OVERDUE FINES: 25¢ per day per item RETURNING LIBRARY MATERIALS: Place in book return tc remove charge from circulauon recon A CC‘ NATIOEJ in pa: Depart .A COMPARATIVE STUDY OF ADMINISTRATION OF NATIONALIZED RAILROADS IN THE UNITED KINGDOM, FRANCE, WEST GERMANY, AND JAPAN BY Le Trung Thuong A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Marketing and Transportation Administration 1980 A CC I NATIOT The in roads are four i.e., governm emissions, 1 he latter mo. greater manag operational e: of this study on the basis 4 France, West ‘ Quanti concepts of p to decompose . cost componen 219 ncies are - I“ v.‘fi . Vuuced ABSTRACT A COMPARATIVE STUDY OF ADMINISTRATION OF ‘NATIONALIZED RAILROADS IN THE UNITED KINGDOM, FRANCE, WEST GERMANY, AND JAPAN BY Le Trung Thuong The institutional arrangements for nationalized rail- roads are founded upon various legal models of organization, i.e., government departments, semi-autonomous boards and commissions, public corporations and joint-stock companies. The latter models are generally postulated as offering greater managerial autonomy which, in turn, promotes higher operational efficiency than do the former. The objective of this study is to test these two hypothesized relationships on the basis of postwar experiences in the United Kingdom, France, West Germany, and Japan. Quantitative measurement of efficiency employs the concepts of production and cost functions as instruments to decompose operational efficiency into its technical and cost components. The indices of technical and cost effi— ciencies are then estimated by comparing the actual output produced, and the cost incurred, by each observation against their corresponding potentials. These potentials are speci— fied by Cobb-Douglas production and cost functions, both Ibeing fitted by a linear programming method of estimation. Efficiency or are analyzed Quali tion begins to of the model railroads st‘, of political cmnwrcial p. Statutory pro 9. E Ch01ce bEt' 10E'ic \ Le‘Trung Thuong Efficiency gaps among the four railroads and their trends are analyzed. Qualitative assessment of legal models of organiza- tion begins with an analysis of theoretical characteristics of the models and the motives behind their adoption for the railroads studied. The motives are found to be primarily iof political and macro-economic nature; the criterion of commercial profitability played only a very secondary role. Statutory provisions on railroad organization were conse- quently geared mostly toward the balance of power among political authorities. The relations between the railroads and political authorities did not correspond closely to postulates of the legal models of organization. Instead, they were shaped by complex blends of forces, i.e., legal philosophies, statu- tory provisions, political and administrative traditions and other historical circumstances. The relations exerted pro- found impacts upon two important aspects of the railroads' internal organization. One was the orientation of the managerial bodies in command of the railroads, i.e., guardians of public interests versus business management; the other was the choice between centralized or decentralized field organization. The performance of the railroads are next assessed in terms of their achievements, or lack of them, in techno- logical modernization, network and labor rationalization and competition with other transport modes. Qualitative empirical evidences ge quantitative that neither lation and : upon railroz to the rail: separation 1 political ax. away from t. failed to h. railroads' 1‘01 by pol interferenc or even par Sional judg Political 5 upon self~3 “Missed from the 1 Th1 relatiOnsh : ‘ b Ibode? Le Le Trung Thuong evidences generally confirm the preliminary findings of the quantitative measurement of efficiency. It is also evidenced that neither a superficial separation between policy formu- lation and implementation nor close supervision and control upon railroad management by political authorities was conducive to the railroads' operational efficiency. Such superficial separation created a confusing no-man's land where both political authorities and railroad management were shied away from taking necessary initiatives and decisions, thus failed to harmonize between broad policy objectives and the railroads' commercial interests. Close supervision and con- trol by political authorities, on the other hand, invited interferences into railroad management and thus constrained, or even paralyzed, management from exercising their profes— sional judgement. Only by a two-way coordination between political authorities and railroad management which rested upon self-restraint from interferences by the former and encouraged initiatives and contributions to policy formulation from the latter could operative efficiency be promoted. This study, in summary, validates the hypothesized relationship between managerial autonomy and operational efficiency and throws some new light into that between legal models of railroad organization and autonomy. The study then makes some preliminary recommendations for American ther hand, the standard of performance only for public policy itself. It is in light of such distinction this strudy will focus only on the asseSsment of operational efficiency which fell impacts an The refer to wt performance as well as inefficiency “0 Possibl Iiotivaticma tion) and (I Should be n. general rat howev’er, Co enterPIiSes business di briefly the tualiZQd in With the te 18 efficiency attained by state enterprises. The discussion which follows will attempt to conceptualize the causes, impacts and posSible magnitude of operational inefficiency. Possible Causes, Impacts and Magnitude ofOperational Inéfficiency The term "x-inefficiency" coined by Leibenstein to refer to what was considered as suboptimal Operational performance of firms reflects the lack of understanding as well as empirical evidences on the causes of operational inefficiency. Scattered studies on this subject suggest two possible causes of x-inefficiency, namely, (a) ex-ante motivational failure (or failure to attempt cost minimiza- tion) and (b) ex-post unsuccessful minimizing attempt. It should be noted that most of these works focus on firms in general rather than exclusively on state enterprises. They, however, contain two elements commonly found in state enterprises: the lack or weakness of competition and business disciplines. It is therefore worthwhile to examine briefly the causes and impacts of x—inefficiency concep— tualized in these works. Managerial models of the firm associate x-inefficiency with the tendency of firms, operating in a non-competitive environment, to pursue goals other than profit maximization without jeopardizing their commercial viability. Crew et a1. (1971) explain x-inefficiency in terms of discretionary power whic competitio zero-cost of firms (1' associates the firm be frequently Beha X-inefficie‘ afim trie: PrOfit maxi: that public °fr6turn Se which these revulated p, Inethods of E X‘ine 19 power which exists in certain markets. According to them, competition rather than profit maximization is a principal zero-cost policing mechanism which drives the cost curves 17 Marris (1967) of firms down to their minimum levels. associates inefficiency with management's desire to enlarge the firm beyond the profit-maximizing size; salaries are frequently assumed to be higher in larger firms.18 Behavioral models, on the other hand, explain x-inefficiency in relation to the decision process by which a firm tries to reach its goal which may or may not be profit maximization. Averch and Johnson (1962) theorize that public utilities which operate under the maximum rate of return set by regulation, would inflate the base upon which these rates are calculated. Consequently, these regulated public utilities tend to use more capital-intensive methods of production (hence being more cost inefficient) than would unconstrained cost minimizers.19 X-inefficiency may also exist in form of resistance to changes, i.e., technological advances. Leibenstein 17. M. A. Crew, M. W. Jones-Lee and C. K. Rowley, "X-Theory versus Management Discretion Theory," The Southern Economic Journal, Vol. 38, No. 2 (October 1971), pp. 173- 184. 18. Robin Marris, The Economic Theory of Managerial Capitalism (London: Macmillan and Co., Ltd., 1967). 19. Harvey Averch and Leland Johnson, "The Behavior of the Firm under Regulatory Constraint," The American Economic Review, Vol. 52, No. 5 (December 1962), pp. 1052- 1069. zational c Similarly, in terms c redressing Perfectly fied custc to ignore mediary of EffeCtivem availabili. f°r exampl be mm 5115 there being satisfied I transPort n ill COm 1 E ria' 6:; " JOHN 21. 20 (1969) introduces the concept of "inert areas" which operate as frictional elements to the rate of diffusion of changes. Accordingly, if the stimulus is not strong enough (i.e., a threat to the firm's survival) to make the potential net gain greater than the cost (i.e., economic or psychological) of change, the firm will tend to maintain a state of "organi- zational or frictional equilibrium," thus, resist to changes.20 Similarly, Hirschman (1970) explains resistance to changes in terms of "voice" and "exit," the two being means of redressing socially undesirable institutional behavior. Perfectly competitive firms respond only to exit of dissatis- fied customers; many state enterprises, being in a position to ignore profit, respond only to complaints through inter- mediary of government. Among these state enterprises, the effectiveness of voice is inversely proportional to the availability of close substitutes. Other things being equal, for example, one would expect a public electric company to be more susceptible to public expression of dissatisfaction, there being more close substitutes for the latter. A dis- satisfied railroad user is more likely to turn to other transport modes, whereas a producer suffering power failure will complain cociferously.21 20.. .H. Leibenstein, "Organizational or Frictional Equilibria, X-efficiency and the Rate of Innovation,"Quar- Eggly Journal of Economics, Vol. 84, No. 4 (1969), pp. 600- 21. Albert Hirschman, Exit, Voice and Loyalty (Harvard University Press, 1970). His exit and voice In motivatiox observed ‘ counterpa. minimizat costs, Cor welfare 1. According its price PIOfit-ma; measured 1 if the Ob: tical C0111 BCEF repr, x'imffic assumP'Cio: loss of e of gain and . demand, t‘ 21 In short, x—inefficiency theoretically exists where motivational forces are weak. The result is the‘firm's observed cost curves being well above their hypothetical counterparts under traditional assumption of cost— minimization. Using observed and hypothetical marginal Icosts, Comannor and Leibenstein (1969) show that the welfare loss due to x-inefficiency could be substantial. Accordingly, when a monopolistic firm is forced to set its price at (observed) marginal-cost, Pmcx’ rather than profit-maximizing price, Pp, an amount of consumer surplus measured by Marshallian triangle, ABC, is gained. However, if the observed marginal cost, Pmcx’ exceeds its hypothe- tical counterpart, Pmc’ due to x-inefficiency, the area BCEF represents the loss of consumer welfare (due to x-inefficiency) which is overlooked under cost-minimizing assumption. Also, the rectangle PmchEPmc represents the loss of efficiency in production (x-inefficiency).22 Of course, the magnitudes of allocative efficiency gain and x-efficiency losses depend upon the elasticity of demand, the levels of observed and hypothetical costs and analysis was originally suggested in his study on the Nigeriam Railway Corporation, Development Projects Observed (Washington, D.C.: The Brookings Institution, 1967). 22. William S. Comannor and H. Leibenstein, "Allo- cative Efficiency, X-Efficiency and the Measurement of welfare Losses," Economica, Vol. 36, No. 143 (August 1969), pp. 304-309. 22 Price/Cost D P A p ‘ excess profit P . . . mcx x-ineffiCiency P F mc Figure 2. Possible magnitude of x-inefficiency. their v suggest signifi associa optimiz or, at is come. 0f law ; flm ente Objecti‘ of these as the j HYpothes HYPOtheE O r» Po L180 . (‘4. "' in 23 their corresponding outputs. Empirical evidence, however, suggests that x-efficiency losses may be significant.23 Research Hypotheses The literature on "x-efficiency" postulates the significance of operational inefficiency and conceptually associates it with the operating environments in which optimization of a firm's economic behavior is constrained or, at least, not encouraged. As far as state enterprise is concerned, such environments are postulated by students of law and public administration as being determined by the enterprise's legal model of organization. It is the objective of this study to provide an empirical testing of these postulates. They are stated in a different form as the following hypotheses: Hypothesis 1: Managerial autonomy is determined by the particular legal model of organization adopted for a state enterprise. Hypothesis 2: The levels of operational efficiency achieved by state enterprises organized under various legal models are measurable and can be traced to the difference in managerial autonomy. Empirical evidences will be drawn from the railroad industry. This industry is selected partially because of the author's personal interest and partially because of the industry's important magnitude within public sectors 23. For example, see: ‘Richard Pryke, Public Enter- prise in Practice, op. cit., pp. 396-400; and C. K. Rowley, Steel and Public Policy (New York: McGraw Hill, 1971). as seen tunity 1 only £01 Kingdom, They are Western are two Railways 991‘ Corp (COnRail the latt Pilblicly Side. T the inte SYStemS ISO preSente of Opera tateS th. only mea but also comPOnen. tion Of I have acc< 24 as seen in Figure 1. Also, in order to provide an oppor- tunity to explore the depth of these reSearch questions, only four nationalized railroad systems in the United Kingdom, France, West Germany and Japan are included. They are among the most important railroad systems in Western industrial economies. Besides these four, there are two other major systems, namely, the Canadian National Railways (CN Rail) in Canada and the National Rail Passen- ger Corporation (Amtrak) and Consolidated Rail Corporation (ConRail) in the United States. The railroad systems in the latter two countries are dual in character where publicly- and privately-owned companies operate side by side. The inclusion of such dual systems would complicate the interpretation of research findings. These dual systems deserve studies of their own. Research Methodology and Outline of Chapters This study consists of two phases. Phase one, being presented in Chapter II, provides a quantitative measurement of operational efficiency and its components. This necessi- tates the development of a model which is capable of not only measuring the relative levels of operational efficiency but also decomposing efficiency into its technical and cost components. Such decomposition facilitates the identifica— tion of major areas of managerial decisions which appear to have accounted for the efficiency gaps and their trends among 1 then be former mnm c objecti being ; activel ment to This ch 0f the behind other f Strativ these f raiera Organizl ( manager: doing, j and the policies COSt and These ma tion, ra- railrOad 25 among the railroads studied. These decision areas will then be investigated in depth in phase two. Phase two consists of Chapters III and IV. The former focuses on the relationship between the legal forms of organization and managerial autonomy. The objective here is to analyze managerial autonomy nOt as being passively defined by legal provisions but as being actively translated into the ability of railroad manage- ment to control the Operations of these enterprises. This chapter will therefore examine the characteristics of the legal models of organization and the motives behind their adOption -for the railroads studied and other factors such as the prevailing political and admini- strative traditions. An analysis is then made as to how these factors have affected the relations between top railroad management and public authorities and the internal organization of the railroads. Chapter IV focuses on the relationship between managerial autonomy and Operational efficiency. In so doing, it will analyze the role played by management in, and the constraints set by public controls and transport policies upon, major decision areas affecting the technical, cost and overall operational efficiencies of the railroads. {these major decision areas include technological moderniza- tion, rationalization of networks and labor forces and railroad pricing and transport coordination. finding mendati 26 Chapter V, the concluding chapter, summarizes the findings of this study and makes some preliminary recom- mendations for the American quasi-government railroads. tive op To do 5 model w into i1;~ Chapter 0f attai Efficien trends, the Prob wE‘YS- T] Prelimim explained CHAPTER II ESTIMATION OF RELATIVE EFFICIENCY The focus of this chapter is on measuring the rela- tive operational efficiency among the railways studied. To do so requires the development of, first, a conceptual model which is capable of decomposing operational efficiency into its technical and cost components as discussed in Chapter I and, second, a method of measuring their levels of attainment by the railways. The measured levels of efficiencies are then analyzed so as (l) to detect general trends, if any, and their implications and (2) to estimate the probable magnitude of efficiency gaps among the rail- ways. These quantitative measurements serve only as preliminary observations which are to be confirmed and explained further through qualitative investigations in the following chapters. Decomposition of Operational Efficiency Attempts have been made to compare the efficiency of railroads internationally by productivity measures. Such an approach contains serious drawbacks. First, a unit of output can be produced by various possible com- binations of several production factors. The meanings 27 of n tivi anal rela‘ factc be mi schem The r uni t- specifi. giVen ti «:7 '0}: 28 of measures such as productivity of labor, capital produc- tivity are therefore equivocal, at least in micro-economic analysis of the firms. They seem to suggest that a causal relationship exists between output and one production factor or more whereas in reality the relationship should be much more complex. Aggregating all factor inputs by some weighting schemes or by monetary values overcomes this shortcoming. The resulting measures, i.e., all-factor productivity or unit-cost, can be used to compare the relative efficiency among railroads. Difficulties arise, however, because a railroad's efficiency consists of technical and cost elements which are determined by factor substitution within the constraints of technical feasibility and cost minimi— zation. Measures of all-factor productivity and unit-cost do not provide any means to separate technical from cost component. Nor do they account for the possible effects of economies of scale if these economies are present in railroad industry. The theory of production function appears to be capable of correcting these defects. The Concept of Production Function The production function for a firm, by definition, specifies the maximum output attainable from a set of inputs wgiVen_the technology available to the firm.1 This maximum 1. Sune Carlson, A Study on the Pure Theory of Pro- duction (New York: A. M. Kelly, 1957), pp. 14-15. output firm 0 indust indust to Sim from ti latter and egg technic frontie functic in that Stant r¢ graphic; (The efj aCCOUnt the mini t° produ 29 output conceptually applies not only to the particular firm of interest but also to all other firms in the same industry. We might call the function, so defined, the industry (frontier) production function (hereafter referred to simply as the production function) to be distinguished from the industry's aggregate production function. The latter expresses the relationship between aggregate output and aggregate inputs of that particular industry.2 The production function, by definition, presupposes technical efficiency. Any observed deviation from the frontier of potential attainment specified by the production function indicates the presence of technical inefficiency in that particular firm. In the case of only two inputs, X and X and con- 1 2’ stant return to scale, the production frontier can be graphically illustrated by an efficient unit isoquant (EUI). (The effects of economies of scale will be taken into account later.) An EUI is a locus of points indicating the minimum quantities of the production factors required to produce one unit of output with varying factor proportion. Technical Efficiency In Figure 3, firms B and C are on the EUI, hence, technically efficient. Firms A and D, on the other hand, 2. D. J. Aegner and S. F. Chu, "On Estimating the Industry Production Function," The American Economic Review, Vol. 58, No. 4 (September 1968), p. 828. (Input unit . put Q Fi 30 Xl/Q (Input x1 per unit of out- Put Q) (Input x 2 per unit of output Q) Figure 3. Graphical illustration of technical, cost and operational efficiency. are u outpu techn. to the effici the ra fore, B and wherea like A isocos1 311 pos Purchas eConomi 31 are using more of both inputs to produce one unit of output, hence, technically inefficient. The relative technical efficiency (TE) index of firm A is equivalent to the ratio OB/OA. Similarly, the relative technical efficiency indices of firms B, C, and D are measured by the ratios OB/OB, OC/OC and Od/OC, respectively. There- fore, the TE indices of technically efficient firms like B and C equals to l (OB/OB and OC/OC, respectively), whereas the TB indices of technically inefficient firms like A and D are less than 1 (OA/OB and OD/OC, respectively). Cost Efficiency Given the relative prices of production factors, isocost lines can be drawn._ Each isocost line represents all possible combinations of factor inputs which can be purchased by the firm at a given total cost. The most economical or least-cost combination of factor inputs which can be used to produce one unit of output is deter— mined by the point of tangency between the EUI and the lowest isocost line. In Figure 3, firms B and C are both technically efficient, but only firm C is also cost efficient. All points on isocost line PCP; represent the same total expenditure. It is therefore possible for a technically and cost efficient firm like C to produce a unit of out- put at the cost of only OR. Firm B spends OB, so the ratio Point. repre: the CI (OR/OE from t of fir indice. The fi average PrOduCi firm c inputs, °Vera11 ShOrtly repress] on i50c( AUC'S t} 0f firm legs tha 32 ratio OR/OB is the cost efficiency (CE) index of firm B. Points A and B are on the same ray from the origin, thus, representing identical factor prOportions. In other words, the CE index of firm A is the same as that of firm B (OR/OB). Similarly, points C and D are on the same ray from the origin, hence, the CE index of firm D equals that of firm C (OC/OC = 1). Operational Efficiency Let us next examine the products of the TE and CE indices: - for firm A : (OB/0A) x (OR/OB) (OR/0A) - for firm B : (OB/OB) x (OR/OB) (OR/OB) x x - for firm D : (CC/OD) (QC/0C) (OC/OD) = (OR/OS) - for firm C : (CC/0C) (OC/OC) = l The first three ratios are equivalent to the ratios of the average unit cost (AUC) of producing at C to the AUC's of producing at A, B, and D, respectively. Also, recall that firm C represents the most economical combination of factor inputs. These ratios can be regarded as representing the overall operational efficiency (hereafter referred to shortly as operational efficiency-~0E).of the respective firms. Firm C lies on the lowest isocost line, PCPé, which represents the lowest possible AUC. Firms A, B, and D lie (on isocost lines which are above PCP; thus incur higher AUC's than theoretically possible. If we define the OE index of firm C as 1, then CE indices of firms A, B, and D are less than 1. pariSt effec' facto; 33 In summary, firm C is operationally efficient, with all of its TE, CE and OE indices being equal to 1. Firms A, B, and D are Operationally inefficient with their respective OE indices being less than 1. However, the inefficiency of firm A are due to both technical and cost effects (both its TE and CE indices are less than 1, too), whereas the inefficiency of firm B is associated with only cost effect (TE = l but CE < 1), and the inefficiency of firm D is explained by only technical effect (TE < l, but CE = 1). As seen, the use of EUI provides not only AUC com- parison but also a means to separate technical from cost effects on efficiency, something that a simple cost or all- factor productivity comparison cannot do. Method of Estimation "Average" Versus Frontier Production Function The indices of technical, cost and operational efficiency are defined above in relation to the frontiers of potential attainments for given input combinations and factor prices. It is conceptually inapprOpriate to esti- mate these frontiers by techniques such as least-square regression which allow negative and positive residuals. A production function estimated as such, for example, represents only an "average" relationship between output and factor inputs since a number of observations lie below and a; actual produc clearl functi or ign. ing pr. ration; techni( for thg than at the me; aVeragE I About 1 means 1 312e," Paramei their E $12e.n 34 and gbgy§_the fitted function. This means that the firm's actual output may either be smaller or greater than the production function permits. Such an average relationship clearly conflicts with the definition of the production function. Such conflict has mostly been either rationalized or ignored in favor of the convenience of statistical test- ing provided by regression technique. It is frequently rationalized that a production function fitted by regression technique is conceptually an "average" production function for the industry; some firms could therefore produce more than average, some less. Aigner and Chu (1968) question the meaning of "average being rationalized as such, i.e., average about what? About output? About some inputs? About technology? Or about something else? If average means that the fitted function is for "firms of average size," then, one must make an unlikely assumption that the parameters of the function are random variables and have their expectations equal to those of the firm of "average size." Similarly, if average refers to "average technology," then, it is difficult to assume that firm which possesses "average technology" with respect to capital also has an "average technology" with respect to labor.3 3. D. J. Aegner, et al., "The Industry Production Function," op. cit, pp. 829-830. spec BEES a pr. effi: thus econ< shori 35 Estimation of Technical Efficiency One possible approach of estimating the frontier specified by the production function is to develop a process function on the basis of engineering data. Such a process function describes the technologically most efficient way of accomplishing the task of production, thus, represents the production function postulated in economic theory. A process function, however, has several shortcomings: 1. Since engineering data encompasses only technical process of production, a process function does not cover all activities of the firm. 2. Engineering process estimation is difficult to obtain especially where the process is a complex one. 3. A process function is established on an ideal labora- tory situation, thus, cannot guarantee that these engineering relations can be realized in actual operations. Instead of estimating the engineering process function, Farrell (1959) suggests fitting an envelope to the sample of observations.4 The frontier specified by Farrell's method represents the best practices observable which may 4. M. J. Farrell, "The Measurement of Production Efficiency," Journal of the Royal Statistical Society (Series A, General), Vol. 120, Part 3 (1957), pp. 243-281. Input 2 unit of PUt Q 36 xl/o Input x1 per 1 unit of out- PUt Q Fitted EUI o XZ/Q Input X2 per unit of output Figure 4. Farrell's method of estimating the EUI. 37 or may not be up to the ideal standard set by engineers. Since our purpose is to measure efficiency only in its relative terms among the railroads studied rather than its absolute level, Farrell's method is judged as appropriate. Fitting an envelope to a sample of observations can be done by either: I - l. estimating an EUI as originally proposed by Farrell (1957),5 further developed by Boles (1966)6 and used by Seitz (1970);7 or i 2. estimating the production function per se as suggested by Aigner and Chu (1968)8 and used by Timmer (1971)9 and Carlsson (1972).lo 5 0 Ibid. 6. James N. Boles, "Efficiency Squared - Efficient Com— putation of Efficiency Indexes," Western Farm Economics Association, Proceedings (1966), pp. 137-142. 7. Wesley D. Seitz, "The Measurement of Efficiency Relative to a Frontier Production Function," American Journal of Agricultural Economics, Vol. 52, No. 4 (November 1970), pp. 505-511. 8. D. J. Aigner and S. F. Chu, "On Estimating the Industry Production Function," 0p. cit. 9. Peter Timmer, "On Measuring Technical Efficiency," Stanford University Food Research Institute Studies in Agri- cultural Economics, Trade and Development, Vol. 9, No. 2 (1970): PP. 99-170. 10. B. Carlsson, "The Measurement of Efficiency in Produc- tion: An Application to Swedish Manufacturing Industry 1968," Swedish Journal of Economics, Vol. 74, No. 4 (1972), pp. 468-488. If.“ [0 38 As previously discussed, an EUI is fitted in an input-input Space and requires an assumption of constant return to scale. Where economies of scale is suspected, the scale effects can be taken into account by a grouping method suggested in Farrell and Fieldhouse (1962).11 Accordingly, the sample of observations is stratified into. sub-samples, each contains only observations of similar scale. Economies of scale, if they exist, will affect all observations in a particular sub-sample the same way. It is then possible to fit several EUIs, one for each sub- sample, as if no economies of scale were present. This grouping method is, however, impractical where the sample is not sufficiently large to create several homogenous sub-samples. Besides, where scale variable is continuous rather than discrete, grouping method can only reduce but never eliminate the problem. Fitting a production function per se in an output- input space, on the contrary, does not require the assumption of constant returns to scale. As graphically illustrated in Figure 4, when the amounts of inputs increase from XA1 and XB1 to XA2 to XBZ' respectively, the volume of output rises from Ql to 02, but at a higher rate than do inputs. When the amounts of inputs increase from XA2 and XB2 to XA3 and 11. M. J. Farrell and M. Fieldhouse, "Estimating Efficient Production Function Under Increasing Return to Scale," Journal of the Royal Statistical Society (Series A, General), Vol. 125 (1962), pp. 252-267. 39 Output 0 Input XA Q3 33 Q 9 XA ---- ...221. 1 ....... 3 ? ' l7 I E / I l 1 Q ” XA ........ --l-L I 2 I‘ I, XA ’/ I, 1 ““““““ ‘ l’ / / I I I , I ’ I 1’ I’ I ’1 [I ’l ’ I ‘1 LII ,’ XB1 XB2 XB3 Input XB Legends: (0) potential output (x) observed output Figure 5. The frontier production surface. in CC SC 115‘ C0] Us: the fur as Par In bet str thdj tio] 40 x33, respectively, the amount of output rises from Q2 to Q3, but at a lower rate than do inputs. In other words, output and inputs do not vary proportionally. How well the effects of variable returns to scale can be taken into account depend upon the shape of the production surface which is, in turn, determined by the functional form of the production function. The primary objective of fitting a production function in this study is to provide a means of estimating technical, cost, and Operational efficiency rather than to determine the shape of the function itself or to estimate return to scale or the properties of factor substitution. It is therefore unlikely to make any significant difference to use one functional form or another. Out of convenience, a Cobb-Douglas form of production function is assumed and used in conjunction with Aigner and Chu's method of fitting the production function. Conceptually, this method of fitting a production function means bringing the production surface as close as possible to the observed outputs without having any particular observed output lying above the fitted surface. In other words, the method minimizes the distances E's between observed and potential outputs subject to the con- straint that no particular observed output could be higher than its potential level as specified by the fitted produc- tion surface. Any violation of this constraint would mean Wh 41 that a particular combination of inputs had produced an output higher than its potential; and this would be con- tradictory to the definition of the production function as containing the maximum outputs attainable. This is clearly an optimization problem. When used in conjunction with a Cobb-Douglas production function in its general form (in logarithm) m 1n Qj = a0 + iglal 1n xij + 1n Ej this method of estimation is equivalent to a linear pro- gramming problem (see mathematical proof in Appendix A): n m A minimize: Z ( Z a. 1n X.) , . 1 1 3:1 1:0 m A subject to: .2 oi 1n Xij 1 ln Qj 1—0 and: mi 1 0 where: Qj : output of firm j a0 : the intercept Xij: input of production factor i by firm j Ej : a random error term that contains technical efficiency term as well a. : factor elasticities (i = 0) H .lalttFSLCECTHF.T.w 42 A Solving this LP problem, we obtain vector 1n Qj-- potential outputs (in logarithm). The ratio Qj/aj is the ratio between actual and potential output, hence by definition, it is the TE index of firm j. It should be noted that the above-described method does not make any assumption on the statistical properties of the error term. Statistical testing is therefore impos- sible. However, the effort to measure relative efficiency among the railroads, as previously discussed, is not aimed at statistical testing of any specific hypothesis, but only serves to identify certain areas for further qualita- tive investigations. Attempts to make statistical assumptions about the error term will only introduce complications unnecessary for the purpose of this study.12 Estimation of Cost Efficiency Cost efficiency index can be estimated by a similar method with the help of the cost function which specifies 12. For those interested in making such statistical assumptions, there are few works which can be referred to. It should also be noted that these few works are still in their early development. For example: Peter Schmidt, "On the Statistical Estimation of Parametric Frontier Production Functions," The Review of Economics and Statistics, Vol. 58, No. 2 (May 1976), pp. 238-239; D. J. Aigner, C. A. Knox Lovell and P. Schmidt, "Formulation and Estimation of Sto- chastic Frontier Production Function Models," Journal of Econometrics, Vol. 6 (1977), pp. 21-37; and P. Schmidt and C. A. Knox Lowell, "Estimating Technical and Allocative Inefficiency Relative to Stochastic Production and Cost Frontier," Econometrics Workshop Paper, No. 7702 (East Lansing: Michigan State University, December 1977). 43 the minimum possible total costs of producing various out- puts at different levels of factor prices. Narlove (1965) derived the cost function from the Cobb-Douglas production function as follows (Logarithm):13 m . = + . ..+ ,+ , 1n TCj Y0 iglyl 1n P13 Ym+1 1n QJ ln UJ where: TCj : the total cost of firm j Qj : the output of firm j Pij : the price of factor i paid by firm j U. : a random error term that contains efficiency term as well The cost function contains both input prices Pij and output level Qj; it therefore can take into account the effects of both upon total cost. In figure 6, for example, the (minimum possible) total costs of producing output Ql at input prices P1 and P2 are TO; and TC', respectively. At a given input price P2, on the other hand, the total costs 2 Note that the increase in total cost from TC of producing outputs Q1 and Q3 are TC and TCS, respectively. I I . 2 to TC3 is not proportional to the increase in outputs from O1 to Q3. In 13. Marc Narlove, Estimation and Identification of Cobb—Douglas Production Functions (Chicago: Rand McNally, 1965). 44 Total Cost TC ’f TC4 Factor I TCGQ TC3: Price P (x I . ITC", ’ ' I ’ ’AI/ TCGP c:2 ,..a'?'c'3,’ 4T 1’ [II I l ,’ /" l ” TC I TC'- I“- - -- “vi 1 ’ I , I / I, / / ’/ .;;- / /’/ l /’ ,’ 1’ ’l ’l 0 Q1 Q3 Q4 Production Quantity Q Legends: (0) potential total cost (x) observed total cost TCGQ: total cost given particular production quantity TCGP: total cost given particular factor price Figure 6. The frontier cost surface. 45 other words, the cost function can take into account the possible effect of variable returns to scale; something which a simple measure of average cost cannot do. As afore-defined, the cost function sets the floor below which no observed total cost could fall whereas the production function sets the ceiling above which no observed output could rise. Estimation of the cost func- tion is therefore the reverse case of that of the production function. In other words, the cost surface is fitted as close as possible from below to the observed total costs, subject to the constraint that, given particular levels of input prices, an output could not be produced as an (observed) cost lower than its corresponding potential level. Mathematically, this is equivalent to the following linear programming problem (see Appendix A for mathematical proof): n m A A Maximize: j£l(i£0Yi 1n Pi + Ym+l Qj) mA A Subject to: igoYi ln Pi + Ym+lQ < 1n TC and Yi 1 0 Solving this LP problem, we obtain vector ln TCj, the potential total cost of producing a particular output at certain input prices (in logarithm). The ratio TCj/TCj is fr in 0:1. pa: Out twc Sho rel 46 the ratio between potential and actual total cost, hence by definition, also the CE index of firm j. Estimation of Operational Efficiency Once the CE and TE indices are obtained from the above LP problems, the OE index can be calculated from the relationship previously shown as: (OE index) = (TE index) x (CE index) This relationship was derived mathematically from the efficient frontiers, and hence represents a conceptual relationship. It is therefore independent of the statis- tical techniques used to fit the production and cost functions. Specification of the Railroad Production and Cost Functions A Method of Dealing with Two Outputs Railroads produce two main categories of services-- freight and passenger--the outputs of which can be measured in ton-kilometers and passenger-kilometers, respectively. One cannot, however, simply add the ton-kilometers and passenger-kilometers produced by a railroad as its total output since factor inputs required for each unit of the two services are unlikely the same. Ideally, the two outputs should be weighted by their relative costs. But since the relative costs are frequently unknown, Munby (1962) suggests 47 weighting the outputs by their Contributions to total reve- nuel4 on an implicit assumption that the charges for these services reflect their relative costs. This assumption is not likely to hold in the case of nationalized railroads. Klein (1953) treats the two outputs as competing for the same pool of resources. Given a combination of factor inputs, the output of freight services is increased at the expense of passenger output, and vice versa. One of the two outputs can therefore be treated as equivalent to a negative input.ls Measurement of Factor Inputs Railroad's capital stock consists of motive power, rolling stock, trackage and line facilities (i.e., railyards, stations and signalling equipment). It is reasonable to assume that motive-power and rolling-stock capacities move closely together and so do track mileage and line facilities. The same thing cannot be, however, assumed between motive- power and rolling stock capacities, on one hand, and track mileage and line facilities, on the other. Track mileage is determined by the geographical configuration of a trans- portation system rather than solely to any specific output 14. D. L. Munby, "The Productivity of British Railways," Bulletin of the Oxford University Institute of Statistics, Vol. 24, No. 1 (February 1962), p. 115. 15. Lawrence R. Klein, A Textbook of Econometrics (Evans- tons, 111.: Row, Peterson and Co., 1953), pp. 227-236. 0r ru me. P02 kiJ 48 level demanded by the market. The latter factor, on the contrary, largely determines the capacity of motive power and rolling stock. It is therefore necessary to approxi- mate these two general categories of capital stock separately. In this study, motive power is selected over rolling stock so as to avoid the complication in totalling passenger and freight car capacities. Similarly, track mileage is chosen over lines facility due to the simplicity of its measurement. Should motive power input be measured in "flow" term, i.e., annual locomotive-kilometers run, or in "stock" term, i.e., the number of locomotives in services? Where output is clearly a flow of services, i.e., ton-kilometers, it seems logical that input of capital equipment should be measured in flow term. A difficulty arises, however, because freight-ton-kilometers and locomotive-kilometers move very closely together. If the input of capital equip- ment is measured in flow term.with a distance dimension, and the measure of output contains the same distance dimen- sion, very nearly the same thing appears on both sides of the equation. One almost relates output against itself, or in other words, distance run cannot "explain" distance run. The selected motive-power variable is therefore measured in this research in stock term, i.e., total horse power available, and, of course, so is trackage, i.e., kilometers. 49 Should these stock variables be adjusted for their rates of capacity utilization? In contrast to the common practice, this study selects not to. As pointed out earlier, the production function, as commonly estimated by regression technique, is actually an industry's aggregate production function expressing an "average” relationship between aggre- gate output and aggregate inputs. Inefficiency is either overlooked or assumed non-existent. Adjusting the stock variables for their utilization rates are therefore logical for such a purpose. This study is, on the contrary, aimed at estimating the efficient frontier Specified by the pro- duction function. Capital stocks are not adjusted for their rates of capacity utilization. Such adjustments simply eliminate or reduce the differences in the efficienty among the railroads which this study attempts to measure. For the same reason, labor input is approximated by the number of employees rather than by man-hours worked. The latter measure is simply the former being adjusted for absentee rate and work schedule. By measuring labor input as the number of employees, the difference in these factors among the railroads, which in turn reflect inefficiency, can be accounted for. The Railroad Production Function In summary, the railroad production function used in this study is as follows: 50 = - , + , + , + , 1n Qth a0 a1 1n QP3t a2 1n th a3 1n TMJt a4 1n th + a In F. +-ln E. 5 3t ]t where QF : output of freight service, measured in ton-kilometers; QP : output of passenger service, measured in passenger- kilometers; M : motive power input, measured in total horsepower available; TM total trackage, measured in kilometers; L : labor input, measured in annual number of employees; F : fuel consumption, measured in BTUs of energy con- sumption for traction efforts. Others are defined as before and subscripts j and t refer to railroad j in time period t. The LP problem of estimating the efficient frontier becomes: 0 a o - A + A A A minimize do cl ln QPjt a2 1n th + a3 ln TM t + a4 ln th + “ . a5 1n th O - A + subject to a0 al In QP t a2 1n th + a3 ln Tth + a4 1n th + as 1n th.i ln Qth dc fr 51 The Railroad Cost Function Since the cost of capital inputs are measured in monetary terms, they can be approximated by one variable rather than two variables as in the production function. The railroad cost function as used in this study is then written as (in logarithm): = + F . + . + , ln TCjt YO Yl 1n Q at 72 ln QP3t Y3 1n KC3t + + + Y4 ln WAjt 75 1n FCjt 1n th where TC : total cost; KC : capital cost, measured as long-term interest rate; WA : wage rate, measured as annual labor cost per rail employee; FC : fuel cost per billion of BTUs; Others : as previously defined. (All cost figures are to be expressed in constant U.S. dollars.) The LP problem of estimating the efficient cost frontier becomes: minimize ?0 + §l 1n Qth + $2 ln QPjt + §3 ln KCjt ln FC. A +A +Y4 1n WAjt Y5 jt 52 . A A . + A . A subject to Y0 + Y1 1n Qth Y2 1n QPjt + Y3 1n KCjt + §4 ln W'Ajt + ?5 ln cht 5,1“ TCjt and §i Z 0 Preliminary Results of Efficiency Measurements Using the afore-described LP technique, the produc- tion and cost functions are estimated from a set of 360 time series and cross-sectional observations. These observations cover twenty plus railroads from industrial and developing economies for the period of the late 19505 to 1974.16 The estimated production and cost functions Specify the frontiers of production and cost possibilities, respectively, against which the efficiencies of each observation are measured. The technical (TE) index of an observation is measured in relation to the potential output as being specified by the production frontier for a technically efficient (but hypothetical) observation which used the same combination (the amount as well as proportion) of 16. This timeframe is a little bit shorter than the one in the following qualitative investigation which goes back to the early postwar years. This limitation is due to the availability of statistical data. During the mid-19503, the International Union of Railways revised some of its methods used in compiling and presenting railways' opera- tional statistics. The time limitation is therefore neces- sary to keep the measurements consistent. Statistical data are from: International Union of Railways, International Railway Statics (Paris: The UIC Statistics Bureau, annually). 53 factor inputs. Thus, the TE index of an observation represents the percentage of the factor inputs actually employed by the observation which could have produced the same output had it been technically efficient. A TE index of 0.70 for observation A, for example, indicates that had A been technically efficient, it could have produced the same level of output with only 70 percent of the amounts of the factor inputs which it actually used. Similarly, the cost efficiency (CE) index of an observation is measured in relation to the potential total cost as being specified by the cost frontier for a (hypo- thetical) observation which were able to substitute freely among factors of production according to their relative prices so as to minimize the cost of producing the same output. A CE index of 0.60 for observation B, for example, indicates that had B substituted efficiently among produc- tion factors, it could have been able to produce the same output for 60 percent of the cost it actually incurred. Caution should, however, be taken in interpreting these indices. First, the frontiers of production and cost possibilities, by virtue of the estimation technique used, contain the best practices observable as opposed to 'the standards specified by the engineering process. An index of 1.00, therefore, only means the best performance among the sample and does‘ggt necessarily indicate perfec- tion. The prices of some production factors varied greatly 54 among the observations. The wage rates paid by some rail- roads in industrial economies, for example, were as high as eight times or more, and also increased at a faster rate than those paid by their counterparts in develOping econo- mies. Given the importance of labor cost, normally amounting to 50 percent or more of total cost, this indicates a very wide range of sbustitution of capital for labor inputs before the railroads in industrial economies could minimize cost ' as effectively as some of their counterparts in developing economies. Such substitution though economically desirable, may be limited by technical feasibility. In other words, even without institutional constraints, it may not.be com- pletely possible for a railroad being plagued by excessively high labor cost to bring its CE index to 1.00 due to techni— cal limitations. It is therefore more meaningful to compare the relative levels of efficiency among railroads with comparable Operating conditions than to examine their absolute levels of efficiency. Technical Efficiency The TE index (Table 1) shows a general trend of improvements, though at markedly different rates, among the railroads' technical efficiency, especially since the mid- 19603. The highest levels of technical performance are exemplified by the Japanese National Railways (JNR), the Canadian National Railways (CNR) and the Canadian Pacific 55 Table 1 Technical Efficiency Indices of Some Major Railway Systems SNCF BR DB JNR CNR FS 1974 1.00 - .64 .97 - .43 1973 .95 - .62 .89 .96 - 1972 .89 .32 .58 .86 1.00 .41 1971 - .34 .57 .83 .94 .41 1970 .87 .37 .59 .80 .89 .43 1969 .84 .34 .58 .74 .82 .42 1968 .71 .33 .48 .70 .83 .42 1967 .61 .28 .42 .58 .80 .39 1966 . .49 .26 .39 .64 .85 .34 1965 .45 .23 - .62 .72 .32 1964 .43 .21 .41 - .75 - 1963 .42 - .42 .66 - - 1962 .41 - .39 .64 - - 1961 .39 .18 .37 .64 - - 1960 .36 - .33 .61 .53 - 1959 .33 - — .56 - - .29 .16 - - - - 1958 56 Railways (CPR) as well as by the French National Railways (SNCF) since the late 1960s. The SNCF shows the fastest rate of technical improvement. At the lower end of tech— nical performance are the British Railways (BR) and the Italian State Railways (FS) whose improvements are also relatively slow. In the middle of the scale is the German Federal Railways (DB) whose technical performance had been comparable to the SNCF’s until the mid-19603, but since have been outspaced by the latter. Are the gaps in technical efficiency mere reflections of the natural environments in which the railways operate, i.e., geographical terrain, network configuration and length of haul? In light of the widening and narrowing of thegaps in technical performance, the answer seems to be negative. These factors are relatively constant over time. Had they accounted for the gaps in the technical performance among the railroads, the gaps should have remained very much the same instead. Cost Efficiency The CE index (Table 2), on the other hand, shows no general trend of improvement; and excluding the JNR, the gaps in the cost efficiency among the railroads are rela- tively narrow. Among the five railroads shown, only the CNR achieves constant increases in cost efficiency, though very slightly. The JNR, the FS, and the DB, which are also 57 Table 2 Cost Efficiency Index BR SNCF DB ps JNR CNR 1974 - .46 .24 .44 .50 - 1973 - .48 .26 - .49 .47 1972 - .49 .28 .30 .52 .48 1971 - - .30 .34 .55 .48 1970 - .54 .35 .37 .60 .47 1969 - .48 .36 .38 .58 .45 1968 - .43 .48 .38 .60 .45 1967 - .40 .31 .43 .70 .46 1965 - .39 - .41 .73 .41 58 ranked in the same descending order with respect to cost efficiency, show constant decline; and so does the SNCF since 1970 after a period of improvement. It is normally expected that improved technical per- formance would be translated into some cost savings, hence higher cost efficiency. The existence Of a sluggish trend in cost efficiency along side technical improvement, how? ever, seems to indicate that the latter has been wiped out by the railroads' inability tO contain costs. This con- clusion is reflected in the operational efficiency (OE) index which combines the effects Of technical and cost performance. Only the SNCF seems to be able to outpace rising costs with its technical improvements (Table 3). In summary, the foregoings seem to validate one of the research hypotheses that the gaps in the operational efficiencies among the railroads are significantly large. The OE index ranges from the 0.105 for the DB and the F5 to the 0.405 for the JNR, the CNR and the SNCF; the gaps are even higher with respect to technical performance. The foregoings also seem to indicate that, among the four railroads to be investigated in this study, the JNR exemplifies the highest level of performance; it is fol- lowed very closely by the SNCF; the performance ofthe DB is significantly lower. The insufficiency in the BR's cost data makes it difficult to estimate the BR's cost efficiency, hence, also its Operational efficiency. However, given the 59 Table 3 Operational Efficiency Index BR SNCF DB FS JNR CNR 1974 - .46 .15 .19 .49 - 1973 - .46 .16 - .49 .45 1972 - .44 .17 .12 .44 .48 1971 - - .17 .14 .46 .45 1970 - .47 .20 .16 .49 .42 1969 - .40 .21 .16 .43 .37 1968 - .31 .16 .16 .42 .37 1967 - .24 .13 .15 .45 .45 1966 - .19 .12 .15 .45 .45 1965 - 018 - .13 .46 .41 60 very low level of technical performance by the BR, it is not unreasonable to expect the Operational efficiency of this railway system to be as low as that Of the DB if not even lower. It should be recalled that the production function which is instrumental to the estimation Of technical efficiency contains the measurements of the railroads' network size (in terms of track mileage), labor force, the state Of technology (as approximated by traction power and energy consumption) and production outputs. The level of. technical efficiency is therefore related not only to modernization efforts but also to the achievements in rationalization Of network and labor force as well as to the ability of the railroads to compete for a desired market share. These factors will therefore be investigated in the following chapters along with the institutional constraints upon the ability Of these enterprises to contain rising costs. CHAPTER III MANAGERIAL AUTONOMY AND LEGAL MODELS OF ORGANIZATION The Objective Of this chapter is to investigate the relationship between the degree Of managerial autonomy and the legal models of organization among the selected railroads. Managerial autonomy in the conduct of state enterprises has two aspects: directive and procedural. The former involves a decentralization of discretion and command while the latter consists Of a relaxation of certain restrictions which apply to ordinary governmental Operations. The procedural aspect may exist without its directive counterpart, especially when the aim Of the creation Of an "autonomous" state enterprise is to provide the public authorities concerned with admini- strative convenience and to relieve them from certain requirements of accountability rather than to delegate to management the powers necessary for business Operation. With this in mind, the following discussion will be divided into two parts. The first part focuses on the statutory arrangements of each Of the railroads studied. Together with this, the motives behind, and the circumstances sur- rounding, the nationalization Of a particular railroad and/or the adoption Of a specific legal model Of organization are 61 he he Wi 5C ha m0 62 also discussed. The second part examines the working of the tOp managerial bodies of the railroads so as to deter- mine where, and to what extent, the powers of policy—making and those of day-to-day management lie. The two parts thus correspond to the procedural and directive aspect Of mana- gerial autonomy, respectively. Legal Models of Organization The Public Corporation in the United Kingdom The current system Of the British Railways (hereafter referred to as the BR) came into being through nationaliza- tion in 1947. Until then, the debate on railroad nationali- zation in Great Britain had been conducted only on questions of broad principles rather than urgent practical necessity due to the fact that the majority Of the railroads had been moderately profitable up to the end of W.W.I. Instead of nationalization, a complex system of government regulation had been established as a safeguard for public interests. With the advent of a Labour Government in 1945, however, some changes took place in public policies. The Motives Behind Nationalization _ AlthOUgh the Labour Party had long advocated the nationalization of transport on the grounds Of the natural monopoly inherent in this industry, political ideology was Ac of 0r SY re: 63 not the main force behind practical policy. Given the. complex system Of government regulation in force at the time, there was hardly any room left for the instrument of public ownership to further the safeguard of public interests against monopoly power. It was the urgent needs for an overhaul and modernization Of the railroad system which prompted nationalization. The railroads had not received prOper maintenance and modernization from their private owners during the pre-war years, and then deterio- rated further through wartime efforts. The costs Of railroad overhaul and modernization were likely to exceed the capability Of private capital. The Labour's policy on public ownership, was, however, directed toward a more comprehensive scheme. Accordingly, railroad nationalization formed only a part of a plan for transferring to public ownership the entire, or at least a very large portion, Of the inland transport system. It was anticipated that by so doing, it would result in more rational allocation Of resouces. Subse- quently, the Transport Act of 1947 nationalized four mainline railways together with their associated lines, docks, steamships, and hotels, the London Passenger Trans- port Board and inland waterways. These would then be placed under a single control of a newly-created public corporation, namely, the British Transport Commission. B< b1 of th in it me. in Wi1 64 The BTC was charged with Operating these undertakings as an integrated system.1 The British Model of Public Corporation The road to nationalization witnessed an important development, namely the creation of statutory public corporations for various public utilities. The principal architect of these contemporary organizations is generally agreed.to have been Herbert (later Lord) Morrison, who, as Minister of Transport in the Labour Government of 1929- 1931, prepared the plan for the London Passenger Transport Board. In the process, Morrison arrived at his conception of a compromise between a public service and a private business at a time when Labour movement was in the throes of inescapable managerialism. A public corporation was then envisaged as a financially autonomous, non-profit-making institution which was managed by non-political appointees; it was created by statute, ultimatelyresponsible to Parlia- ment through the supervisory Minister for policy decisions in the public interests, but Operated along commercial lines without being subject to close governmental control over day-to-day management. Morrison was powerfully supported 1. See, for example: Guilbert Walker, "The Transport Act 1947," Economic Journal, Vol. 58, NO. 1 (March 1948), pp. 11-30; Edgar Milhaud, "Nationalization in Great Britain," Annals of Collective Economy, Vol. 19, NO. 2 (April-June 1948). Pp. 65-157; and Derek H. Aldcroft, British Transport Since 1914 (London: David and Charles, Inc., 1975), pp. 70-73. 65 by the Labour Prime Minister, Lord Attlee, whose experience in running the postal service as a government department had convinced him of the need for managerial autonomy.2 It Should, however, be noted that the post—war public corporations deviated in several respects from the pre—war Morrisonian concept. While the post-Keynesian recognition Of the needs for coordinated macroeconomic planning was gaining its popularity, the illusion that such planning process could be non-political had been dis- pelled by wartime experience. It then became more acceptable to grant Ministers with more comprehensive powers over the corporations than what had been envisaged in the pre—war years. SO, in addition to the formal powers Of issuing general directions on matters Of national interests, there emerged from most nationalization acts six major groups Of specific ministerial powers. They were: 1. Patronage power in appointing the Board Of Directors, consumer councils and auditors for each corporation as well as to determine their tenure and remuneration; 2. Power to give Specific directions on certain matters, i.e., safety; 3. Discretionary power to approve various matters such as capital investment programs and the general conduct Of finance, training, and research; 4. Power to coordinate various public corporations under the Minister's control; 2. J. E. S. Hayward, "Recent British Nationalization Theory and Practice in Retrospect," Annals of Public and Co- -opera- tive Economy, Vol. 33, NOS. 1- 2 (January-June 1962), pp. 33- 49. 8) ti da th en p01 in - 30m Wan 66 5. Power to Obtain information about the activities Of the corporations, not merely for the Minister's en- lightenment but also for that of Parliament and the public; and 6. Power to hold public inquiries Of quasi-judicial character. - It was these powers which led to the growth Of a secondary area of non-statutory intervention by Ministers. This was not the case where the Ministers abused their statutory powers but rather that the statutory powers had enabled the Ministers to extend their influence beyond the area envisaged when the incorporating statutes were enacted. Non-statutory interventions of this type seriously challenged the doctrine Of parliamentary accountability. In fact, it had been a common practice among British ministers to excuse themselves from answering many parliamentary ques- tions by referral to these questions as matters Of day-to- day management. SO, although Parliament is theoretically the ultimate guardians of the public interests being entrusted in nationalized industries, their supervisory powers over these organizations have been greatly reduced by the concept of public corporation.3 3. See, for example: J. F. Garner, "Public Corporations in the United Kingdom," in W. G. Friedman et. al., ed., Government Enterprise, op. cit., pp. 3-25; J. E. S. Hay- ward, "Recent British Nationalization," op. cit.; and A. H. Hanson, Parliament and Public Ownership (London: Cassell, 1961). IH 96 wi We gr me Al Ge: TAG OJ PP. ’1. cha 67 The National Company in France Quite in contrast to the British case, railroad nationalization in France was characterized by a less ambitious plan and an absence of enthusiasm for any parti- cular legal model Of organization. In addition, by the time of nationalization in 1937, the French government had gained considerable experiences in direct ownership and Operation of some railroad networks. The Road to Nationalization Direct government Operation of railroads dated back to 1908 when the government created the Etat network out of the bankrupt Western Railway. This government network was placed directly under the supervision of the Minister of Public Works who approved its budget and accounts; a general manager was appointed by decree to run the network with the assistance of a board Of management whose members were also appointed by the_government; and employees were granted civil servant status. In 1919, the span of govern- ment participation was expanded with the creation of the Alsace-Lorraine network which had been recovered from the Germans.4 The two government networks Operated along with ........ 4. For details on these State networks, see: Pierre Levy, The.Rai1roads in France," The Annals of the American Academy of Political and Social Sciences, NO. 187 (September 1936), pp. 184-192; and Kimon A. Doukas, The French Railroads and the State (New York: Columbia University Press, 1945), Chapter 7. fi‘ la1 gue fox rig CdI W.W sid rai qua. of ' rate Were up I rate 68 five other concessionary railroads. The finance of the latter had been supported by a system Of government guarantee of interest and dividend since 1882. In return for this public aid, the companies had been subject to rigid control Of rates, services and finance as well as to carrying out the government's rail construction program.5 The deficits borne by the government during the W.W.I., were so substantial that they necessitated recon- sideration of the relations between the government and the railroads. The resulting Convention Of 1921 replaced the guarantee system with a Common Fund into which the surpluses Of the private and public networks over the fixed dividend- rates on their stock were to be paid and from which deficits were to be covered. A deficit in the Fund was to be made up by rate increases while a surplus would be followed by rate cuts. The Operation of the Common Fund required that the railroads should act as uniformly as possible. The 1921 Convention therefore provided that although each net- work was allowed tO keep its own internal organization and individual management, together the seven networks would be subject to the supervision of a Supreme Railway Council (Council Superieur des Chemins de Fer) and an Executive Committee (Comité de Direction). The Council was purely 5. H. E. Dougall, "Public and Private Operation Of Railways in France," The Annals of the American Academy of Political and Social Sciences, Vol. 201 (January 1930), p. 211. ad1 was 1m; in In as By giv 90V1 hav: Com: at 1 0n 1 mear out; Char when dtta ment Rail. and . C00r< V01. 69 advisory to the Minister Of Public Works. The Committee was, on the other hand, an executive body ruling on all important questions Of common concern to all networks. The 1921 Convention and its Amendment Of 1933 proved unworkable. Deficits mounted steadily since 1929 because, in face of depression, rate could not be adjusted upward. In addition, the Executive Committee drew heavily criticism as serving private interests at the expense of public needs. By virtue of the Committee's composition, each railroad was given one vote; with five private companies versus two government networks, the private interests were assured of having a constant majority in the deliberations Of the Committee.6 These factors played an important role in the debate at the time about the future of the railroads in France. On the basis Of past experience, rate adjustments as a means to equilibrate the railroads' finances were ruled out; under such principle, the railroads would have to charge higher rates when traffics were down and lower rates when demand was up.7 It then appeared that relief could be attained only through very substantial Operating economies. 6. For details on the 1921 Convention and the 1933 Amend- ment, see: Georges Harcavi, "Nationalization of the French Railways," The Annals of the American Academy of Political and Social Sciences, Vol. 201 (January 1939), pp. 217-226. 7. H. E. Dougall, "Railway Nationalization and Transport Coordination in France," The Journal of Political Economy, 'Vol. 46, No. 2 (April 1938), p. 224. 70 The Socialist, which came into power in 1936, insisted that concentration Of control through nationalization constituted the only way to eliminate wasteful duplications which the concessionary system had produced. Despite some Opposition, this view was successfully put into effect through a decree law of August 31, 1937 which unified the seven railroads into the National Company Of French Railroads (Société Nationale de Chemins de Fer Francais, hereafter referred to as the SNCF). SO, while railroad nationalization in the United Kingdom was motivated by an ambition to integrate transport via unified public ownership, that in France was primarily, in the words Of an Observer at the time, a "centralization" scheme.8 The road-rail coordination scheme which accompanied nationalization contained only some minor alterations to the existing policy legislated in 1934. Since its conception, transport coordination policy in France which rested on capacity control, had been motivated by neither a desire for fair play nor efficient resource allocation between com- peting modes but primarily by the urgent needs of doing something about the financial strains on the Treasury as a result Of mounting railway deficits.9 8. Ibid., p. 224. 9. H. E. Dougall, "Some Lessons Drawn from European Experi- ence," American Economic Review, Supplement, Vol. 25 (1935), p. 122. ' _T_h_e 0ve mod pol is 1 of 1 Nati impc exis were cont or t the ' Wherc say . righi legaj rail: which 10 ‘ Frand coho; 11. 1 PP. 21 71 The Legal Status of the SNCF State enterprises in France, in the absence Of any overriding system, have been placed under various legal models Of organization, each owing its existence to the political and economic circumstances Of the time.10 This is quite in contrast to the British case where the device Of public corporation dominated the trend Of thoughts. Nationalization Of private railroads in France raised an important legal question. Under the concessionary system existing up to 1937, private companies and the government were tied together by contracts. TO be freed from these contracts the government had either to repurchase the lines or to abrogate the contracts. The financial condition Of the Treasury at the time did not permit the former course, whereas the latter course was considered as unethical, to say the least.11 The result was a compromise between out- right nationalization and concession, at least from the legal angle. Effective on January 1, 1938, the five private railroads denounced their rights under various concessions 'which would have expired between 1950 and 1960. Their 10. André Hirschfeld, "The Role Of Public Enterprise in the Franch Economy," Annals of the Public and Co—operative Economy, Vol. 44, NO. 3 (July-Sept. 1973), p. 259. 11. H. E. Dougall, "Public and Private Operation," 0p. cit., pp. 213-214. ' ass ret enc at Dec. pri‘ gove SNCE shax Cont cred comp lega taki: admil admi: ment: ViCeg 12. 239 0f Ra abate t10n F.25‘. 72 assets and liabilities were transferred to the SNCF. In return, the private companies were given 49 percent prefer— ence shares in the SNCP, the interest of which was fixed at 6 percent annually and payable semiannually until December 31, 1982. By that date, the shares owned by private companies would have been fully amortized and the government would have succeeded as the sole owner Of the SNCF. Until then, the government owned 51 percent Of the shares Of the SNCF which beared the same interest rate Of 6 percent. These shares represented both the total assets contributed by the former government railroads and the credits which the Treasury had made available to the private companies operating hitherto.12 As a "National Company," the SNCF was differentiated legally from other models Of public and semi-public under- takings, namely, the public Offices (regies), the public administrative establishments (establissements publics administratifs) and the commercial and industrial establish- ments (establissements commercials ou industriels). "Regies" are public Offices in charge of those ser- vices which are directly performed by the government or 12. K. A. Doukas, The French Railroads, op. cit., pp. 238- 239; and EurOpean Economic Community Commission, Legal Status of Rail, Road and Inland Waterway Transport in the Member States of the European Economic Community (Paris: Publica- tion Services Of the EurOpean Communities, July 1962), pp. F.25-F.26. Se S u 73 its territorial subdivisions. They do not possess any separate legal personality. Public administrative establishments are public bodies untitled to render administrative decisions of self- executory character, subject, of course, to the control of supervising authorities. The personnel Of these organiza- tions are normally considered as public servants. The public administrative establishments are not entirely integrated into the general administration as are the regies. The former enjoy a considerable degree of finan- cial autonomy but are still subject to the budgetary rules and principles Of public accountability. The purpose of each establishment is narrowly defined. These bodies Operate in practically every field Of public conern, i.e., agriculture, national education and arts, public health, city planning, information and communications, etc. The public commercial and industrial establishments come close to the British model Of public corporation. They possess a separate legal personality which, together with a clear concept of the "specialty“ of their industrial or commercial functions, permits a considerable degree of managerial autonomy. Personnel Of these bodies are employed by private contracts, thus, do not have the status Of civil servants. Similarly, their contracts with their users and suppliers are private law agreements. Thepublic commercial Cl i: be 81 la co ta co. pr. meJ COI las dur CO!) and Was 13 Pub V01. 74 and industrial establishments are, however, subject to financial and administrative controls by public authorities. The national companies, theoretically, come closest to regular commercial enterprises. They are joint-stock companies in character. Even in cases where their creation is a result of nationalization, the economic identity of the companies has Often been left untouched. The relations between the national companies and the employees, users, suppliers or third persons are entirely ruled by commercial law and thus fall within the jurisdiction of the general courts. Their relations to the State, however, still par- takes the dichotomy between private Operation and public control. The national companies are subject to various priori and posteriori controls Of the Ministers and Parlia- ment; the degree Of which varies considerably among companies.13 The original statutory arrangements Of the SNCF had lasted very briefly before they underwent important changes during the post-war years which, in effect, imposed tighter controls over all State enterprises in France. The causes and effects of these changes will be examined in turn. The administration of each State enterprise in France was generally vested with an administrative council and a 13. For details, see Winfried Dallmayr, "Public and Semi- Public Corporations in France," Law and Contemporary Problems, Vol. 26, NO. 4 (Autumn 1961), pp. 755-793. 75 chairman-managing director. Following the pattern advocated by the General Confederation of Labor (Confedera- tion Generale du Travail--CGT), the councils were tri- partite with members representing workers, consumers or users and the government. From the beginning it had been extremely difficult to appoint representatives Of the con- sumers. For one thing, it was not easy to determine which persons or institutions were in fact consumers Of the out- put Of certain undertakings. Also, consumers were not an organized group. Taking advantage Of these, the Communist Party secured for its members and supporters the seats reserved for consumer representatives. Together with its powerful CGT, the Party successfully swamped the administrative councils of state enterprises with its nominees. Communists' gains over the control Of state enterprises reached such an alarming proportion that the matter was investigated and then denounced by several parliamentary commissions of inquiry in 1947. The reports Of these commissions served as a catalyst for government actions taken to regain its control over these undertakings, something the government had always longed for.14 14. For details on the rise and fall Of Communist domina- tion in State enterprises, see Mario Einaudi et al., Nationalization in France and Italy (New York: Cornell University Press, 1955), pp. 100-105. t}. fc 0V of 15, 76 New legislations were enacted to enlarge the scope of ministerial controls over State enterprises. Personnel policies of these undertakings, for example, could not be changed until the appropriate Minister had received the advice Of an interministerial committee dealing with these matters. These were additional to a list Of ministerial powers which had been brought into being by the virtue Of national planning. Examples Of the latter included the creation of a Committee on Investment (Commission de Invest- issements) in 1948 to examine and give advice on investment programs Of State enterprises, their timing and financing methods; the investment programs were then coordinated by the General Planning Commission (Commissariat General du Plan) with other State enterprises in the same industry and then presented by the government with the general budget for authorization through legislation.15 Next, in May 1953, ministerial supervision was extended over technical matters Of state enterprises through the media of Commissaire du Gouvernement and Controleur d'Etat. 1. Each Commissaire du Gouvernement served as a liason between the supervising Minister and the enterprise. In general, his task is to watch over the major de- cisions Of the enterprise, to satisfy himself on the general standard Of the management and to be ready to keep the Minister fully informed and alert to develop- ing difficulties. 15. M. Einaudi, et al., Nationalization, op. cit., p. 112. 77 2. Each Controleur d'Etat was appointed by the Minister Of Finance. His function was to provide the Minister with Opinions on proposals from the enterprise which required the Minister's approval as well as to report annually to the Minister on the enterprise's economic and financial situation. For the discharge of their duties, both the Commissaire and Controleu were empowered to attend the meetings Of the Administrative Council and other committees Of the enter- prise as well as tO examine the enterprise's records at their wish.16 Parliamentary supervision and control over state enterprises were also strengthened via several means: 1. The creation Of sub-committees of the finance com- mittees Of each chamber; 2. The modifications of statutes and decrees by which the original administrative frameworks of the enter- prises were established; 3. The control Over the budgets Of state enterprises; and 4. The creation Of extra-parliamentary supervisory bodies. This was quite a contrast to the original statutes where no machinery had been set up to provide direct parliamentary supervision Of state enterprises. As for the sub-committees, they were specifically empowered to "follow and assess the management Of the 16. D. M. B. Lévy, "Control Of Public Enterprises in France," in W. G. Friedmann et al., ed., Government Enter- prise, op. cit., pp. 123-132; and NEDO, Government and Public Enterprises, op. cit., p. 8. 78 nationalized enterprises" via the medium of gathering in- fonmation. Their powers were later enlarged to include "the most extended investigatory power" not only in gather- ing technical information on the management of state enter- prises but also in examining their past and future. Parliamentary modifications Of statutes and decrees creating state enterprises were generally piecemeal. Their effects were, however, quite profound, in some occasions, in restraining the powers Of the Ministers over State enterprises. For example, Parliament explicitly abrogated the decrees giving Commissaires du Gouvernement priori powers over certain aspects of the production programs, investment policy, personnel policy and future budgetary requests Of state enterprises. Accordingly, once decisions on these matters had been made, the management was to be left to act as it saw fit without further priori approval Of the Commissaires; otherwise the management could hardly be expected to work effectively. As far as budgetary control was concerned, parlia— mentary scrutiny were statutorily limited to the approval of loans to the enterprises and capital advances for new construction. This part of the budget was, however, very vital to the enterprises; and the powerful influence of this parliamentary power on the undertakings were documented. Most important among the last aSpect Of parliamentary powers was the Commission for Verification Of the Accounts 79 of State Enterprises (Commission des Virification des Comptes). 'The Commission was created by statute in January, 1948; hitherto it had been independent Of the Court des Comptes but closely associated with it. The Commission's function was twofold: To audit the accounts Of state enterprises and to report on the efficiency with which they were managed. The Commission had no powers of its own to enforce its recommendations. It therefore had to depend on Parliament and Ministers for their pressures on the enterprises to implement its reforms.17 The German Federal Railways The current administrative framework of the German Federal Railways is governed by the Federal Railway Law of December 13, 1951. The legislation contained, to a great extent, a long tradition of social Obligations, Of efforts to balance between government control and autonomy and Of efforts to reconcile the interests Of the Federal (Bund) and individual states (Lander). In order to compre— hend the administrative framework Of the DB today, it is important to appreciate the role Of this socio-political tradition. 17. Edward Lewis, "Parliamentary Control of Nationalized Industry in France," The American Political Science Review, Vol. 51, NO. 3 (September 1957), pp. 669—683. For details on the Verification Commissions see NEDO, Government and Public Enterprises, op. cit., pp. 72-76. 80 Organizational Tradition The federal system Of railroads in Germany originated in the Weimar Constitution of August 11, 1919, when seven networks previously owned by individual states were conso- lidated to form the Deutsche Reichsbahn. Earlier attempts to centralize railroad ownership and administration had failed due to strong Opposition from individual states which had tried to safeguard their sovereign rights and the surpluses generated by their railroads. It took a general deterioration in the financial conditions Of the railroads and the changes in the political system resulting from.WW I as well as the experience Of wartime unified administration to create the circumstances favorable to railroad consolidation. One important aspect of the consolidation scheme Of 1919 was the idea Of semi-autonomous organization. Article 92 Of the Constitution stated that the national railway system, not withstanding the inclusion of its budget and its accounts in the general budget and general accounts, shall be ad- ministered as an autonomous economic enterprise, responsible for its own expenditures including interest and amortization of the railway debt and the accumulation Of a railway reserve fund. The climate Of the time and thereafter, unfortunately, was hardly favorable to autonomy. Various political forces resisted any autonomous arrangement, i.e., the Lander tended to rely on their influence in the central political organi- zations to protect certain surviving stakes which they had 81 in the new federal railroad system. More important, mount- ing deficits made financial aids from the Treasury and, hence,the departmental controls which accompanied them, unavoidable. The foregoing conditions drew heavy criticism from various sources especially business groups. The outcome was a compromise in which government ownership remained while steps were taken to move toward managerial autonomy. A decree in 1924 gave the Deutsche Reichsbahn (DR) the legal status of Sondervermoegen, meaning that its assets were separated from the properties of the State and that its budget was taken out of ordinary government budget. This arrangement, however, did not last_long. Late in the same year, the DR were reconstituted SO as to carry out their new Obligation as a mortgage to secure war reparation payments in the Dawes Plan. At this point, foreign philOSOphy had its impacts upon the reconstitution Of railway administration. A group Of foreign experts, appointed by the Reparations Commission, called for a complete break with the Old tradition of rail— way administration in Germany which had been placing social Obligations above commercial profitability. The group recom- mended the transformation Of the DR into an independent joint-stock company (Gesellschaft). The recommendations were subsequently adopted in an Act dated August 30, 1924. Accordingly, the DR were renamed as the Deutsche Reichsbahn 82 Gesellschaft (DRG). The Company was charged with Operating "the railways in accordance with business principles, due regard being paid to German economic interests." In addi- tion, so as to facilitate international control, the Administrative Council (Verwaltungsrat) was created. The Council had half of its members being appointed by repara- tions trustee. It was vested with important managerial powers although most Of these powers were expected to be delegated to a Directorate. The use of the term Verwal- tungsrat reflected a departure from the law and custom Of German (private) corporations which were marked by the separation between oversight (conducted by an Aufsichsrat in the behalf of the stockholders) and management proper (responsible for by Vorstand). The powers of the Admini- strative Council raised serious doubt on some German minds whether the "autonomous" organization was intended to provide managerial autonomy or simply to facilitate foreign intervention. (From such a distant origin, the Administrative Council of the Federal Railways Of today plays practically only an advisory role.) I Such doubts apparently accounted for the emphasis on managerial oversight by the government after foreign participation in railroad administration was abolished in the Young Plan of 1930 (the railroads still continued to make reparation payments, however). This emphasis was furthered in subsequent years. First, the Lander was 83 deprived, in 1934, Of their power to name their representa- tives on the Administrative Council. Then in 1939, the DRG were integrated into the administrative system of the Nazi government. The word Gesellschaft was dropped from the name Of the railways. The post Of the general manager of the federal railroads was to be filled by a cabinet minister (in practice, by the then Minister Of Transport). Also, the Administrative Council was replaced by a purely advisory body. The principle of financial autonomy in the sense of separate property (sondervermoegen) was, however, maintained. In practice, the railroads had become an instrument for the implementation Of the transport policy Of the Nazi;18 a policy under which government intervention was probably unparalleled in European transport history.19 The Legal Status Of the German Federal Railways_ Immediately after WW II, the central organization of the railroads in Germany was broken up into separate centers according to the partition of the country into four occupation zones. Not until the passage Of the Bonn Constitution in 1949 18. For details on the German Federal Railways during the pre-WW II period, see: Arthur W. Macmahon, "Autonomous Public Enterprise. The German Railways," Political Science Quarterly, Vol. 54, NO. 4 (December 1939), pp. 481-513; Alfred W. Leyen, "The German Federal Railway Company,“ Annals of Collective Economy, Vol. 2, NO. 4 (NOV.-Dec. 1926), pp. 321-344; and Arno Gunther, "The German Railways and the Collective Economy,‘ Annals of Public and Co-Operative Economy, Vol. 31, NO. 1 (Jan.-March 1960), pp. 24-90. 19. N. S. Despicht, Policies for Transport in the Common Market (Kent: Lambarde Press, 1964), p. 85. 84 that the task of unification of the railroads in the Western zones could be completed. The Constitution divided the legislative and administrative powers between the Bund and the Lander. Accordingly, the Bund was empowered tO legis- late and manage the federal railroads; non-federal railroads were managed and controlled by the Lander. During the debates leading toward the Federal Railway Act Of December 13, 1951, the fundamental principles behind the constitution of the Federal Railways (Deutsche Bundes- bahn--hereafter referred to as the DB) were largely unchal- 1enged, undoubtedly because of the strength of administrative traditions. Although the DB was given certain measures of financial and administrative autonomy, the Old concept Of the railroads as public services and other administrative traditions were still evidenced. As for the public service concept, there are three distinct meanings of the term in German: 1. It is used to denote the internal activity Of agencies of public authority as Opposed to that of private bodies; 2. It is Often used as synonym to the concept Of "col- lective economy," i.e., the performance of the tasks for the benefits Of the community without regard to any profits motive; or 3. It is used as a literal translation of the concept Of "public utilities," i.e., the provision Of uninterrupted services to all comers. In the Federal Railway Act Of 1951, the first meaning Of the term public service—-Offentlicher Dienst--was used, implying 85 that the DB had the standing Of government authority.20 In practice, hOwever, it was the socio-economic rather than the legal meaning which had shaped the relationship between the DB and the government. This was evidenced in an Official commentary on the 1951 Act. The Act laid down an Obligation to Operate the Federal Railways "in accordance with commer- cial principles while safeguarding the interests Of the German economy . . . ." The commentary, however, made it clear that "the Bundesbahn must not be a private undertaking solely concerned with earning the profits; it should steer a safe middle course between public utility and its economic requirements" and "while respecting sound principles Of management, the Objective of the Bundesbahn must retain the service of German people and the German economy. Its approach must not be that of a commercial firm whose main Objective is to make profits." It followed that the Federal Railways should in the first place, serve the general economy Of the country (the second meaning Of the term public service) while its obligation to follow sound commercial principles was mainly a matter Of organization.21 The tradition Of public service set the Federal Rail- ways apart from virtually all other state enterprises in Germany. Although the DB contained in its constitution and 20. A. Gunther, "The German Railways,‘ op. cit., pp. 30-32. 21. Ibid., p. 70. 86 organization certain elements normally found in a public corporation prOper, it was basically a quasi-government departmental agency. Organizations of this type (consisting Of only the DB and the Postal Service--the Bundespost) were characterized by the following: 1. They were created by special statues; 2. They had nO separate personality vis-a-vis the State but had all intents as a separate entity with most of the attributes Of "legal capacity”_vis-a-vis third parties; 3. Their staffs were employees Of the Federal government with about one-half Of them having civil servant status (the other half was covered by collective agreements between management and non-civil-service unionts); 4. The property Of the agencies was legally separated from the property of the Federal government (sondervermoegen); 5. Their accounts did not form part Of the Federal budget; and 6. Their constitutions, functions, and powers were as pro- vided in their statutes.22 Given the legal status of the Federal Railways as a quasi- government departmental agency, the Minister of Transport would have a right Of veto in virtually all general matters affecting the interests of the State and the Bundesbahn. Parliamentary supervisory powers over the Federal Rail- ways, On the contrary, were indirect and limited. Vis-a-vis the DB, the Bundesrat (the upper house) did not have any direct supervisory powers. The approval Of the Bundesrat, 22.‘ NEDO, Government and Public Enterprises, op. cit., p. 17. 87 however, was required not only for legislation but also for most government regulations. The Bundesrat could therefore exercise its influence over ministerial policy. The powers of the Bundestag (the lower house) were also limited. Of course, the Minister of Transport was responsible to the Bundestag for the working Of the Federal Railways. It should be, however, noted that the Minister was no longer a member Of railroad management as in the time of the Deutsche Reichsbahn, although he was not an outsider to anything like the extent that a British Minister was to a public corporation in the United Kingdom. The fact that the German Minister was no longer a member of railroad management weakened parliamentary controls. Like his British counterpart, though probably to some less extent, the German Minister could excuse himself from answering certain parliamentary questions on the grounds that these were related to the day-to-day matters of railroad management. Another tradition was still influential upon the working Of the Federal Railways, namely, the tradition Of administrative decentralization (as is going to be discussed later in this chapter, this tradition accounted for the decentralized organizations of the Federal Railways). His- torically, the Federal Railways were amalgamated from individual networks Of the sovereign states. The Lander governments had therefore always sought for means to protect their regional interests in the Federal system. Under the 88 Bohn Constitution, the Lander governments did not have any powers vis-a-vis the DB, but they had the right to be consulted. Accordingly, the management and the regional Officials of the DB on one hand, and the Lander transport authorities on the other, had a duty to keep each other informed of all major developments. The Lander Ministers of Transport could demand information in the same way as their federal counterpart. More specifically, the Lander authorities had to be given an Opportunity to state their views on such matters as closure Of lines, stations or workshops, changes in the location of regional Offices. They had to be consulted on tariff changes, timetables and development plans. If there was any dispute about whether their prOposals had been sufficiently considered by rail- road management, they could take the matter to the Federal Railways' Administrative Council and from there to the government for decision. Either their approval was required or their views were to be consulted on certain regional appointments.24 The Public Corporation in Japan While the administrative frameworks Of the nationalized railroads in Western countries were the products of their own legal and pOlitical traditions, the current public 24. F. Ridley, "The German Federal Railways--A State Admin- istered System," Parliamentary Affairs, Vol. 17, NO. 2 (Spring 1964), PP. 191-192. 89 corporation status of the Japanese National Railways (here- after referred to as the JNR) was an alien concept. It was superimposed upon the existing nationalized railroad system by the Supreme Commander for the Allied Powers (SCAP) in 1949. Until then the system had been a government depart- mental undertaking within the Ministry Of Communication since 1906, and then, an independent Railway Ministry since 1920.25 The Motives Behind Incorporation In an overall attempt to democratize post-war Japan, the SCAP repealed all supressive laws against trade union movements. The Trade Union Law of 1946 abolished differen- tial treatments between private and public sector labor relations by granting workers three fundamental rights: the right to organize, the right to bargain collectively, and the right to strike. It should be noted that the trade union movement in Japan had had a political orientation dating back to the Meiji Era. However, SCAP Officials, being preoccupied with democratization of Japan, did not Oppose union political activities. As a matter Of fact, 25. For details on the railway organization in Japan before W.W.II, see: Watarai Toshihara, Nationalization of Railways in Japan (Ph.D. Dissertation, Columbia University; New York, 1915); Key Whan Kim, The Price and Output Policies of the Japan National Railways, 1949—1963 (Ph.D. Dissertation, Uni- versity of California, Berkeley; Berkeley, 1971), Chapter 1. 90 SCAP Officials had hoped that the trade unions would evolve and coalise into something like the British Labour Party. The real problem with this policy, as Johnson (1973) pointed out, was not that the unions were political but that the Communists, not being plagued by the same frictional rival- ries as non-Communist labor leaders, had been able to capture the control Of the labor movement and were running it in a decidely undemocratic manner.26 With post-war inflation running high, the living conditions of railway employees were deteriorating rapidly. The trade union movement became more radical and large- scale strikes broke out in the latter half Of 1946. The SCAP came to regard the strikes in the public sector as anti-occupation, and accordingly, abandoned its non- interference position in 1947. The National Public Service Law of 1947 was passed so as to withdraw from public sector employees the right to collective bargaining and the right to strike. When this law was enacted, workers in state enterprises, particularly the railways, violently Opposed it. They viewed the law as an infringement on the rights which they had gained earlier in state enterprises. On July 22, 1948, in an attempt to molify the workers in state enterprises as well as to prohibit strikes in the 26. Chalmers Johnson, Conspiracy at Matsukawa (Berkeley: University Of California Press, 1972), pp. 21-22. 91 public sector, General MacArthur sent Prime Minister Ashida a letter demanding that immediate steps be taken to effect a comprehensive revision of the 1947 law. The outcome was the passage in 1948 of the Public Corporations and National Enterprises Labor Relation Law. This new law granted the workers in five direct government enterprises and three Kosha (a form Of public corporation to be explained later) the rights to organize and to bargain collectively, but it denied them the right to strike. The law Of 1948 had always been resented by the employees of these organizations, and illegal strikes by them were still one Of the major sources of labour strife in Japan more than 25 years later. SO beyond the three original kosha designated by General MacArthur, no further public corporations were added to the kosha category. All public corporations established later have been put into other categories which fall under the nation's basic Trade Union Law which grants the workers all three fundamental rights. Attempts to do otherwise have been blocked by the powerful Sohyo Labor Federation unless the government would repeal the 1948 law and extend to workers in direct government enterprises and kosha the same rights which have been enjoyed by workers in other state enterprises and in the private sector.27 27. Chalmers Johnson, Japan's Public Policy Companies (washington, D.C.: American Enterprise Institute for Public Policy Research, 1978), p. 30. For more details on indus- trial relations in Japan's public sector, see: 92 Meanwhile, SCAP's policy had switched from democrati- zation to economic recovery for Japan. New policies, designed to stop inflation, stabilize the Japanese yen and lower unit production cost, led to a large number of workers being laid Off, with over one hundred thousand in the rail- roads alone. In some localities rail workers resorted to sabotage in retaliation against layoffs; the mysterious death Of the then President of the JNR in 1949 has been believed to be connected to such sabotage. The conditions Of the railroads deteriorated to a point where the Japanese, in 1949, referred to their railroad system as a kotsu jigoku ("transportation hell").28 Since the beginning of the occupation, SCAP had encouraged the establishment of semi-autonomous state enter- prises as a measure to democratize Japan. With the labor relations in state enterprises getting out of control, this policy was extended to the nationalized railroads, the tobacco and salt monopoly and the telephone and telegraph undertaking. Such a move was pressed by SCAP as an effective means to break the trade unions of these three organizations Tadashi Mitsufuji, "Industrial Relations in the Public Sector in Japan," in The Japanese Institute of Labor, The Changing Patterns of Industrial Relations in Asian Countries (Pro- ceedings Of the Asian REgional Conference on Industrial Relations, Tokyo, 1969), pp. 20-42. 28. Chalmers Johnson, Conspiracy, op. cit., pp. 397-398. 93 away from those Of other state enterprises, thus weakening the militancy Of their labor movements. Consequently, these three undertakings were quickly converted from direct government undertakings into kosha-type public corporations, beginning with the railroads in 1949. The General Characteristics Of Kosha State enterprises in Japan can be divided into two major legal groups: private law establishments and public law establishments. The former group consists of mixed enterprises. The latter group can be further divided into (a) public law establishments without corporate personality such as government departmental undertakings (i.e., Postal Service, the Mint Bureau) and municipal enterprises, and (b) public law establishments with corporate personality such as public corporations and semi-independent agencies. The majority Of important state enterprises in Japan belong to the public corporation group (tokushu hojin). This legal form of organization has several variations; the differences among them are more in their fields Of Operation than in their legal characteristics. Most important among the public corporation group are kosha, kodan, kOkO, ginko, and kinko. 1. Kosha consists of three important natural and poli- tical monopolies, i.e., the JNR, the Nippon Telephone and Telegraph (NTT) and the Salt and Tobacco Mono- polies. 94 2. Kodan are primarily in the field Of construction. 3. KOkO engage in financing risky business ventures which lack alternative credit sources. 4. Ginko are generally public finance corporations. 5. Kinko provide loans and finances to co-Operatives.29 The public corporations in Japan are statutorily subject to stricter parliamentary and ministerial controls than are their counterparts in the United Kingdom; in particular, the Kosha are the most public of all. Strict controls are accounted for by several reasons, among them are an absence of a clear concept of public corporations, non-commercial motives behind the Acts Of Incorporation, and a long tradi- tion Of pre-war government departments. As a matter of fact, during the drafting of the JNR law, Opinions concerning the degree Of independence to be granted to railroad management differed among various sections Of the General Headquarters Of the Allied Forces. The Civil Transport Section insisted upon a high degree Of managerial independence for the JNR, while the Economic and Scientific Section pressed for strong governmental control Of the National Railways. The Japanese Ministry Of Finance, Ministry Of Transport and Parliament shared the same View with the Economic and Scientific Section. Given such strong 29. K. Yoshitake, An Introduction to Public Enterprise in Japan, op. cit., pp. 1-28. 95 alliance among several public authorities, heavy govern- mental control naturally prevailed. In the early years of the JNR, the General Head- quarters Of the Allied Forces guided the National Railways informally so that management could be as flexible and had as much initiative as public corporations in the United States or the United Kingdom. Accordingly, the Headquarters ordered JNR management to concentrate on managerial efficiency without paying any concern to parliamentary accountability; the latter was to be left to the Minister Of Tran5port. In other words, JNR management should avoid having any contact with Parliament or other authorities. This intention Of the Headquarters was not Observed by the Japanese side. From the beginning, Parliament required the attendance of JNR officers for answering the questions asked by Members of Parliament on the National Railways; activities in the same manner as they had done during the pre-war years when the Railways had still been a government department. Yoshitake (1973) explained the failure Of the Headquarters in resisting the tendency toward heavy control by Japanese authorities in terms Of a possibility that SCAP Officials might have been biased toward government corporate enterprises in the United States. This type of enterpriSe in the United States was theoretically subject to higher degrees Of control than the public corporations in the United Kingdom.30 30. K. Yoshitake, Public Enterprises in Japan, op. cit., pp. 270-271. 96 Most serious and unique among the control measures have been those concerning the finance of the JNR (Table 4). The JNR law requires the National Railways to submit budget estimates to the Minister Of Transport together with development and financial plans and other documents related to the estimates. The Minister Of Transport, after neces- sary consultation and coordination with the Minister of Finance, has to get the approval Of the Cabinet. The latter then submits the estimates Of the JNR budget to Parliament along with the general budget Of the government. Although it has been pointed out that there is some degree Of flex- ibility granted to JNR management in spending the appropria- tion, in practice the budgetary process does not differ much from.that of a government department. SO, in order to accommodate both the requirements of government control and the need of business Operations, the JNR has to maintain Simultaneously two accounting systems as well as having its transactions internally audited by JNR Auditing Committee in addition to governmental auditing.31 As far as pricing is concerned, the JNR has to seek the approval of Parliament and the Minister Of Transport in most cases. The rates and fares Of the JNR are divided into three categories, namely, "statutory," "approval" and "autonomous." 31. Ibid., p. 129. 97 Table 4 Financial Control by the Government on Main Groups of Public Enterprise in Japan Kosha Koko Kodan (JNR, NTT, Salt (Loans and _(Japan Road and Tobacco Finance Public Corporation, Monopoly Corporations) Japan Housing Corp., etc.) Budget Voting Of Voting Of Approval of Estimates Parliament Parliament Minister Annual Examined Examined Corporate Together with Together with Planning Budget Budget Estimate Estimate Bond Issue Ministerial Ministerial Approval Approval Long-Term Ministerial Borrowing Approval (Non- Applicable) Short-Term Borrowing Price Parliamentary Parliamentary Ministerial. Determination Voting or Voting or Approval Ministerial Ministerial Approval Approval Source: K. Yoshitake, Public Enterprise in Japan, op. cit., p. 131. 98 1. "Statutory" category requires parlimentary approval. It includes non-commuter passenger fares, ferry passenger fares and carload freight rates. 2. "Approval" category needs only the approval Of the Minister of Transport. It consists of commuter passenger fares, less-than-carload freight rates, baggage and express rates, minimum passenger fares and freight rates. 3. "Autonomous" categOry is left up to the NJR's dis- cretion. It contains only rate and fare adjustments of a routine nature which do not affect the total revenue Of the JNR to any extent, i.e., demurrage charges, promotional fares. As such, the JNR's freedom of pricing has been firtually nonexistent. In fact, in 1962, "autonomous" category accounted for approximately only 2 percent Of the total . 32 Operating revenue. Besides these statutory controls, the JNR is also subject to frequent parliamentary interventions into its day-to-day management. Elsewhere, this could have been recognized as an interference which clearly exceeded the authority Of Parliament. In Japan, due to an absence Of a clear concept Of public corporations, there has been no such recognition and parliamentary interference into day- to-day matters are considered routine.33 32. Japanese National Railways, Japanese National Railways. A General Description. (Tokyo: Nihon Kokuyu Tetsudo, 1963), p. 16; and K. W. Kim, The Policies of JNR, op. cit., pp. 45-46. 33. K. Yoshitake, Public Enterprise in Japan, op. cit., p. 272. 99 Railroad Organization Besides being a state enterprise, a nationalized railroad system is also a business enterprise. The former status raises questions concerning the relationship between government and railroad management to which the railroad's legal model Of organization is an attempted answer. The latter status calls for an internal organization which pro- vides an effective chain of command essential for rail operations. It is related, at the tOp heirarchical level, to the position of executive management vis-a-vis the Board Of Directors and, at lower levels, to patterns of authority, namely, centralization versus decentralization. A railroad is a supplier of services (as Opposed to physical products) the production and consumption Of which- must take place concurrently in the same location. A rail- road is therefore dispersed geographically and thus neces- sitates organizational structures specially designed to deal with this unique feature. A centralized structure takes the form Of departmental organization with authority moving in a line from the headquarters to Operating depart- ments at the district and divisonal levels. An alternative model of organization is decentralized with the district superintendent deciding a number Of issues affecting all Operating departments at the district level, thus tending to Operate a small railroad within a large one. In general, 100 departmental organization centralizes authority and facili- tates systemwide control and coordination of resources. The divisional organization decentralizes decision-making and facilitates faster and more exact responses to specific local traffic requirements.34 The Organization Of the British Railways The British Transport Commission As previously mentioned, the Transport Act of 1947 nationalized four mainline railroads and a large portion Of the inland transport system and placed them under the newly-created British Transport Commission. The general duty of the BTC was to "exercise their powers under this Act as to provide, or secure or promote the provision Of an efficient, adequate, economical and properly integrated system Of public transport and port facilities within Great Britain." The BTC would be placed under the supervisory responsi- bility of the Minister Of Transport. The latter was empowered to give the Commission directions Of a general character in matters appearing to him as affecting the national interests. Several Executives, each for the Railways, Docks and Inland Waterways, Road Transport and London Transport, were 34. John L. Hazard, Transportation: Management, Economics, Policy (Cambridge: Cornell Maritime Press, Inc., 1977), p. 143. 101 to be set up to assist the BTC in the discharge of the Commission's duties. Each Executive was to act as an agent for the Commission and to exercise such functions of the Commission as were from time to time delegated to it or under a scheme made by the Commission and approved by the Minister.35 Such modal orientation in the organization Of the BTC clearly did not lend itself to transport integration as intended by the Act. Modal loyalty by management and staff Of each Executive persisted. The Railway Executive, for example, saw its functions as to manage the Railways as an unified undertaking, and as rapidly as possible to achieve economies through standardized procedures and techniques; transport coordination was a very secondary Objective.36 As a matter Of fact, each Executive organized its regional structure differently, i.e., the Railway Executive had six regions, the Docks and Waterways Execu- tive had five and the Road Executive had twelve. Later in a White Paper in 1954, the BTC defended this modal organization as being essential in making the transition 35. For detailed summary Of the 1947 Act see: Edgard Milhaud,"Nationalization in Great Britain," Annals of Collective Economy, vol. 19, no. 2 (April-June 1948), pp. 94-112. 36. Micheal R. Bonavia, The Organization of British Rail- ways (London: Ian Allan, 1971), p. 56. 102 from private to public ownership with the least possible disturbance to normal day-to-day working.37 The attitudes of the trade unions also tended to hinder transport integration. While ideologically sup- porting the concept Of integration, they Opposed, in practice, any measure that might injure the interests of their members even locally or temporarily. The National Union Of Railwaymen saw integration as a means to influence back to the Railways the traffic which had been lost to road transport; the Transport and General Workers (the British equivalent to the Teamsters in the United States) saw matters in exactly the Opposite light.38 As for the Railways, the Act prescribed neither the mode of internal organization nor the exact way in which managerial tasks were to be carried out below the Executive level. The Ministry of Transport Officials apparently favored a functional setup where each of the RE members would be responsible for a separate department. Each Of these departments would have regional Offices being respon- sible to it although there would also be a Chief Regional Officer in each region being charged with departmental coordination. With the first group of RE members (appointed 37. Ibid., p. 63. 38. Michael R. Bonavia, "British Rail's New Deal," Trans- portation Journal, vol. 18, no. 2 (Winter 1978), p. 3. O 103 by the Minister Of Transport in accordance with statutory provision) being drawn from among former railroad depart- mental Officers, the functional mode of organization was put into practice. From the beginning, there had been constant disagree- ments between the BTC and the RE on the scope Of their authorities. The RE contended that the Chief Regional Officers were to be responsible for the day-to-day manage- ment of the regional networks while all questions Of policy and forward planning were reserved for the RE itself. Furthermore, the RE considered the powers reserved for the BTC as merely certain exceptions to the overall scheme Of delegation, which in fact transferred the managerial tasks Of the Railways to the Executive. The BTC, on the contrary, regarded the RE and the Commission itself as partners in the task Of running the Railways. The BTC did not accept the notion that the specific reservation to the Commissions Of certain powers as constituting the limit Of its scope in policy formulation.39 In particular, Sir Missenden (the first RE Chairman), publicly Opposed the idea Of a functional RE acting as the agent of the BTC where he was unable to play the supreme control and coordinating role of a general manager as he 39. M. R. Bonavia, The Organization of British Railways, op. cit., p. 51. 104 had been with the former Southern Railway. Equally, he resented the presence Of the BTC as intermediary between the Railways and the Minister. Apparently, Missenden was not sold on the theory of the BTC acting as a buffer between the Railways and the ministerial circles. During the tenure of John (later Sir) Elliot, who succeeded Missenden in 1951, the relations between the BTC and the RE were improved somewhat.40 Unfortunately, the advent in 1951 of a Conservative Government, which was com- mitted to changing drastically certain provisions Of the 1947 Act and the organization of the Railways, created a very uncertain situation. The prospect of its abolition discouraged the RE from planning for the future as well as put it in no position to command. So although the RE lingered on until 1953, it had long before lost its authority and initiatives. In other circumstances, the BTC could have taken over these responsibilities, but its Chairman--Lord Hurcomb--had come to the end Of his term in office in 1952 and was reappointed for just one year as caretaker. The change Of Government in 1951, therefore, meant two years Of uncertainty amounting to inability to act.41 40. "Seven Chairmen," Modern Railways (May 1976), p. 174. 41. R. Pryke, Public Enterprise in Practice, op. cit., pp. 56-57. 105 The Abolition Of the Railway Executive The new Government was critical Of the little progress toward transport integration and Of what it considered as excessive centralization Of railroad management under the 1947 Act. As for transport integration, the Conservative Government took a view that this would be best secured by competition. Subsequently, the Transport Act of 1953 set out to denationalize road haulage. On reorganization, the Act abolished all the Executives except that Of the London Transport. An interim organization was set up for the Rail- ways under which the primary function Of the BTC was to control major policy and general direction; day-to-day management was delegated to regional management. The func- tional setup of the former RE was also abolished with departmental authority being transferred to the regions; the Chief Regional Officers were renamed Chief Regional Managers to reflect their new SOOpe of authority under which all regional departmental Officers reported to them. The newly-found freedom Of management for Regional Managers lasted for only fifteen months. On January 1, 1955, the Area Boards, whose creation was provided in the 1953 Act, were appointed. Now, the Chief Regional Managers were subordinate to the overall direction of both the BTC and the Area Boards. Yet the two control bodies were linked together very loosely by a simple provision in the 106 1953 Act that the Area Board Chairmen were also members of the BTC (although mostly on a part-time basis). The link between the Chief Regional Managers and the BTC was also loosely dependent upon the periodic attendance of the former at the latter's meetings to report on their activities. Consequent to the interposition of the Area Boards in the chain Of command, the BTC was deprived of its control. The chain of command was further complicated by reorganization at the BTC'S headquarters. The BTC was then under the chairmanship Of Lord Robertson who had been selected for the post on his record as an outstanding military administrator with a brief career in business. His background apparently led him to regarding the Railways as service department rather than a commercial undertaking (more on this in the next chapter). Also, lacking a rail- road background, Robertson Often failed to detect miscal- culations Of his technical Officers. This was evident in his costly Modernization Plan Of 1955 (also, more on this in the next chapter) as well as in his large and complex organization at the BTC's headquarters. The organization might look logical on paper but was, in practice, utterly lacking in flexibility and aptitude for quick decision- making.42 42. "Seven Chairmen," 0p. cit., p. 174. 107 Under the new headquarters organization, advisory bodies, committees and sub-commissions proliferated. Chief Regional Managers had to deal with not only their own Area Boards but also these bodies. This was quite a contrast to the simplicity under the preceding interim 43 Ineffective chain Of command created a organization. dilemma. The BTC was forced to take decisions of day-to- day management and at the same time had to impose upon itself the obligation to refrain from taking such decisions and from ensuring their implementation. In other words, the policy formulated by the BTC was made ineffective by the lack Of an organization to carry it out.44 The organization of the BTC came under the scrutiny of the Selected Committee on Nationalized Industries and a Special advisory group in 1960. The latter was headed by Sir Stedeford and reported privately to the Minister of Transport (its report was not published). The Stedeford group was apparently critical Of the organization Of the BTC, far more so than the SCNI (Selected Committee on Nationalied Industries). These criticisms weighed 43. For details on the complexity of the BTC's headquarters organization, see: M. R. Bonavia, The Organization of British Railways, op. cit., pp. 65-79. 44. R. Pryke, "Public Enterprise and the Economics of Large Scale Production," Annals of Public and Cooperative Economy, vol. 41, no. 1 (Jan.-March 1970), pp. 51-52; and Public Enterprise in Practice, op. cit., pp. 246-247. 108 heavily with the Minister in reaching the decision to make drastic changes. The Creation of the British Railways Board The main raison d'etre for the existence of the BTC had been its powers to integrate transport services under unified public ownership. Once this Obligation was removed by the denationalization Of road haulage under the 1953 Act, the continued existence Of the BTC became anomalous. Besides, its existence had been regarded by the Conservative Govern- ment as a hinderance to the decentralization of railroad management. The Transport Act of 1962, therefore, abolished the BTC and replaced it with separate BOards for each of the main activities previously managed by the BTC. They were: the British Railways Board, the British Docks Board, the British Waterways Board and the London Transport Board; the remaining activities Of the BTC (the section Of road haulage not yet denationalized, bus and coach services) were grouped under a newly created Transport Holding Company. A Nation- alized Transport Advisory Council was created by the Act, apparently so as to counter any criticism of the 1962 Act by its political opponents as destroying all possibilities of coordination among various nationalized transport undertakings. As far as the BR was concerned, the 1962 Act furthered the principle Of decentralization of management. The 109 newly created British Railways Board, though assumed the responsibility for running the Railways as a national system, would perform only those central functions which were essential to running the railways as such. Notably, the British Railways Board would be responsible for national staffing or appointments and wage negotiations, overall control over finance and investment, policies for safety, training and research and the determination of the future size and shape Of the BR. Other functions would be dele- gated to the Regional Railways Boards. These bodies were created by the Act to replace the Area Boards of the now- dissolved BTC. They were charged with the responsibility to manage and Operate their regional railroad systems. SO as to secure co-ordination between national policy and regional railroad Operation, one representative from each of the Regional Railways Boards would assume membership on the British Railways Board. These reorganization measures were, however, soon overshadowed by the policy and manage- ment style Of Lord Beeching, the first Chairman of the Railways Board. Beeching brought to the BR a looser and more flexible approach to management, akin to that prevailing in large industrial firms. There followed a series Of changes, i.e., more Board members were made full-time, they were again assigned functional responsibilities, all being aimed at strengthening team management; the chain Of command at 110 regional level was shortened, to name a few.45 Beeching enjoyed a great advantage over his predecessors at the BTC, namely, a strong support and friendship from the then Minister Of Tran5port. Both, being career industrialists, shared an understanding and intention Of revitalizing the commercial viability of the BR quickly.46 Beeching was most remembered for his reports which, for the first time since nationalization, chartered clearly the course toward a profitable system. Accordingly, operations were to be concentrated on sectors where railways could compete profit- ably; the size Of the BR was to be contracted until what remained was profitable. Unfortunately, the year 1964 brought another change of government and, inevitably, a promise Of new transport policy. In 1965, Beeching returned to the Imperial Chemicals Industry after only four years in the chair Of first the BTC then the Railways Board. Reorganization Under the 1968 Transport Act The new Labour Government criticized strongly the transport policy Of its predecessor as paring down the rail- road system tO a socially and economically unacceptable level without producing any markedly improved financial turnout for the BR. The Government renewed the emphasis on transport 45. See: M. R. Bonavia, The Organization of British Rail- ways, op. cit., pp. 95-105. 46. "Seven Chairmen," op. cit., p. 175. 111 coordination and as assurance that the transport system as a whole would take full account Of the social and economic requirements Of the country. Mrs. Castle, then Minister Of Transport, made it clear in her evidence to the Selected Committee on Nationalized Industries that she was an "inter- ventionist"; after all, as she pointed out, She propped the nationalized Boards with all kinds Of grants and they were in many ways dependent on her decisions.47 Subsequently, the government set out to redefine the size and shape Of the basic railroad network in accordance with a mixture of social needs and commercial profitability instead of the latter alone. For the first time since nationalization, however, the 1968 Act provided that the government would assume financial responsibility for losses incurred on services retained for social reasons. The Rail- ways Board could then be given realistic financial Objectives to enable it to move as soon as possible to a fully economic basis of Operation. As for transport coordination, the 1968 Act seemed to take a view that if separate managerial attention were given to the component parts Of the task, overall efficiency would increase. Transport integration would then have tO be achieved by a statutory directive to the corporations. Accordingly, the BR, the Docks and the Waterways were to 47. Quoted in R. Kelf-Cohen, British Nationalization 1945- 1973 (New York: St. Martin's Press, 1973), p. 84. 112 remain as before under their respective Boards, but a number Of new agencies were created, namely, the National Freight Corporation, two National Bus Companies and the Passenger Transport Authorities whereas the Transport Hold- ing Companies was abolished. The National Freight Corporation (NFC) ended the competition between the road haulage vehicles of the former Transport Holding Company and the equivalent rail services (i.e., the Railways' Freightliner services, Sundries Divi- sion and cartage vehicles( by combining the two. The NFC would assume the contractual responsibility for all rail traffic which originated with a road collection; whereas British Rail should retain the responsibility for traffic originated from rail-connected sources. SO as to ensure that the use of rail facilities should not be prejudiced against by the fact that "joint" transits began and ended on the road, the NFC was instructed by the Act to secure that, in the provision Of integrated services, goods would be carried by rail whenever such carriage was effi- cient and economical. The Chairmen Of the Railways Board and the NFC were made members Of the new Freight Integration Council whose task was to advise the Minister of Transport on freight integration. As for passenger services, the 1968 Act empowered the Minister of Transport to create Passenger Transport Authorities (PTA) in the country. The PTAs were designated 113 as virtually autonomous monopolists within their reSpective areas since all municipal bus undertakings were transferred to the PTA executives. Outside the major conurbations, bus transport was to be controlled by a National Bus Company (with a separate Scottish Transport Group for Scotland) which would acquire the bus interests Of the former Transport Holding Company. These two corporations would take over the responsibility from the Railways for securing the pro- vision of bus services to replace withdrawn rail services. The foregoing shows that the responsibilities for transport integration was moved up closer to the ministerial level, especially in freight transport. Given this and the "interventionist" intention Of the Minister, Kelf-Cohen (1973) concluded that, in essence, the Railways Board had become a hybrid between an autonomous public corporatiOn and a government department.48 On the subject Of the internal organization of the Railways Board, the 1968 Act required the Board to report their review and conclusions on organization to the Minister within twelve months Of the passage of the Act. Pursuant to this provision, the Board, with the assistance of manage- ment consultants (McKinsey and Company, Inc.), conducted an extensive review Of the BR's organization. A number of important conclusions was reached as the following. ..... 48. Ibid., p. 84. 114 First, the Railways Board itself would assume a mainly non-executive role and give greater emphasis to overall corporate planning, policy-making and longer term direction of British Rail and each of its other businesses while ensuring that each business was effectively managed. Although Board members would not take any line management responsibility, each was to be assigned to a specific area of expertise and expected to play a leading role in influ- encing Board decisions in that particular area. As for the subsidiary businesses other than rail, the report proposed to set up subsidiary Boards which were non-executive and chaired by members from the main Board. Second, the Board intended to delegate the respon- sibility for the management Of British Rail and each Of its non-rail businesses to a Chief Executive, one for each business. As for regional railway management, the 1968 Act had already provoked the statutory requirement to appoint Regional Boards. The report, however, recommended that these Boards were to be continued, but only in advisory capacity. The Regional General Managers would therefore report in a direct "line" relationship to the Chief Execu- tive (Railways). They, together with the Chief Executive (Railways) and the Executive Directors and the Planning Manager at the headquarters, would form the Railway Manage- ment Group. The preceding two changes were intended tO strengthen the relationship between the regions and the headquarters. 115 Third, the organization of British Rail below the Chief Executive level was tO be restructured for more effective management. This would consist of two phases. Phase one aimed at restructuring and strengthening the headquarters departments under Executive Directors so that they could support the Chief Executive and focus on planning and controlling key elements of business more effectively. This and the foregoing became effective on January 1, 1970 (Figure 7). Phase two would be to review the regional organizations in the light Of the revised approach to rail- way management. This phase would not begin the implementa- tion of phase one was well under way (probably in a year).49 It then took three years to complete the detailed study and planning for the implementation Of changes brought about by phase two.50 The BR's Field Organization For a period Of less than a quarter of a century, the BR had undergone four major reorganizations, each brought with it a virtually total change. Subsequently, the BR had to spend half Of its life either in anticipation 49. British Railway Board, Report on Organization (pre- sented to Parliament in pursuance of Section 45(4) of the Transport Act, 1968; London: HMSO, December 11, 1969); summarized in "Reorganization of British Railways Board," Railway Gazette, vol. 126, no. 1 (January 2, 1970), pp. 9-12. 50. M. R. Bonavia, The Organization of ER, op. cit., p. 143. 116 eoeumuacmmno cameo m.mOZm one .5 madman .umo mo s.>c3Haum u uo :Oauunaccuuo acucuosuum. .uawac .n :« nowadauouoo or» so canon “cannon sagucuooo .zamo caduceus can xooum agendas .zamz nuances "one HuuuuoEEou auto enema Adazmxmmmmxu unwound Acuocou can Hoscouuom ”coca (\III oxen: can can: unocoeuom “zazm \i( ll . 3 8 tho 3a3h _ Band ZBLO a: A . i— — «L n F n _ IIIIIII TL _ Asazm .aamc .zaaov _ BUHmBMHD . _ P‘ L _ _ zaza_ _eamm __mea __ qzo _ zone N n 15“ V I >1 . \ I c 1‘ \ \\\ . ,4 . ~ _ \\ . " q43 i where Qj represents the potential (or maximum obtainable) outputs from given input combinations. Only technically efficient firms satisfy the final equality. All other firms have a smaller actual output, Qj than would be achieved if they were efficient by the standards specified by the production function. An infinite number of sets of &i will satisfy equa- tion 2.4. To force the estimated production surface to lie as close as possible to the observed set of points, the linear sum of the errors is minimized. Also, note that no firm can produce more than its potential output for a given input combination it actually used. The estimation technique is then equivalent to: n A Minimize 2 1n E. m Subject to: Z a. 1n Xij 1.1n Qj and a 270 This is a linear programming (LP) problem. Since 1n E. = ln . - ln Q. Q] J the objective function for the LP problem can be rewritten as: Minimize II Mb ...-I :3 ts > II II M5 :1 For any set of data the term (. _g ln Q.) 3—1 3 is a constant. Any set of ai that minimizes for one value of 3 ln Q.) 3-1 3 will minimize for any other value including zero, so the term can be drOpped from equation 2.5 without consequence. The remainder is suitable as the objective function for the LP. The objective function then becomes: a. 1n X. Minimize O 1 1 "Ma 1 271 The LP method of estimating the production function can be written in an expanded form as: n A " +" +...+" M1n1m1ze j;l(a0 + alln Xl a2 1n X2 am 1n Xm) Subjecttoa +8, lnX +&zlnx +...+& lnX :Q and a. > 0 . 1— Solving this LP problem, we obtain vectors 1n Q3. and A 1n Ej (the slack variable in the LP problem). Since In E. = 1n A. - 1n . J Q3 Q3 taking the antilogs of this equality, we get: [11 ) ll pets > Inverting this ratio, we get: UPI H II {O’LP The ratio Qj/Qj is the ratio between actual and potential outputs, hence, by definition, it is the technical efficiency 272 A (TE) index of firm j. In other words, the vector l/Ej con- tains the TE indices with one index for each observation. 2. Estimation of the Cost Function Nerlove (1965)1 derived the cost function from the Cobb-Douglas production function as follows (in logarithm): "“43 = + . 1n TCj Y0 Y1 ln P ln Q. + 1n U. 1 J . J .. + Y i 1] n+1 where: TCj : the total cost of firm j Q : the output of firm j P. : the price of factor i paid by firm j a random error term that contains efficiency C.‘ term as well This equation can be written in more compact form as: m-l - 1n TC. = . 1n P.. + 1n U. (2.8) 3 iéo Y1 1] 3 where the first column of the matrix 1n Pij is a vector of ones to allow for the intercept, and the (m + l)th column is a vector containing ln Qj' 1. Marc Nerlove, Estimation and Indentification of Cobb- Douglas Production Functions (Chicago: Rand McNally, 1965). 273 Since the cost function specifies the minimum TC in contrast to the maximum output defined by the production function, estimating the former represents the inverse case of estimating the latter. In other words, the cost function is estimated by fitting an "envelope from below" to the observed points rather than an "envelope from above" as in the case of the production function. Also, one constraint of the cost function is that no firm can achieve an actual total cost, TCj, lower than its potential total cost, TCj. In order to be a minimum frontier, the cost function expressed in equation 2.8 should be estimated such that: + A 2 y ln P.. = ln TC. < ln TC. (2.9) i=0 1 13 3 " J This can be done by: 1'1 A Minimizing 2 1n U. j=l 3 m+1 A Subject to 2 Y1 ln P 5_ln TC. i=0 and Yi :_0 Also note that: 1n U. = 1n TC. - 1n TC. 3 J J m+1 A = 1n TC. - 2 y. In P 3 i=0 1 13 274 The same mathematical process as in Chapter 2 under A Method of Dealing with Two Outputs, can be used to derive the objective function for the LP problem: m+1 A Minimize - X 7. 1n P. . 1 1 1=0 m+1 or Maximize 2 7. 1n P. i=0 1 1 The LP method of estimating the cost function can be written in an expanded form as: Maximize A A A + . . . + A A (Y0 + Y1 In P + Y2 1n P Ym 1n P + Ym+ 1n Q) 1 ll M23 j 1 Subject to: A A +...+ ' 21 Ym 1“ Pm1+'Ym+1 Y0 Y1 ln Pll Y2 ln P 1n legln TC A + A + Y0 Y1 1“ Pln and Yi‘: 0 Solving this LP problem, we obtain vectors 1n TCj and In Uj' Taking the antilogs of the equality A 1n U. = 1n TC. - 1n TC. 3 J J 275 we get: U. = TC. TC. 3 3/ 3 By inverting this vector, we also get: 1 U. = TC. TC. / J J/ J The ratio TCj/TCj is the ratio between potential and actual costs, hence, by definition, it is the cost efficiency (CE) index of firm j. In other words, the vector l/Uj con- tains the CE indices with one index for each observation. BI BL IOGRAP HY Books Aldcroft, Derek H. British Transport Since 1914. London: Davis and Charles, Inc., 1975. Allen, G. Freeman. British Railways After Beeching. London: Ian Allen, 1975. Bonavia, Micheal R. The Organization of British Railways. London: Ian Allen, 1971. Braunthel, Gerard. The West German Legislative Process. A Case Study of Two Transport Bills. Ithaca: Cornell University Press, 1972. Carlson, Sune. A Study on the Pure Theory of Production. New York: A. M. Kelly, 1969. Despicht, Nigel S. Policies for Transport in the Common Market. Kent: Lambarde Press, 1964. Dogan, Mattei, ed. Top Mandarins of Western Europe. The Political Role of Top Civil Servants. New York: John Wiley and Sons, 1975. Doukas, Kimon A. The French Railroads and the State. New York: Columbia University Press, 1945. Einaudi, Mario et a1. Nationalization in France and Italy. New York: Cornell University Press, 1955. Foster, C. D. The Transport Problem. London: Blackie and Son, ltd., 1963. . Politics, Finance and the Role of Economics. London: George Allen and Unwin, ltd., 1971. Friedman, Wolfgang G., ed. Government Enterprises. A Comparative Study. New York: Columbia University Press, 1970. Gwilliam, K. M. Transport and Public Policy. London: George Allen and Unwin, ltd., 1970. 276 277 Hanson, Albert H. Public Enterprises and Economic Develop- ment. London: Routledge and K. Paul, 1959. Hazard, John L. Transportation: Management, Economics, Policy. Cambridge: Cornell Maritime Press, Inc., 1977. Hirschman, Alber. Exit, Voice and Loyalty. Cambridge: Harvard University Press, 1970. . Devel0pment Projects Observed. Washington, D.C.: The Brookings Institution, 1967. Humpidge, C. Relations Between Road, Rail and Water Trans- port. Its Control and Direction in the Different Countries Since 1945. Brussels: International Union of Public Transport, 1965. Japanese Institute of Labor. The Changing Patterns of Industrial Relations in Asian Countries (Proceedings of the Asian Regional Conference on Industrial Rela- tions). Tokyo, 1969. Johnson, Chalmers. Japan's Public Policy Companies. Wash- ington, D.C.: American Enterprise Institute for Public Policy Research, 1978. . Conspiracy at Matsukawa. Berkeley: University of California Press, 1972. Joy, Steward. The Train that Ran Away. London: Ian Allan, 1973. Katzarov, Konstantin. The Theory of Nationalization. The Hague: Martinus Nijhoff, 1974. Kelf-Cohen, R. British Nationalization 1945-1973. New York: St. Martin's Press, 1973. Kim, Key Whan. The Price and Output Policies of the Japan National Railways, 1949-1963. (Ph.D. dissertation) University of California, Berkeley, 1971. Klein, Lawrence R. A Textbook of Econometrics. Evanston, Illinois: Row, Peterson and Co., 1953. Marris, Robin. The Economic Theory of Managerial Capitalism. London: MacMillan and Co., ltd., 1967. Morrison, Herbert (Lord). Socialisation and Transport. London: Constable and Co., ltd., 1933. 278 Narlove, Marc. Estimation and Identification of Cobb-Douglas Production Functions. Chicago: Rand McNally, 1965. Polanyi, George. Contrasts in Nationalized Transport Since 1947. London: Eaton House, 1968. Pozen, Robert C. Legal Choices for State Enterprises in the Third World. New York: New York University Press, 1976. Pryke, Richard. Public Enterprise in Practice. London: MacGibbon, 1971. Reid, Graham L. and Kevin Allen. Nationalized Industries. Baltimore: Penguin Books, 1970. Robson, William, ed. Public Enterprise. London: George Allen and Unwin, Ltd., 1937. . Problems of Nationalized Industries. New York: The Oxford University Press, 1952. , Nationalized Industry and Public Ownership. New York: University of Toronto Press, 1960. Rowley, C. K. Steel and Public Policy. New York: McGraw Hill, 1971. Splawn, Walter M. W. Government Ownership and Operation of Railroads. New York: MacMillan Co., 1928. Thompson, Slawson, ed. The Railway Library, 1913. Chicago: R. R. Donelly and Sons Co., 1914. Toshihara, Watarai. Nationalization of Railways in Japan. New York: Ph.D. dissertation, Columbia University, 1915. Turvey, Ralph, ed. Public Enterprise, Selected Readings. Baltimore: Penguin Books Yoshitaka, Kiyohiko. An Introduction to Public Enterprise in Japan. London: Sage Publications, 1973. Articles Aigner, D. J. and S. F. Chu. “On Estimating the Industry Production Function." The American Economic Review, Vol. 58, No. 4 (September 1968), pp. 8-16. 279 Aldcroft, D. H. "Innovation on the Railways. The Lag in Diesel and Electric Traction." Journal of Transport Economics and Policy, Vol. 3, NO. 1 (January 1969), pp. 96-107. . "Book Review on 'The Organization of British Railways' by M. R. Bonavia." Journal of Transport Economics and Policy, Vol. 6, No. 1, p. 95. Alias, J. "Structural Organization of a Railway System." Rail International, April 1976, pp. 191-207. Averch, Harvey and Leland Johnson. "The Behavior of the Firm Under Regulatory Constraint." The American Economic Review, Vol. 52, NO. 5 (December 1962), pp. 1052-1069. Boles, James N. "Efficiency Squared: Efficiency Computa- tion of Efficiency Indexes." Western Farm Economics Association, Proceedings, 1966, pp. 137-142. Bonavia, Micheal R. "BR's New Field Organization Proposals." Modern Railways, July 1971, pp. 5-7. Carlsson, B. "The Measurement of Efficiency in Production: An Application to Swedish Manufacturing Industry, 1968." Swedish Journal of Economics, Vol. 74, NO. 4 (1972), pp. 468-488. Comannor, William S. and H. Leibenstein. "Allocative Effi- ciency, X-Efficiency and the Measurement of Welfare Losses." Economica, Vol. 46, NO. 143 (August 1969), pp. 304-309. "Commercial Drive Puts SNCF into Growth Spiral." Railway Gazette International, Vol. 131, No. 4 (April 1975), pp. 133-135. ' Crew, M. A., M. W. Jones-Lee, and C. K. Rowley. "X-Theory versus Management Discretion Theory." The Southern Economic Journal, Vol. 38, NO. 2 (October 1971), pp. 173-184. Dallmayr, Winfried. "Public and Semi-Public Corporations in France." Law and Contemporary Problems, Vol. 26, NO. 4 (Autumn 1961), pp. 755-793. "DB Oeftering: Economist at the Helm." International Rail- way Journal, Vol. 4, No. 5 (May 1964), pp. 28-29. 280 DeBrulle, Daniel. "Economic Democracy in Nationalized Undertakings in France and Great Britain." Annals of Public and Co-operative Economy, Vol. 36, No. 4 (October—December 1965), pp. 461—492. Delacarte, Louis. "Tariff Policy and Operating Technique." International Railway Congress, June 1964, pp. 32-36. Diamant, Alfred. “A Case Study of Administrative Autonomy: Controls and Tensions in French Administration." Political Studies, Vol. 6, No. 2 (June 1958), pp. 147—166. Dimock, M. E. "Government Corporations: A Focus of Policy and Administration." American Political Science Review. Part I, Vol. 43, NO. 5 (October 1949), pp. 899-921 and Part II, Vol. 43, No. 6 (December 1949), pp. 1145-1164. Dougall, H. E. "Public and Private Operation of Railways in France." The Annals of the American Academy of Political and Social Sciences, Vol. 201 (January 1969), pp. 211-216. . "Railway Nationalization and Transport Coordi- nation in France." The Journal of Political Economy, . "Some Lessons Drawn From EurOpean Experience." American Economic Review, supplement, Vol. 25, NO. l (1935): pp. 111-127. . "Railway Rates and Rate Making in France Since I921." The Journal of Political Economy, Vol. 41, No. 3 (June 1933), pp. 289-233. Dreze, Jaques H. "Some Postwar Contributions of French Economists to Theory and Public Policy." American Economic Review, Vol. 54, NO. 1 (June 1964), pp. 1-54. Farell, M. J. "The Measurement of Production Efficiency." Journal of the Royal Statistical Society (Series A, General), Vol. 120, Part 3 (1957), pp. 243-281. and M. Fieldhouse. "Estimating Efficient Produc- tion Function Under Increasing Return to Scale." Journal of the Royal Statistical Society(Series A, General), Vol. 125, No. (1962), pp. 252-267. Fioc, Albert. "Traffic Evolution and Modernization in France." International Railway Congress, June 1966, pp. 6-9. 281 "German Federal Railway." International Railway Journal, Vol. 4, NO. 5 (May 1964), pp. 25-27. "The German Federal Railway Sinks Deeper into Trouble." Modern Railways, March 1976, pp. 112-113. Guilbert, R. "The Reform of the French Railways: Its Principles, Its Objectives, the Conditions for Its Success." Rail International, Vol. 1, NO. 7-8 (July—August 1970), pp. 449-458. Gunther, Arno. "The German Railways and the Collective Economy." Annuals of Public and Co-operative Economy, Vol. 31, No. l (January-March 1960), pp. 24-90. Hanson, Albert H. "Parliamentary Questions on the Nation- alized Industries." Public Administration, Vol. 29 (Spring 1951), PP. 51-66. Harcavi, Georges. "Nationalization of the French Railways." The Annals of the American Academy of Political and Social Sciences, Vol. 201 (January 1939), pp. 217-226. Harrison, A. J. "Transport in the E.E.C." Parts 3 and 4. Modern Transport, March 2, 1963, p. 12 and March 16, 1963, p. 10. Hirschfield, Andre. "The Role of Public Enterprise in the French Economy." Annals of the Public and Co—operative Economy, Vol. 44, No. 3 (July—September 1973), pp. 225-270. "Japan's Troubled Railroads." Business Week, April 24, 1971, pp. 38-400 "Japanese Railways Move Ahead." International Railway Journal, Vol. 4, No. 10 (October 1964), pp. 45-50. "Learning to Survive in the Market." Railway Gazette Inter- national, Vol. 135, No. 1 (January 1979), pp. 43-46. Leiberstein, Harvey. "Allocative Efficiency vs. "X- Efficiency.'" The American Economic Review, Vol. 56, NO. 3 (June 1966), pp. 392-415. . "Organizational or Frictional Equilibria, X- Efficiency and the Rate of Innovations." Quarterly Journal of Economics, Vol. 48, NO. 4 (November 1969), pp. 600-623. 282 Lewis, Edward. "Parliamentary Control of Nationalized Industry in France." The American Political Science Review, Vol. 51, NO. 3 (September 1957), PP. 669- 683. Levy, Pierre. "The Railroads in France." The Annals of the American Academy of Political and Social Sciences, No. 187 (September 1936): Pp. 184-192. Leyen, Alfred V. "The German Federal Railway Company." Annals of Collective Economy, Vol. 2, No. 4 (November- December 1926), pp. 321-344. MacMahon, Arthur W. "Autonomous Public Enterprise. The German Railways." Political Science Quarterly. Part I: Vol. 54, No. 4 (December 1939), pp. 481-513; Part II: Vol. 55, No. 1 (February 1940), pp. 24-52; Part III: Vol. 55, NO. 2 (1940), pp. 176-198. Marschak, Thomas. "Capital Budgeting and Pricing in the French Nationalized Industries." The Journal of Business, Vol. 33, NO. Milhaud, Edgard. "Nationalization in Great Britain." Annals of Collective Economy, Vol. 19, NO. 2 (April- June 1948). pp. 65-157. Moreau, D. "French Railmen Fight against Dismantling of the Railways." The Railway Review, January 11, 1963, pp. 6-7. Munby, D. L. "The Productivity of the British Railways." Bulletin of the Oxford University Institute of Statistics, Vol. 24, No. 1 (February 1962), pp. 113- 145. . "Mrs. Castle's Transport Policy." Journal of Transport Economics and Policy, Vol. 2, No. 2 (May 1970): PP. 145-156. Oeftering, H. M. "Modern Management and the German Federal Railways." Rail International, Vol. 1, No. 3 (March Okado, Kiyoski. "Transport System, Transport Policies." The Wheel Extended, Vol. 2, No. 3 (Sinter 1972), pp. 28-36. Okano, Yukihedi. "A Century of Transport Policies." The Wheel Extended, Vol. 2, NO. 3 (Winter 1972), pp. 3-11. 283 Pryke, Richard. "Public Enterprise and the Economics of Large-scale Production." Annals of Public and Co- operative Economy, Vol. 41, No. 1 (January-March 1970). PP. 43-62. "The Reorganization Game." Modern Railways, May 1969, pp. 345—346. "Reorganization of the British Railways Board." Railway Gazette, Vol. 126, No. 1 (January 2, 1970), pp. 9-12 0 Ridley, F. "The German Federal Railways--A State Admini- stered System." Parliamentary Affairs, Vol. 17, No. (1964): Pp. 182-194. Rockefelt, Werner. "Internal Organization (The DB)." The Railway Gazette International Railway Congress Number, Ruggles, Nancy. "The Welfare Basis of the Marginal Cost Pricing Principle." The Review of Economic Studies, Vol. 17, No. 42 (1949-1950), pp. 29—46. . "Recent Developments in the Theory of Marginal Cost Pricing." The Review of Economic Studies, Vol. 47, No. 43 (1949-1950). PP. 107-126. Schenker, Eric. "European Transport Policy and the Common Market." Public Utilities Fortnightly, Vol. 75, NO. 2 (June 10, 1965), pp. 37-45. Schmidt, Peter. "On the Statistical Estimation of Parametric Frontier Production Functions." The Review of Econo- mics and Statistics, Vol. 58, No. 2 (May 1976), pp. 238-239. Seidman, Harold. "The Theory of the Autonomous Government Corporation: A Critical Appraisal." Public Admini- stration Review, Vol. 12, No. 2 (Spring 1952, pp. Seitz, Wesley D. "The Measurement of Efficiency Relative to a Frontier Production Function." American Journal of Agricultural Economics, Vol. 52, No. 4 (November 1970): Pp. 505—511. "Seven Chairmen." Modern Railways, May 1976: PP- 173'174- "Sharp Upswing in Container Loadings Could Help Beat Freight Challenge." International Railway Journal, Vol. 12, No. 10 (October 1972), pp. 40—47. 284 "SNCF: A Masterpiece in Railroad Technology." International Railway Journal, Vol. 3, NO. 11 (November 1963), pp. 33-47. "The State in the Market." The Economists, December 30, 1978, pp. 37-55. "Ten-Year Plan to Make JNR Viable." Railway Gazette, Vol. 26, No. 6 (June 5, 1970), pp. Thomas, William E. "A Look at German Rail." Passenger Train Journal (Summer 1975), pp. 10-13. Timmer, Peter. "On Measuring Technical Efficiency." Stand— ford University Food Research Institute Studies in Agricultural Economics, Trade and Development, Vol. 9, No. 2 (1970): PP. 99-170. Trench, Sylvia. "Transport in the Common Market." Planning. (July 8, 1962), pp. 225-281. Turvey, Ralph and A. R. Prest. "Cost-Benefit Analysis: A Survey." The Economic Journal, Vol. 75, NO. 300 (December 1975), pp. 683-735. Walker, Guilbert. "The Transport Act 1947." Economic Journal, Vol. 58, No. 1 (March 1948), pp. 11-30. Walters, A. A. "The Long and the Short of Transport." Bulletin of the Oxford University Institute of Statistics, Vol. 27, NO. 2 (May 1965), pp. 97-101. Wickham, Sylvian. "The Development of French Railways Under the French Four-Year Plans." Bulletin of the Oxford University Institute of Statistics, Vol. 24, No. 1 (February 1962), pp. 168-184. Reports and Documents European Economic Community Commission. Legal Status of Rail, Road and Inland Waterway Transport in the Member States of the European Economic Community. Paris: Publication Services of the European Communi- ties, July 1962. The German Federal Railways. DB Today. Frankfurt, 1964. Japanese National Railways. The Japanese National Railways. A General Description. Tokyo: Nikon Kokuyu Tetsudo, 1963. 285 International Union of Railways. International Railway Statistics. Paris: The UIC Statistics Bureau. Annually. National Economic Development Office, The United Kingdom. Relationship of Government and Public Enterprises in France, West Germany and Sweden. London: National Economic Development Office, 1976. National Economic Development Office, The United Kingdom. A Study of UK Nationalized Industries. London: National Economic Development Office, 1976. Takatori, Yoshiaki. Modern Management in the Japanese National Railways (mimeographed report, 1972). United Nations, Economic Commission of Europe. Annual Bulletin of Transport Statistics for Europe. Geneva: annually (E/F/R. T3.II.E.19). - United Nations. Report of the United Nations Seminar on Organization and Administration of Public Enter- prises. New York, 1967 (ST/TAO/M/35).