The Willamette River Case: Gains 237 



tributors. The sales of the municipalities, PUD's, and private 

 electric utilities - to domestic-residential and commercial-industrial 

 customers — which virtually exhaust total sales to ultimate cus- 

 tomers ^ — will indicate any significant differences in distribution 

 patterns. Table 47 provides the breakdown for each type of 

 distributor among the two categories of sales.* 



The data from Table 47 also can be used in combination with 

 the information on BPA's energy sales to determine what propor- 

 tions of the latter's sales end up in industrial and domestic uses. 



TABLE 47. Per cent Distribution of Kilowatt-Hour Sales by Type 

 of Use in the Bonneville Power Administration 

 Marketing Area 



Class of Domestic- Commercial- 



distributor residential " industrial " 



Municipalities 57.3 36.6 



Public utility districts 52.6 42.5 



Private electric utilities 44.3 51.2 



* Failure of the figures for each class of distributor to total 100 per cent is 

 accounted for by miscellaneous sales not included in our categories. 



^ The sales of rural electric co-operatives have not been broken down for two 

 reasons. First, they are not required to report annually to the Federal Power 

 Commission, and therefore the data on their distribution are not readily avail- 

 able. Second, despite their number, they accoimt for only about 3 per cent of 

 the total of BPA's power sales (4 per cent of the increment between 1954 and 

 1955) and therefore do not appear to justify the added effort required to obtain 

 data from them comparable to that available for the other distributors. 



' Interchange transactions with other utilities are excluded in the analysis. 



* The data on which these statistics are based were taken from those annual 

 reports of distributors in the BPA area to the FPC which were available for 

 analysis during the first week of December 1956. These represent 82 per cent 

 of the annual reports from municipalities, 96 per cent of the PUD reports, and 

 78 per cent of the total from private utilities. In this sense, we can regard the 

 data as a relatively complete enumeration of the total statistical population. It 

 may be of interest, however, to regard these data as samples from a possibly 

 larger universe of municipalities, PUD's, and private utilities which conceivably 

 could be supplied with power for resale by BPA. In that case, we would wish to 

 know whether or not the differences among our summary statistics were sta- 

 tistically significant. We have thus analyzed the variances about the means for 

 each group and have determined that tlie variance among group means exceeds 

 the variance within groups indicating that the differences which are shown to 

 exist in Table 47 can be regarded as significant at the 1 per cent level. 



