COM/LYN/CXW/epg ALTERNATE DRAFT Agenda ID #2677

Alternate to Agenda #1866

Ratesetting

Decision ALTERNATE PROPOSED DECISION OF COMMISSIONERS LYNCH AND WOOD (Mailed 9/4/03)

BEFORE THE PUBLIC UTILITIES COMMISSION OF THE STATE OF CALIFORNIA

Commission's Own Motion to Assess and Revise the New Regulatory Framework for Pacific Bell and Verizon California Incorporated.

Rulemaking 01-09-001

(Filed September 6, 2001)

Order Instituting Investigation on the Commission's Own Motion to Assess and Revise the New Regulatory Framework for Pacific Bell and Verizon California Incorporated

Investigation 01-09-002

(Filed September 6, 2001)

(See Appendix A For List of Appearances)

INTERIM OPINION REGARDING PHASE 2B ISSUES
SERVICE QUALITY OF PACIFIC BELL AND
VERIZON CALIFORNIA, INC.

Table of Contents

Title Page

INTERIM OPINION REGARDING PHASE 2B ISSUES SERVICE QUALITY OF PACIFIC BELL AND VERIZON CALIFORNIA, INC. 1 1

I. Summary 2 2

A. Introduction: Major Finding and Scope of Study 2 2

B. Verizon Generally Met Commission Standards and Bettered Its Peers on Most Measures 3 3

C. Pacific Generally Met Commission Standards and Often Bettered Its Peers, But Had Some Significant Problem Areas 4 4

D. Trend Analysis: Under NRF, Verizon's Service Quality Has Generally Improved; Pacific's Quality Has Improved Somewhat For Business Customers, But Has Remained Stable or Deteriorated for Residential Customers 5 5

E. Other Information Consistent with Quantifiable Data 9 9

F. Areas for Improvement 10 10

II. Scope of This Phase and Methodology 10 10

III. California Measures of Service Quality and Standards 15 15

A. GO 133-B Measures, Standards, and Caveats 15 15

1. Definition of "Primary Line" is Unclear 18 18

2. Automated Response Units (ARU) 18 18

3. Busy or Abandoned Calls Not Counted Under GO 133-B 20 20

4. Commission Recognizes Need to Revise GO 133-B 20 20

B. Performance of Pacific and Verizon Against GO 133-B Measures 21 21

1. Held Primary Service Orders 21 21

2. Installation-Line Energizing Commitments 27 27

3. Customer Trouble Reports 29 29

4. Toll Operator Answering Time 31 31

5. Directory Assistance Operator Answering Time 34 34

6. Trouble Service Answering Time 36 36

7. Business Office Answering Time (BOAT) 39 39

C. Summary of Empirical Assessment of Pacific's and Verizon's Performance on GO 133-B Measures 46 46

IV. Federal Measures of Service Quality - ARMIS and MCOT Data 47 47

A. ARMIS Measures 47 47

B. Accuracy of Data 48 48

1. General Issues with Pacific's Data 48 48

2. Pacific's Data Concerning Installation Orders Require Clarification 51 51

3. Allegation that Pacific's Reports Contain Erroneous Duplicate Records Has No Factual Basis 52 52

4. Allegation that Pacific's Reports Contain Erroneous "Anomalous Records" Has No Factual Basis 53 53

5. Verizon's Data Are Accurate 53 53

C. Comparison of ARMIS Results With Other Carriers 54 54

D. Summary Table of ARMIS 43-05 Measures 56 56

1. The Number of Initial Trouble Reports per 100 Lines. Pacific: Residential - No Trend, Business - Improving Trend; Verizon: Residential and Business - Improving Trend 63 63

2. The Number of Repeat Trouble Reports per 100 Lines. Pacific: Residence - No Trend, Business - Improving Trend; Verizon: Residence and Business - Improving Trend 66 66

3. The Number of Initial Out-of-Service Trouble Reports per 100 Lines. Pacific: Residence and Business - No Trend; Verizon: Residence - No Trend, Business - Improving Trend 69 69

4. The Number of Repeat Out-of-Service Trouble Reports per 100 Lines. Pacific: Residence - No Trend, Business - Improving Trend; Verizon - Residence and Business - No Trend 72 72

5. The Number of Subsequent Initial Trouble Reports and Subsequent Repeat Trouble Reports: Insufficient Observations 74 74

6. The Number of Initial All Other Trouble Reports per 100 Lines. Pacific: Residential - Declining Trend, Business - Improving Trend; Verizon: Residence - No Trend, Business - Improving Trend 75 75

7. The Number of Repeat All Other Trouble Reports per 100 Lines. Pacific: Residential - Declining Trend, Business - Improving Trend; Verizon: Residential - No Trend, Business- Improving Trend 77 77

8. Initial out of service repair interval (in hours). Pacific: Residential - Sharp Declining Trend followed by Improving Trend, Business - No Trend; Verizon: Residential and Business - No Trend 80 80

9. Pacific: Residential - Sharp Declining Trend followed by Improving Trend, Business - No Trend Verizon: Residential and Business - No Trend 84 84

10. Initial all other repair interval (in hours). Pacific: Residential - Sharp Declining Trend followed by Improving Trend, Business - Improving Trend; Verizon: Residential - Declining Trend; Business - No Trend 87 87

11. Repeat all other repair interval (in hours). Pacific: Residential - Sharp Declining Trend followed by Improving Trend, Business - No Trend; Verizon: Residential - Declining Trend, Business - No Trend 89 89

12. Average Installation Interval. Pacific: Residence and Business - No Trend; Verizon: Residence and Business - Large Deterioration in Service Followed by Even Larger Improvement 92 92

13. Switch Downtime. Pacific - No Trend, But Consistently Short Downtime; Verizon: Deteriorating Trend 96 96

14. Switches Down per Switch 98 98

15. Number of Switch "Occurrences" 98 98

16. Installation Commitments Met. Pacific: Residential - No Trend, Business - Deteriorating Trend; Verizon: Residential and Business - No Trend 99 99

E. Summary of Empirical Assessment of Pacific's and Verizon's Performance on ARMIS 43-05 Measures 102 102

F. MCOT Data 111 111

1. MCOT Data - Pacific Shows No Service Diminishment Following Amertech Merger 111 111

2. MCOT Data - Verizon California (GTE) Shows No Diminishment of Service Quality Following Merger 115 115

V. Survey Data and Customer Satisfaction 118 118

A. Customer Satisfaction and Service Quality Surveys - Pacific 118 118

1. ORA Survey 118 118

2. Pacific's Surveys 123 123

3. Summary 134 134

B. Customer Satisfaction and Service Quality Surveys - Verizon 135 135

1. ORA Survey 135 135

2. Verizon's Surveys 135 135

VI. Other Direct Measures of Service Quality 135 135

A. Informal Complaints for Pacific 136 136

B. Verizon Informal Complaints 141 141

C. Formal Complaints - Pacific 144 144

D. Formal Complaints - Verizon 150 150

VII. Other Issues In This Proceeding 151 151

A. Allegation by TURN that Pacific Inappropriately Aggregated Data 152 152

B. Technological Change and Effects on Pacific's Service Quality 152 152

C. Growth - Pacific 154 154

D. Staffing - Pacific 155 155

E. Weather - Pacific 156 156

F. Marketing - Pacific 159 159

G. Still Other Issues - Pacific 159 159

H. Movement of Functions to Unregulated Affiliates 160 160

I. Service Performance Guarantee - Verizon 162 162

J. Technological Change - Verizon 163 163

K. Growth - Verizon 163 163

L. Staffing - Verizon 163 163

M. Weather - Verizon 164 164

N. Marketing - Verizon 165 165

O. Mergers and Structural Changes - Verizon 165 165

VIII. NRF Incentives, Service Quality, and Competition 166 166

A. NRF Incentives and Service Quality, Positions of Parties 166 166

B. Discussion 168 168

C. Effect of Competition on Service Quality - Positions of Parties 169 169

D. Competition and Service Quality - Discussion 170 170

IX. Comments on Proposed Decision 172 172

X. Assignment of Proceeding 172 172

Findings of Fact 172 172

CONCLUSIONS OF LAW 208 208

ORDER 210 210

I. Summary

A. Introduction: Major Finding and Scope of Study

This proceeding has conducted a comprehensive investigation into the quality of telecommunications service offered to Californians by Verizon and Pacific under the New Regulatory Framework (NRF) mode of incentive regulation. We find that Verizon offered generally improving service during the period under study that compared favorably with other large carriers on most measures, with just a few areas of weakness. We find that Pacific's quality was stable for the majority of measures and compared favorably to other large carriers on many measures, but that Pacific exhibited several important areas of weakness, especially for residential customers.

This investigation assessed the performance of Verizon and Pacific in meeting the six California-adopted performance standards contained in General Order (GO) 133-B. In addition, we used standard statistical methods to analyze the trends in service quality for Verizon and Pacific under NRF regulation. The investigation also examines federal service quality data. Since there are no standards adopted by the Federal Communications Commission (FCC) for these service quality measures, we compare the performance of each company against a reference group of the ten largest national utilities (excluding Pacific but including Verizon-California). While we recognize that differences in data methodology among carriers may limit the comparability of the federal data among carriers, we include this analysis to provide context to the other information in the record and are mindful of its potential flaws. As with the California data, we also use statistical methods to determine the trends in service quality over the NRF period. In addition, the investigation reviewed survey data, regulatory proceedings, and informal complaint data to supplement the picture developed through our data analysis.

We note that there are limitations with virtually each of the various measures of service quality that we examine in this decision. We therefore endeavor to base our conclusions regarding service quality on the totality of information provided in the record.

B. Verizon Generally Met Commission Standards and Bettered Its Peers on Most Measures

1. GO 133-B

On the GO 133-B service quality measures, Verizon complied with four of the six service quality standards adopted by this Commission for all years covered in our study.¹ On the remaining two measures, trouble report service answer time and business office answer time, Verizon complied with the standard for most years, but failed to meet the standards for certain years in the early to mid-1990s. Verizon has complied with all six GO 133-B standards since 1998.

2. ARMIS

When evaluated on the Federal Communications Commission's (FCC) Automated Reporting Management Information System (ARMIS) service quality measures, we find that Verizon exceeded the performance of a reference group on eight measures for both residential and business lines, and on two measures for residential lines only. (See summary table below.) Verizon had statistically indistinguishable performance on two measures for both residential and business lines and on two measures for business lines, only. Based on our statistical analysis, Verizon did not fail to meet the performance of the reference group on any measure. Thus, on all significant Federal measures of service quality, Verizon met or exceeded the performance of the reference group of large utilities.

VERIZON ARMIS TO REFERENCE GROUP SUMMARY

Measure Type	Worse	Better	Inconclusive
Residential	0	10	2
Business	0	8	4
Bus. and Res.	0	0	1

C. Pacific Generally Met Commission Standards and Often Bettered Its Peers, But Had Some Significant Problem Areas

1. GO 133-B

On the GO 133-B service quality measures, Pacific complied with four of the six service quality standards adopted by this Commission for all years covered in our study. However, Pacific failed to meet the trouble report answer time standard for the first eight of the eleven years under review and failed to meet the business office answer time standard for three of the ten years under review. Pacific has complied with all six standards since 1999.

2. ARMIS

When evaluated on the FCC's ARMIS service quality measures, we find that Pacific exceeds the performance of a reference group on six measures for both residential and business lines (primarily trouble report measures), and on one measure for business lines only. (See summary table below.) Pacific had statistically indistinguishable performance on two measures for residential lines, and on five measures for business lines only. Pacific failed to meet the performance of the reference group on four measures for residential lines related to repair intervals.

PACIFIC ARMIS TO REFERENCE GROUP SUMMARY

Measure Type	Worse	Better	No Change
Residential	4	6	2
Business	0	7	5
Bus. and Res.	0	1	0

As these data suggest, Pacific had several areas of service weakness specific to residential services. Compared to the national reference group, Pacific had far fewer incidences of service trouble or outages, but once this occurred, Pacific was slower to resolve the trouble than its peers. Pacific had difficulty meeting answer time requirements, particularly for repair calls, but also for business office calls, a situation that caused the Commission in a 1997 decision to warn Pacific to improve its service. Pacific was also slow to answer customer billing queries, a service quality indicator not systematically measured and for which there is no current standard.

D. Trend Analysis: Under NRF, Verizon's Service Quality Has Generally Improved; Pacific's Quality Has Improved Somewhat For Business Customers, But Has Remained Stable or Deteriorated for Residential Customers

In addition to measuring the level of service for each company, we statistically examined how service changed during the years for which data is available for the period of NRF regulation. In particular, we sought to determine whether service quality had improved or deteriorated (or a mixture of both) over time under NRF regulation.

Concerning the effect of NRF regulation on service quality, we find it impossible to determine based on the record whether NRF itself caused a positive or negative effect upon service quality. Because data is only available during a portion of the NRF period, the data itself is insufficient to make a causal determination of service quality pre- and post-NRF. However, the data is sufficient to indicate the level and direction of service quality during the period under review.

We find that, in general, service has improved or remained stable during the NRF years, but not for Pacific's residential customers, who have experienced declining service for two measures (trouble reports) and significant periods of declining service for four other measures (repair intervals). In addition, Pacific's business office answer time statistics have deteriorated since 1998.

To reach this conclusion, we reviewed 7 GO 133-B measures and 16 ARMIS measures in this study to determine whether they showed a statistically significant increase or decline in service quality. Twelve of the ARMIS measures were examined separately for residential and business lines. This yields a total of 35 measures that we examined to determine if there significant trends in service quality over the NRF years.²

1. Pacific Performance Trends

With respect to the five GO 133-B measures for which we can make findings,³ all five show no statistically significant trend of improvement or decline.

With respect to the ARMIS measures, Pacific showed statistically significant improvement on 6 business measures and one combined business/residential measure during the NRF period; it showed no statistically significant change on 19 of them; and it showed statistically significant declines on 2 of the residential measures and one business measure. (See summary chart below.) Although they showed no statistically significant change, each of the four residential repair interval measures exhibited a similar trend, where repair intervals doubled over a four to five year period, followed by a gradual return almost to prior levels. In a 2001 decision, the Commission found that SBC's pattern of increasing residential repair intervals violated the Public Utilities Code and Commission requirements.

PACIFIC ARMIS TREND SUMMARY

Measure Type	Worse	Better	No Change
Residential	2	0	10
Business	1	6	5
Bus. and Res.	0	1	4

In sum, for Pacific, more business measures show improvement than show decline during the NRF period, and most business and residential measures show no significant change. However, the lack of any improvement for a single residential measure, the doubling of residential repair intervals, and the fact that the only two statistically significant changes in trend for residential service indicate deterioration does raise concern regarding residential service quality relative to business service quality during the NRF period.

2. Verizon Performance Trend

Verizon showed statistically significant improvement on 12 of the 35 total measures during the NRF period; it showed no statistically significant change on 19 of them, and it showed statistically significant declines on 4 of the measures. In contrast to Pacific, the disparity between the business and residential service is much less dramatic, especially since Verizon shows an improvement in both ARMIS business (5 measures) and residential services (2 measures). Thus, more variables showed improvement than showed decline. The only ARMIS measures showing deterioration were two residential repair interval measures and two combined business/residential measures. Of note is the steep deterioration in Verizon's switch downtime per downed switch measure. While both of Verizon's business and residential installation intervals data showed significant increases followed by decreases, the improvement in both measures exceeded the deterioration.

VERIZON ARMIS TREND SUMMARY

Measure Type	Worse	Better	No Change
Residential	2	2	8
Business	0	5	7
Bus. and Res.	2	2	0

3. Summary of Trend Analysis

These outcomes - greater improvement in business service quality relative to residential service, more overall improvements than declines in service quality, and better performance of Verizon than Pacific - are not possible to reconcile with the proposition that NRF caused either a systematic decline or improvement in service.

We expect the parties to present recommendations in Phase 3B of this proceeding concerning how to build on the record of generally stable or improving service quality produced under NRF and to improve on those areas of weakness in service quality exhibited during the NRF period.

E. Other Information Consistent with Quantifiable Data

As part of our investigation, we also reviewed survey data, informal complaint data, and formal Commission investigations of Pacific and Verizon. This information, which is more difficult to interpret quantitatively, presents a qualitative picture that supplements our statistical assessment. We encountered questions regarding the relevance or statistical reliability of most of the survey data related to Pacific. Customers are generally pleased with Verizon's service quality. Regarding complaint data filed with the Commission, Verizon's informal complaint numbers are proportionately lower than Pacific's.

The Commission has aggressively pursued lapses in service quality or marketing standards and for Pacific has seen an increased pace of formal proceedings finding violations of service quality requirements since 1995. Though formal enforcement proceedings have been necessary to correct problems, such proceedings are resource intensive for the Commission and parties and inevitably require 12 months or longer to resolve. We will pursue in Phase 3B whether changes to our NRF incentive mechanism and monitoring program are necessary to ensure timely resolution of service quality problem areas.

F. Areas for Improvement

As with any investigation, we find areas for improvement. In particular, we have identified areas of service where utilities can and should improve both their performance and their measurements of performance. In addition, we have identified several areas where regulation requires clarification and better measures of service quality. We note that Rulemaking (R.) 02-12-004 was opened to adopt revisions in GO 133-B and that is the appropriate forum for modifying these standards. Nevertheless, we are confident that the findings of our current investigation show where variables need clarification, where measurement is lacking, where standards may be necessary, and where no change is warranted. These findings should prove helpful to R.02-12-004.

¹ We examine the years 1993 through 2001 for all GO 133-B measures. Where data exists, we examine 1990-2001. ² 35 measures were common to both Pacific and Verizon. Pacific had data on one additional measure, outages over 2 minutes per switch. ³ As explained below, we are unable to make a trend finding for Held Orders because Pacific used an incorrect methodology for tracking this measure, and we are unable to make a trend finding for Business Office Answer Times (BOAT) because Pacific changed its method of calculating this measure in the middle of the study period.