unmediated access to facts, but inadequately understood analytical choices.
His positive empirical program combines longer chronological coverage with finer historical resolution. Researchers should secure every accessible series, whether or not its causal significance is immediately apparent, and push evidence further into the past. Early capitalist development in Spain and Italy, together with Spanish price series beginning in 1500, suggests that cyclical phenomena may be much older than commonly assumed.
Wir brauchen ferner quantitative im Gegensatz zu bloß institutioneller Wirtschaftsgeschichte.
English translation: We also need quantitative economic history, as opposed to merely institutional economic history.
Institutional history remains indispensable, but quantitative histories of industries and individual firms offer particularly direct access to cyclical mechanisms. Aggregate figures can conceal compensating movements and consequently give a false picture. Industrial monographs can illuminate processes where time-series analysis fails; isolated crisis narratives no longer provide an adequate research framework.
Statistical technique must likewise develop from knowledge of its object. Trend elimination requires an account of what a trend represents. Choices between levels and logarithms, absolute differences and percentage deviations, cannot be justified solely by computational convenience. Schumpeter directs an equally sharp criticism at theorists: even the construction of price and production indices lacks sufficiently clear foundations concerning what those indices measure. The needed advance is reciprocal—an economic theory of statistical concepts and a statistical theory of economic concepts.
Auch heute noch sind jene Theoretiker eine kleine Minderheit, die begreifen, daß die Wechsellagenforschung der Nährboden der allgemeinen Theorie der Zukunft ist.
English translation: Even today, those theorists who understand that business-cycle research is the fertile ground for the general theory of the future remain a small minority.
The article consequently makes cycle research central to the renewal of economics. Analytical machinery must be rebuilt to accommodate oscillations instead of dismissing them as inessential deviations. Failures to apply existing theory should prompt criticism of its practitioners, not wholesale rejection of analytical reasoning. Schumpeter closes with a concrete shared task: identifying and representing the waves of different lengths contained in the evidence. The article’s significance lies in this demanding conception of explanation, in which historical detail, measurement, and theory continually correct one another.
This work was divided into 1 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 1 sections and cites the passage.
Ask the Librarian