A quantity-based annual index of United States industrial production, 1790--1915: An empirical appraisal of historical business-cycle fluctuations

The dissertation consists of three related chapters on the historical and cyclical properties of the U.S. macroeconomy. The initial chapter documents the construction of completely new time-series data, while the latter two examine its business-cycle implications.; As a remedy for the notorious deficiency of pre-Civil War macroeconomic data, the first chapter introduces an annual index of U.S. industrial production consistently defined from 1790 until the First World War. The index incorporates 39 annual physical-volume series in the manufacturing and mining industries. The index is consistent with the infrequent broad benchmarks for the pre-Civil War period, and is markedly less volatile than earlier indexes constructed for the post-Civil War era.; Chapter 2 examines whether U.S. business-cycle volatility moderated over the 1790–1915 period using the index unveiled in Chapter 1. I find that pre-Civil War volatility, however defined, is statistically indistinguishable from that observed in the post-Civil War, pre-WWI economy. This salient feature of the long-span data is noteworthy because the nineteenth-century economy underwent profound structural changes while countercyclical Keynesian policies were non-existent. Further analysis shows that accentuated cyclical co-movement across broad sectors attenuates any tendency toward volatility moderation. This last property coincides with the ongoing market integration of the nineteenth-century American economy.; Chapter 3 reassesses the accuracy of the pre-WWI NBER business-cycle chronology. The fresh macroeconomic data either remove or revise close to one-half of the NBER peak and trough years between 1796 and 1914. Consequently, nineteenth-century recessions (expansions) are systematically shorter (longer) and occur less frequently than conventionally maintained. Indeed, revisions to the first half of the two-hundred-year chronology generate a profound implication: the frequency and duration of pre-WWI business cycles are identical to their annual post-WWII counterparts. That is, the appealing story of modern U.S. macroeconomic stabilization reads increasingly like a fairy tale.