Staff Paper P73-1 Revised August 1973 PRODUCTIVITY GROWTH IN GERMAN AGRICULTURE: 1850 TO 1970 Appendix DATA ON PROD"CTIVITY GROWTH IN GERMAN AGRICULTURE: 1850 TO 1970 Adolf Weber Staff Papers are published without formal review within the Department of Agricultural and Applied Economics. The research on which this paper is based was supported in part by the University of Minnesota Economic Development C-nter. The author is indebted to Barbara B. Miller for assistance in the organization and processing of statistical materials and to Vernor W. Ruttan for critical review and editorial suggestions. PRODUCTIVITY GROWTH IN GERMAN AGRICULTURE: 1850-1970* ADOLF WEBER The process of agricultural development can be usefully interpreted as a dynamic process of induced technical and institutional innovation and factor substitution in response to growth in demand, changes in resour:!e endowments, and changes in relative factor and product prices. 1 'he purpose of this paper is to describe the long-term trends in German agricultural development for the period since 1850 and to test the "induced innovation" hypothesis against the German experience.2 The German case is of considerable interest in attempting to under­ stand the agricultural development process. At the beginning of the nineteenth century Germany was more than a generation behind Britain in industrial and agricultural development. Public support for advances in science, technology, and education was undertaken for the deliberate purpose of overcoming the gap in agricultural and industrial technology and in economic power between Germany and Great Britain. The publicly supported agricultural experiment station was a German institutional in­ novation. It was the model for similar developments in both Japan and the United States [8, pp. 136-138]. Germany was successful in achieving relatively high rates of partial and total productivity growth in agriculture in the 19th century (Table 1). And agricultural productivity growth in Germ,any compares favorably ADOLF WEBER is Professor at the Institut f~r Agrarpolitik and Marktlehre, Christian-Albreohts-Universittt, Kiel, Germany. 2 Table 1. Trends in factor productivity, Germany, 1850-1968 (five-year averages centered on year shown) Annual compound rate of change 1850 1880 192 5a 1950 a to to to to 1880 1910 1935 196g Output (net of seeds and feed) 1.5 1.7 2.7 3.5 Total inputs 0.8 0.6 3.0 1.6 Total productivity (output/total inputs) 0.7 1.1 -0.2 2.0 Numbs, of male workersc 0.5 <0.1 -2.0 -3.6 Output per male worker 1.0 1.7 4.8 7.3 Agricultural land area 0.0 <-0 .d 0.1 <0.1 Agricultural land area per male worker -0.5 -0.1 d 2.1 3.6 Output per ha. agricultural land 1.5 1 . 8 d 2.7 3.6 Arable land are (0.0) <-O. 1 d -0.1 -0.2 Arable land area per male worker (-0.5) <-O.id 1.7 3.4 Output per ha. arable land 1.5 1 .8d 3.1 3.5 aYear shown. bWest Germany only. cMale workers in agriculture, forestry and fishery dLand data is for 1883 rather than 1880. Source: FromF9] and C3, various issues-. 3 with the rates achieved by Japan and the United States in the 20th century. This pattern of productivity growth in German agriculture was more "balanced" than in the United States or the United Kingdom, where until 1925 productivity growth was dominated by growth in output per worker (Figure 1). Germany started its productivity growth from a lower output per hectare but a higher output per agricultural worker than Japan. The pattern of productivity growth in agriculture in Ger­ many was, however, remarkably similar to that in Japan. The German case is also of significance since it provides an oppor­ tunity to explore the role of a small scale livestock sector in the development process. The livestock sector in German agriculture, specifically in the western and northwestern parts of the country represented a small scale labor intensive crop and crop residue processing activity that added a vertical dimension to the size of a farm in an environment where the potential expansion in land area was severely constrained. Much of the contemporary literature in agri­ cultural development has been preoccupied with crop agriculture and has ignored the potential contribution of livestock production, particularly small scale livestock production, in the agricultural development process. Induced Technical Change in Agriculture3 It is generally agreed that technical change has represented an important source of growth in agricultural output. It is also increas­ ingly recognized that agricultural technology is relatively location specific and that there are multiple paths of technological development 4 (Figure 1). Technology can be developed to facilitate the substitution of relatively abundant (hence cheap) factors for relatively scarce (hence expensive) factors. In the United States, for example, it was primarily progress in mechanical technology which facilitated the expansion of agricultural production and productivity by relieving the technical constraints on the area that could be cultivated per worker. In Japan it was primarily progress in biological technology, represented by seed improvements which increased the yield response to higher levels of fertilizer applications, which permitted rapid growth in agricultural output in spite of severe constraints on the supply of land. It has been increasingly clear that the ability of a country to achieve rapid growth in agricultural output and productivity depends on its capacity to generate an ecologically adapted and economically viable agricultural technology. The generation of an efficient path of technical change is viewed primarily as endogen­ ous rather than exogenous to the total development process. The process by which an efficient path of technical changes is induced involves a complex relationship between factor endowments, relative factor prices, and the innovative behavior of farms, private agribusiness firms, and public sector agricultural experiment stations. Efficient adaptations by the agricultural sector to the growth of de­ mand and to changes in relative factor endowments involves both movement along a fixed production surface and the creation of new production sur­ faces which are optimum under the set of factor and product prices. This process is illustrated, using examples of both mechanical and bio­ logical technology, in Figure 2. 5 0. .. .... '.. I: /"1 ... qe. .........- - - - ........ -- A.. 4' . .... .. ...... - -. i - - r - - ­ 5 °PI 7 I 30 JO40 SO0000UV, 5 Ark'vltwe Ouk4 fIWteet Uau Mgb ei b. Figure 1. Historical growth paths of agricultural development in Denmark, France, Germany, Japan, United Kingdom, and the United States, 1880-1965, five-year averages Soiuameu: For Germany: Weber ... For Japan, Denmark, France, United Kingdom (UK), and the Unitjd States: Hayami and Ruttan 8, pp. 67-81; 6 LL > . N .... .....uai, .......... * a *: a > 0 ........ o .... ONV (kOO. ONHD31 lVD1001019) a EOV1(ADOONtDBl 1 aIVI]) l:AO ........ Le..... t oes N I 0 I " 7 The process of technical innovation can be described as a movement along a "metaproduction function" or "innovation possibility frontier." '4 In Figure 2 (left) U represents the land-labor isoquant of the metapro­ duction function, which is the envelope of less elastic isoquants such as uo and uI corresponding to different types of machinery or technology. A certain technology represented by uo, a reaper for example, is invented when a price ratio, Po, prevails for some time. When this price ratio changes from po to pI, another technology represented by ul, for example the combine, is indicated. Similar inducements in the livestock sector might be imagined by the invention of chaff-cutters and automatized animal feeding systems. The new technology represented by ul, which expands the land area per worker or the capital invested in livestock, irrigation systems, crop trees, greenhouses per unit of land, generally corresponds to higher animal or mechanical power inputs per worker. This implies a complementary relationship between land and power, which may be illus­ trated by line EA, M]. It is hypothesized that mechanical innovation involves the substitution of land and power for labor in response to a change in the wage rate relative to land and machinery prices. The process of advance in biological technology is also illustrated in Figure 2 (right). V represents the land-fertilizer isoquant of the metaproduction function. The metaproduction function is the envelope of less elastic isoquants, such as vo and vl, which correspond to crop varieties characterized by different levels of fertilizer responsiveness. A decline in the price of fertilizer is regarded as inducing a response 8 by plant breeders to develop more fertilizer-responsive crop varieties and by farmers to adopt the new varieties as they become available. The complementary relationship between biological technologies and fertilizer use, represented by EF,B], also extends to the protective chemicals (insecticides, herbicides) and the institutional innovations associated with the marketing and delivery of chemicel inputs and ser­ vices. Similarly, in livestock production a decline in the price of concentrated feedstuffs (oilcake, fish meal, urea) has induced animal nutritionists and breederE to direct their efforts to the development of feedstuffs which incorporate a higher percentage of the lower cost proteins and to select and breed for lines which have a more rapid rate of gain when fed the new rations. Complementarity between breeders and nutrition also extends to related biological and chemical technologies in the area of animal health. The hypothesized relationships between changes in relative factor prices and changes in factor use generated by the model outlined above are summarized in Tabl 2. Before proceeding to the btatistical tests, it will be useful to review the long-term trends in productivity growth and factor use in German agriculture. Factor Endowments, Prices, and Productivity During the first half of the 19th century the level of economic development in Germany was not substantially different from that in many less-developed countries today.