used principally in the analysis of short-run supply response.'^ In recent 

 regional adjustment studies conducted in cooperation with the USDA, 

 supply functions have heen estimated through linear programming. The 

 procedure is described briefly as follows: 



(1) Stratify the region into areas based upon production oppor- 

 tunities and natural resources. 



(2) Sample each area to provide a basis for constructing represen- 

 tative farms. 



(3) Construct representative farm linear programming models and 

 solve them with a series of product prices to obtain step supply 

 functions. 



(4) Sum the supply functions of the individual representative 

 farms over areas or regions. 



This section describes each of the above steps. The procedures 

 employed in this study are for the most part similar to those used in 

 the other regional adjustment studies. A notable exception is the meth- 

 od adopted for the specification of representative farms which is de- 

 signed to reduce bias in the aggregation of supply functions. 



Selection of Areas Within the Northeast 



The objective of stratification is to divide the region into farming 

 areas that are homogeneous in terms of such factors as climate, topo- 

 graphy, soils, or marketing outlets. Thus, the farms within the area 

 might be expected to have the choice of similar enterprise alternatives, 

 and to be faced with similar yield potentials, prices, and costs. The 

 stratification of the region into areas is followed by the classification of 

 all farms into a smaller number of representative farm groupings within 

 areas. The task of selecting areas and representative farms within areas 

 is closely related. 



The Northeast region was divided into 20 such areas based upon the 

 above considerations. These areas are shown in Figure 2 and are identi- 

 fied as follows r''^ 



1. Central Maine 



2. Southern Maine-Southern New Hampshire 



3. Southeastern New England 



4. Northeastern Vermont-Northwestern New Hampshire 



5. Southeastern Vermont-Southwestern New Hampshire 



6. Southwestern New England 



7. Northwestern Vermont 



8. Southwestern Vermont 



9. Hudson Valley 



10. Northern New York 



11. Oneida, Mohawk, and Black River Valleys 



6 Effort has been made by Nerlove and others to estimate long-run supply elasti- 

 cities using time series analysis. See Marc Nerlove, The Dynamics of Supply, the 

 John Hopkins Press, Baltimore, 1958. 



'' Urban centers and forest or mountain lands are not considered as production 

 areas. 



11 



