[BETA] Fortnightly

UNIVERSAL SERVICE ATTRACTS MUCH ATTENTION these days, both in energy and telecommunications. But how do you measure success? Do regulators decide when goals are met by looking across an industry, or should management make the call company by company?

Consider the Telecommunications Act of 1996. It identifies the maintenance of affordable, or "universal," service for low-income consumers as an explicit statutory goal. In the electric industry, virtually every piece of restructuring legislation and every regulatory decision to date has included a universal service provision.

Yet while setting a goal of universal service is easy to do in principle, it is much more difficult in practice. For instance, the concepts of universal service and affordability are often confused. Promoting universal service is an end; promoting affordable service is the means to that end. The terms are not synonymous.

The biggest problem by far comes in measuring whether the programs will actually promote universal service.

Where is the infrastructure (intellectual, data collection, or evaluation) to assess whether such programs are or will be effective? Legislation and regulation on electric restructuring in states such as California, New Hampshire and Massachusetts have established programs to promote and maintain universal service without establishing any mechanism to measure the programs' success. Five years from now, it will be impossible to answer the question: "Are we better off today because of these programs?"

What is needed is a set of data that allows policymakers to review not that x amount of money has been spent, or that y number of low-income customers have been reached, but that certain performance goals have been achieved.

Given the increasing interest toward performance-based management, both in the utility industry and elsewhere, I propose a similar method for measuring universal service and apply the method to existing data for a set of electric and natural gas utilities operating in one state, Pennsylvania.

The method employs five objective performance criteria: (1) the frequency of service terminations; (2) customer compliance with deferred payment agreements; (3) money at risk, or the utility's financial exposure to nonpayment of bills; (4) customers in arrears (and not part of a payment plan); and (5) weighted arrears, reflecting the size of monthly residential bills and indicating the extent to which residential service lies in jeopardy of termination. These indicators can show whether or not a company has attained and maintained universal service (see sidebar, "The Model Defined").

Regulators should design and fund universal service programs that are affordable and available. Moreover, regulators, planners and company administrators should be interested in measuring universal service over time. I believe that a system such as that proposed here is appropriate for telecommunications, natural gas and electric companies interested in universal service in a competitive environment.

Assessing Performance

Performance measurement increasingly is being applied to both public and private programs. Perhaps the best-known application is the Government Performance and Results Act of 1993. GPRA was designed to address the same conceptual issues that a competitive utility must address for its universal service programs: "to grapple with how to best improve effectiveness and service quality while