UNIVERSITY OF ILLINOIS LIBRARY AT URBANA-CHAMPAIGN BOOKSTACKS b/4 fc /-5 N0J7CE: Return or renew all Library Materials! The Minimum Fee for each Lost Book Is $50.00. The person charging this material is responsible for its return to the library from which it was withdrawn on or before the Latest Date stamped below. Then, mutilation, and underlining of books are reasons for discipli- nary action and may result in dismissal from the University. To renew call Telephone Center, 333-8400 UNIVERSITY OF ILLINOIS LIBRARY AT URBANA-CHAMPAIGN BEC i o 1995 L161— O-1096 :^s I *-?. FACULTY WORKING PAPER NO. 875 Determinants of Regulatory Policies Toward Competition in the Electric Utility Industry Walter J. Primeaux, Jr., John E. Filer, Robert S. Herren, and Daniel R. Hollas ** 4fc College of Commerce and Business Administration Bureau of Economic ana Business Research ersUy of Illinois. Urb?na-Champaign BEBR FACULTY WORKING PAPER NO. 875 College of Commerce and Business Administration University of Illinois at Urbana-Champaign June 1982 Determinants of Regulatory Policies Toward Competition in the Electric Utility Industry Walter J. Primeaux, Jr., Professor Department of Business Administration John E. Filer University of Mississippi Robert S. Herren University of Mississippi Daniel R. Hollas University of Mississippi Digitized by the Internet Archive in 2011 with funding from University of Illinois Urbana-Champaign http://www.archive.org/details/determinantsofre875prim Abstract Direct competition in the electric industry does actually exist in some cities; its impact on firm performance has been analayzed in a number of previous studies. Control of entry into a market is an important method regulatory commissions use to protect the monopoly status of an existing firm, yet there have been no previous attempts to investigate and identify the factors which ultimately influence regulatory policy toward permitting direct competition in this industry. That is one of the main objectives of this study. The model specified is a logit transformation with multiple explanatory variables. The statistical results are generally consistent with the Stigler-Peltzman theory of regulation. The data also tend to refute the notion that regulatory policy is an exogeneous variable which results from ad hoc political and administrative factors. The findings suggest, instead, that regulatory policy is a direct result of economic factors. DETERMINANTS OF REGULATORY POLICIES TOWARD COMPETITION IN THE ELECTRIC UTILITY INDUSTRY INTRODUCTION Control of entry into markets is an important method regulators use to ensure monopoly in the electric utility industry. Though unusual, direct competition in the electric industry does actually exist in some cities and its impact on performance has been analyzed in a number of different studies by Primeaux (1974, i975b ,' 1977 ,' 1978) . However, no previous studies have attempted to invest igate and identify the factors which ultimately influence regulatory policy toward permitting direct competition in this industry. George Stigler (1971) presented a revisionist theory of regulation which challenges the notion that regulation is designed to serve the public interest. Stigler's model was formalized and extended by Sam Peltzman (1S76). Peltzman concluded that special interest groups incur costs and realize benefits from regulation; total realized net benefits, however, are not as great as they would be in the absence of regulation.- Benefits accrue in. the form of transfers of wealth and arise through control of entry, establishment of rate structure (price fixing), control over price and production of substitutes, and cash payments. This study is concerned with the use of entry control as a manifestation of use of regulation to transfer wealth. Using Peltzman's theory of regulation, an attempt is made to identify factors which indicate hostility toward competition on the part of regulators. The statis- tical results are generally consistent with the Stigler- Peltzman theory of regulation. The data also tend to refute the notion that regulatory policy is an exogeneous variable which results from ad hoc political and adminis- trative factors. The findings suggest, instead, that regulatory policy is a direct result of economic factors. Regulatory Objectives Following Peltzman's generalization (1976) of a Stiglerian model of political wealth transfers, we assume the regulator wants to maximize a majority M, of voters, generated by (1) M = n-f - (N-n) -h where n = number of potential voters in the beneficiary group , f = (net) probability that a beneficiary will grant support N = total number of potential voters h = (net) probability that he who is taxed {ewery non-n) opposes. Although beneficiaries pay with both votes and dollars, the productivity of the dollars to a regulator lies in mitigation of opposition. Hence the regulator has an ob jecti ve f uncti on of the form (2) M = M(W r W 2 , ..., W n ) 1 where W. = wealth of group i, and where M .> This objective function is maximized subject to a constraint on total wealth (V): (3) V = Wj + w 2 + + w n = v(w 1 , w 2 , ... w n ) where V->0, but where V, 2 <0 > i.e., the total wealth to be distributed is limited. Among the groups vying to achieve benefits or mitigate losses from the regulatory process are consumers of the regu lated good, producers (both of the regulated good, and pro- ducers of substitutes), and the regulators themselves. In the case of electric utility regulation by PSC's, we can specialize the majority generating function, (2) as (4) M = M(P, n £ , n s ) where P = price of electricity, n E = wealth of electricity producers, and n<- = wealth of producers of substitutes (who may be regualted by the same commission). We assume the regulators have powers to affect the market which result in M > , M > 0, (i.e., greater wealth n E n s generates support) and M p <0 (i.e., higher electricity rates generate opposition from residential customers, commercial and industrial customers, and "consumer action 2 groups ). If we introduce cost and demand conditions, summarized by a profit function (5) n = f(P,c) where c = c(Q) = production costs as function of quantity, the formal problem for a successful utility regulator is to maximize the Lagrangian (6) L = M(P,n E ,n s ) + x [ n- f(P,c)] with respect to P, II-, n_ and A which yields This is the Sti gl er-Pel tzman conclusion that the marginal political product of a dollar of profits (Mn-» Mn<-) must equal the marginal political product of a rate cut (-M p ) that also costs a dollar of profits (f p is the dollar profit loss per dollar of rate reduction.) Or, in terms of the present study of electric utility regulation, we generate a testable hypothesis that, through regulation of electricity rates and entry, PSC's attempt to maximize M by balancing marginal gains to interested groups. Winners and Losers from Entry Control Regulated electric utilities tend to utilize discriminatory rate structures (Primeaux and Nelson, 1980; Hollas and Friedland, 1980). Entry control is a necessary means of perpetuating this discrimination because duopoly, or increased competition through entry, would reduce the extent of the discrimination. Increased competition reduces the ability of a firm to price dis- criminate profitably. Residential customers are probably the biggest losers in a discriminatory rate structure, due to their relatively inelastic demand, consequently they should benefit the most from competition. An impor- tant constraint upon residential customers, in particular, is that it may be costly to organize into a coalition in order to exert political pressures. However, if competition is active and the natural monopoly theory does not hold, as reported in Primeaux (1974), all customer groups will receive lower average rates due to the competition. The discriminating utility will be the only clear loser due to reduced profit poten- tial from discrimination. Of course, profit reduction may also originate from other sources. Consumers of substitute fuels such as natural gas benefit from competition for two reasons. First, competition in the electric industry lowers electric rates and increases competition between electric and 3 gas utilities; this will directly benefit gas customers. Second, since some gas users will then switch to elec- tricity because of its lower price, this will tend to reduce gas prices for those who continue to use natural gas . Natural gas producers would, in general, oppose a policy of direct competition in electricity. Natural gas and electricity are direct substitutes for many residential and commercial markets (e.g. electric home heating versus natural gas heating). Because all elec- tricity is not produced by using natural gas as a fuel, one would expect the increase in natural gas used by electric producers from increased competition in the electric industry to be less than the decrease in natural gas used by households and commercial users from lower electric rates. Retail natural gas utilities should also be adversely affected by direct electric utility competition because this would reduce the price of electricity, which is a substitute for the natural gas which they sell. Thus commissions in states with relatively more powerful natural gas distributors will be less likely to encourage ■ competi tion . A MODEL OF REGULATORY POLICY TOWARD COMPETITION IN THE ELECTRIC UTILITY INDUSTRY The general factors which may explain the tendency of Public Service Commissions (PSC) to be hostile toward competing electric utilities have been outlined in the previous section. This section develops and tests a model of regulatory policy toward comptition in the electric utility industry. The dependent variable in the model is unobserved. It consists of an assessed probability of hostile poli- cies toward competing utilities which lies between zero and unity. What is actually observed is whether the particular state PSC is hostile toward competition or not; and the sample results in a qualitative, dichoto- mous dependent variable, taking a value of zero or one. The policy of the commisssion in each individual state is classified as hostile or non-hostile toward direct competition. The information used to classify policies was obtained through questionnaires sent to commissioners in May 1971. The policy was classified as hostile if either direct competition was not permitted legally or if the commission stated that it would not allow competition. On the other hand, the commission was classified as non-hostile either if the commission does not have jurisdiction (it can not prevent competition) or if it contends that direct competition would be allowed. States without commissions were classified as non-hostile toward competition. Table 1 presents a summary of the 8 5 the classifications of regulatory policy. The major hypothesis of this paper is that economic variables explain costs and benefits (potential) of regulation to the interest groups which support or oppose entry control. A group derives power either from their numbers or their (potential) wealth. We now turn to a discussion of the explanatory variables of the model . Consumers of Electricity The mean income of residential consumers should be inversely related to their desire for direct competition between electric utilities. For example, the price elasticity of demand for electricity of a given customer would be lower, the smaller the fraction his utility bills are of his total budget. Lower income consumers would probably be more concerned than high income consumers about paying marginally higher fuel bills, because they buy from a monopolist not facing direct competition. Another factor which could indicate the strength or power of consumers is whether utility regulators are elected or appointed. Peltzman (1976, p. 215) argues that the method of regulator selection does not make any difference. Jack Hirshleifer (1976, p. 242) however, maintains that it does make a difference. Whereas the governor or state legislators which appoint a commis- sioner may campaign on several issues in a given plat- form, an elected commissioner will campaign on a very narrow platform. Thus, voters elect a governor or legislator for reasons in addition to the performance of their commission appointees. However, elected com- missioners are selected based upon only their regula- tory platform. In addition to the other questions examined, this study also tests the hypothesis that elected officials respond more readily to constituency pressures than appointed officials. Value added by manufacturing fimrs ia a proxy for wealth of industrial customers. As this measure of wealth increases, one would expect a more favorable PSC policy toward competition. Gas Producers Because electricity and gas are substitutes, natural gas producers would not benefit from increased competition in the electric industry (see discussion above). Therefore, as the importance of natural gas production ri ses relative to that of electric utilities, we would expect PSCs to become more hostile toward competing electric utilities. This outcome would be expected because producers of 10 substitute products for electricity prefer higher electric prices to reduce competition with their products. Electric Utilities The more monopoly power an electric utility has, the more wealth it can lose if competition takes place. Therefore, in states where utilities currently exercise substantial monopoly power, PSC's should be more hostile toward competition; there is, however, a problem of simultaneous cause and effect. Policy is certainly affected by the power or influence of electric producers, however, the PSC policy and attitude toward direct compe- tition should also affect the power arid wealth of the electric utilities. This problem is discussed in more detail be! ow. Empirical Results The model specified is a logit transformation with multiple explanatory variables of the type introduced by M. Nerlove and S. J. Press (1973) and take'the form: (3) P = e 3 i X i+ u i 1+e 6 i X i+ u i which when transformed, gives (9) In (yip) = s i X. + u i 11 where P is the probability that the relevant PSC is p "hostile" toward competition, (In(-rrp) is the "log of the odds" of a PSC being hostile), the X. are vectors of explanatory variable values, and u. is an error ter m As u. is either (1-6. X.) or (-e,X.)» u- is heterosce- dastic, and maximum likelihood estimation of the g. must be used. Specifically, the model to be tested is: (10) In(yTp-) = 6 1 (ELECTRICITY) + ^(ELECTED) + 6 3 (INCOME) + e 4 (VALUE ADDED) + 5 (NATGAS) + u where ELECTRICITY measures the influence of electric utilities, ELECTED, INCOME, and VALUE ADDED represent consumer interest variables, and NATGAS represents a substitute producer's interest variable. The nature of the variables and data is presented in Table 2, however,- a few additional comments on the ELECTRICITY variable may be in order. If the ratio equals unity, then the utility is charging the monopoly price. Consequently, a larger ratio indicates less realized monopoly power. Two complications arise with measurement of the ELECTRICITY variable. First, the 1969 estimates of this variable for residential consumers presented in a recent article by Robert Meyer, and Leland Hayne (1980) 12 were used in the analysis. Second, simultaneous equation bias exists because the power and strength of the utility affects policy but the effect is not one way only; policy also affects the power and strength of the utility. Since data processed by equation (3) showed a (expected) high correlation between ELECTRICITY and the error term, an instrumental variable (IV) tech- nique was utilized in which estimates of ELECTRICITY were obtained by regressing ELECTRICITY on a subset of exogenous variables. The following analysis should clarify the technique: (11) and (12) Y l = 6 1 Y 2 + e 2 Z l + 3 3 Z 2 + 6 4 Z 3 + U l a l Y l + a 2 Z 4 + a 3 Z 5 + V 2 whe re Y, = policy (1,0) Y ? = influence (power) of electric producers Z, = influence of natural gas producers Zp = influence of residential consumer interest - Z 3 = influence of industrial consumer interest etc . and u, and v~ are error terms. A small increase in lt. would result in a small increase in Y,. However, equation (12) shows that a change in Y. results in a change in Y~- Therefore, a change 13 in u, results in a change in Y^; that is, one of the explanatory variables and the error term are correlated (highly correlated in this specific case). An instru- mental variable was created to remove correlation by regressing Y- on a subset of the exogenous, variables. Specifically, the power of an electric utility Y ? measured by the ratio of residential monopoly price to actual price, was regressed on the following exogenous variables: number of retail establishments in 1972, 1971 state per capita income, number of municipal electric utilities in 1971, 1971 total revenue of municipal utilities, 1971 percent of population in SMSA areas, 1971 state bituminous coal production per capita, and retail sales per establishment. From this equation a predicted value of the ratio of the residential monopoly price to actual price, Y,,, was obtained. These estimates of Y 2 were then substituted into equation (11) for the log it estimation procedure. Table 3 presents the results of the logit analysis. The coefficients are defined as the marginal impact on the dependent variable of a one unit change in the explanatory variable, holding other variables constant at mean values. Care must be taken in interpreting the coefficients; in this case the depedent variable p is 1n (7Tp)- To assess the effect on the probabil i ty 14 of hostile policy, P, the effect on ln(y^p) must be solved for P. (The relationship between a unit change in an independent variable and the value of P is, hence nonlinear, and depends on the level of P . (See Footnote 9.) Column 1 presents the estimated co- efficients and column 2 reports the t-ratios. All coefficients have the correct sign and all coefficients are statistically significant at the .05 level or better except for the NATGAS variable, which is significant at about the .07 level. The Stigler- Peltzman hypothesis that numerous groups will benefit from regulation is generally supported, however, it appears that elected commissioners are more likely to favor a competitive policy. This result is at odds with Peltzman's argument that the method of regulator selection should be unimportant. The estimates of the coefficients of the logit model can be used to assess the relative influence of the in- terested groups vying for political favor from the commissions. Table 4 below gives the effect on the assessed probability of a given PSC being "hostile" toward competition, given a change in the circumstances of interested groups. The probability is calculated for both elected and appointed commissions. The mean 15 value of each explanatory variable, X., was inserted in the regression equation and a predicted value of ln(-T^T5") , and hence P, was obtained. We then changed each expl ana tcry variable from X. to (X. + s.) where s. is the sample standard deviation of the explanatory variable, X., and solved for P, holding the remaining g explanatory variables constant at their means. The third and fourth columns give the percentage change in the probability of observing an anti -competi ti ve commission, given a one- standard-deviation change in the value of the explanatory variable. From this we note a number of conclusions, most of which reinforce the Stigler-Peltzman theory. The regulated industry itself has the greatest influence on commission policies, regardless of whether the commission is elected or appointed. If the variable measuring the relative strength of natural gas interests is increased, there is a significant impact on policy in the predicted direction, and elected commissions are significantly more sensitive than appointed 10 Large industrial users of electricity have the third highest level of influence on commission policies after the utilites themselves and producers of substitutes, while the more diffuse residential consumer interest has the least impact. The only result inconsistent with Peltzman's predictions is the different impact 16 on elected versus appointed commission, as noted above Summary Statistical results from this study generally support the Sti gl er-Pel tzman theory of regulation. The only finding not consistent with Peltzman's formu- lation of the theory is that elected officials are more likely to favor pro-competitive policies than appointed officials. Pel tzman has argued this should be an unim- portant factor. For public policy purposes, however, it seems that any movement toward elected regulatory commissions would tend to foster pro-competitive policies, at least in the short run. Data reveal that an increase in realized monopoly power of the utility increases the probability of hostile PSC policies toward competition. An increase in average value added in manufacturing and a decrease in the state's per capita income increased the probability of favorable PSC policies toward competition. Also, the more power- ful are natural gas interests, the more hostile are commission policies toward competition. These empirical findings refute the hypothesis that regulatory policy is somehow an exogenous variable which results from ad hoc political and administrative factors. Instead, it appears that regulatory policy is a direct result of economic factors. Table 1 Public Service Commission Policies Toward Competition in 1971 State Al abama Alaska Ari zona Arkansas California Col orado Connecticut Del aware Florida Georgia Idaho Illinois Indiana Iowa Kansas Kentucky Louisiana Maine Maryl and Massachusetts Michigan Missouri Montana Nevada New Mexico New York North Carol ina North Dakota Ohio Oregon Pennsyl vania Rhode Island South Carolina South Dakota Tenenssee Texas Utah Vermont Virginia West Virginia Wisconsin Commi ss i on Pol i Direct Compe cy Toward t i t i o n Ca teg of onzation Pol icy Wyomi ng no pel icy seeks to elimin not permitted not permitted probably not pe no policy actively di scou no jurisdiction not permitted no pol i cy not permitted not permitted not permitted no jurisdiction not permitted a 1 1 owed i n some not permitted no established pol no response not permitted opposed to exte no jurisdiction competition be municipal util no jurisdiction not permitted not permitted would allow com considered use North Carol ina that municipal systems are fr not a 1 1 owed not sanctioned no response not encouraged legislature wou not permitted b util i t i e s no regulation not permitted no regulation not permitted not permitted not permitted permitted if ex is inadequate permitted if ex is inadequate not permitted ate rmi t rages cases icy nsion of competition over existing tween private and i t i e s petition if ful courts have held ly owned electric ee to compete or advocated Id not permit e tween private non hos hos hos hos non hos non hos non hos hos hos non hos non hos non- hos hos non non -hostile tile tile tile tile -hostile tile -hostile tile -hostile tile tile tile -hosti 1 e tile -host i 1 e tile ■hostile tile d tile -hostile -hos tile lsting service i s t i n g service non -hostile hostile hostile non-hostile non-hostile hostile non-hosti 1 e hostile 13 non-hostile hostile non-hostile non-hosti le c hostile non-host ile c hostile hostile hostile non-hostile non-hostile hostile a This commission was categorized as hostile due to corr- spondence with utility official in Hagerstown, Maryland. This commission was categorized as hostile due to studies by Primeaux (1975a) and Galat (1971). c These states were categorized as non-hostile toward compe- tition as 1) they had no state regulatory commission in 1971 and 2) we observed the existence of cities with competing electric utilities. Table 2 Variables Included in the Analysis Yj - 1971 PSC policy toward competition in . the state; 1 = hostile toward competition, = favors competition. ELECTRICITY = 1969 ratio of the monopoly price to actul price for residential customers. ELECTED = method of selecting commissioners; 1 = elected, = appointed. INCOME natural log of 1971 state per capita inc ome VALUE ADDED = 1972 state value-added per firm in manufacturing (in millions) NATGAS = 1971 ratio of cubic feet of natural gas produced (in millions) to KWH (in millions) of electricity sold in the state. Sources: Y. was obtained through questionnaires mailed by Primeaux to utility commissions; ELECTRICITY was taken from Table 2 of Meyer and Hayne (1980, p. 560); INCOME and VALUE ADDED are from Statistical Abstract ; NATGAS is from Gas Facts and Stati stical Yearbook. "~~ .Table 3 Results of Logit Analysis ELECTED VALUE ADDED NATGAS ELECTRICITY INCOME Coefficient -0.933 (0.533) -2.721 (1.213) 0.022 (0.014) -0.112 (0.042) 1.431 (0.520) t- ratio ■1.75 ■2.24 1.52 2.64 2.7 5 Standard errors in parentheses b Signif i cant at the .01 level Significant at the .05 level Significant at the .10 level -Table 4 Relative Influence of Interested Groups on Commission Policies 3 (Elected (Appointed Mean Predicted Probability 13 .25 9 Commercial /Industrial .094 Electricity Consumers (VALUE ADDED) Natural Gas Interests .540 (NATGAS) Residential Consumers/ . 300 Consumer Action Groups (INCOME) Electric Utility Interests .062 (ELECTRICITY) Commission) %a P Commission) %aP .471 -64% .209 103% .749 16% .521 -76% 143 -56% 5 9% 11% -70% Effect of a one-standard deviation increase in explanatory variable value on P, holding other variables constant at their means . Mean values of continuous explanatory variables inserted in estimated equation. 'Value of 1 inserted (if commission elected), and inserted (if commission appointed) in estimated equation. Footnotes differs from Peltzman's in two i respects explicitly incorporate Hi rshl ei f er ' s suggestion Our function 1 ) we (1976) that "the regulators themselves constitute an interest group"; and 2) we implicitly relax Peltzman's assumption that M..=0, if j, or, "there are no inter- group dependencies:" An example of such an interde- pendency would be the PSC regulation of both natural gas prices and electricity rates, when natural gas and electricity are substitutes (in residential usage, say) and natural gas is used as fuel for electricity generation . The interests of consumers and "consumer action 1 will not be, in general, identical. groups 3. This analysis assumes that the electric and gas com- panies are competing firms and not combination utili- ties providing both energy services. 4. There may also be a second order effect: producers of fuels used by the utilities may gain or lose depending upon output effects. That is, second degree price discrimination probably increases output, creating additional demand for fuels such as coal, oil, and natural gas. Output effects of third degree price discrimination are not as clear and depend upon the shape of the demand curve (Robinson, 1933). On the other hand, increase competition will also increase output due to lower prices of electricity. This also creates increased consumption by utilities for fuels to generate electricity. This information was obtained through questionnaires sent to commissioners in May 1971 (Primeaux, book manuscript in progress). The states not included were those which Primeaux had no responses, or all utilities were publicly owned, or which did not have commissions and competing utilities. It is our intuitive feeling that Peltzman is correct in long-run equilibrium. The observed difference in the policies of elected versus appointed regulators reported later in this paper may be analagous to the (seemingly) contradictory theories of the consumption function reconciled by Friedman's "permanent income hypothesi s . " 8 9 10 As i for corr dard For Econ s st this elat err a di omet anda equ ed w cr o sens r i cs vi ae the (in Year d th numb mill book el ec util of s trie i t i e hort r d with ati on a f th Y~, f the^e s i o n of , pp. 4 ta. St f retai of do! the so 1 i t i e s Mi neral an IV technique, the regressors re uncor related with u. and highly and chosen to minimize the stan- stimate and the covariance matrix. this technique, see Maddala, 76-477. The following works pro- i c a 1 Abstract, 1976 provided shments and retail sales dison Electric's Statistical e da er of retai 1 es tab 1 i ions of dollars). E was the source for utilities and total s . Mineral Yearbook tons of bituminous coal and lignite the number of municipal revenue of municipal 1971 supplied number We are indebted to Sam Peltzman for this method of assessing "relative contribution." p Due to the nature of the In(y^-p-) function, ( p [1n( ^ )] = P ( 1 - P ) and is not independent of the level of P), one cannot take any single mean p effect of a varaible on W^rpO by itself and solve for the mean effect of the explanatory variable on P, the probability of a commission being hostile toward competition. Values for allother explana- tory variables must be included before p Aln(j T Q) can be solved for aP References American Gas Association, Gas Facts 1971, 1976 (Arling- ton , Va . ) . Edison Electric Institute, Statistical Year Book of The Electric Utility Industry 1971 (New York). John R. Galat, "Direct Competition Between Regulated Public Utilities: A Case Study," in Joseph E. Haring and Joseph F. Humphrey (eds) Utility Regulation During Inflation (Occidential College: 1971) . Jack Hirshleifer, "Toward a More General Theory of Regu- lation, Comment," Journal of Law and Economics , 19:241- 244 (August 1976). Daniel R. Hollas and Thomas S. Friedland, "Price Dis- crimination in the Municipal Electric Industry," Research in Law and Economics , 2:181-198 (1980). G. S. Maddala. Econometrics (McGraw-Hill, 1977). Robert A. Meyer and Leland Hayne , "The Effectiveness of Price Regulation," Review of Economics and Statistics , 62:555-66 (Nov. 1980). Mark Nerlove and S. J. Press, "Univariate and Multivariate Log-Linear and Logistic Models," Rand Corporation , R-1305-EDA/NIH (December 1973). Sam Peltzman, "Twoard a More General Theory of Regulation," Journal of Law and Economics , 19:211-240 (August 1976). Walter J. Primeaux, Jr., "A Duopoly in Electricity: Competition in a 'Natural Monopoly 1 " Quarterly Review of Economics and Business , 14:65-73 (Summer 1974). Walter J. Primeaux, Jr. "The Decline in Electric Utility Competition." Land Economics , 51:144-148 (May 1975a). Walter J. Primeaux, Jr., "A Reexamination of the Mono- poly Market Structure for Electric Utilities" in Pro - moting Competition in Regulated Industries (Washington, D.C.: The Brookings Institutions, 1975b). Walter J. Primeaux, Jr., "An Assessment of X-Efficiency Gained Through Competition," Review of Economics and Statistics , 59: 105-103 (February 1977) . Walter J. Primeaux Jr. An Examination of Direct Electric Utility Competition: Some Persepecti ves on the Natural Monopoly Myth (Book manuscript in progress). Walter J. Primeaux, Jr. "The Effect of Competition on Capacity Uti 1 i ti za t i on in the Electric Industry," Economic Inquiry , 26:237-68 (April 1978). Walter J. Primeaux, Jr. and Randy A. Nelson, "An Exa- mination of Price Discrimination and Internal Subsidi- zation by Electric Utilities," Southern Economic Journal 47:84-99 (July 1980). Joan Robinson, The Economics of Imperfect Competition (London: Macmi 1 1 an , 1933 ) . Statistical Abstract of the U.S. 1976 (U.S. Dept. of Commerce , 1976 ) . U.S. Bureau of Mines. 1 (U.S. Government Printing Office, Minerals Yearbook, 1971 vol . Washington : 1973) KMAN U. RY INC. |§ JUN95 Masf N. MANCHESTER, INDIANA 46962