¹ See, for example, Parker, William N. and Klein, Judith L. V., “Productivity Growth in Grain Production in the United States, 1840–60 and 1900–10,” in Output, Employment, and Productivity in the United States After 1840, Studies in Income and Wealth, Vol. 30, National Bureau of Economic Research (New York: Columbia University Press, 1966), pp. 523–82Google Scholar, and Loomis, Ralph A. and Barton, Glen T., Productivity of Agriculture, United States, 1870–1958, United States Department of Agriculture Technical Bulletin 1238 (Washington: U.S. Government Printing Office, 1961)Google Scholar. For a recent study of an important dairy state during this period, see Lampard, Eric E., The Rise of the Dairy Industry in Wisconsin (Madison: State Historical Society of Wisconsin, 1963)Google Scholar.

² Parker and Klein, “Productivity Growth.

³ Research is in progress to develop labor input estimates that can be combined with this article's yield data to measure changes in labor productivity in dairying during the period 1850 to 1910.

⁴ The relationship between these influences may be thought of in terms of a simple model. At one point in time, geographical and economic conditions are outside the control of the individual farmer; he must take them as given. He adjusts to these conditions by adopting techniques (breeds, feeding quality and quantity, and animal care) consistent with his economic goals. As the geographic and especially the economic conditions change over time, the farmer reacts by changing his techniques— adopting existing ones or developing new ones. As he improves his techniques, the average yield of his cows increases; as many farmers react similarly throughout a state, the state average yield rises. In this article the term “techniques” (or “practices”) refers to choice of breeding, breeds, of quality or quantity of feed, animal care, or other operations designed to increase milk yields.

⁵ Despite the aggregate importance of dairying in the American agricultural economy, no continuous data for total milk production were made in nineteenth-century sources before 1890. The only continuous figures available are butter and cheese output and number of dairy animals reported in the U.S. Census of Agriculture beginning in 1850. For this article, the butter and cheese data were converted to estimates of total milk production using conversion factors (for the amount of milk required to produce a pound of butter or cheese and for relative consumption of fluid milk during this period) derived from an examination of nineteenth-century sources such as farm journals, reports of Dairymen's Associations, and state Agriculture Board reports. A description of the sources and techniques will be provided to any interested reader. For additional dairy statistics, see Strauss, F. and Bean, L., Farm Income, Farm Production and Farm Prices, U.S. Department of Agriculture Technical Bulletin No. 73 (Washington: U.S. Government Printing Office, 1941)Google Scholar;Vial, E. P., Production and Consumption of Dairy Products, U.S. Department of Agriculture Technical Bulletin No. 722 (Washington: U.S. Government Printing Office, 1940)Google Scholar;Report of the Commissioner of Patents for 1861 (Washington: U.S. Government Printing Office, 1861)Google Scholar;Alvord, Henry E., Statistics of the Dairy (Washington: U.S. Government Printing Office, 1903)Google Scholar;Willard, X. A., Practical Dairy Husbandry (New York: D. Moore Company, 1872)Google Scholar; and Elwood, Robert E., Lewis, Arthur A., and Strubel, Ronald A., Changes in Technology and Labor Requirements in Livestock Production, Works Progress Administration, National Research Project, WPA Report No. 2–14 (Washington: U.S. Government Printing Office, 1941)Google Scholar.

⁶ See “Extension of the Annual Milking Season,” below.

⁷ Gates, Paul W., The Farmer's Age: Agriculture, 1815–1860 (New York: Holt, Rinehart & Winston, 1960), p. 239Google Scholar. The Erie Railroad was extended into the Orange County dairy area of New York in 1842, shifting some of that region's butter and cheese producers to fluid-milk production.

⁸ Merritt, E., Production ana Consumption of Dairy Products, U.S. Department of Agriculture Bulletin No. 177 (Washington: U.S. Government Printing Office, 1915), p. 12CrossRefGoogle Scholar.

⁹ Throughout the analysis a state's average annual yield is accepted as at least an approximate quantitative indicator of the level of techniques being used in the state at that time. See footnote 11 for the basis of this assumption.

¹⁰ Specific comments on the pace, direction, and importance of the growth of western dairying can be sampled in Missouri Agricultural Experiment Station Bulletin No. 56 (1902)Google Scholar;Wisconsin Agricultural Experiment Station Bulletin No. 209 (1911)Google Scholar; and Wisconsin Dairymen's Association, Sixth Annual Report (1878), pp. 93–95Google Scholar.

¹¹ Since state average yields are taken as approximate indicators of the degree of use of techniques existing in the nation, one can assume the areas of (1) to be high-yield states, areas in (2) to be medium-yield, and those in (3) to be low-yield states. Experiment station studies have shown that the level of yields is closely related to the level of techniques used in a region. Natural surroundings affect the cost of feed and care, but dairying can be performed profitably—assuming that demand conditions otherwise justify it—in most inhabited areas of the United States by improving feeding, breeds, or handling techniques. Because of this close relationship between techniques (as defined previously) and yield levels, the outer limit of the national production possibility frontier is defined throughout this article as the level of techniques in use in the highest-yield states at any point in time. This should not suggest that social environment—increased urbanization, specialization by farmers—was not causally related. Yield levels, however, are quantitative indicators of farmers’ reactions to opportunities growing from environmental changes.

¹² This statement has both a mathematical and historical meaning. If the incremental yields exceed the average (case 1), the average increases; if they are the same (case 2), it remains constant; and if they are less than the average (case 3), the average declines.

There is an historical logic as well. If dairying becomes relatively more important (1) in areas where there is lack of knowledge or of acceptance of best available techniques, the animals will not be bred, fed, or cared for as well as if dairy activity expands (2) in areas where they will immediately be fed and managed in the best possible manner. The result in the first instance will be average or below-average yields by these animals (until there are advances in techniques or adoption of better existing ones) and the national average will consequently remain steady or decline. In the second, the marginal yields will exceed the average—pulling the average upward.

¹³ Obviously the westward movement and changes in techniques occurred simultaneously. To assume that they did not introduces an element of artificiality regarding historical phenomena. For example, if dairy farmers really had been confronted by a static technology and restraints on diffusion of techniques, dairy activity probably would not have spread into the West and South as it actually did between 1850 and 1910, but would have intensified in the northeastern states—the areas of best practices in 1850. This method is used in this article, however, simply to provide an analytical device that is helpful in disentangling, even if only crudely and artificially, the complex of historical events. The conclusions produced by this method should be interpreted accordingly.

¹⁴ High-yield states are those in which the average annual milk yield in 1850 exceeded 3000 pounds per cow; medium-yield states are those in which the average was between 1600 and 3000 pounds, and low-yield states are those in which the average was below 1600 pounds. No-yield states are those in which no milk production was reported in 1850 or in which the percentage of the national dairy cows in the state was less than one-tenth of 1 percent. For composition of each category see note below Table 4.

¹⁵ The only increase in the simulated average occurred in 1870 because of a relative rise in high-yield states reinforced by a substantial decline in low-yield areas, largely reflecting the effects of the Civil War on southern dairying. New York's percentage, for example, rose from 13.08 percent in 1860 to 15.11 percent in 1870. Alabama's share dropped from 2.68 to 1.90; Georgia's from 3.49 to 2.59. Altogether seven low-yield states showed a decline in 1870. The shift into medium-yield areas before 1890 reflects the growth of dairying in the Middle West. The 1910 data show a slight increase in the low-yield areas indicating the beginnings of the twentieth-century expansion in the South. During the last decade of the nineteenth century, the Department of Agriculture and southern experiment stations had begun to exhort southern farmers to concentrate more attention on dairying. See U.S. Department of Agri-culture Farmers’ Bulletin No. 151 (1902)Google Scholar;Texas Agricultural Experiment Station Bulletin No. 47 (1898)Google Scholar;South Carolina Agricultural Experiment Station Bulletin No. 19 (1895)Google Scholar; and Louisiana Agricultural Experiment Station Bulletin No. 34, Second Series (1895)Google Scholar.

¹⁶ The changes in the data reflect farmers’ reactions to new conditions outside the northeastern states. Developing urban centers, improved transportation, and population growth generally created conditions in some western areas similar to those in parts of the East in 1850. The medium- or no-yield states showing the largest percentage growth before 1910 included Iowa, Nebraska, Kansas, Minnesota, and Wisconsin, where dairying offered an alternative or supplement to other products. In some areas, dairying replaced wheat as a farmer's principal activity when the profitability of wheat or the fertility of soil had declined. Dairying had largely supplanted wheat-growing in southern Wisconsin and northern Iowa by 1890 and was beginning to do so in southeastern Minnesota. Dairying was also better suited to the more intensive mixed-farming methods being adopted by many farmers during this period. Milk production could be valuable both as a commercial product and as a source of family food to the farmer. Milk cows also provided an outlet for home-grown feed crops, a supply of calves that could be sold, a source of fertilizer, and, especially where the animals kept were “dual-purpose” beef-dairy cattle, a source of meat. On many farms, pork, fed partly on buttermilk, provided a profitable by-product of the dairy.

¹⁷ This can be seen in data in U.S. Department of Agriculture Bulletin No. 25 (Washington: U.S. Government Printing Office, 1929)Google Scholar.

¹⁸ All milk-producing animals, whether inherently high-yielding or low-yielding ones, require a basic ration for body maintenance. Feed consumed above this amount is converted by the animal into milk. Beyond the animal's inherent milk-producing range, additional feeding chiefly produces additional body weight. The productive advantage of a specialized dairy breed lies in the wider range of its milk-producing proclivity. Dual-purpose or “native” stock reach their maximum potential output at lower-yield levels than dairy breeds.

Examples of the discussion about the preference for nonspecialized breeds can be found in: The 14th Annual Report of the Wisconsin Dairymen's Association (Madison, 1886)Google Scholar;Annual Report of the Nebraska State Board of Agriculture, 1895 (Lincoln, 1896)Google Scholar;28th Annual Report of the Missouri Board of Agriculture (Jefferson City, 1895)Google Scholar; and the Report of the Board of Directors of Iowa State Agricultural Society, 1880 (Des Moines, 1881)Google Scholar. The national prevalence of dual-purpose breeds was discussed for example, in the Transactions of the Wisconsin State Agricultural Society, 1872–73 and in Stewart, Henry, The Dairyman's Manual (New York: Orange Judd Company, 1888)Google Scholar. Also see Report of the Commissioner of Agriculture, 1863 (Washington: U.S. Government Printing Office, 1940), p. 371.Google Scholartin No. 55 (Washington: U.S. Government Printing Office, 1903)Google Scholar.

¹⁹ For example, see Transactions of the Wisconsin State Agricultural Society, Vol. II (Madison, 1852)Google Scholar and Corbin, John P., Practical Hints on Dairying (Whitney's Point, New York, 1871)CrossRefGoogle Scholar. For comment, see Proceedings of Convention of Agriculturists Held in the Department of Agriculture, January 10–18, 1882 (Washington, 1882)Google Scholar.

²⁰ U.S. Department of Agriculture, U.S.D.A. Agricultural Statistics: 1939 (Washington: U.S. Government Printing Office, 1940), p. 371Google Scholar.

²¹ Sample comments on the progress of changes in feeding methods among regions can be found in: North Carolina Agricultural Experiment Station Bulletin No. 106 (1894)Google Scholar;Ohio Experiment Station Bulletin No. 50 (1898)Google Scholar;Texas Experiment Station Bulletin No. 47 (1898)Google Scholar; and Indiana Experiment Station Circular No. 13 (1908)Google Scholar.

²² A rising national average yield is compatible with the absence of important breed improvements, particularly if practices are changed in areas in which techniques have been poor or average. There is a distinction between a milk cow's potential milk yield and its actual yield. An increase in the potential yield, which is basically an inherited characteristic and a function of breed, results from improved breeding and selection techniques. A rising actual yield, however, which the data in Tables 1 and 2 measure, can result from farmers’ efforts to bring their cow's yields nearer to the inherent potential through better and more liberal feeding and care. If native or dual-purpose animals are fed so poorly that even their relatively low potential is not reached—as was true in most states in 1850—additional feeding can produce greater actual yields. As feeding and care techniques were improved in the West and South, actual yields rose in these states—and consequently so did the national average—even though there was little national breed improvement. In fact, the data in Tables 1 and 2 suggest this. The results of breed improvement, if accompanied by feeding and care improvement, would have been a significant increase in the upper Limit state yield, but this did not occur. The rising average and lower limit that the data show suggest a closing of the gap between actual and potential yields with little change in the national level of potential yields before 1910.

²³ The regional milking period estimates for 1850 and 1910 were computed on the basis of evidence found in the agricultural literature of these periods.

²⁴ The seasonal price variation for fluid milk in five major cities in 1910 can be seen in Cassels, John M., A Study of Fluid Milk Prices (Cambridge: Harvard University Press, 1937), pp. 271, 275–77, and 279.Google Scholar For a sample of discussions of the growth of winter dairying, see U.S. Department of Agriculture, Bureau of Animal Industry Bulletin No. 169 (Washington: U.S. Government Printing Office, 1913)Google Scholar; Annual Report of the New York Dairymen's Association, 1878; Transactions of the Wisconsin State Agricultural Society for 1883–84; 14th Annual Report of the Wisconsin Dairy-men's Association, 1886; 28th Annual Report of the Missouri Board of Agriculture, 1895; and the 16th Annual Report of the Bureau of Animal Industry (Washington, 1899)Google Scholar.

²⁵ Not only was fuller use of dairy capital an important source of the expansion of annual output, but research in progress on dairy labor productivity tentatively indicates that growth of output was achieved from the reduction of farm labor under-employment as well. Reduction of capital and labor unemployment was particularly noticeable in noneastern areas. It is possible, of course, that a large portion of economic improvement and growth is gained in this way in all economies emerging from an underdeveloped state as wider markets develop.

²⁶ Speculation on reasons for the lag in development of agricultural biology and chemistry behind mechanical innovation can be found in Parker and Klein, “Productivity Growth,” p. 546.