The Willamette River Case: Costs 217 



by those earning $5,000 or less annually. If we assume the appro- 

 priate model to be the one favoring high-income groups and 

 investment, the two low-income groups would bear only 20 per cent 

 of the total, while those in families earning 1 10,000 and over would 

 bear approximately 47 per cent of the total. 



A similar approach could be employed, in principle, in connec- 

 tion with the differential state and local tax liabilities. However, 

 because of the complexity of questions involving an appropriate 

 tax model for state and local bodies, and the paucity of published 

 data useful for quantitative analysis, we do not attempt to analyze 

 the intra-regional incidence. 



In the case of the nonfederal public body, funds would be raised 

 through voluntary subscription to revenue bonds; thus, there is 

 less reason to question the correspondence between the interest rate 

 and the marginal sacrifice of savers subscribing to the bonds.^^ In 

 one particular, however, the interest on these bonds overstates 

 their price relative to similar securities whose yields are not 

 exempted from federal taxes. Thus, the rate of 2.85 per cent used 

 in our analysis may be understated relative to securities which 

 enjoy no advantages conferred by public policy. However, our 

 assumption of power rates set to return revenues adequate to cover 

 debt service charges one and a half times should compensate for a 

 tendency to diverge on the low side in estimating the appropriate 

 annual capital charges.^^ The only significant difference in inci- 

 dence involves the shifting of federal tax liabilities; that is possible 

 because of the doctrine of intergovernmental tax immunity. This 

 shifting of tax incidence from power customers to others is of the 

 same order of magnitude, $705,000 annually, as in the case of 

 federal development of the reimbursable project features. The 

 incidence of shifted tax burdens from power consumers to the 

 general public is symmetrical with that shown for the federal case 

 and can be obtained from Tables 39 and 40. 



"Voluntary subscription to bonds promising a specified yield distinguishes 

 the case from that in which the coercive powers of a public body are employed 

 to raise funds through taxation. 



^Annual capital charges imputed to the operation ($1.3 million) actually 

 approximate the level which is indicated to be necessary to cover opportunity 

 costs in the case of federal development. Where funds are raised through taxa- 

 tion, rather than subscribed to voluntarily, social costs will probably be higher, 

 and we may actually have erred on the high side in our estimate in this case. 



