Performance-Based Regulation: Efficiency and the Measurement of Productivity Offset B. F. Roberts Performance Based Regulation (PBR) is emerging as a key element of the restructuring and re-regulation of the electric utility industry. PBR is being considered as a replacement for rate of return regulation for electric customer access (transmission, distribution and customer service), and as a regulatory transition into unregulated competition for the electric generation sector. The PBR is a key element of the California Public Utilities Commission's restructuring plan. The PBR has the appeal of simplifying utility regulation by virtually eliminating command and control micro-management by regulators. The simplification is obtained by imposing a simple price cap formula and allowing utilities the freedom to manage their operations, provided their electricity prices do not exceed the levels dictated by the formula. The more enlightened regulators have come to recognize that command and control micro-management of electric utilities is untenable in today's energy environment -- it overwhelms regulatory resources, destroys utility management's incentives for efficiency, and has been a major contributor to high electricity prices. The PBR is looked to as a means of simplifying regulation by setting the appropriate incentives for utility management while retaining general oversight and control. The PBR also offers regulators an easy regulatory mechanism for unbundling electricity services as the industry moves to vertically disaggregated markets including open competition in generation. Utilities see the PBR as a means of lifting the heavy hand of command and control regulation and allowing them to creatively manage their businesses. Utilities see the opportunity for shareholder gains through aggressive improvements in their operating efficiency. Utilities also see the PBR as a protective device for the collection of stranded generation assets in the transition toward competitive generation markets. Much of the appeal for PBR regulation is its apparent simplicity. However, the simplicity of the PBR formula is problematic if not carefully and correctly given numerical values. In spite of its simplicity, the PBR formula can have substantial efficiency and equity implications, especially when regulators attempt to use a "one-formula fits all" approach to regulating multiple utilities. It is fundamental that PBR formulas be uniquely tailored for each and every utility. The PBR typically includes a price cap formula and a profit/risk sharing agreement. The price cap formula frequently used in connection with electric utility Performance Based Regulation is: Pt = (1 + INFt - Xt + Zt) * Pt-1 where: Pt is a measure of the system average rate or price of electricity, INFt is a measure of inflation relevant to the cost of inputs to electric service, Xt is the productivity offset, and Zt is an adjustment for other exogenous factors. The price cap formula implies that the annual growth rate of electricity price should increase with the rate of electric utility industry cost inflation less the rate of growth of electric industry total factor productivity, plus the rate of growth of other costs outside the control of the electric company such as customer growth, disaster recovery, etc.1 While each element of the formula can be the source of controversy in regulatory proceedings, the setting of the productivity offset typically gains the most attention. Issues concerning the development of the productivity offset for the electric utility industry is the focus of this paper. The productivity offset is intended to be a reasonable measure of electric utility industry total factor productivity growth and a target for individual company total factor productivity growth. It provides incentives (in combination with a profit sharing formula) for the utility company to adopt new technologies and reduce costs to meet the target and, to beat the target to the benefit of shareholders and ratepayers. If the productivity offset is too high, the price cap will fail to cover the utility's production costs. If it is too low, then the company will earn extra-ordinary profits at the expense of ratepayers. Industry total factor productivity studies provide the foundation for estimating the appropriate productivity offset. Individual company total factor productivity studies provide a cross check to evaluate historical company productivity growth relative to that of the industry. Implicit in the methodology for these total factor productivity studies (using either the index number approach or the econometric approach) is an assumption that electric utilities are efficient (except for small deviations from efficiency due to random events). That means, the utilities truly achieve cost minimization in serving their customers. In other words, the utilities all operate on the "efficiency frontier" for the industry. The truth of this fundamental analytic assumption is certainly not evident from recent developments in the electric industry or studies that have focused on efficiency. The persistence of wide differences in productivity ratios across utilities suggest the probability of systematic, sustained inefficiency in some utilities. Given the importance of the productivity offset to the desired operation of the PBR price cap formula, critical examination of the efficiency of utilities adopting PBR's and the methodologies supporting the setting of the productivity offset is certainly warranted.2 Productivity is simply the ratio of a firm's output to its input. To calculate total factor productivity, multiple outputs and multiple inputs are aggregated such that output and input are each represented by scalars. Graphically, the productivity ratio can be interpreted as a ray from the origin intersecting input-output (X, Y) coordinates as shown on Chart 1.3 Productivity will vary depending on technology and many other factors. The input-output combinations of (X, Y2) (representing utility 2) is more productive than is the combination of (X, Y1) (representing Utility 1), because Y2 > Y1 implies more output per unit of input and a lower cost per unit of output. If Utility 2's combination (X, Y2) provides the highest output per unit of input among all feasible combinations, then this combination is "efficient" and is a point on the "efficiency frontier". Firm 1's combination (X, Y1) is clearly "inefficient" and falls below the efficiency frontier. In this simple case, (Y2 - Y1) / Y2 is a measure of the inefficiency of Firm 1's input-output combination (X, Y1). Whether or not a utility actually operates on the efficiency frontier and the distance from the frontier that it operates should be examined before setting a productivity offset value for the price cap formula for the utility. The productivity offset is intended to be a measure of industry total factor productivity growth, reflecting the industry's adoption of technological advances that facilitates increased output for any given level of production input. In other words, the growth rate of total factor productivity is the difference between the growth rate of real output and the growth rate of real factor input. If a firm is operating on the efficiency frontier, then the firm's total factor productivity growth rate and the growth rate of technical change are the same. However, if a firm operates below the efficiency frontier, then its total factor productivity growth rate is composed of both technical change and efficiency change. This will be illustrated on Chart 2. For simplicity, assume that the efficiency frontier is linear and, during the initial time period, the frontier intersects the point (X, Y2). During the next period, technical advance moves the efficiency frontier upward allowing more output for any given level of input. Assume further, that Firm 2 exploits the technical advance and operates on the new efficiency frontier. With input level X, Firm 2's output changed by the amount Y2' - Y2 and its productivity (Y2' / X) - (Y2 / X). This change was entirely a function of technical progress. Firm 1, on the otherhand changed its level of output (given input level X) by the amount Y1' - Y1 and its productivity ratio (Y1' / X) - (Y1 / X). This change is composed of two parts (shown by the brackets on Chart 2): Y2' - Y2 technical progress (same as firm 2) and (Y1' - Y1) - (Y2' - Y2) efficiency change. In other words, Firm 1's change in productivity was composed of both technical progress and change in efficiency (achieved by moving closer to the efficiency frontier). Firm 1 realized a much larger productivity change than did Firm 2 because of it's initial inefficient operation, below the efficiency frontier. Inefficient firms have the potential of greater productivity change than do efficient firms.4 Under cost of service rate regulation, Firm 1's system average rate would have started out higher than Firm 2's because of Firm 1's higher unit cost (i.e. X / Y1 versus X / Y2). Applying the same productivity offset to the price cap formula's for the two firms, would perpetuate the rate difference. In effect, using the PBR price cap formula with a technical progress productivity offset would provide a perpetual relative price reward for inefficiency. Or, a relative price penalty for efficiency. Both utilities will have incentive to remain or to become efficient under the price cap formula, but use of a uniform productivity offset has several non-neutral equity effects. It transfers wealth from the ratepayers to the shareholders of the inefficient utility in proportion to the utility's degree of inefficiency. In so doing, it also enhances the financial position of the inefficient utility relative to the efficient utility, which could subsequently reduce the value of the efficient utility under open access competition. In other words, the inefficient utility could, under the umbrella of the price cap formula, improve efficiency, while holding prices at the formula level, and build up cash reserves to become a more formidable competitor when access is opened. From the ratepayers point of view, this funding of a transition to efficiency may, to some degree represent an alternative to payment for stranded assets. From the point of view of the efficient utility, however, there is no comparable trade-off, the ratepayer funding of the transition would represent a ratepayer subsidy to a competitor. An alternative price cap formula that explicitly accounts for the possibility of inefficiency includes a decomposition of the productivity offset into two terms (Xt = XTPt + XEFt): XTPt measuring productivity growth due to technical progress and XEFt measuring the annual target change in efficiency. This approach, in effect, tailors the price cap formula to the efficiency position of each utility. Implementation of the approach requires electric industry total factor productivity studies utilizing efficient frontier estimating techniques. The studies will yield both the annual position and shift of the industry efficiency frontier (XTP - the industry growth rate of total factor productivity) and the degree of inefficiency of each utility. XEFt can be set from each company's efficiency index, and the dates at which various degrees of efficiency are targets to be achieved. Total factor productivity studies that do not use frontier estimation methods may not be of use for the setting of the productivity offset components. Since such studies typically use data for most large utilities versus using only the data for efficient utilities, the TFP estimates are most likely biased upward. This will occur if utilities in the sample experienced increases in productivity that are greater than the shift in the efficiency frontier.5 In light of industry concern about emerging competition in recent years, it should be expected that data will include substantial changes in inefficiency. The superficial simplicity of the price cap formula makes it a very appealing alternative to cost of service command and control ratemaking. The appropriate implementation of the price cap formula, however, is analytically demanding. And, the potential implications of the form actually implemented on ratepayers, and shareholders (of all utilities affected) need to be thoroughly examined by the regulators. Endnotes: See Morin, Roger A., Regulatory Finance: Utilities Cost of Capital, Public Utility Reports, Inc. The recent cost cutting (without reduction in output) by many electric utilities in preparation for increased competition, and estimates of so-called stranded assets, strongly suggest that many firms have been operating at cost levels far above minimum feasible costs. The recent study Performance Evaluation: California Investor-Owned Electric Utilities, Economic Sciences Corporation, 1994, found wide ranges of efficiency across the industry. Subsequent announcements of 25% rate reduction targets by two of California's major utilities verify the existance of substatial inefficiency. Such large changes in productivity are simply not feasible from shifts in the efficiency frontier. B.A. Holden, "SCEcorp, Anticipating Deregulation, Plans to Cut Rates 25% by Year 2000", Wall Street Journal, March 28, 1995. R. Mitchell, "PG&E: One Step Ahead of Future Shock", Business Weekly, November 14, 1994. This section draws heavily on, S.O. Grosskopf, "Efficiency and Productivity", The Measurement of Productive Efficiency, H.O. Fried, C.A.K. Lovell, S.S. Schmidt, Oxford University Press, 1993. The British experience with electric price cap formulas demonstrates the implications of misjudging the potential for efficey improvement. As reported in the Wall Street Journal article, "Utility Privatizations Backfire in the U.K.", March 30, 1995, the ability of utilities to cut fat and boost productivity was severly underestimated by regulators. Soaring profits and executive bonuses have caused ratepayer pressures for revisiting price regulation policies. This issue has been examined in the study, "Efficiency and Productivity Growth in U.S. Banking", by Paul W. Bauer, Allen N. Berger, and David B. Humphrey, in The Measurement of Productive Efficiency, op cit. The authors state, "While some prior studies have investigated technical change in banking, they have done so using the data of all banks, rather than those on the efficiency frontier. Such a procedure may confound technical change on the frontier with fluctuations in inefficiency that alter the average distance from the frontier." ESC ELECTRIC UTILITY ANALYSIS REPORT 95-1 Copyright ©1995 Economic Sciences Corporation To comment on this paper, send e-mail to: comments@econsci.com Economic Sciences Corporation