Oskar Morgenstern’s journal article outlines a research program for studying the international transmission of economic fluctuations, accompanying his historical and statistical investigation for the National Bureau of Economic Research. Its six sections move from criticism of existing theory through distinctions among shocks, cycles, and channels of contact to problems of measurement, institutional change, and the “world economy.” The central argument is methodological: transmission cannot be inferred from the coincidence of national cycles or explained adequately by extending a theory of a closed economy. It requires detailed, institutionally grounded evidence about particular markets and their changing relationships.
Morgenstern’s opening criticism concerns the distance between plausible explanations and established theory. Existing accounts of international propagation inherit the contradictions of competing national-cycle theories; mathematical “model sequences,” including foreign-trade multipliers, often substitute extreme simplification for rigor. His objection is not to mathematical economics as an aspiration, but to treating preliminary constructions as demonstrated explanations, especially when applying them to policy.
At best, some of the current theories purporting to explain the fluctuations of a closed national economy have the character and undoubtedly significant value of heuristic theories; but from the invention of a preliminary heuristic procedure to the establishment of an exact theory is still a big step.
This distinction governs his own proposals. Following Wesley C. Mitchell’s empirical example, Morgenstern judges research by its capacity to establish new facts and stimulate further investigation. Selection of material is unavoidable, but rigid theoretical preselection risks excluding phenomena whose importance becomes apparent only during inquiry. Statistical collection and comparative economic history must therefore help formulate the problem, rather than merely supply tests for an already settled explanation.
The second section separates three relationships commonly grouped under transmission: an external shock absorbed as a disturbance; a noncyclical shock that induces a cycle in the receiving country; and sustained, phase-by-phase dependence between complete cycles. Each demands different evidence. Similar timing may reflect independent domestic forces, a common external cause, or international influence; conversely, transmission may occur without simultaneous movements. Turning-point comparisons are useful beginnings, but amplitude, duration, location, and the receiving economy’s cyclical phase are also essential.
The decisive conceptual move is to replace contact between supposedly unified national economies with contact among particular economic spheres.
Instead, we conceive of the possibility that certain spheres of two national economies, say, because of their similar functions, may represent much more of a common structure than the structure which the same sphere forms with some others of its own economy.
Financial markets offer the clearest starting point, but trade, lending, direct investment, migration, and imitation provide distinct channels. Their importance and speed need not remain constant. Financial expectations may propagate before physical transactions become visible, while different capital movements can run in opposite directions simultaneously. Aggregates can consequently conceal precisely the relationships an explanation needs to recover. Morgenstern also distinguishes increased use of an existing transmission mechanism from institutional transition to another mechanism: discovering a channel would not establish its permanence.
The fourth section asks how observed changes might be divided into autonomous national and internationally induced components. Morgenstern regards unambiguous quantitative separation as a distant goal, given unresolved difficulties even in decomposing domestic time series. Removing apparently random disturbances can erase events responsible for subsequent changes elsewhere. Likewise, a country’s foreign-trade dependence does not measure its importance to other countries: the United States can be relatively independent of trade yet exert substantial external influence.
In order to classify reactions as more or less "sensitive" to given influences, the strength and continuity (or frequency) of the force operating upon the elements whose "sensitiveness" is to be determined must be described and measured.
Sensitivity is thus relational, not simply a name for large or frequent fluctuations. An apparently stable variable may have responded strongly to weak impulses. Without evidence about those impulses and counteracting forces, rankings of sensitivity remain uncertain; provisional classifications must rest on historically informed knowledge of possible interaction.
Institutional comparability determines the proposed investigation’s initial scope. Morgenstern favors the financial markets of the United States, Great Britain, Germany, and France before 1914, particularly the period from about 1870. The gold standard and relatively unrestricted international movements provide a more homogeneous setting than the postwar years, despite the latter’s better statistics. He distinguishes similarly developed economies from complementary relationships between countries with different productive structures. Similar institutions, however, do not guarantee similar outcomes or uses of those institutions.
The final section rejects a “world economy” as a substitute for identifying concrete connections. Different sectors of one country may face different sets of foreign markets. Internationally traded shares illustrate how arbitrage and psychological leadership can connect economies, but do not establish a single world cycle. Foreign national or specific cycles may serve as reference cycles where dependence is sufficiently well established.
By itself, covariation does not give us information about a possible transmission, because the lags and leads may be due to faults in the material or they may prove to be too shifting.
Morgenstern concludes by defending close historical scrutiny alongside long statistical series. Extreme crises can expose relationships obscured in ordinary periods, provided institutional discontinuities are respected. The article’s enduring relevance lies in this demanding account of causal identification: establish channels, direction, timing, magnitude, and institutional context before claiming transmission. Its research program remains explicitly heuristic, designed to make stronger explanations possible rather than announce a completed theory.
This work was divided into 9 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 9 sections and cites the passage.
Ask the Librarian