The estimation of the number of cows permitted by the forage 

 supply is somewhat more cumbersome than for other resources. The 

 forage supply depends upon such factors as the type of forage produced, 

 length of stand, level of fertilization, as well as the possible substitu- 

 tion of forage for grain in the dairy cow ration. Estimation of the most 

 likely forage supply for farms with alternative resource bases can be 

 made by some preliminary programming and determining the types 

 of results obtained. By varying cropland, rather close estimates of 

 forage production and utilization are possible for farms with different 

 resource combinations. 



Farms 3 through 6 have been classified into Groups 3, 4, 5, and 6, 

 respectively. Farm 3 would be limited to 28 cows by the forage supply, 

 with cow numbers above that of existing dairy barn capacity but less 

 than total dairy building capacity. Farm 4 would be limited to about 

 22 cows by the quantity of winter labor, which is less than the existing 

 dairy barn capacity and the forage capacity on this farm. Farm 5 would 

 have enough winter labor for 34 cows, which would allow six more 

 cows that the present dairy building capacity but fewer cows than the 

 forage supply or the total dairy Ijuilding capacity. Farm 6 would be 

 limited to 20 cows by the total dairy barn capacity, or fewer cows than 

 permitted by the winter labor or forage supplies on this farm. 



The farms in different States were classified in this manner. There 

 were variations in such items as length of winter season, yields, and 

 value of property between many areas, resulting in different coefficients 

 than those used for the computations in Table 4. Also, some researchers 

 first classified farm by size of dairy herd before grouping according to 

 the most limiting resource. 



Development of Restrictions and Machinery Complements 



After the farms were classified, a single benchmark farm was 

 selected for each group excluding the nondairy group. Resource re- 

 strictions were then developed for the benchmark farm which were 

 merely averages of the farms within each group. For example, averages 

 were computed for acreages of cropland and permanent pasture, tons of 

 silo capacity, borrowing capacity, and existing dairy housing space. In 

 developing the machinery complements for each benchmark farm, the 

 model incidence on farms in each group was used. 



The Linear Programming Models 



Linear programming models were designed to compute the opti- 

 mum organization of representative farms from the standpoint of max- 

 imizing profits. Implicit in the construction of the models was the 

 hypothesis that farmers adjust output in response to prevailing milk 

 prices, factor costs, and the state of technology, so as to maximize re- 

 turns from available labor, land, and capital resources. The linear pro- 

 gramming models were coordinated with respect to four major con- 



19 



