4. Preliminary Considerations

4.1 Necessity of Hearing

Almost all of the comments agreed with the Commission's determination that this case be classified as quasi-legislative and that no hearings are required. AT&T and others suggest that, in light of the technical nature of some issues, a workshop may be necessary. While the Consumer Coalition does not believe a workshop is necessary if a zero error rate is set, it does not oppose a workshop so long as the resulting schedule does not prevent the Commission from meeting its July 1 deadline.

The only party that believes hearings are necessary is the trade association, ATA. It is unclear whether ATA also disagrees with the classification of this as a quasi-legislative proceeding, but compared with the other two categories - adjudicatory and ratemaking - clearly quasi-legislative is most applicable. As the Attorney General notes, the California Supreme Court has held that hearings, while permissible, are not required in a quasi-legislative proceeding. (San Diego Gas and Electric Company v. Superior Court (1966) 13 Cal.4th 893, 950-051; see also, Franchise Tax Board v. Superior Court (1951) 36 Cal.2d 538, 549.)

ATA fails to provide sufficient evidence of a material fact in dispute to warrant hearings. Of the four "material disputed facts" that it suggests, two focus on the Commission's jurisdiction and call for a legal conclusion. While the other two points are relevant, other parties have included data on these issues, and ATA could and should have done the same. ATA also suggests that Pub. Util. Code § 1708 requires hearings. But Section 1708 is not applicable here because the Commission's action will not serve, in the words of that statute, to "rescind, alter or amend" a prior order or decision of the Commission.

The OIR's conclusion not to hold hearings is consistent with due process, public policy and statutory requirements.

4.2 Applicability of Rules

The Attorney General and the DCA, as well as industry parties, urge the Commission to make it clear that the rules adopted in this proceeding apply only to predictive dialers, also referred to as automatic dialing devices, and are in no way intended to affect the current law governing the use of automatic dialing-announcing devices. (See Cal. Civ. Code § 1770(v); see also Pub. Util. Code § 2874.) These devices disseminate a prerecorded message to the telephone numbers called, provided generally that the caller has a business relationship with the called parties or has the consent of the parties to receive such calls. We agree that the rules adopted in this proceeding apply only to automatically dialed calls that are intended to connect the called party to a live agent or telemarketer rather than to a prerecorded message.1

4.3 Percentage of `Live Calls'

As many of the comments point out, an error rate must be defined to specify whether it is a percentage of "live calls" or "all calls." It must also specify the time period of the measurement. Most of the parties agree that the error rate is properly measured as a percentage of live calls - calls where a person answers. AT&T comments that in its experience only about 20% of predictive dialer calls are answered by a live person. The rest are answered by a machine or are not answered at all. The Consumer Coalition, agreeing that the measure should be made with respect to live calls, notes that this measurement will result in a lower total number of abandoned calls, thus getting closer to accomplishing the goal of the legislation. Our decision today adopts this interpretation in considering the error rate.

4.4 Monthly Measuring Rate

The Commission must also define the amount of time over which the error rate is measured. Parties have made numerous proposals: per day, per month, per quarter and per six months. Pacific and Sprint suggest that the error rate should be a percentage of live calls over a monthly period. The Consumer Coalition agrees. The industry guidelines suggest the measurement should be on a "per day" basis, but this applies to telemarketing campaigns that may or may not last for more than a month. We believe that measuring the error rate on a monthly basis is reasonable. We agree with the Consumer Coalition that anything longer than that could allow too much fluctuation in the level of abandoned calls over the reporting period.

4.5 Definition of `error'

Virtually all parties suggest that the Commission must define an "error" before setting an appropriate error rate. In the context of the legislation, the Commission must determine what it means to make a telephone connection when an agent is not "available" to take the call. Does "available" mean that an agent must respond the moment after the called party says "Hello," or does it mean that a certain amount of "dead air" time is permissible while the predictive dialer determines that a live person has answered the phone and connects that person to an agent? (When the predictive dialer encounters a string of words when the call is answered, it assumes that the answerer is a recording device; when the predictive dialer encounters a short greeting ("Hello..."), it assumes that the answerer is a live person and connects the call to the telemarketer.)

All parties agree that if a predictive dialing device disconnects a call after the called party has answered, that constitutes an "error," or, in industry parlance, an "abandoned call." The question, however, is how much time should the predictive dialer take to disconnect. The DCA states:

It probably is not sufficient to state, without more, that the "error rate" should be zero. In practice, there is always some delay in responding after a called party answers a marketer's call. The key issue is what delay (measured by seconds or milliseconds) is permissible before the machine-caller responds. Without some standard, there may be no change in the actual practices of some marketers, who may contend that an operator is "available" but first must respond to others who have answered resulting in indeterminate delays. To deal with foreseeable evasions of that kind, the rules might state that once the called party answers the telephone, a live operator must respond with a human, non-recorded, non-machine generated voice within two seconds. (DCA Comments, at 2.)

Guidelines of the industry's Direct Marketing Association (DMA) suggest that the consumer should not be placed on hold for longer than two seconds before being connected to an operator or being disconnected. Sytel's comments point out that the United Kingdom DMA has adopted a one-second standard. The DCA proposes a two-second standard but urges a formal study. WorldCom and AT&T state that they need at least four seconds to connect a call to an agent because the predictive dialer must first determine if a live person has answered the call and then switch the call to the agent. Neither WorldCom nor AT&T, however, provides any data showing that their dialers cannot be programmed to meet a two-second standard. The Consumer Coalition contends that companies using these devices have had an opportunity in their comments to provide detailed data of the capabilities and limitations of their predictive dialing equipment, but that most simply offer unsupported conclusions. In the face of this lack of data, the Coalition urges the Commission to adopt the United Kingdom one-second rule as the minimum amount of time allowed by the industry as a benchmark for reasonableness.

The primary purpose of AB 870 is to reduce the number of calls received by consumers in which they say "hello" and are greeted with silence and the "click" of disconnection. In an industry where answer and disconnect times are measured in milliseconds, we believe that the U.S. industry guideline of two seconds is a reasonable one, and that is what we initially proposed. Comments of parties and further review of technical requirements persuade us that the standard should require that an agent or telemarketer respond to a called party within two seconds of the called party's completed greeting or, alternatively, an agent or telemarketer must respond within four seconds from the time the called party's phone goes off-hook.

We prefer the standard of a response within two seconds of a called party's completed greeting. However, some users of predictive dialer equipment state that they are not capable of complying with the two-second standard because of the manner in which their equipment is now programmed. In the interest of providing these predictive dialer users a transition period in which to reprogram, we will permit them to report errors based on the four-second off-hook standard for a period of six months. As of January 1, 2003, the four-second off-hook standard will sunset and errors will be measured on the basis of an agent or telemarketer responding to a called party within two seconds of the called party's completed greeting.

In summary, we define an "error" as a call made by predictive dialing equipment and answered by a live person in which (1) the predictive dialer disconnects the call after the called party has answered, or (2) the called party does not receive a response from the calling agent or telemarketer within two seconds of the called party's completed greeting, or, alternatively, no agent or telemarketer is available within four seconds of the called party's telephone going off-hook. The four-second off-hook standard is a transitional one that will be phased out effective January 1, 2003.

1 See also Pub. Util. Code § 2872 for specific exemptions from automatic dialer rules.

Previous PageTop Of PageNext PageGo To First Page