affect the traditional local exchange business. We do know, however, that major segments of the local exchange market-those that carry the highest volumes of traffic and are the most profitable are being diverted to alternative networks.

To the extent that these alternative services compete with the services offered by local exchange companies, they should be covered in the scope of this legislation.

IS THE BILL COMPETITIVELY NEUTRAL?

In other words, the legislation should not distinguish between one group of service providers and another. If the intent of this legislation is to facilitate government surveillance of electronic communications, then public policy should provide for the government to be able to capture such communications wherever they may be taking place, not just on the traditional local exchange network. Otherwise, steadily increasing amounts of communications that bypass the local exchange network will also circumvent surveillance efforts.

By escaping the requirements of this legislation, these alternative telecommunications networks are given a competitive advantage in both cost and marketing. They do not have to spend considerable sums in modifying equipment or services to comply with the Act. They do not have to be hauled to court or the FCC whenever there is a dispute. They can market their "enhanced security" over public switched network carriers. And security is a major component of consumer confidence, trust and loyalty in choosing a telecommunications provider.

The FBI contends that its "problems" do not lie in these alternative networks; thus, the legislation need not cover these new entrants on the telecommunications scene. But that view embraces an historic perspective. Since the legislation is intended to address future telecommunications technologies, it is important that its coverage also takes a forward-looking view.

LEGISLATION SHOULD COVER SERVICES, NOT "ENTITIES"

USTA continually has proposed language which would apply the requirements of the Act evenly to any service which involves switching or transmission of electronic communications for hire to unaffiliated parties. By applying the bill to services, and not to companies (i.e. common carriers), the bill would attain competitive neutrality. The Sectional Summary accompanying this bill includes a reference to a functional definition of "local exchange carrier" in Mouse-passed M.R.3636. This support language should be put into the legislative language.

The legislation does provide the FCC with authority to determine when or if an alternative service provider should be covered by the bill, "when such service is a replacement for a substantial portion of the local exchange service and that it is in the public interest to deem such person or entity to be a common carrier for purposes of this Act." (emphasis added.)

But clearly, there's already difference of opinion in interpreting "substantial portion" since many of the above-mentioned networks are not covered.

2. Reasonableness and deployment of new services and technologies

Senator Leahy is quoted in the August 10 edition of Communications Daily as saying that the goal of this legislation "is to assist law enforcement needs without

* frustrating the development of new communications technologies or the competitiveness of America's high-tech industry."

But reasonable minds still are debating whether the actual provisions of this legislation accomplish that goal. Doubts still exist as to whether companies can deploy technology that cannot currently be tapped without facing the threat of court action and possible withdrawal of a technology, service or feature.

Even worse, the existence of this potential threat could hinder investment and development efforts in the nation's telecommunications infrastructure. If a company is faced with investing millions of dollars in a new product, but finds in the middle of its development efforts that a surveillance capability will be very difficult-and expensive to develop, the company may decide to "cut its losses" and abandon an otherwise valuable product development effort.

Further, if a technology is achievable, the bill requires that companies comply with the capabilities requirements. Does this mean that a small telephone cooperative in rural Vermont must modify or retrofit its facilities even though there is little or no likelihood that its surveillance capabilities will ever be needed?

For example, the FBI's REQUIREMENTS document declares law enforcement's desire to "follow" calls through the network to their ultimate destination. This may be possible as an application of advanced intelligent network (AIN) technology, not yet deployed. One conservative estimate of the cost of deploying this feature would

be $100 million per region. To support the FBI's requirements, AIN would need to be deployed universally, which is not currently being contemplated.

To satisfy these lingering doubts, the bill must provide a reasonableness standard and not guarantee absolute compliance with law enforcement's requirements. Good faith efforts to comply should be sufficient to find a company in compliance with the Act.

In this regard, it must include economic feasibility as well as law enforcement needs in determining compliance. Again, the Sectional Summary accompanying the bill includes a discussion of economic reasonableness in describing the bill's intent. This language needs to be placed in the legislation itself.

The relative costs and benefits of the bill's requirements placed on all companies must also be a factor in determining compliance. Otherwise, ubiquitous deployment of all "publicly available" capabilities would appear to be required by the bill.

REASONABLENESS AND TIMELY ACTION

The Act states that companies must comply with the capabilities requirements only if: “(i) alternative technologies or capabilities or the facilities of another carrier are not reasonably available * * * and (ii) compliance with the requirements of this chapter is reasonably achievable through the application of available technology or would have been achievable if timely action had been taken." (emphasis

added.)

As discussed above, the term "reasonably achievable" leaves much to the imagination, especially without specific reference to economic feasibility or cost-effective deployment considerations. Moreover, what "timely action" should have been taken? There simply is too much ambiguity for determining whether this provision could permit a court retroactively to order a company to withdraw or ban deployment of services or technologies at some time in the future.

Both of these terms need clarification before any company will feel safe investing the considerable money and resources needed to deploy new technologies and serv

ices.

RESPONSIBILITY OF MANUFACTURERS

USTA believes that the legislation now requires full cooperation and compliance of manufacturers with the requirements of the Act. Without shared responsibility, it would be possible for a carrier to face a court injunction against deploying a new technology or service when a manufacturer fails to provide it with the necessary capabilities which may be technologically or economically achievable, just not available.

INTERNATIONAL COMPETITIVENESS

In this global marketplace, what we do in the United States immediately affects market opportunities elsewhere in the world. Barriers to technology or service development here can Significantly affect the competitiveness of American companies visa-vis their international competitors.

This legislation effectively forces foreign companies to develop the same surveillance capabilities as AT&T. To the extent that the United States creates unique or expensive requirements for suppliers, it could limit technology alternatives available to American consumers or create offshore opportunities for American companies.

Similarly, what if only one company can develop the capabilities required by the Act? Will carriers be forced to purchase and deploy that company's equipment, regardless of the countless other factors that go into such multi-million dollar decisions?

3. Cost reimbursement

The FBI has estimated (in a "classified" analysis) that the government's costs in reimbursing industry for developing and deploying the modifications necessary to existing facilities to bring them into compliance with law enforcement requirements will be $500 million.

Communications Daily, in an August 4, 1994 article, noted it had obtained the FBI's confidential cost/benefits analysis which demonstrated several flaws. For example, the "FBI figures didn't include cost of upgrading or maintaining switch software and didn't account for new technologies, such as PCS. Nor did it factor in costs that cable companies would have to pay when they began to provide telephone services."

In testimony earlier this year before the House and Senate Judiciary subcommittees, USTA noted that the estimated cost of effecting software changes for call for

warding alone could be as much as $1.8 billion. One large USTA member has estimated that it would cost over $200 million to modify its call forwarding features to accommodate government requirements in this bill. And BellSouth requested a vendor's "inquiry-level" estimate for complying with the FBI's requirements on its wireline business only. The estimate: $138 million to $247 million.

Additional estimates can be made by looking at similar recent network upgrades made to accommodate_government requirements to provide equal access to all communications carriers. Two of USTA's members each estimate spending $250 million in this regard. Further, an August 5, 1994 Communications Daily article notes that the FCC has estimated that implementing another government mandate for enabling customers to chose their long distance carrier automatically from pay phones— will cost $1.8 billion.

In short, $500 million, as far as we can tell, will not even come close to the mark. USTA's principal concern with this legislation, therefore, is what happens when, or if, the $500 million fails to cover the costs of complying with the government's requirements.

"GOLDPLATING"

The record for government cost reimbursement for existing law enforcement activities does not give cause for great comfort with industry members, either. One USTA member company, with only moderate law enforcement obligations, spends $3.7 million a year accommodating over 100,000 law enforcement subpoenas, many of which require boxes of documentation.

These expenditures of time and personnel are borne by the company alone, without government reimbursement, even though the company frequently requests compensation for its efforts.

One can only presume that obligations are only a fraction of those imposed on such companies where law enforcement activity is more concentrated.

And, perhaps because there is no mechanism that limits government's "free" access to any information it wants, its demands for customer information have been increasing dramatically on an annual basis.

In fact, in June of this year, at least one of USTA's large company members received a subpoena from the Drug Enforcement Agency requesting "the identity of published and unlisted [presumably meaning nonpublished numbers] residential and business telephones *** by area code, and telephone number, together with first name, last name (company name), address, city and state for all published and unlisted residential and business telephones in the service area *** (See attachment.)

Besides the sheer audacity of this request, and its obvious implications regarding privacy and security of innocent American citizens, there naturally was no offer to compensate the company for complying with the Agency's demands, despite the fact that commercially available data bases (purged of nonpublished numbers) exist on the market today.

There must be a mechanism for preventing the government from making excessive demands, when a more modest and less expensive solution may satisfy most of the government's objectives.

THE OPPORTUNITY COSTS OF SURVEILLANCE REQUIREMENTS

It should be noted that whatever costs are being discussed, they represent time, effort, investment, research, and implementation of capabilities to meet government demands. There is no market incentive or cost recovery involved in developing such capability. That is, companies do not benefit from installing surveillance capabilities. (In fact, there are market disincentives if one takes into account the fact that companies are being required to build the ability to compromise security into their networks.)

To the extent that such time and investment is dedicated to meeting law enforcement's surveillance demands, and not developing advances in network enhancements for the improvement of private and commercial communications capabilities, then this Act has to be measured in terms of its opportunity costs; i.e., resources lost or diverted from other pursuits.

This is not to say that there are not significant societal benefits to meeting law enforcement demands. But the relative social benefits of law enforcement objectives need to be weighed against the costs of compliance.

FOUR (OR SIX) YEAR COMPLIANCE DEADLINE

Coming up with new features is no easy feat. Developing standards alone takes time, human effort and resources, and money. Even then, standards are only the beginning of a continuum which can last for years before eventually resulting in the deployment of new features or services in a manner that ensures their interoperability with existing network facilities.

To illustrate, after standards are developed, the following steps must be taken: 1) Develop technical requirements or specifications to 24 months;

2) Solicit industry input and finalize the requirements-6 to 12 months); 3) Vendor implementation-9 to 24 months

4) Trial and testing to determine actual viability and interoperability with existing network facilities and features;

5) Deployment (subject to cost/benefit analysis and funding availability);

6) Operation and maintenance.

Much of this process is not within the control of telephone companies, particularly as more and more services and features are a function of central office switching equipment.

It is clear to USTA members that there is both insufficient funding and quite possibly unreasonable time requirements in this legislation to cover the substantial costs and efforts needed to comply with the Act even within the first four years following its enactment.

COSTS AFTER FOUR YEARS

It is not reasonable to assume that costs will, as alleged by the FBI, be merely "de minimis" after four years.

Costs associated with developing and deploying new features simply are not "de minimis." As with the development of any feature, it takes time and money to meet additional requirements, particularly requirements as complex as those being posed by the FBI. And to the extent that this legislation develops a market for noncovered communications providers in a competitive environment, these costs will discriminate against carriers and their customers and investors.

In other words, there is and will be a continuing need to develop new standards as new technologies are introduced and features, functions and other applications are built around the new technology platforms.

An additional safeguard needs to be added to the bill to prevent the government from seeking modifications it could not obtain during the first four years (of six, if an extension is granted) after enactment.

The legislation commits the government to paying for increased capacity costs indefinitely. This provision should apply to capability compliance costs, as well, and should include the costs of operating and maintaining these government-imposed capabilities.

A SURVEILLANCE BUDGET: "PAY-AS-YOU-GO"

If it is only for the first four years that the government is given its $500 million authorization, then nothing in the bill prevents the government from asking a company in the fifth year after enactment to modify its facilities to comply with the Act's requirements.

To prevent retroactive application of the requirements, the legislation should clarify that the government cannot come back to a company in the fifth year, or later, and demand compliance for that which it could not afford, or did not properly plan for during the first four years.

What is needed is a means to prevent government from making unlimited demands at the expense of the American consumer. In short, government needs cost containment incentives.

This can be accomplished by strictly limiting law enforcement demands to the amount ($500 million) authorized in this Act. Companies should be found in compliance with the Act if the government fails to reimburse them. In other words, the requirements of the Act should not attach to the company, its customers and its shareholders.

The $500 million authorized for the government's implementation of the Act thereby would become a "surveillance budget” under which law enforcement would be forced to develop appropriate priorities and reasonable requirements. If, in fact, the government's estimated cost of compliance is $500 million, then it should be able

to accomplish all of its objectives within this budget. If, however, $500 million is too small an amount, then the government will need to determine where and how best to spend its budget.

There are several advantages to this approach. First, it prevents "goldplating." Law enforcement will have to concentrate its efforts where they are needed most, rather than ask for ubiquitous capability deployment where there may be little or no need. Pay-as-you-go will force the government either to limit its spending requests or modify its technical requirements.

Second, consumers will not be burdened from attempts to pass through significant costs associated with unreimbursed government requirements. This becomes even more problematic in the changing regulatory environment facing telephone companies today. The introduction of competition into telecommunications markets undermines a carrier's ability to recover nonmarket costs such as surveillance costsfrom its customers.

Third, small companies will be disproportionately affected by the costs of complying with this Act. Most USTA members can ill-afford million dollar fixes (that may be considered under the Act as "reasonable") for building capabilities that may never be used.

Fourth, USTA noted in its testimony earlier this year that a poll conducted by Yankelovich Partners in March and reported in Time Magazine found that “twothirds of Americans said that it was more important to protect privacy of phone calls than to preserve the ability of the police to conduct wiretaps." By limiting government surveillance appropriated funds, Americans will know as a matter of public record exactly how much the government is spending on surveillance activities. This "pay-as-you-go" solution should be applied to all capabilities and capacity requirements after four years following enactment, as well.

REIMBURSEMENT PROCEDURES

Companies need assurance the government will pay for what it demands. This legislation gives the FCC authority to settle any disputes that may arise in debating "reasonable costs" to be reimbursed by the government. But the timing of payment remains in question. The Act should ensure that companies need not comply until or unless they are paid.

This bill also contains a provision granting the Attorney General authority "to establish any procedures and regulations deemed necessary to effectuate timely and cost-efficient reimbursement * *" At least two USTA member companies are victims of nonpayment of long-outstanding bills.

USTA recommends adding language to this provision that ensures a process by which industry has an opportunity to effect these regulations, and that requires issuance of such regulations in a timely manner.

WHY SHOULD THE GOVERNMENT PAY?

Some argue that the government has no obligation to compensate private industry at all for meeting social obligations.

First, as noted above, there is no consumer advantage or market value to building surveillance capabilities into telecommunications equipment. Reimbursement is not paying for some corporate benefit. It is paying for the government's law enforcement requirements, a legitimate taxpayer expense.

Second, as noted above, imposing significant cost obligations on companies of vastly different sizes and financial capabilities would impose substantially dispropor

tionate economic burdens.

Third, because the bill as introduced fails to treat all telecommunications service providers equally, failure to compensate those companies covered by the bill will cause substantial competitive advantages for those companies not obligated to comply with and pay for law enforcement requirements.

Fourth, as aforementioned, failure to compensate for government surveillance activities effectively hides the government's expenditures from public scrutiny. Finally, a government obligation to pay for what it demands, under strict budgetary constraints, prevents it from "goldplating" its demands.

THE AUSTRALIA EXAMPLE

With the introduction in 1991 of competition in its telecommunications market, the Australian government granted to carriers the right to charge for assistance they provide to police authorities.

The government's decision was based on the following factors: