VII. Conclusions Regarding Slamming

A. Summary

The evidence demonstrates that Qwest has committed extensive and multiple violations of § 2889.5. The LECs' and Qwest's PIC dispute numbers indicate widespread slamming in violation of § 2889.5 in 1999 and 2000. Since October 2000, Qwest has begun tracking "realized" PIC disputes, and has recorded over 3,420 disputes through March 2001. Moreover, the evidence demonstrates that Qwest falsified some third-party verification tapes and letters of authorization, and had problems retrieving third-party verification tapes and making them available to customers. In some instances, Qwest failed to maintain its independence from its third-party verification agents and sales agents. The slamming rate for Spanish and Asian preferred language residential customers for 1999 and 2000 was consistently higher than the rate for English language preferred customers; thus, Qwest's slamming activities have a disproportionate impact on ethnic communities. Finally, although Qwest has recently improved its policies and procedures to prevent slamming, it continues to have problems in this area.

B. The PIC Dispute Data Indicate Widespread Slamming

1. Commission Precedent on Evidence of Slamming

As stated above, a PIC dispute is a customer allegation that his or her telephone service was switched without permission. Section 2889.5 prevents a telephone corporation from changing the provider of any telephone service until the telephone corporation: (1) thoroughly informs the subscriber of the nature and extent of the service being offered; (2) specifically establishes whether the subscriber intends to change his or her telephone service provider and explains charges associated with that change; and (3) confirms the subscriber's decision to change by an independent third-party verification company. A customer's credible allegation of a PIC dispute, standing alone, constitutes compelling evidence that Qwest has violated § 2889.5 by failing to ensure that each of the steps outlined above was followed prior to the switch.

In this case, the Qwest-related PIC dispute reports of the LECs and Qwest total over 70,000. The Commission previously has considered the reliability of PIC dispute data to demonstrate the number of violations of § 2889.5. In CTS, 72 CPUC2d at 633, the Commission found that that CTS switched 56,000 customers without their authorization. The Commission imposed a $500 fine for 39,200 of these violations, yielding a $19 million fine, all but $2 million of which was stayed. In relying on PIC dispute data, the Commission reasoned:

"The PIC dispute data maintained by the LECs is the most comprehensive data set available which relates to unauthorized customer transfer. While not purporting to conclusively prove any specific underlying fact, there is sufficient evidence to support an inference of wide-spread unauthorized customer transfers. Absent evidence directly undermining the credibility of the LEC's process and record keeping, or suggesting an alternative data source, the Commission will continue to rely on this data."

More recently, in the Vista Communications OII (Vista), D.01-09-017, the Commission held the PIC dispute reports regarding Vista to be reliable and admissible. The Commission found that the 10,773 PIC disputes attributable to Vista for 1997-1998 constituted sufficient evidence to find that number of slamming violations occurred. In Vista, the Commission reduced the number of violations to 7,000 on CSD's recommendation to correct for miscalculations, and imposed a $7 million fine, representing $1000 for each violation. The Commission reaffirmed its holding in CTS on the reliability of PIC dispute data.

"Since the customer interviews represented customers who complained to LECs and Vista's-affiliated carriers throughout the state and involved slamming allegations during the period of this investigation against Vista, and since there is no showing of inaccuracy of any carriers in recording or categorizing slams, we are convinced that the customer complaints investigated are representative of the thousands of PIC disputes recorded in this proceeding. Therefore, we accept the previously discussed total of PIC disputes during 1997-1999 (10,773) as the number of unlawful incidents during this period." (Vista, slip op. at 18-19.)

2. Qwest Challenges the PIC Dispute Numbers

Qwest critizes the LECs' process and record keeping regarding PIC disputes, first, because the LECs do not investigate each complaint. However, Sandy McGreevy, General Manager of Consumer Protection and Communication at Pacific, explained that it did not investigate the complaints primarily because Qwest chose the "no-fault" PIC dispute payment plan, under which Qwest paid a $9.98 PIC change charge for each unauthorized switch without investigating the PIC dispute.⁹

Qwest also argues that many of Pacific's recorded PIC disputes are the result of Pacific's "winback" campaign rather than a reliable indication of invalid PIC changes. However, Pacific does not compete in the long distance market and CSD states it did not use Pacific's intraLATA (i.e., local) PIC dispute reports in this case.

Qwest asked the Commission to set aside submission to take official notice of CSD filings in another Commission proceeding (made after the record closed in this proceeding.) Qwest believes these filings demonstrate the unreliability of the LECs' PIC dispute reports. Alternatively, Qwest requested the Commission to strike CSD's testimony of uninvestigated PIC dispute reports, or to order an issue sanction, finding that the PIC dispute reports do not prove legal violations. These motions focus on the reliability of Pacific's PIC dispute report, as opposed to Verizon's. The presiding officer set aside submission in order to consider Qwest's motions.¹⁰ We summarize these motions in the sequence of their filing.

Qwest's August 17 motion requests the Commission to take official notice of CSD's compliance filing in another Commission proceeding.¹¹ This CSD filing recommends, in relevant part, hiring an independent auditor at Pacific's expense to investigate the accuracy of Pacific's tracking and billing of intraLATA disputes. CSD recommends this because its preliminary investigation finds that Pacific's billing system improperly reported legitimate consumer switches to AT&T as intraLATA PIC disputes. Later, Qwest asked the Commission to take official notice of Pacific's response to CSD's compliance filing, and particularly to Pacific's statement that Pacific's intraLATA disputes historically have been coded by the "same mechanized ordering and billing system Pacific has used successfully for interLATA PIC changes and disputes." Qwest argues that this statement contradicts CSD's argument that the evidence is irrelevant, and demonstrates the unreliability of Pacific's tracking system for both intra- and interLATA disputes. Finally, Qwest moved to strike reports of uninvestigated and unverified complaints, including PIC dispute reports, or alternatively, for the Commission to impose an issue sanction and rule that such reports do not prove violations of law.

CSD opposes these motions on procedural and relevancy grounds. In particular, CSD believes its critique of Pacific's performance regarding intraLATA PIC disputes, and Pacific's rebuttal, are irrelevant to this OII because CSD did not count intraLATA disputes attributable to Qwest in enumerating violations.

We deny Qwest's motions because they concern filings that have no evidentiary value here.¹² CSD did not rely on Pacific's intraLATA PIC dispute process (the subject of the filings for which Qwest asks official notice.) More fundamentally, the reliability of the interLATA PIC dispute reports (including Pacific's, Verizon's, and Qwest's own reports) has been exhaustively tested in this investigation, and has been found to be high. Specifically, CSD investigated a sample of the LECs' and Qwest's PIC dispute complaints and obtained the declarations of 61 customers alleging that they were slammed. Based on the declarations, we find that 60 of these customers were slammed. CSD also completed interviews with an additional 115 customers alleging that they were slammed. We find that all but 16 of these customers were slammed.¹³ The customer declarations and interviews demonstrate an approximate 10% error margin in the PIC dispute reports by the LECs and Qwest. We find that over 70,000 PIC disputes reported to the LECs and Qwest, even if discounted by 10%, credibly and reliably show widespread violation of § 2889.5. Finally, until recently, Qwest itself relied on the LEC PIC dispute reports to determine whether its sales agents should be terminated, thus demonstrating that Qwest believed them to be reliable for this purpose.¹⁴ Consequently, it would be highly inappropriate to draw from the fact that CSD has criticized Pacific's intraLATA PIC dispute process any inference that CSD lacks credibility or that the PIC dispute reports of Pacific and the other companies are unreliable. Moreover, the evidence of which Qwest requests we take official notice is still being debated in another proceeding.

We also deny Qwest's multiple motions for sanctions against CSD. The gravemen of Qwest's sanction requests is that CSD failed to disclose to Qwest the information Qwest now seeks to officially notice. Qwest also accuses CSD of withholding in discovery from Qwest Pacific's actual PIC dispute reports, which contain a disclaimer that all reported disputes are the result of unsubstantiated customer claims to Pacific. However, this disclaimer did appear in an attachment to one of CSD's witnesses testimony; thus Qwest had it prior to the evidentiary hearings. Moreover, Qwest asked Pacific's McGreevy about this issue during cross-examination. We base our conclusion on the credibility and reliability of the PIC dispute reports on CSD's independent verification, and do not infer that Pacific independently verifies these complaints. For the reasons set forth above, Qwest's motions for sanctions against CSD are denied.

3. Qwest's Credibility and Burden of Proof Defenses

Qwest did not produce any customer witnesses, nor did Qwest depose or cross-examine any of CSD's customer witnesses. However, Qwest questions the credibility of both the customer witnesses and CSD's own investigators who interviewed many customers. Qwest also alleges that CSD did

not meet its burden of proof in demonstrating that many of the customers were slammed. As stated above, we find that CSD confirmed that about 90% of a sample of 176 complaining customers were slammed. We address Qwest's specific arguments to the contrary in this section.

Qwest challenges witnesses' credibility, for example, by arguing that the fact that some interviewees did not also file a declaration affects the validity of their complaint. We disagree. CSD presented 61 declarations (as opposed to testimony from CSD investigators summarizing these complaints) because the investigators who interviewed these declarants no longer worked for the Commission, not because the declarants were more credible than other customers.

Qwest believes that CSD has not met its burden of proving a slam has occurred if CSD has not offered evidence on any possible contingency that might have occurred. According to Qwest:

"For example, a PIC dispute could be made by one family member who did not know another family member (including the subscriber) had authorized the switch. Some customers forget that they themselves authorized the switch. Others suffer from buyer's remorse and falsely report PIC disputes." (Qwest Opening Brief at p. 17.)

This argument essentially asks us to require CSD to anticipate all possible defenses. Such is not CSD's burden. Section 2889.5 requires the long distance telephone company to establish, among other things, the intent of the subscriber to switch long distance carriers. (CTS, 72 CPUC2d at 635-36.) Qwest's current practice is to require its third-party verification agent to verify a subscriber's authorization by asking whether the person attempting to make the switch is in fact authorized to make the switch.¹⁵ Once CSD establishes by credible evidence that Qwest has switched a subscriber's telephone service without the customer's authorization, the burden shifts to Qwest to demonstrate that another adult authorized this switch on a customer's behalf. To the extent a third party, such as Qwest, wishes to rely on the agency relationship to bind the principal, the third party has a duty to ascertain the scope of the agency. (Id. at 636.) CSD does not have the burden of disproving a negative, namely, that no adult living or answering the telephone in the household was ever authorized to switch the telephone service.

Similarly, if CSD establishes through another adult living in the household that the subscriber has been slammed, the Commission must assess the credibility of this evidence to determine if it was more likely than not that a slam occurred. CSD does not always have to produce testimony from the subscriber of record in order to prove its case. In fact, this would be impossible in some situations, as when Qwest switched the long distance service of a subscriber, Mr. Nakashima, who died 11 years before the telephone service was slammed.¹⁶

In short, CSD is not required to speak directly to the subscriber regarding the complaint, because other household members may file the complaint. In contrast, Qwest is required by law to obtain verification of authorization for a change in service from the subscriber or the subscriber's agent, in cases where, for example, the subscriber is deceased.

Qwest also argues that CSD did not meet its burden of proof where (1) it did not ask for a third-party verification tape, or (2) if it had a verification tape, it did not play the tape for the subscriber. Although playing the verification tape for the customer is preferable when CSD is able, CSD is unable to do so in all instances. The failure of CSD to do so does not mean it has not demonstrated a slam occurred. In fact, Qwest itself admits to a forgery in instances where CSD did not play a third-party verification tape. (See, e.g., Exhibit C19, number 13 (Shu) and number 15 (Lim).)

Qwest details 50 instances where it believes CSD did not meet its burden of proving a slam. In all but one instance, we find a slam occurred.¹⁷ (See Appendix C for a summary of the Commission's conclusions for each of these 50 instances.) Examples of Qwest's arguments include the following:

McRae: Qwest says it has a third-party verification tape for this customer's wife and CSD did not play the recording. Qwest says the wife represented she had authority to make changes to the account. In a rebuttal, McRae's declaration states, "Qwest sent me a disk that supposedly contained a voice record file of my wife authorizing the switch. The file is in a language neither my wife or I speak and was very fast. It was obviously faked! I still have the file for evidence."¹⁸




Sabo: Qwest argues the subscriber never testified that no other authorized adult placed the order or that the person whose name appears on the letter of authorization is unknown to him or lacks authority to make changes in the telephone service. However, the Qwest letter of authorization survey signed by Sabo states that the letter of authorization is not in his name, nor does he know a Xuan Ho, and the address on the letter of authorization is not, and has never been, his address. Sabo further states, "Is Xuan Ho a real person? Why is he using/claiming my phone number?" ¹⁹




Nguyen: The subscriber (Vu Nguyen) stated that his brother, Dung Nguyen, who (according to Qwest) authorized the change in service, moved to Seattle in February 1999, prior to the switch which occurred on August 18, 1999. Qwest states CSD never established that Dung Nguyen lacked authority to make the changes, and that Dung was not in town at the time the order was placed. The CSD investigator also reported that Vu Nguyen's father, who pays the household bills, stated that Dung Nguyen has never been an authorized subscriber to the household's telephone service.²⁰

Qwest also argues that PIC disputes may have been caused by customer confusion arising from Qwest's acquisition of LCI International Telecommunications, Inc. However, no customer witness testified as to any confusion. Qwest states that resellers were responsible for some of the illegal switches, but fails to provide persuasive evidence of this, or any explanation regarding which reseller is responsible. Furthermore, when CSD asked Qwest in discovery to differentiate between complaints against itself and complaints against its resellers, Qwest stated it did not maintain records that would provide a breakdown of which PIC disputes are attributable to which companies.²¹

C. Qwest's Admissions

Since October 2000, shortly after adoption of the FCC Consent Decree, Qwest investigates all reported PIC disputes and tracks those that it defines as "realized." Qwest classifies a PIC dispute as "realized" if (a) Qwest cannot produce a third-party verification tape within 14 days of the complaint or (b) the customer denies that the voice on the third-party verification tape is the voice of someone authorized to make changes to the service. From October 2000 through March 2001, Qwest recorded 3,420 PIC disputes. The number of "realized" PIC disputes has decreased in 2001.

Qwest states that, since it has begun tracking "realized" PIC disputes, less than 25% of all PIC disputes have become "realized." This number is conservative. Qwest derived it by dividing the number of "realized" PIC disputes for December 2000 through February 2001 by the total reported PIC disputes for those months listed in Exhibit 28. When the same computation is made for October 2000 through February 2001, the percentage of "realized" PIC disputes jumps to 32% of all PIC disputes.

Qwest's "realized" PIC disputes are compelling evidence of § 2889.5 violations, because the complaining customers have not only alleged their telephone service was switched without authorization, but they have also either denied that the voice on the verification tape is that of someone authorized to make the switch, or Qwest has been unable to timely retrieve the verification tape. However, Qwest argues that a significant number of "realizations" are the result of Qwest's inability to produce a verification tape within 14 days, and that California law does not require verification retrieval within 14 days of a complaint.

The record is silent on how many "realized" PIC disputes are the result of Qwest's inability to produce a third-party verification tape within 14 days. The record is also silent on why late-retrieved verification should be considered reliable. Also, § 2889.5(a)(7) requires Qwest to make the verifications available to the subscriber upon request. For these reasons, the Commission is fully justified in regarding the slam as established when the company fails to produce timely verification. In the future, Qwest should produce the record of verification at or near the time of a customer's complaint.

Qwest also argues that an unknown number of "realized" PIC disputes result from customers who in fact authorized the switch, but nonetheless denied recognizing the third-party verification tape. However, Qwest's evidence on this point is anecdotal, not customer specific, and does not support an inference that this is a widespread problem.²²

D. Falsified Verification Tapes, Forged Letters of Authorization

Section 2889.5(a)(3) requires that Qwest confirm a residential subscriber's decision to change his or her telephone service provider by an independent third-party verification. The record demonstrates that in 1999 and 2000, Qwest had a serious problem with its sales agents or telemarketers forging or falsifying letters of authorization or third-party verification tapes.

CSD interviewed many customers who complained about verifications that were either not their handwriting or their voice. For example:

Mrs. H.Y. Nakashima listened to a verification tape produced by Qwest with what purported to be Mrs. Nakashima's deceased husband's voice. Mrs. Nakashima's husband passed away eleven years ago and she never changed the telephone service to her name.²³

Millie Jung listened to a verification tape produced by Qwest with what purported to be Mrs. Jung's recently deceased husband's voice. Mrs. Jung did not recognized the voice on the tape. Also, the voice on the tape provided the wrong birth date for Mr. Jung.²⁴

Tak Ming Chan reviewed the letter of authorization Qwest sent to the Commission's CAB in response to the complaint. Chan states he did not fill out any portion of the authorization and the signature is not his signature.²⁵

Helen Duce, who set up telephone service for her daughter when she moved into an apartment, and discontinued it about five months later when her daughter moved, stated the letter of authorization appears to contain her daughter's former address, telephone number and signature, but the signature is not her daughter's.²⁶

Greenlining/LIF also presented two witnesses who stated Qwest had slammed members of their family and falsified the verifications. Luis Arteaga, who also testified on broader immigrant issues, stated Qwest had slammed his own limited-English speaking parents twice in 2000 and that they did not know how to resolve the problem or to leave the Qwest billing system. According to Arteaga, both the letter of authorization and two third-party verification tapes were forged. Although the Arteagas had a PIC freeze on their long distance telephone account for interLATA service, Qwest switched their intraLATA account. It took Arteaga until early 2001 to resolve the problem, and cost him much aggravation and time.

Calvin Dong, who complained on behalf of his mother who speaks limited English, testified that in October 2000, her long distance telephone service was switched to Qwest without authorization, after he had resolved a similar problem in 1999. Dong called to complain for several months, and it took Qwest until March 2001 to send him a third-party verification tape. According to Dong, the verification was forged. It took until April 2001 to remove Qwest from his bill, and Dong spent about 6 to 10 hours trying to resolve the problem.

The problem of falsified tapes and letters of authorization is not limited to Qwest' s operations in California. In 1999, the FCC proposed a $2 million fine on Qwest, stating that it believed Qwest's activities to be egregious because Qwest apparently relied on forged or falsified letters.

Qwest admits that 58 of the complaints investigated by CSD in this case were "apparently caused by forgeries" by representatives employed by Qwest's sales agents. Nonetheless, Qwest questions the credibility of this evidence because CSD's investigators chose not to provide their own opinion as to whether or not the recordings were forged. We find credible the many customer representations of forgery of letters of authorization or third-party verification tapes, and are not persuaded otherwise by the absence of CSD's personal observations.

Qwest also challenges CSD's assertion that a forgery has occurred in any case where the third-party verification tape or letter of authorization is not in the name of the subscriber. Qwest believes the more likely problem is either a recycled telephone number, a customer inadvertently providing the wrong telephone number, or a third-party verification vendor filing or retrieval error. However, Qwest's belief is based on speculation unsupported by credible evidence.

Whenever forgeries occurred, according to Qwest, the third-party sales agent breached its contract with Qwest, Qwest was defrauded by the offending individual, the sales representative responsible for the slam was terminated, and in most instances, the sales agent responsible for the sales representative was also terminated. Qwest may have remedies against its third-party sales agents. However, it is still responsible for their actions. (See Vista at pp. 14-15, holding that Vista may be punished for the acts of its independent contractors.)

E. Failure to Retain Verification Tapes and Make Them Available

Section 2889.5(a)(7) requires Qwest to retain a record of the subscriber's verification for at least one year, and to make the records available to the subscriber, the Attorney General, or the Commission upon request. The record demonstrates that in 1999 and 2000, Qwest had serious problems with maintenance and retrieval of third-party verification tapes, although its retrieval rate has improved. (See also "Qwest's Admissions" discussed above, regarding the problem of delayed retrieval.)

Qwest admits it had problems retrieving third-party verification tapes in 1999, with an overall success rate only a little above 50%. An internal Qwest memorandum indicated that Qwest had a 32% retrieval rate for its third-party verification tapes in September 1999. According to CSD, the retrieval rate for Qwest's third-party verification vendors (ADC, CyberRep.Com, Teltrust, Quintel/West Interactive) ranged between 42.86% and 97.77% between March and December 2000. Qwest states that during this same time period, its retrieval success rate was 89.3%. Qwest has improved further, with a 96.1% retrieval rate in January 2001.

As already discussed, Qwest alleges that CSD did not meet its burden of proving slams for many customers, in part, because Qwest stated it had a third-party verification tape for these customers. (See also Appendix C to this decision.) In some of those cases, Qwest relies on Pitchford's testimony²⁷ indicating the verification tape was found, but does not state when. In many instances, Qwest was unable to retrieve the third-party verification tapes at or near the time either the customer or the Commission requested them.²⁸ Because of this problem, in the future, we are directing Qwest to produce such record of verification no later than 10 business days after a customer's or the Commission's request, in order to fulfill the requirements of § 2889.5(a)(7) to provide such verification to a subscriber or the Commission upon request.

F. Disproportionate Impact on Ethnic Communities

In the OII at p. 7, staff alleged that Qwest slammed residential customers who indicated Spanish or one of the Asian languages as their preferred language at a much higher rate than English speaking residential customers. For 1999 and 2000, the interLATA PIC dispute rate for Spanish and Asian language preferred residential customers was higher than the rate for English language preferred customers. This is so, even though the PIC change percentages for these two groups is smaller than for English language preferred customers.

Language Preferred	PIC Dispute Rate - 1999	PIC Dispute Rate - 2000
English	2.5%	2.2%
Spanish	6.8%	7.3%
Asian	9.8%	7.2%

Qwest claims that these higher PIC dispute rates were not caused by targeting, but rather by the unusually high proportion of face-to-face sales in ethnic communities. Because Qwest discovered a high percentage of PIC dispute rates resulting from such sales, it has eliminated them. According to Qwest, this higher PIC dispute rate may also have been caused by mistranslated Spanish scripts, which Qwest states it has corrected.

The record is insufficient to make a finding of the exact cause of these higher rates. However, if Qwest is correct as to the causes, the higher rate should no longer exist, because Qwest has eliminated this marketing behavior.

As in CTS, we emphasize here our commitment to ensure full and fair customer choice to all customers regardless of their primary language preference. We will not tolerate singling out any group of customers due to "industry norms" or any other spurious reason. We also require that, commencing with its next quarterly compliance report required by D.00-06-079 (this report is discussed below), Qwest shall disaggregate the actual PIC dispute numbers in California, as well as the PIC dispute rate, for English, Spanish, and Asian preferred language customers. If the PIC dispute rate for Spanish and Asian preferred language customers continues to be higher than the PIC dispute rate for English preferred language customers in any quarter, Qwest shall explain why and how it proposes to abate the problem.

Greenlining/LIF also presented three expert witnesses²⁹ who testified that Spanish and Asian preferred language customers are much less likely to complain than customers whose primary language is English. Greenlining/LIF suggest that complaints from these populations are more numerous and represent a much larger problem than are actually reported. According to Greenlining/LIF, these Spanish and Asian customers are less likely to complain because of cultural differences, their disinclination to question authority, their generally lower income level often associated with lower education and less knowledge of consumer rights, and their deep fear of being reported to the Immigration and Naturalization Service if they "cause trouble" by reporting a telephone problem.

Qwest attacks the credibility of these three experts on the ground that their testimony is based upon their years of experience with ethnic issues, rather than on academic or empirical research. Qwest's expert witness, Armando Gutierrez, believes the empirical evidence proves the opposite to be true, and that the individuals who choose to leave Mexico and Central America tend to be the most ambitious and assertive members of their society; otherwise, they would remain content with their home lives.³⁰

The empirical data relied upon by Gutierrez is largely census data showing that nearly 25% of California's population is foreign born. The census data also addresses the percentage of immigrants who speak English well within designated time periods after arriving to the United States. Gutierrez then extrapolates certain points from other studies on legal and illegal immigrants from Mexico, namely, that even some recent legal immigrants have completed high school and about 15% are college graduates. The primary study cited by Gutierrez addressing consumer behavior describes Hispanics as brand conscious, but not necessarily loyal consumers who can be persuaded to buy a different product over the current one.

Gutierrez' testimony is no less subjective than the testimony offered by Greenlining/LIF. The difference is that Gutierrez includes in his testimony data from which he draws questionable inferences, particularly concerning those low-income people who have immigrated during the last ten years. For instance, Guteierrez contends that because other academics describe these immigrants as the most ambitious and assertive in their home country, these immigrants will display the same characteristics in the United States. However, it does not follow from that assumption that immigrants will be more likely to "challenge authority" in their new country, and the experience of three of Greenlining/LIF's experts suggest the contrary.³¹

The record supports a finding that PIC disputes from Spanish and Asian preferred speaking customers may be understated. It does not support a finding as to the amount of such understatement, nor whether it is significantly higher for these particular communities than for the general population.

G. Independence of Qwest's Third-party Verification Agents

Section 2889.5(a)(3)(A) requires the third-party verification company to meet each of the following criteria: (1) be independent from the telephone corporation that seeks to provide the subscriber's new service; (2) not be directly or indirectly managed, controlled, directed, or owned by the telephone corporation that seeks to provide the new service; (3) operate from physically separate facilities from those of the telephone corporation that seeks to provide the new service; and (4) not derive commissions or compensation based upon the number of sales confirmed.

In the past, several of Qwest's third-party verification vendors have not been independent. For example, Qwest's contract with its third-party verification vendor Teltrust expands the scope of Teltrust's services beyond verification services. Qwest's contract with Teltrust states that Teltrust will also provide a "PIC Freeze Resolution Service" and a "PIC Freeze Installation Service." Under the contract, Teltrust will also provide inbound order taking for Qwest media campaigns and inbound customer service. Because it provides these sales and customer services, Teltrust is not independent from Qwest.

Qwest argues that § 2889.5 does not say that Qwest cannot retain the third-party verification vendor to provide other services. However, the statute requires these vendors to be independent from Qwest, and by providing other marketing and customer-related services to Qwest, these vendors are not independent because they may want to ensure sales and customer satisfaction with Qwest.

CSD also demonstrated that Qwest did not contract for third-party verification services directly with West Interactive in 1998 and 1999, but rather let one of its sales vendors do so. Because of this arrangement, Qwest had difficulty retrieving the third-party verification tapes from West Interactive; it had a retrieval rate in the year 2000 of 42.86%. Again, Qwest argues that the statute does not preclude this relationship. However, the verification vendor has a contractual relationship with the sales agent (presumably whose calls it is verifying), thus thwarting the statutory mandate of independence. Qwest also violated § 2889.5(a)(7) where, as here, it was unable to make the verification of the sale available to the subscriber. Qwest states that it has discontinued these practices.

H. Qwest's Historic vs. Ongoing Practices

Qwest describes its current "anti-slamming" process as consisting of the following elements: verification procedures; monitoring employees of independent, third-party sales distributors and agents³²; annual training of third-party sales agents and distributors; immediate termination of individual third-party sales representatives with a single "realized PIC dispute"; remedial action against third-party sales agents whose realized PIC dispute rate in any month is greater than 2% of the PIC-change orders the agents submit; modified compensation structures for Qwest's outbound telemarketers from a per sale to hourly basis; maintaining a "stay away list"; and customer education. Qwest argues that these procedures are effective because its recent PIC dispute rate is decreasing.

Although some policies and practices now in effect at Qwest differ from those in place when many of the complaints occurred, "we do not condone some sort of customer service and statutory compliance `learning period'" for public utilities. (See CTS, 72 CPUC2d at 629.) In CTS, we stated that the purpose of distinguishing between historic and future time periods is to align the sanctions with the behavior to which they are directed. Here, we take into account Qwest's new procedures in determining sanctions, but balance that against the continued problems with these procedures.

First, Qwest has reduced the number of recent PIC disputes, in large part, by investigating complaints and counting as a PIC dispute those that are "realized". Arguably, Qwest has reduced the field of PIC disputes by this definition. While we understand that Qwest wishes to determine the validity of customer complaints against it, we expect Qwest to put as much, if not more effort toward preventing slams by ensuring that a customer order to switch to Qwest is valid to begin with, rather than by concentrating on proving the invalidity of customer complaints. In short, Qwest cannot solve its slamming problems by redefining slamming.

Qwest's new policies require a sales agent with a "realized" PIC dispute rate that exceeds 2% to be terminated after several phases of probation. However, this means that Qwest will tolerate a sales agent with a "realized" PIC dispute rate of 1500 if it has about 90,000 confirmed sales. Because a "realized" PIC dispute is a narrowly defined category, this policy does not appropriately protect California customers and we modify it in the conditions we adopt below.

Second, Qwest's current policies require its third-party verification vendors to retrieve 98% of the third-party verification tapes requested. This policy violates Pub. Util. Code § 2889.5(a)(7), which requires Qwest to retain the record of verification for one year, and to make the record available to the subscriber or the Commission upon request. We require a 100% retrieval rate in the conditions we adopt below.

Third, Qwest's dealings with its sales agent, Results Marketing, Inc., demonstrate that its current policies and practices have not fully remedied its past behavior. Greenlining/LIF presented as a witness Felipe Rubio of Results Marketing, Inc., a sales agent marketing exclusively to Latinos that was under contract with Qwest to provide face-to-face marketing. Qwest terminated its relationship with Results Marketing in June 2000 due to a high level of PIC disputes. However, in late May 2000, on the eve of Qwest's termination of Results Marketing, Qwest knowingly hired Results Marketing as a sub-agent of Snyder Direct Services, Inc., which also had a high PIC dispute rate.

Qwest's witness Pitchford knew of this subcontracting arrangement. Pitchford testified that Qwest believed Snyder Direct had the best process in place to control PIC disputes, and that Qwest agreed to move Results Marketing under Snyder Direct to remedy the problems with Results. Qwest argues that if it had not given Results this opportunity to cure its processes, it might be accused of discriminating against a minority-owned business. However, we are not convinced this was the reason, because Qwest also determined in 2000 that Snyder Direct had an unacceptable level of PIC disputes and terminated its relationship with Snyder Direct in October 2000. Also, Results Marketing had a high number of sales in 2000. Qwest's recent actions in this matter demonstrate its continued difficulty in enforcing its "zero tolerance" policy with respect to slamming and its poor supervision over its sales agents as late as mid-2000.