Industrial production and industrial productivity in the German Empire

"In a series of papers and in a major book, Stephen Broadberry investigates the comparative productivity development of the British, US, and German economy. A central part of his analysis is the comparison of industrial productivity. Broadberry shows that the comparative industrial productivity is remarkably stable between the three economies over time and that the major reason for an “overtaking” of Britain by Germany in the late 19th and early 20th century is structural change from agriculture towards industry in Germany. Yet, industrial productivity in Germany was not far ahead of Britain’s in 1913. A precondition for such an international comparison is reliable industrial production data. In a recent paper Burhop / Wolff (2005) present new estimates of the German net national product and of Germany’s industrial production for the years 1851-1913. These findings could have a major influence on international comparisons since a higher German industrial production and industrial productivity for 1913 and the decades before are hypothesized. The current paper further investigates the calculation of production indices for 13 industry groups (including mining and construction) in Germany during 1871 and World War I. In addition, labour productivity estimations based on sectoral studies (metal production, metal processing, chemical industry, textiles, building materials and mining) are included. Furthermore, the assumptions made by Hoffmann (1965) to extrapolate his 1913 production values backwards to 1870 are discussed and partly improved by using evidence from more recent historical findings and by employing contemporary data for single industries. Additional information is included into new time series extrapolations, confirming a higher level of industrial production and industrial productivity in Germany. Especially, Hoffmann (1965) significantly underestimated production in metal processing, but also in metal production and chemical industry. On the other hand, Hoffmann’s estimates for textiles, building materials, and mining are confirmed. Overall, the level of industrial production was higher in 1913 Germany than suggested by Hoffmann’s data, but less than suggested by Burhop / Wolff (2005). However, the time-series extrapolation of industrial production backwards to 1871 confirms the index constructed by Burhop / Wolff. Thus, the level of industrial production was higher than the traditional estimates by Hoffmann (1965) and Wagenfuhr (1933) suggest. Furthermore, the cyclical behaviour of the new index of industrial production and of its components is investigated employing the Hodrick-Prescott-Filter. The results indicate a stable pattern of business cycle peaks and bottoms throughout the period under investigation. The major boom during 1871-74 and the following downturn are mainly driven by the construction sector; other industries do not exhibit extraordinary fluctuations. In addition, a prolonged stagnation between 1873 and 1896 – the so-called “Great depression” – is not confirmed by the new index. Finally, the VAR based Toda-Yamamoto causality test is employed to detect leading sectors of Germany’s industrialization process. So far, especially Railways and joint-stock credit-banks are considered as leading sectors during Germany’s early phase of industrialization. We will test the leading sector hypothesis for twelve industry groups using non-structural VARs and by employing impulse-response-analysis."