This paper reconsiders the contribution of Henry Ludwell Moore to dynamic economics through the use of harmonic analysis. We show that Moore’s analysis is innovative in its use of the Fourier transformation for the identification of cycles with different periodicities. This enables Moore to identify cycles of longer length with more precision than would be the case for the standard methodology. We are able to replicate the main features of his results and confirm the existence of a rainfall cycle with a periodicity similar to that of the business cycle (eight years). However, we find that the evidence for a longer (thirty-three-year) rainfall cycle is weaker than Moore indicates. We also argue that a central theme of Moore’s analysis—the relationship among rainfall, agricultural productivity, and the business cycle—marks an early precursor of the “real business cycle” approach. George Stigler’s (1962) dismissal of Moore’s work on cycles as “a complete failure” is therefore, in our opinion, unfair. Instead, we argue that, although his work is certainly flawed, it nevertheless deserves a place in both the history of business cycle theory and empirical economics.
Business Cycle Fluctuations in U.S. Macroeconomic Time Series
