252 MULTIPLE PURPOSE RIVER DEVELOPMENT 



as the net gains from the railways' share of the savings. We divide 

 these gains functionally as before — 55 per cent representing the 

 before-tax share of capital, 12 per cent labor's gains, and 33 per 

 cent the share of general consumers of railway services. To deter- 

 mine the locus of these gains, we assume that 50 per cent of the 

 total before-tax gains to capital appears as reductions in general 

 tax liabilities, distributed regionally for convenience by the weights 

 given by tax Model A; the other 50 per cent appears as dividend 

 payments or increased net worth of shareholders in the railways. 

 We assume that 10 per cent of these are distributed in the Pacific 

 Coast region, the remainder in other regions. Moreover, we assume 

 that the gains going to labor are distributed about equally between 

 the Pacific Coast and all other regions; while the gains to general 

 consumers of railway services are assumed to share 10 per cent in 

 the Pacific Coast region and 90 per cent in the remainder of the 

 nation. 



Table 54 presents in schematic fashion the distribution of gains 

 by regions to provide a rough approximation of relative magni- 

 tudes. The assumptions employed in this model do not have a care- 

 fully grounded empirical basis, but were selected rather on the 

 basis of rough judgment as to realistic possibilities to provide some 

 insight into the nature of the regional locus of gains. ^^ 



To locate the gains beyond the point at which they first come to 

 rest as gains to households, we would have to embark on an inordi- 

 nately complex piece of empirical and institutional analysis. For 

 example, an increased demand in the Northwest for labor in alumi- 

 num reduction would represent a relative decline in the labor mar- 

 ket from which the firm would have drawn in the Ohio Valley. 

 Labor in the Northwest would gain, whereas in the Ohio Valley it 

 would suffer relative losses. Increased orders for petroleum coke from 

 the Wilmington, California, plant of the Great Lakes Carbon Cor- 



" We hold no brief for the model employed here. Students of regional eco- 

 nomics in the Northwest more intimately familiar with the structure of North- 

 west industry, the interregional commodity flows, rates of return to different 

 industrial undertakings, the participation of co-operating factors in the shares 

 of gains, and the regional distribution of equity securities in Northwest business 

 enterprises may improve on the assumptions to provide a more refined model. 

 The regional distribution of gains resulting from such a model, however, on 

 balance is not likely to differ in any startling degree from those summarized in 

 Table 55. 



