The Following Testimony, by Linette Young, dated July 19, 2002, can be sumarized by the her last statement:
" It is my conclusion that Pacific has provided misleading and inaccurate service quality data and reports to the CPUC and to the FCC. This Commission must not tolerate this behavior from any utility it regulates. Only an exhaustive and independent audit of Pacific’s service quality reporting can determine if Pacific’s inaccurate reporting is deliberate."
BEFORE THE PUBLIC UTILITIES COMMISSION OF THE STATE OF CALIFORNIA
REPLY TESTIMONY OF LINETTE YOUNG
(SBC PACIFIC BELL)
OFFICE OF RATEPAYERS ADVOCATES
R.01-09-001/I.01-09-002: PHASE 2B
Date: July 19, 2002
Q.1. What is the purpose of your reply testimony?
A.1. The purpose of my reply testimony is to refute the testimonies of Professor John Hauser, Mr. Rick Resnick, and Dr. Robert Harris as they pertain to SBC Pacific Bell’s (Pacific) service quality as related to installations.
Q.2. Please summarize your findings in as regarding their testimonies.
A.2. Professor Hauser, Mr. Resnick, and Dr. Harris base their conclusions about Pacific’s installation service quality on Pacific’s erroneous and self-reported results in GO-133B filings and ARMIS 43-05 reports. The three areas covered by these witnesses include reported held orders in GO-133B filings, Commitments Met in both GO-133B and ARMIS 43-05 filings, and Average Installation Intervals as reported in ARMIS 43-05 filings. The reported results in these three categories are questionable and inaccurate, as revealed by ORA’s analysis of Pacific’s self-reported service quality results and the raw installation data Pacific provided ORA in response to data requests.
Q.3. Harris states that “held primary service orders have decreased and that ‘[b]eginning in 1998, Pacific has had no more than three Held Primary orders in any single year .” Is this an accurate statement about held primary service orders?
A.3. Unfortunately, no. Based on its analysis of a mere sampling ofCAB complaint data, ORA has shown that there were more held orders (33) than Pacific reported to the Commission in its GO-133B filings . ORA believes these 33 unreported held orders represent the proverbial “tip of the iceberg.” ORA found that Pacific’s method of counting held orders, the use of inaccurate commitment dates, and the use of inaccurate completion dates results in self-reported performance reports that makes Pacific’s installation service quality appear better than it actually is. Furthermore, the results of ORA’s statistical analysis has shown that Pacific ranks very low compared to other SBC companies on held orders. This raises further questions about Pacific’s accuracy in reporting held orders in GO-133B filings.
Q.4. Pacific’s witnesses have stated that Pacific’s record of met commitments (Installation-Line Energizing Commitments) in GO-133B filings and ARMIS 43-05 (% Commitments Met) reports demonstrate that Pacific “has consistently provided outstanding performance in meeting it’s GO-133B Installation – Line Energizing Commitments (Commitments Met) requirements on a company-wide basis .” Do these self-reported results accurately reflect Pacific’s performance in meeting installation commitments?
A.4. No. First of all, the ARMIS and GO-133B met commitments do not measure exclusively commitments met for orders for basic service, which is defined as including an order for an inward access line. Both the ARMIS and GO-133B measures also include orders for vertical services, such as Caller-ID, which are normally provisioned within the same day that they are ordered. The inclusion of these services, when measuring the percentage of commitments met, gives a misleading picture of a carrier’s performance in the provisioning of basic service.
ORA calculated the percentages of met commitments for basic service installations for the years 1999, 2000, and the first six months of 2001. The calculation was based on an interval that was derived by subtracting the completion date from the commitment date. An interval greater than or equal to 1 day was considered a missed commitment. The data sets were used after duplicate records, outliers, and the anomalous records containing no commitment dates removed. ORA found that for these years the percentages of met commitments fell far below the GO-133B 95% benchmark. In 1999, 2000, and 2001, Pacific met only 90.8%, 90.8 %, and 89.6% of its commitments for installation of basic service. The actual performance may be even lower as ORA discovered evidence of multiple records of orders for the same installation; these result in the appearance of met commitments and completed orders while, in fact, the customer waits for telephone installation.
Q.5. In his Opening Testimony, Mr. Resnick states, “Both commitments met and installation intervals continued to improve particularly during the past 2 years .” Dr. Harris, in his testimonystates, “ What is more remarkable, though, is that installation intervals have fallen since then, in spite of the continuing increase in the number of installation orders .” What factors have contributed to this “remarkable” improvement in Pacific’s average installation interval?
A.5. The table below illustrates some of the factors that are contributing to Pacific’s decreased average installation interval as reported to ARMIS. The left side of the table show’s Pacific’s performance in provisioning basic telephone service for the measures: average installation interval, percent of commitments met, and orders for basic service completed in less than 5 working days.
Table 1: Pacific’s Performance in the Provisioning of Basic Service and the Data Irregularities Contained in the Pacific Installation Data
This table shows that Pacific’s performance in provisioning basic telephone service has essentially remained the same in the years 1999 and 2000 with some marginal improvement shown in the first six months of 2001. If Pacific’s performance is relatively unchanged, what is accounting for the “remarkable” improvement that Harris mentioned? It appears that the data irregularities found in Pacific’s data are contributing substantially to Pacific’s “remarkable” performance. In 1999 the number of anomalous records, duplicate records, and multiple records for the same order were small enough not to affect Pacific’s overall installation interval. However, in the two years that Mr. Resnick notes as showing improvement, the percentages of these irregularities increased enough to affect the overall installation interval in Pacific’s favor .
While ORA acknowledges that decreased orders for inward access lines and vertical services orders may improve the overall installation interval, the fact remains that the provisioning of basic telephone service to ratepayers is not improving over time.
Q.6. Mr. Resnick asserts that the SBC/Ameritech merger service quality data filed with the FCC “should not be used for year-over-year comparative purposes. As with the ARMIS reports, there are no standards associated with the FCC’s Service Quality Report.” Is this accurate?
A.6. Absolutely not. MCOT service quality data does indeed have standards for reporting service quality data and record retention. The NARUC Service Quality White Paper and subsequent auditors reports set the business rules for SBC’s reporting responsibilities and record retention requirements . Pacific’s initial set of filings for the period July 1999 through June 2000 are imperfect, as demonstrated by the auditor’s report, but they are still a very useful tool for analyzing Pacific’s performance in certain measures .
ORA’s analysis of MCOT data was limited to 3 line items: Total Orders for Installation, Installation Orders Held Over 30 Days, and Orders Completed Within 5 Working Days. ORA’s statistical analysis of held order performance utilized the ratio of Installation Orders Held Over 30 Days divided by Total Orders for Installation. Since all SBC companies reported total orders in the same manner, this line item is a useful comparative number for deriving the ratio. Since Pacific Bell, Nevada Bell, and Southwestern Bell undercounted held orders during the FCC’s evaluation period, a recount of those companies held orders would only serve to make these three companies’ service appear to be worse than was initially reported. Since Pacific Bell and Nevada Bell were the two worst performers in the statistical analysis, an increase in the number of reported held orders should not change the ranking of the companies. The percentage of Orders Completed Within 5 Working days was used for comparative purposes from year-to-year for Pacific only.
On January 4, 2002, SBC refiled its MCOT reports for the period 2000 and 2001. An examination of this data received from the Enforcement Division of the FCC shows that Pacific is not able to produce installation data for the months of January through September of 2000. The January through September reports appear to have been filed earlier on November 12, 2001. Not only is this lack of installation data a direct violation of the FCC’s record retention requirement, it appears Pacific may be attempting to mislead the FCC with this report. Pacific recently provided ORA with disaggregated data for this same time period. As the data provided to ORA is from the same database that is used to produce the MCOT reports, this data should have been reported to the FCC.
Q.7. After reviewing the Opening Testimonies of Pacific’s witnesses, and considering ORA’s subsequent analysis what are your conclusions?
A.7. It is my conclusion that Pacific has provided misleading and inaccurate service quality data and reports to the CPUC and to the FCC. This Commission must not tolerate this behavior from any utility it regulates. Only an exhaustive and independent audit of Pacific’s service quality reporting can determine if Pacific’s inaccurate reporting is deliberate.
Q.8. Does this conclude your testimony at this time?
See Ernst and Young “Report of Independent Accountants”, Attachment A p.1-3., dated January 3, 2001. The auditors noted that all companies were including all new (N), transfer (T) and change (C) orders, including orders for vertical services, in these report items during the evaluation period. SBC agreed to modify it’s processes to exclude all orders for vertical services from the report items. The auditors also “obtained different results in total on a disaggregated basis” from Pacific Bell for Orders for Basic Service, Orders Completed within 5 Working Days, Missed Installation Commitments, Out of Service Over 24 Hours. (Attachment A. p2). Furthermore, the auditors identified Pacific as undercounting Installation Orders Delayed Over 30 Days.
more body text.