(50x20 = 1,000). Sample segments were drawn randomly from within 

 blocks. Therefore, the construction of blocks within areas provided a 

 further guarantee of uniform geographic distribution of segments. The 

 sampling rate varied between areas, being 2^/2 percent in Area 15, 2 per- 

 cent in Area 16 and 17, and 1^4 percent in Area 19. 



Farm Survey Data 



Farms in the samples were surveyed in the summer of 1961. A total 

 of 2,739 schedules were usable for analysis. The survey was designed to 

 obtain information to permit a description of the resources available for 

 use by the individual farmer. Data were collected regarding cropland 

 and its use, the capacity and use of farm facilities, labor supply, and 

 capital structure. In addition, information was obtained concerning pro- 

 duction practices. The survey data provided the basis for developing 

 representative farm models. 



Other Data 



To complement the farm survey data, production, cost, and price 

 data were assembled from secondary sources. ^^ These provided a con- 

 sistent basis for developing model coefficients which reflect the real 

 difference in the productivity of resources and price differences be- 

 tween the 20 areas studied. Two sets of production relationships were 

 specified. An average set of coefficients was established representing 

 yields with average management and technology used in 1960-61. In 

 addition, a similar set of coefficients was developed representing manage- 

 ment and yields associated with the top 25 percent of farmers in 

 1960-61. These superior production relationships were used in the 

 analysis done in this study. 



Long-term real estate credit borrowing capacities had to be de- 

 veloped. The survey schedules provided information concerning farm 

 debt. Real estate appraisals of representative farms by areas were made 

 by members of the Farmers Home Administration. Net worths for study 

 farms were then calculated and borrowing capacities computed. 



The price projections used in the Lake States Dairy Adjustment 

 Study provided the basis for the level of prices. ^^ Historical price rela- 

 tionships were investigated for each of the 20 areas in the Northeast. 

 Since the milk price was varied in the linear programming analysis, this 

 was the only historical price that was not built into the analysis. Input 

 costs whose levels were not specified were estimated to correspond with 

 the regional price relationship of other items. The intent of these pricing 

 procedures was to simulate historical relationships at an expected 1965 

 level. 



'^^Agricultural Planning Data for the Northeastern United States, Pa. Agr. Expt 

 Sta., A.E. & R.S. 51, July 1965. 



11 Sundquist, W. B., et. al.. Equilibrium Analysis of Income — Improving ad- 

 justments on Farms in the Lake States Dairy Region. Minn. Agr. Expt. Sta. Tech. 

 Bui. 246, 1963. 



15 



