Joseph A. Schumpeter, Business Cycles and Forecasting. Third Edition [review] (1950)

Schumpeter’s review of Elmer Clark Bratt’s revised textbook combines a strong pedagogical recommendation with pointed methodological criticism. He praises its incorporation of new evidence and approaches, its usefulness for undergraduate and introductory graduate study, and its balanced treatment of economic planning and full employment. His central judgment is that Bratt provides an unusually serviceable synthesis, though its historical coverage, statistical procedures, and conceptual classifications sometimes obscure the processes they seek to explain.

The review moves from the book’s educational merits and historical survey to trend analysis, the classification of cyclical movements, and the organization of competing theories. Schumpeter values historical narrative as a means of giving students something approaching experience of economic fluctuations. Yet Bratt’s account is uneven before 1929 and pays insufficient attention to individual industries. For Schumpeter, the economy-wide reach of business cycles does not justify excluding industrial growth and decline from their explanation. This criticism establishes a recurring concern: aggregate descriptions must not displace investigation of underlying mechanisms.

His discussion of trend analysis makes that concern explicit:

This fitting of trends to time series is a dangerous procedure that makes sense only under very restrictive assumptions, both of a statistical and of a theoretical nature.

Although Bratt carefully defends total physical output as a measure of long-period growth, Schumpeter finds the warnings about trend fitting inadequate. He asks that future editions introduce students to autoregressive stochastic processes and their potentially misleading effects. The objection is methodological rather than merely technical: apparent patterns in a series require scrutiny of the assumptions through which they become visible.

A related problem arises when Bratt distinguishes relatively short business cycles from longer “secondary trends.” Schumpeter accepts that terminology can reflect expository convenience, but argues that this distinction has explanatory consequences:

I submit, however, that the conceptual arrangement preferred by the author obscures the important similarity of mechanism that exists between his cycles and his secondary trends.

The important issue is therefore not whether a long movement deserves the name “cycle,” but whether differently classified movements share causal mechanisms. Schumpeter also questions the consistency between Bratt’s definition of cycles averaging approximately three and a half years and the evidence presented elsewhere in the book, while commending his commonsense treatment of amplitudes, phases, and international comparisons.

The final section welcomes Bratt’s theoretical pluralism. Distinguishing originating causes from conditions of self-generating oscillation roughly separates impulses from propagation mechanisms, but Schumpeter cautions that the two interact: impulses depend on situations produced by propagation, and propagation is affected by impulses. Likewise, the distinction between endogenous and exogenous explanation needs greater precision. His footnote identifies a consequential ambiguity:

This will exclude—and make exogenous—some purely economic factors that arise from within the business sphere.

Here Schumpeter distinguishes what is external to a formal model from what is external to business activity. A factor generated within the economy may nevertheless be exogenous to a model whose oscillations follow solely from specified relations and initial conditions. He suggests that identifying exogenous factors with influences originating outside the business sphere might be more useful.

Despite these reservations, Schumpeter strongly endorses Bratt’s presentation of business-cycle theories as complementary emphases on different aspects of a shared phenomenon. Reserved criticism and breadth of selection make the textbook valuable even to established economists working in other fields. The review’s enduring interest lies in this combination of theoretical openness and analytical vigilance: synthesis is welcome, provided that statistical conventions and classificatory choices do not conceal causal relationships.