10. Customer Accounts Expenses

For the non labor portion of four accounts -- Account 902, meter reading; Account 903.200, credit; Account 903.500, billing; and Account 903.800, call center, SCE used a three-year trend of 2001-2003 data to forecast test year expenses. SCE expects non-labor costs to increase at a faster rate than customer growth. For the non-labor portion of these accounts, SCE's test year forecasts amount to $30,291,000.

ORA opposes the use of the three-year trend noting that while SCE provides five years of data, it only trends accounts where recent three-years of data show an upward trend. Where costs are declining over the last 3-5 years, SCE did not use trending of non-labor costs but instead increased 2003 costs by the 2003-2006 customer growth. Rather than trending, ORA used the customer growth method for the non labor portion of theses four accounts, as both it and SCE did for the labor costs portion of these accounts. For the non-labor portion of these accounts, ORA's test year forecasts amount to $26,422,000.

For Account 903.200, Aglet supports ORA's adjustment pointing out that r-squared value of 0.81 for three data points is unimpressive. Aglet also notes SCE testimony that states credit related activities were lower than normal in 2001, the first year in SCE's trend, and the effects of the 2000 - 2001 financial crisis stabilized in 2002, the second year in SCE's trend.

Use of a trend based on three-years of data is suspect in that it has a very limited number of data points and minor variations in any of the years could cause wide variations in the projection of the trend. This potential problem is magnified when considering the time frame of the recorded data used in SCE's trend. In general, due to the 2000-2001 energy crisis, SCE significantly reduced its 2001 expenditures when compared to 2000 costs. 2002 and 2003 were recovery years and the associated increases over 2001 may not be indicative of normal growth in costs. In its testimony, SCE provided reasons for using the three-year trend. We will consider those explanations, the recorded data36 and our concerns regarding three-year trending of 2001 to 2003 data in addressing each of the accounts at issue.

From 1999 to 2003, the recorded adjusted non-labor expense for this account increased from $5,819,000 to $10,139,000, or 74%. Expenses increased for each of the years during that timeframe. SCE's test year 2006 request of $11,661,000 is 15% higher than the 2003 recorded amount. SCE explains that the primary driver for the non-labor trend in this account is vehicle related costs, including fuel. Based on this information, we find SCE's requested increase to be reasonable and will adopt $11,661,000 for the non-labor portion of this account.

From 1999 to 2003, the recorded adjusted non-labor expense for this account decreased from $4,501,000 to $4,162,000, or 8%. Expenses fluctuated during that timeframe, with the lowest amount being incurred in 2001. SCE's test year 2006 request of $5,157,000 is 24% higher than the 2003 recorded amount. SCE explains that the primary drivers for the non-labor trend in this account are vehicle related costs and IT support. This is not demonstrated by the decrease in costs from 1999 to 2003. Also, our concerns regarding three-year trending of 2001 to 2003 data are applicable here. The first point in the trend, $3,429,000 for 2001, is the lowest of any of the recorded amounts and appears to have been affected by the energy crisis. ORA's test year forecast of $4,350,000, based on customer growth, is more reasonable and will be adopted.

From 1999 to 2003, the recorded adjusted non-labor expense for this account increased from $2,950,000 to $3,434,000, or 16%. Expenses fluctuated during that timeframe, with the lowest amount being incurred in 2001. SCE's test year 2006 request of $5,412,000 is 58% higher than the 2003 recorded amount. SCE explains that the primary driver for the non-labor trend is the automation of systems in recent years due to regulatory mandates. There may be merit in SCE's explanation, but the recorded data does not support its request. Again, our concerns regarding three-year trending of 2001 to 2003 data are applicable here. The first point in the trend, $2,036,000 for 2001, is the lowest of any of the recorded amounts and appears to have been affected by the energy crisis. ORA's test year forecast of $3,680,000, based on customer growth, is more reasonable and will be adopted.

From 1999 to 2003, the recorded adjusted non-labor expense for this account increased from $8,141,000 to $8,359,000, or 3%. However, expenses during this timeframe fluctuated from a low of $7,778,000 in 2001 to a high of $10,433,000 in 2000. SCE's test year 2006 request of $8,061,000 is 4% lower than the 2003 recorded amount. SCE explains that the primary driver for the non-labor trend in this account is new automated systems. Also outsourcing has caused non-labor increases while providing labor decreases. Because of the wide variance in recorded costs, the non-labor portion of this account does not appear to directly relate to customer growth. While we have concerns with the use of the three-year trend, SCE's test year estimate of $8,061,000 is low compared to the 2003 recorded adjusted amount of $8,359,000 or the 5 year (1999 - 2003) averaged amount of $8,540,000. Based on this information, we find SCE's requested increase to be reasonable and will adopt $8,061,000 for the non-labor portion of this account.

In its update testimony, SCE identified a U. S. Postal Service requested postage rate increase of 5.4% to be effective as early as January 2006. On November 1, 2005, the U. S. Postal Commission issued its decision recommending the adoption of the postage increase. On November 14, 2005, the Postal Service Board of Governors voted to approve the Postal Rate Commission's recommendation, with an effective date of January 8, 2006. It is reasonable to reflect this known change in postage rates in the calculation of the forecasted test year postage expense. The test year forecasted postage expense will therefore be increased by $1,018,000, for a total of $20,233,000.

For Customer Service Application Services, SCE used a three-year trend method to forecast both labor and non-labor expenses for Information Technology Application Services. In support of its trend analysis, SCE states it has provided evidence that growth for this activity will increase at a rate greater than customer growth. SCE's test year forecast for this account is $22,600,000.

ORA recommends the use of 2003 recorded expense, $18,468,000, escalated by 4.53% customer growth to forecast the test year expense. ORA's forecast for this account is $19,304,000.

Our concerns regarding the trending of 2001 to 2003 data, as expressed in the previous section concerning three-year trends of non-labor costs, apply here also. The pattern of reduced spending in 2001 exists for this account also. Recorded 2001 amounted to $15,776,000, the lowest of any of the recorded years. Although SCE has provided explanations for reductions in costs from 1999 to 2001 and increased costs from 2001 to 2003, many of the explanations are not quantified. It is difficult to determine whether the energy and financial crisis of 2000 - 2001 had any effect on spending patterns during the 2001 - 2003 timeframe that was used by SCE for trending purposes.

As mentioned previously, SCE gave explanations of the increased costs during the 2001 - 2003 timeframe. The $1,100,000 labor increase from 2001 to 2002 was due to two factors: transfer of employees from the eBusiness operational unit to IT Application Services for CSBU and full implementation of the Call Workflow Optimization/Computer Telephony Integration project. The $1,000,000 labor increase from 2002 to 2003 was due to: Client-driven small enhancement requests, full implementation of three projects and realignment between the Transmission and Distribution and Customer Services business units. For non-labor, SCE indicates that expenses remained relatively constant for this account from 2001 to 2003.

SCE explains the 2003 to 2006 increases for this account relate to anticipated regulatory driven initiatives.37 While regulatory driven costs recorded during the 2001 to 2003 timeframe have not been quantified, certainly some of the costs incurred during that timeframe relate to regulatory driven initiatives. However, other historic costs such as realignment of business units and transfer of employees do not, and no reason has been given to justify trending the effect of these costs to the test year. Also, we do not know whether future regulatory initiatives will increase the need for funding in this account or whether supplemental funding will be provided as part of the initiative. In short, we do not feel comfortable in determining that the three-year trend of 2001 to 2003 recorded data reflects costs that should be extrapolated to 2006 to reflect appropriate future regulatory driven initiatives.

In rebuttal, SCE also attributes increases in test year expenses to systems applications to support new customer-related initiatives and adequate maintenance of existing applications, as well as an increase in work requests resulting from regulatory requirements.38 In addressing efficiency concerns, SCE states that IT applications, by their nature, usually result in productivity savings and refers to Table I-1 in Exhibit 50, which lists historical and future productivity savings in customer service operations. Associated productivity is reflected in other FERC accounts, as was the case for the Authorized Payment Agency project and the Meter Process Automation project. The productivity effect of completed IT applications appears to be reflected in the test year results of operations. However, productivity related to undefined applications developed within the three-year trend does not appear to be reflected in SCE's forecasts. While SCE is requesting expenses related to future IT projects, it does not appear that its showing reflects potential productivity savings related to such projects.

Given our concerns regarding trending, quantification of regulatory impacts, and productivity, we will not adopt SCE's request for this account. We will instead adopt ORA's customer growth methodology, which incorporates recorded increases through 2003 and allows for 4.53% additional growth through the test year. The adopted test year forecast is $19,304,000.

SCE's revised forecast for the uncollectible factor for the test year is 0.278%. This is the sum of the 1999 - 2003 five-year average of 0.266%, plus an additional 0.007% for uncollectible other operating revenue, plus an additional 0.005% for a PROACT adjustment. The PROACT adjustment reflects the impacts of the PROACT-related redistribution of revenues between residential and non-residential customers. SCE's revised forecast reflects agreement with Aglet's PROACT calculation resulting in a modification of the PROACT adjustment from 0.015% to 0.005%.

ORA recommends an uncollectible factor of 0.2708%. This is the sum of the 1999 - 2003 five-year average of 0.266%, plus 0.004% for the late payment charge and approximately 0.001% for the field assignment charge. ORA opposes the PROACT adjustment.

Aglet recommends an uncollectible factor of 0.220%. Aglet averaged the last two years, 2002 and 2003, equaling 0.215%, in developing its uncollectible recommendation. In support of its forecast, Aglet states that the 2004 unadjusted recorded uncollectible factor was 0.212%. Aglet also added 0.004% for the late payment charge and 0.001% for the field assignment charge, as proposed by ORA. Aglet reviewed SCE's calculation of the PROACT adjustment, and while opposing it, calculated that the maximum upward adjustment would be 0.005% rather than the 0.015% requested by SCE.

In SCE's last GRC, the Commission adopted an uncollectible rate of 0.319% based on a five-year (1996 - 2000) average as proposed by SCE. While that factor is significantly higher than the 2003 recorded adjusted factor of 0.202 and the 2004 unadjusted recorded factor of 0.212%, based on the data that indicated a rise in the recorded factor from 0.283% in 1996 to 0.348% in 1999 and a decline to 0.311% in 2000, an average appeared to be appropriate and was adopted.

In this proceeding, a five-year average does not appear to be appropriate. The recorded adjusted uncollectible factor has declined from 0.348% in 1999, to 0.311% in 2000, to 0.242% in 2001, to 0.227% in 2002, and to 0.202% in 2003. Aglet calculates the correlation coefficient to be 0.946, meaning that 94.6%of the variability of the results can be explained by the trend over time. We do not anticipate that decline will continue, since recorded information for 2004 indicates that the decline may be flattening out. Aglet's proposal to average the 2002 and 2003 recorded uncollectible factors is reasonable and will be adopted.

SCE asserts that a major reason for the recent downward trend in uncollectibles is the drop in mortgage interest rates and the boom in refinancing, which has increased customers' disposable income and their ability to pay. However, SCE has not demonstrated the relationship between ratepayers who refinance and ratepayers who are unable or decline to pay their utility bills. Also, while SCE relates uncollectibles to the Federal Funds Rate over the 1999 - 2003 timeframe, the correlation coefficient that indicates that as much as 75% of the historical variation in uncollectible is explained by this relationship is significantly less than the correlation coefficient that indicates 94.6% of the variability of uncollectibles, over the same timeframe, can be explained by just a time trend.

Also, Aglet asserts, and SCE disputes, that the decline in uncollectibles is caused by the impacts of SCE's credit and collections actions. The record is insufficient to quantify the effects of SCE's credit and collections actions and relate the effects to declining or rising uncollectible costs. Interest rates, unemployment, or the economy in general may also have effects on uncollectible expense. More than likely, uncollectibles are affected by a combination of these and other factors. Without more convincing evidence, it is reasonable to evaluate the data over time and consider averaging, trending or use of last recorded year, similar to our evaluation of other expense items.

Regarding the PROACT adjustment, SCE reasonably argues that PROACT changed the balance between residential and commercial and industrial customers in the percentages of SCE's revenue they each represent. We will reflect the 0.005% adjustment as calculated by Aglet and agreed to by SCE.

SCE, ORA and Aglet agree that an increase of 0.004% for the late payment charge is reasonable and it will be included in the adopted uncollectible rate. Also, based on our resolution, in this decision, of the field collection charge, the associated effect on uncollectibles of 0.001% will also be included in the adopted factor. The adopted uncollectible factor is therefore 0.225%.

For Market Research and Communications, SCE forecasts incremental test year expenses of $1,570,000. The request would fund four programs: Residential Services and Outreach for $464,000, In Language Communications for $275,000, Customer Process Based Satisfaction Survey for $431,000, and Internet Improvements for $400,000. SCE states the incremental increase in funding for these programs will result in improved customer service delivery, increased utility program customer participation and avoided cost savings, such as for postage.

ORA supports the Language Communications activities for $275,000, but does not support funding for the remaining three programs. As discussed below, we also adopt increases of $464,000 for Residential Services and Outreach and $200,000 for Internet Improvements. In total the test year increases amount to $939,000, as opposed to SCE's request of $1,570,000.

SCE proposes increased spending of $464,000 to help it more effectively provide basic customer services to residential customers. SCE plans to conduct market research to determine preferences for basic customer care programs and tell it how residential customers want to interact with the company. SCE could then determine if current programs and service should be enhanced or discontinue or if new offerings are needed. SCE also plans to broaden its outreach campaigns to help customers become more aware of various programs and service they find valuable and provide information about important issues that impact them.

ORA is concerned that the benefits to customers are not commensurate with the costs. One of the identified benefits of the program is a $64,200 reduction in postage related to customers switching to online billing. While ORA indicates this is far short of the $434,000 spending request, we believe some programs may be justified for reasons other than cost/benefit. In general, the purpose of the program, which is to provide better basic service to residential customers, appears to be appropriate. Assuming some savings such as for postage, the net cost to customers is not substantial. We see overall benefit for this program and will include it in the adopted expense for this account.

SCE plans to create a company-wide method for assessing customer satisfaction with the service provided by SCE from an end-to-end process perspective, rather than simply measuring individual transactions. ORA believes SCE should more efficiently utilize its current budget which ORA characterizes as quite substantial.

We will deny SCE's requested funding for this item. We are not convinced that there is an urgent need to change the method for assessing customer satisfaction, as it relates to customer service operations. While the new method would be more comprehensive, SCE has not explained any problems with the current method other than that the overall satisfaction with SCE as a company remains lower. There does not appear to be substantial dissatisfaction with customer service operations. Dissatisfaction with SCE as a company may be the result of many things, such as high rates, that are not customer service related. Changing the method for determining customer satisfaction related to customer service operations may not be the relevant course of action. If truly desirable, SCE may be able to transition into the new method, at a reduced scale, using existing funding levels. It can be expanded in the future based on analysis of achieved benefits, monetary or otherwise.

SCE requests an increase of about $800,000 (split between accounts 905 and 908) to fully fund its internet design and development function. SCE's last major modification of the website occurred in 2000. SCE indicates that since that time business conducted over the internet has increased and customer expectations for conducting business electronically have continued to evolve. SCE adds that research has also established that customers expect self-service functionality from websites.

ORA opposes SCE's request, characterizing it as more of a wish list with highly speculative savings.

SCE provides a reasonable rationale for expanding the functionality of its website. Use of the internet continues grow and evolve. However, while indicating that it recorded approximately $580,000 for this function in 2003 and that it needs an additional $800,000, SCE's testimony does not provide any information that shows how the approximate $1,380,000 will be spent. Rather than totally deny the additional funding, we will include 50% of SCE's incremental request, or $400,000, split evenly between Accounts 905 and 908. We include the increased funds to recognize the value to customers of expanded website capabilities, as generally described in SCE's direct and rebuttal testimony.

TURN asserts that the current pool of Direct Access (DA) customers can decrease, but it cannot increase under the Commission's current rulings. Therefore, there should be no growth in DA related costs for incremental expenses derived as a result of customer growth or three-year trends. TURN recommends that the overall forecast of 2006 DA costs should be capped at the 2003 recorded/adjusted level with no increase to 2006.

SCE does not agree with TURN's proposal. SCE states that DA related costs are now part of SCE's budgets in each operating area (e.g., billing) and are no longer separately tracked. Also because some functions may have relatively fixed costs at any given point in time and other functions' costs will increase faster than customer growth, the average of all functions' costs will increase at around the rate of customer growth. Lastly, SCE states that DA related costs are continuing to increase due to the complexity and uncertainty relating to current DA issues, such as switching rules, load growth rules, relocation rules, and DA cost responsibility.

10.10 Discussion

We will not adopt TURN's adjustment for this proceeding. First of all, for costs that are driven by customer growth, we are reluctant to impose TURN's proposal without a better understanding of how the customer growth adjustment is developed. It is not clear whether or not DA customers were included in the analysis of customers used to determine customer growth. If they are, the overall customer growth rate derived for adjusting expenses would appear to be appropriate. In that case total costs for DA and non-DA customers would reflect the correct overall customer growth. If direct access customers were excluded from the customer growth analysis, the derived customer growth may be too high since there should be no growth related to DA customers. In that case the total costs may be overstated. However, even if DA customers were excluded from the customer growth analysis, expenses in Accounts 901 - 903 may increase for reasons other than customer growth and it may be appropriate for DA customers to pay their share of those increased costs. Because DA costs are no longer tracked, the recorded 2003 DA related costs are not part of this proceeding's record. For the reasons given by SCE, determining the appropriate separate DA related costs and capping those costs would be difficult at best. We also note that the proposed adjustments are not large. For example, TURN proposes a $38,000 adjustment for meter reading, in contrast to SCE's approximate request of $40,000,000 for that account.

SCE's current discretionary DA service fees contained in SCE Rate Schedules CC-DASF and ESP-DSF were implemented in 1999. According to TURN, while SCE's other labor and non-labor costs have increased with inflation, its DA service fees have not. TURN therefore proposes a 25% increase in discretionary DA service fees to account for inflation. TURN states that inflation rates for 1999 - 2006 customer accounts labor is forecast to increase by approximately 29% and distribution labor is forecast to increase by approximately 28%. TURN's proposed adjustment results in an increase to other electric revenues amounting to $227,000 for the test year.

SCE states that TURN's proposal to increase DA service fees by 25% is arbitrary. SCE indicates that it is currently in the process of reviewing the costs related these rate schedules and anticipates filing an application with the Commission to update the associated fees. SCE believes that it would be most fair to all affected parties to make such changes through a separate, properly noticed proceeding specifically for that purpose.

TURN's proposal to update the DA service fees to reflect inflation from 1999 to 2006 is reasonable. SCE has not provided any reasons to not do so other than to argue that it would be more appropriate to address this issue in a separate proceeding. While we agree with SCE that it would appropriate to consider changes to the rate schedules in a separate proceeding, it is not clear when SCE will file its application or when any changes will become effective. In the meantime, assuming that some increased cost over the 1999 level is appropriate, all other customers will be subsidizing the direct access customers. Also, such subsidization may have occurred historically and may be occurring currently, for the same reason. Therefore, rather than waiting for the results from a proceeding that has not even started, we will reflect, for GRC purposes, TURN's proposed inflation adjustment to reflect a 25% increase in discretionary DA service fees in 2006. Until the Commission issues a decision on SCE's anticipated application regarding DA service fees, the proposed 25% increase is reasonable when considering historic escalation.

36 The record evidence in this proceeding does not provide the specific data that was trended. "Recorded adjusted" data contained in SCE's testimony was apparently adjusted further for productivity before trending. We will however examine the "adjusted recorded" data in determining the reasonableness of the trends.

37 See Exhibit 56, page 101.

38 See Exhibit 99, pages 19 to 22.

Previous PageTop Of PageNext PageGo To First Page