8. Four-Factor Allocation - Parent
We find that the record does not support using the Commission's preferred method, a four-factor allocation, as proposed by Suburban for general cost allocation to affiliated companies and other unaffiliated entities served by Suburban's parent company. The data provided this time by Suburban is incomplete, inconsistent, and therefore unreasonable to distribute general costs incurred to benefit multiple entities. We find only two factors, operating expense and payroll, are sufficiently accurate and therefore reliable as a means of allocating costs. This results in an adopted allocation factor of 15.85% and not 32.7% as requested by Suburban for parent company costs to affiliated and unaffiliated entities. We do find that the use of the four-factor allocation method proposed by Suburban within the utility group, excluding the unaffiliated entities, is reasonable.
8.1. Background
The Commission staff, in 1956, issued the still used standard practice "Allocation of Administrative and General Expenses and Common Utility Plant" (Allocation Practice) which is used to allocate indirect costs that cannot otherwise be reasonably charged directly to different departments within a company, between related companies or different regulatory jurisdictions. The Allocation Practice uses a weighted average of four common factors to allocate indirect costs. These are: (1) direct operating expenses, excluding uncollectible revenue, general expenses, depreciation and taxes; (2) gross plant (3) number of customers (or subscribers of telephone service); and (4) number of employees (using direct operating payroll, excluding general office payroll. (DRA Opening Brief at 9 - 10 citing to the Allocation Practice.)
There are four groups of entities which have a cost allocation from Suburban's parent: (1) Suburban; (2) 11 affiliated Texas Utilities; (3) five affiliated Southeast Utilities; and (4) 547 unaffiliated entities serviced under a contract. Thus there are 564 separate entities5 allocated costs and whose data is used by Suburban for its proposed four-factor allocation of parent costs.
As shown in Table 1, Suburban's California Operations have four widely varied factors, ranging from a high of 57.8% factor for customers to a low of only 13.5% for payroll, for an average of 32.7%. Table 2 shows Suburban's proposed derivation of allocation factors for all entities. The only entity or group with a larger allocation than Suburban's is the entire group of 547 unaffiliated entities serviced under a contract. The next largest single entity is Monarch Utilities with 16.4%.
Table 1 Suburban Proposed Test year 2012 Allocation Using Suburban 2009 Data | ||
Direct Operating Expense |
$27,221,789 |
18.2% |
Gross Plant |
$181,567,267 |
41.4% |
Customers |
75,392 |
57.8% |
Payroll |
$6,361,682 |
13.5% |
Weighted Average - Suburban |
32.7% | |
(Source: DRA Opening Brief at 10 citing Suburban's work papers.) | ||
8.2. Suburban's Position
Suburban argues that its proposed four-factor allocation of 32.7% is calculated consistently over time in compliance with the Allocation Practice and recent decisions in Suburban proceedings. Suburban states the only changes since the last general rate case are a reduction in the number of non-regulated customers served by affiliates (i.e., the unaffiliated entities served under contract). There has been little or no change in the operations of SouthWest (Texas) and the non-regulated affiliates, other than a contraction in the client base of non-regulated utilities, since the Commission issued its decision in the last general rate case. (Sub Opening Brief 15 - 17.) Suburban further argues that the contracts (at least the exemplars used in examining DRA's witness) do not include the number of customers served by the unaffiliated entities.
8.3. DRA's Position
DRA objects to the data used by Suburban in calculating the four-factor. DRA argues the gross plant data is incomplete and inconsistent because it does not reflect the gross plant of unaffiliated entities serviced under a contract. (DRA Opening Brief at 21.) DRA argues that without the correct data this calculation is skewed and therefore it excludes the gross plant factor.
DRA also objects to the customer count used by Suburban. For example, Suburban uses 75,392 for its own customer count, 27,110 customers for Monarch Utilities, the next largest affiliated utility in Texas, but uses only one customer for each of the 547 unaffiliated entities serviced under a contract. As a result, Suburban and Monarch Utilities are calculated to have 57.8% and 20.8% of all customers while each entity served under a contact is counted as one for each of the 547 separate entities. These 547 separately serviced entities are calculated by Suburban to be 0.4% of all customers. Through data requests and other research DRA believes the more accurate customer count for these 547 entities would be approximately 350,000 customers and it therefore recalculates the weighted factor using this figure. (DRA Opening Brief at 18.) DRA thus argues that the meaning of "customer" in the Allocation Practice must be consistent and the 547 unaffiliated entities serviced under a contract should not be counted as a single customer each when Suburban is counted as 75,392 customers.
Finally, DRA objects to Suburban's imputation of no payroll to three of the Southeast Utilities because Suburban states these companies are operated by other affiliates and do not have their own employees. DRA argues this "grossly skews the allocations toward other affiliates including Suburban." (DRA Opening Brief at 24 quoting Ex. DRA-1 at 8-49.)
8.4. Discussion
Most costs are directly allocated to an activity or event that precipitates the cost. This is the basic "matching concept" in generally accounting principles. For example, the cost of meter reading (labor, equipment, etc.,) is separately identified as an operating expense just as the cost of digging a new well (engineering, labor, equipment, materials, etc.,) is separately identified and recorded as plant in service. Other less direct costs are allocated on the most reasonable method. Administrative costs related to employees for example can be allocated based on the direct labor costs of different activities. In a similar fashion, the four-factor allocation method is intended to be the method of last resort, i.e., when no other allocation is more accurate.
We find Suburban's arguments and calculations to be unreasonable because its interpretations applied to the plant and customer factors are not credible. In order to include any data in a calculation the Commission must have complete faith in the truthfulness and accuracy of the data and cannot rely not a selective reading of the Allocation Practice.
Most problematic is the customer count. It is not likely that Suburban receives 75,000 times the service received by each one of the unaffiliated entities who are counted as a single customer. Thus Suburban's proposed customer count calculation is not a credible allocation factor.
DRA objected in the prior proceeding but was not found persuasive at that time. Therefore this time DRA tried to determine the more likely number of customers who are served by the 547 unaffiliated entities and derived an estimate of 350,000 customers. But it is not DRA's burden to find and develop the data. Suburban's burden of proof is to provide the whole truth in a fair and consistent manner. Counting each contract as a single customer, while counting Suburban as over 75,000 customers, is not credible, therefore it is not fair or reasonable. To the extent that Suburban has shown there is no need to include the number of customers in the individual contract, we conclude that if the number of customers is irrelevant to the services provided to the unaffiliated entities then the number of customers (either 1 or 75,000) is an irrelevant factor for allocation purposes.
We could simply treat Suburban as one of 564 (sum of all entities) which would result in an allocation of 0.0017%, which is effectively zero. Suburban argues that the parent company, "SouthWest, has a legal relationship" with 547 client non-affiliated entities. (Transcript at 130 - 131.) And that Suburban has a legal relationship with 75,000 customers. Both are true. But they are not the same thing. SouthWest, whose costs are allocated to Suburban and all the other entities affiliated or unaffiliated, does not serve 57,000 customers: Suburban does. SouthWest only has a single entity relationship with Suburban. By a sleight of hand the witness attributes Suburban's customers to SouthWest when SouthWest directly serves only the single entity of Suburban, not Suburban's 75,000 customers - who are directly served only by Suburban. This is misleading and we find Suburban's testimony and argument unpersuasive.
By comparison, while the 547 unaffiliated entities are only attributed 0.4% of customers, they constitute 63.8% of direct operating expenses. We note that Suburban seems to have direct operating expense data while it claims in data responses and in evidentiary hearings that it does not have customer data. It is not likely that the unaffiliated entities group, which has 63.8% of total operating expenses, would only serve 0.4% of the total customers. We find Suburban not to be credible on this point.
We also find that Suburban's calculations for gross plant are not reliable. As with customers, Suburban is not reasonable in its application of the Allocation Practice. It is trying to rely on prior decisions which allowed the use of a four-factor method. We are not bound by those prior decisions: we are bound to use the evidentiary record before us today.
DRA persevered in this proceeding and demonstrated the one necessary point: the data offered by Suburban is not consistent for all entities. Because the data is not consistent any proposed allocation based on that data is not fair.
We find DRA may also be correct that three of the companies do not have any employees. When we look at the financial impact, however, we find that any allocation of employees/payroll to these three companies does not materially alter the remaining allocation compared to using the two factors of payroll and direct operating expense. The operating expenses for the three Southeast Utilities with no payroll are each 0.1% of the total. We therefore need not make an adjustment for personnel because it would be immaterial.
We will not use DRA's alternative calculations of a four-factor. Although DRA tried to correct the inconsistent data offered by Suburban we find it simpler and more accurate this time to use only two factors, gross operating expense and payroll. Using Suburban's data in Tables 1 and 2 the adopted allocation is the average of 18.2% for gross operating expense and 13.5% for payroll, for an allocation factor of 15.85%.6
In the next general rate case Suburban may either use these two factors (with accurate and consistent data) or the preferred four-factor allocation, but only if it can conclusively demonstrate the data is accurate and consistent across all entities.
5 (1 + 11 + 5 + 547 = 564.)
6 (13.5% + 18.2%)/2 = 15.85%.)
