Empirical Case 7 ESTIMATING THE COST OF SERVICE FOR AN ELECTRIC UTILITY

Transcription

1 Empirical Case 7 ESTIMATING THE COST OF SERVICE FOR AN ELECTRIC UTILITY Regulated industries must pay particular attention to their cost structure. Estimates of their operating costs must be presented to rate-making authorities in support of any request by the regulated fm for a rate increase. As a consequence, these industries rely on rate researchers to empirically estimate the cost of providing their services. As one would expect, the estimated cost of service has many components. Economic theory states that the short-run cost of providing a good or service consists of both variable and fixed costs. From a practical standpoint, the rate researcher attempts to operationalize these classifications. First, there are explicit variable costs of providing the product. These include all expenses directly related to the production of the good or service, such as labor, energy, and raw-materials expenditures. Second,, there are fixed costs, including depreciation expenses, taxes, and capital investment expenditures. The standard formula for cost of service (COS) can be represented by the following: where COS = TVC + 'DEP + TAXES + R-BASE TVC = total variable costs, DEP = depreciation expenses, TAXES = all appropriate tax expenditures, R = allowed rate of return, and BASE = capital investment expenditures required to produce the service The electric utility industry has been the subject of intense investigation regarding its cost of service. Recall from introductory economics that electric utilities are natural monopolies. As such, an unregulated price of electricity would be detrimental to society. The utility would earn economic profits at the expense of consumers, and the presence of extensive economies of scale would prohibit entry by potential competitors. The solution to this problem is for the government to regulate the price such that the natural monopolist earns just a normal rate of return. That is, revenues of providing the service are controlled to be exactly equal to the cost of providing the service. To illustrate how a rate researcher might estimate the cost of service for a proposed utility in the New England region, you collect the following data on thirteen regional plants for existing utilities. To hold technology constant, you restricf the data to include only fossil-fueled steam-electric plants as opposed to hydro or nuclear plants.

2 356 Pan 3 / Production and Cost Analysis For simplicity, the proposed utility will operate in the same region and will produ the average amount of output, incur the average amount of depreciation expense ($ million) and taxes ($7 million), and undertake the average amount of capital investme variable cost that will be incurred. To es-te the variable costs of providing the service, you specify the average variable cost equation as a quadratic function of given as follows: AVC = a0 + alq + a@ where AVC is the average variable production costs and is the sum of labor and fu expenditures divided by output. Estimate the preceding AVC function with the plant-level data given in Tab1 using least-squares regression analysis. Use your results to answer the following qu tions. Maine Massachusetts Graham Station Walter F. Wyman New Boston Kendall Cannon Street Brayton Point Salem Harbor B. F. Cleary Output = power generation (million kilowatt-hours) Labor expenditures = total production expense, exclusive of fuel (WOO) Fuel expenditure = fuel production expense ($000) Source: Historical P h Costs and A d Information Administration. Production Expenses for Selected Elecrn'c Plants

3 Chapter 8 1 Cdsr Estimution and Forecasting 357 QUESTIONS 1. Does your estimated AVC function conform to economic theory? Explain. 2. Use the parimeter estimates to obtain estimates of TVC. 3. Given your to& variable cost estimation results, determine the estimated cost of service for this proposed utility. 4. Given this cost of service estimate and assuming that the utility sells the average amount of output, what is the price per kilowatt-hour the authorities will allow this fm to charge?

4 Empirical Case 8 ESTIMATING THE COST FUNCTION FOR FREIGHT TRUCKING In the United States, the 1980s have been called the decade of deregulation. Transportation industries witnessed a major relaxation of government controls as part of this trend toward a restoration of market forces to the business sector. In particular, regulation of entry conditions and price levels have been removed in both air and land transport industries to let the market determine the number of competitors and the pricetquantity allocation of services. An interesting outcome associated with this process is the altered decision making of high-cost f ms formerly protected under regulation from potential entrants. Once entry restrictions were relaxed, these finns had to consider the threat of entry by more costefficient fms and the associated decline in market price. Without significant cost reduction, these f ms would face short-run losses and eventually exit the market in the long run. Consequently, an accurate assessment of cost conditions is of particular importance in the post deregulatory market.' TABLE It Firm Variable Cost 1 102, , %, %, , , , , , ,666 1,413, ,082 2,071, ,680 2,195, ,133 1,670, , ,466, ,446 1,367, ,6% 1,701,125 Variable cost = total operating expenses, measured in $000 Output = ton-miles. measured in thousands Source: All data are given in the 1983 edition of Trinc's Blue Book of the Trucking Industry, published by Trine Transportation Consultants.

5 < Chapter 8 1 Cosr Estimation and Forecarling 359 You are interested in assessing the short-run relationship between price and quanity supplied for the general freight trucking industry in the deregulated period. To accomplish this, you estimate the industry's short-run cost structure using the following quadratic form: AVC = ao + a,q +,Q2 where AV& represents average variable production costs and Q is a measure of output. Table 1 presents the data you have collected for the general freight industry. Use this data to estimate the short-mu cost struchm. Then use your results to answer the following questions. QUESTIONS 1. Does the estimated average variable cost function conform to economic theory? Explain. 2. Use the resulting parameter estimates to obtain estimates of AVC and MC. 3. Graph your estimates of the AVC and MC Curves. At what level of Q does AVC reach a mi~urn1-t is!- oultbmph& What is the market-determi& price per ton-mile if a finn suppiied the amount of output found in question 3? Explain what this price means.