L.S.E. Essays on Cost

Edited by: Buchanan, James M. and George F. Thirlby
(1919- )
Display paragraphs in this book containing:
First Pub. Date
New York: New York University Press
Pub. Date
Collected essays, various authors, 1934-1973. First published as a collection 1973 for the London School of Economics. Includes essays by Ronald H. Coase, Friedrich A. Hayek, Lionel Robbins, and more.
11 of 15

The Ruler

A paper read, in part, to the Cape Town branch of the Economic Society of South Africa on 11, October 1946. First published in full in the South African Journal of Economics (December 1946).



This is a paper*48 concerning rules based upon the idea that the output of an industrial unit ought to be adjusted so as to secure that revenue bears a certain relationship to cost: so that marginal revenue is equal to marginal cost: or that price is equal to marginal cost; or that total revenue is equal to total cost. It is not intended to raise the question whether there is theoretical justification for saying that one or other of these relationships is the one that ought to be achieved. The question that it does raise is whether the actual cost-revenue relationship is an objective something in the sense that it can be scrutinized, in order to discover whether it conforms to the desired relationship, by an ad hoc authority external to the industrial unit concerned.*49 Somewhere or other I have seen the suggestion that, for failing to carry out instructions to achieve the desired relationship, the 'manager' of the unit should be dismissed. It is obviously implied that somebody can look and see whether the relationship is being, or has been, achieved. Throughout my paper I shall refer to the ad hoc authority as 'the Ruler'.*50 The expression 'the rule' will refer to these rules in general unless I say that I am discussing a particular one.


This then is the suggestion emerging from my paper: that the relationship has not the objectivity that is by implication attributed to it; consequently that the application of the rule is impracticable. It follows too that the proposal to apply the rule betrays a grievous failure to allow for the nature of the administrative task. This is a remarkable conclusion, because, whatever else the rule advocates wish to do, their main object seems to be to secure the appropriate administration of resources: they seem to want to get things in the right places so as to satisfy everybody as much as possible in the best way.


I shall first discuss and illustrate at length the nature of the administrative task and how the cost-revenue relationship emerges with the performance of that task and then consider the application of the rule on the supposition that the terms 'cost' and 'revenue' are to be taken to have the meaning implied in the opening discussion, and also on the supposition that they are to be taken to have other meanings.

Administration in planning stages. Location, layout and choice of equipment


Administration occurs in planning stages. The cost-revenue relationship and the final budget relating to intended operations emerge in planning stages. Important aspects of planning are the choice of the location, layout and technical equipment of the plant. The most efficient structure is not given technically, but depends upon an administrative judgement about the forthcoming market conditions. The choice and the cost-revenue relationship are inextricably interconnected. The choices are a matter of selection from alternative combinations of resources: an infinite number of variations from which the administrator selects a few for more deliberate consideration before he makes a final choice. A contemplated variation to achieve revenue in one way may involve contemplated losses of revenue that might be achieved if the variation did not occur: the sacrifice of potential alternative revenue appears as cost. A rule purporting to secure a certain cost-revenue relationship could have no claim to secure the optimum combination and spatial distribution of resources unless it took account of opportunities to be displaced. But alternative opportunities are elusive even to the person whose job it is to look for them, and are never susceptible to precise measurement even by him.



During the deliberations concerning the planning and running of the Cape Town and Wellington Railway, much discussion and dispute arose over the question of how to connect the line with a place called Stellenbosch, which lay south of the direct route between the terminii. One of the alternative methods proposed was to allow the trunk to deviate in order to take in Stellenbosch. But it was pointed out that this deviation would be likely to cause the loss of traffic which a direct trunk would obtain from important corn-growing districts north of the line. In other words, at this level of deliberation, one element in the cost of this variation in structure in the sense of layout (or, alternatively, location) would be the anticipated loss of revenue from this traffic from the north, a cost which would have to be set against the expected revenue from the Stellenbosch traffic. There was an opportunity of earning certain revenue—from the corn traffic—by one variation of structure. That opportunity would be lost by adopting the other variation instead.


To prevent this loss of traffic from the corn-growing districts, it was suggested that the line should follow the direct route for a certain distance from Cape Town (so making it unnecessary for people in the north to cross some shifting sands in order to get to the line), and then make the deviation. But, as this arrangement would add somewhat to the total length of journey by rail from Wellington to Cape Town, an objection to it was raised on the ground that people near to the Wellington terminus would prefer to send their goods direct by road. At this stage of the discussion, the anticipated loss of revenue from the Wellington traffic can be regarded as an element in the cost of retaining the corn traffic by a projected modification of the technical structure in the sense of layout (or location).


Another method of providing for Stellenbosch which was calculated to avoid losing either of these traffics was that of constructing a short branch from about the middle of a direct trunk. This proposal again introduced the question of reduced traffic—e.g. from people south of the line who would have a further distance to travel to Cape Town via the branch should they use it. However, I have sufficiently indicated how such influences must be presumed to be interrelated with the choice of structure in one sense of the expression, namely, in the sense of layout or location, and at this point only wish to suggest how they are interrelated with it in the other sense of choice of technical equipment. It was proposed that, if a branch were substituted for the deviation, it might take the form of a tramway (presumably a horse-tramway) instead of being equipped to the standard intended for the trunk line. Obviously, the point of the proposal was that revenue might not be adequate to justify the introduction of resources of higher cost than those required for the tramway.


The decisions reached need not concern us: all that is necessary is that we should see the influences at work, and understand their meaning in relation to the problem of the appropriate distribution of resources; in particular, the meaning of the omission of lost opportunities from 'cost'.

The omission of lost opportunities from 'cost'


I know the thoughts that will be in some minds as I say this: 'But it is not this aspect of cost that we are interested in. What we are interested in is cost in the sense of the amount of liquid resources which it is finally decided to bring into the enterprise: the money which will flow in'. I must at this point be brief in attempting to allay this impatience. I shall simply assume (pretend) that whoever it was who had the disposal of the liquid resources for the railway undertaking was at the same time considering as an alternative investment the disposal of exactly the same liquid resources by erecting and running a chain of saloon bars. Then the cost of bringing the liquid resources into the railway undertaking must be related to the yield in revenue expected to accrue from this alternative opportunity. But to confine one's attention to this level of deliberation allows to escape from notice all the fallible judgements connected with the planning of location, layout and technical equipment which it was my purpose to bring to the fore.


It seems fairly obvious that, even if, at any particular level of deliberation, the rule could be, and was being, observed, its observance would not imply that people's preferences were not to be defeated through inappropriate combination or spatial distribution of factors of production, unless into 'cost' were added lost opportunities which the decision-maker had never observed, or, having observed, had ignored. For example, if the application of the same factors and money in one manner or place A were expected to yield 100, in another manner or place B 100, and in another manner or place C 150, but if C were ignored or not noticed, A being chosen (for no apparent reason!) instead of B, cost (in the sense of anticipated alternative revenue) would be 100 and cost equal to revenue. If, however, C were taken into account in considering A, cost would be 150, i.e. higher than revenue 100; if the relative strength of anticipated money demand were taken as the criterion, the factors and money would be more 'efficiently' used not at A but at C, where cost would be 100 and revenue 150. It is not suggested, of course, that the 'inefficiency' involved in ignoring opportunities would be confined to undertakings subject to the rule; only that the rule may not be able to ensure that it does not occur. It no doubt occurs anyway, because nobody is omniscient. As a safeguard against it, one relies, within limits, upon somebody else's taking advantage of one man's oversight.

The railway illustration is not an exceptional case


It may be suggested that the problem of settling the railway route to which I have referred is one special to railway undertakings and has little relevance to other undertakings. This is certainly not true. The connection will be more obvious if the problem is thought of as analogous to that of settling the location, rather than the layout, of a different kind of industrial plant: the location problem is one of judging the strength of conflicting pulls, some of them different factor supply prices, some of them different product demand prices, at different geographical points and—when changes over time are being forecasted—at different points of time. The other matter—choice of equipment—is a question which in many industrial plants is likely to be constantly recurring, not only when new (or renewed) equipment is required, but also when projected changes in kind or quantity of output raise the question of what old or new equipment to use.

The propensity to avoid the issue


It may be sought to avoid the issue by saying that anyway this matter of planning the technical structure is one which would be settled by somebody other than 'the manager' whose cost-revenue relationship is to be scrutinized; that for him the plant would be a technical datum, outside the range of his responsibility. This might in some cases and to some extent be true; but to make it so would be merely to shift the burden of cost-revenue calculation*52 on to the shoulders of a higher authority, and leave his activities, and the relations between him and 'the manager', still to be discussed. The economist who abstracts from these questions abstracts to the same extent from business administration.


There is a tendency for economists to avoid or ignore the problem of choice of structure.*53 Some of them regard case discussion of deliberations upon such matters as being of technical rather than of economic importance.*54 In 'short-run' theoretical cost-revenue discussions, the choice of structure tends to be hidden behind a vague something termed 'the entrepreneur's fixed costs'. The mere assumption, for 'long-run' discussion, that the 'firm' or 'entrepreneur' selects, or has selected, an optimum structure, does not carry us far in the direction of understanding the issues rendering the choice, and the cost-revenue relationship bound up with it, highly subjective, indefinite and fallible. I shall refer to these concepts of 'fixed costs', 'short run' and 'long run' again. But I must first call attention to another 'inefficiency' factor which might preclude the application of the rule: that arising out of loose coordination of the diffused administrative elements inside a large-scale industrial undertaking—i.e. an undertaking in which administration is split among a number of people and has to be coordinated somehow.*55

Loose coordination


This appears as failure to bring one man's knowledge to bear upon the activities of another so as to influence a decision which is supposed to be made by them jointly.*56



Let us suppose that in a certain undertaking the buying of a raw material, manufacture and the selling of the product are controlled by three different people, who are heads of separate purchasing, production and sales departments respectively, the activities of the three people being coordinated by a committee of the three, the committee being superordinate to its subordinate members. Purchasing (at a definite price, or within certain price variations), manufacture per unit of time and sales (at a certain price) have been planned, and are proceeding under standing orders laid down by the committee to implement the plan. But for some reason sales are likely to fall below the level anticipated by the committee.


Now it is possible that the sales department, negligently or otherwise, remains unaware of the impending fall until it actually occurs. Or the sales department may be in various senses aware of it, although its awareness is not appropriately coordinated. Salesmen may know but not report. They may report, while their reports are left unread. The reports may be read by somebody, while the information is not communicated to the head of the sales department. The information may be communicated to the department head, but he may fail to report to the committee. For concision, let us suppose that the knowledge resides in the head of the sales department, but that for some reason it does not get communicated to the committee; and let us abstract from the possibility that the committee had, as a deliberate act of administrative judgement in laying down standing orders, decided not to consider changes in the selling market until an actual fall in sales occurred. Purchase of the raw material and manufacture continue under the existing standing orders, yielding a product which will not fetch the planned price.

Loose coordination failing to reflect impending market changes in the production plan


The illustration that I have chosen is one in which there is a failure to communicate a knowledge of impending changes in market conditions, and cause it to be reflected in the production plan. Where production is for 'the market', production must in any case precede sale: a risk that market conditions will change unexpectedly in the interim is always present. The expression 'efficient administration' might be said to imply the acquisition and use of knowledge of impending changes—to reflect, as it were, in the plan of operations, the changes which will have occurred by the time of sale. The best-administered organization is not omniscient: it is possible for market conditions to change against its latest anticipations. But it is possible to think of an organization being more or less lax or weak in revising its anticipations and securing the reflection of the changes in its production arrangements,*57 although we may not be able to devise a measure of the degree of its laxity or weakness. Possibly the test of efficiency here would consist of a comparison of an ex post account of operations with an ex ante plan of operations. In the illustration which I have given the weakness would be shown up by a falling short of realized below anticipated revenue—unless a sufficient safety margin had been allowed in the plan, a possibility which might hide the weakness, if weakness it would be when a safety margin had allowed for it.

Can the rule control the weakness?


It is difficult to see how the mere application of the rule could in this respect either add strength to the administration or operate as a test of its weakness. The conditions of the rule might well have been satisfied in the ex ante calculations, although the administration failed then and subsequently to allow for the change in the market conditions. That is to say, the administration might have planned, and begun to produce, that output which appeared to observe the approved cost-revenue relationship, and failed to adjust for the market change before receiving its impact. Weakness of corrdination is something not susceptible to measurement, though it would have its effect upon the capacity of the enterprise to stand on its own feet and, in so far as it was not allowed for by safety margins in forecasted results, upon the eventual divergence of results from the original forecast. I should perhaps say here, parenthetically, that if realized revenue has to be equal to expected revenue, it seems to me that the rule is being widened to accord (in respect of revenue) with a test that assumes that accounts ought to correspond with budgets.


It was consciousness of the importance of the acquisition of knowledge of impending changes, and of the importance of replanning to secure the reflection of changed anticipations in production operations, which led me to stress loose coordination as a possible inefficiency factor, and which now leads me to point out that replanning following changed anticipations involves a recalculation of cost; that at this replanning stage the cost of using acquired assets really depends upon the opinion of the administration of the undertaking about the alternative markets which it forecasts.

Replanning. Recalculation of cost. Cost dependent upon the opinion of the administration


If a market change which was not anticipated by the existing standing orders has occurred, or is now expected to occur, whether the late anticipation or non-anticipation of it is attributable to loose administrative arrangements or not, and whether the rule has been applied or not, it is apparent that replanning is called for: cost has to be recalculated. In the circumstances in which the firm finds itself with accumulated stock which it cannot sell at the planned price, the cost of getting rid of the stock immediately in the intended market is determined by the opportunities which the administration foresees for its disposal immediately in different markets or later on in the same or different markets. Very similar remarks apply to the equipment used in production. It is fairly apparent that what cost will be is a matter of the opinion of the administration of the firm. This opinion is dependent upon their forecasts of their market opportunities.

The problem of structure-variation recurs


At this replanning stage the administration is faced with choices similar to those between alternative combinations (or spatial distribution) of factors with which it was faced at the original planning stage. The choice might for example be one between continuing production at a reduced level with the same equipment or (if the equipment had sufficiently important competing uses) switching to other equipment. Clearly cost and revenue calculation may again be interconnected with opinion about appropriate structure-variation.


The rule is concerned with planning stages or it is not concerned with cost. To discover cost of production continuing under standing orders, it is necessary to trace back to planning stages. The terms 'fixed costs' and 'variable costs' should be reconsidered. If those who advocate the enforcement of the rule plead that they are not concerned with these planning stages, but only with interim periods between planning stages, it must be replied that then they are not concerned with cost. Cost occurs only when decisions are made, that is, in planning stages. In the interim periods resources (including money) flow under standing orders, but the cost involved in the flow was incurred by the decisions which settled those standing orders. Machlup has extended the familiar saying 'All costs are variable in the long run' to 'All costs are variable in the long run, that is, in a planning stage'.*58 My remarks go a little further: costs occur only in planning stages. To discover what the cost of the flow of production occurring between two planning stages had been (and, a fortiori, what the cost of its intra- or extra-marginal unit had been)*59 it would be necessary to trace back to, and inquire into, the decision which started or changed the flow—i.e. to trace back to the planning stage and to the administrator's mind. It might be necessary to trace back to several planning stages. In doing this we should discover that if the distinction between 'fixed' and 'variable' costs is a distinction between what occurs in a planning stage and what occurs in an interim during which no planning occurs; the distinction is a false one.



Suppose that in January it had been decided to use certain plant for the production of a certain commodity throughout the year, without considering further the cost of using the plant, irrespective of whether output was changed during the year; and that in March the output (and quantity of materials, etc. applied to the plant) was changed by a new decision which, as arranged in January, was reached without reconsidering whether to transfer the plant to another use. Examination of the flow of resources in April would show certain quantities of factors being applied to the plant, and possibly certain money being paid for the factors—even for the use of the plant. But these would be merely objective flows (with, incidentally, no distinguishable marginal unit of product).*60 They ought to correspond with the anticipated objective flows embodied in the budget of January, subject to the revision of March; but to discover the costs which had been involved in settling these flows (and, incidentally, the cost which had been involved in the production of the marginal unit of product)*61 it would be necessary to trace back and do two things: 1) in order to discover the alternative opportunities of using the materials, etc. applied to the plant, and/or money paid for such things, to examine the narrower budgeting process and decision of March which had determined the volume of April production, and 2) in order to discover the alternative opportunities of using the plant throughout the year (or of using the money required for its hire), to examine the wider budgeting process and decision of January which had determined the use of the plant in April as well as the volume of production before March. In doing the second thing we should be looking for what are called 'fixed costs', which are supposed to be 'variable in the long run',*62 that is (as Machlup puts it) 'in a planning stage'. In doing the first thing, we should be looking for what are called the 'variable costs'. But to find them too it would be necessary to go to a 'planning stage': March was a planning stage.

The use of 'fixed' equipment. 'Bygones', user cost and renewals. The qualifications of the person(s) reckoning user cost


In an earlier part of my paper I referred to the propensity to avoid the issue concerning the administrative act of choice involved in planning. The avoidance of the issue occurs if it is assumed that plant has already been erected and its cost is therefore a 'bygone'.*63 A matter which requires elucidation here is the nature of the 'bygone'.


It is of course true, as I have indicated, that the administration of a firm may by an earlier decision, whose implementation becomes part of the firm's standing orders, rule out the cost of using 'fixed' equipment from consideration at later planning stages. But where the cost of using it has not been ruled out of consideration, or where the period during which it was so ruled out (but not the 'life' of the equipment itself) has expired, and a new decision as to the equipment's use is about to be made, the (anticipated) use will usually or often*64 have a cost which comes up for consideration in the deliberations preceding the impending decision: the cost dependent upon the administration's contemplation of contemporaneous or intertemporal alternative opportunities of using the equipment. In what seems to me to be alternative terminology coming from Keynes via Mr Bauer: 'The user cost of a unit of output in the short period is the reduction in the discounted value of expected future quasi-rents of a piece of equipment through using it for the production of that unit of output rather than leaving it unused.'*65


The same remarks apply not only to 'fixed' equipment but also to materials, unexpired labour contracts, other factors and money within the ownership of the firm at the planning stage. It is then not admissible to assume that the cost of using 'fixed' equipment and other owned resources is a 'bygone' that does not have to be taken into account in planning. What is a 'bygone' that has become irrelevant is the loss of whatever has already been given up (or contracted to be given up)—usually money—in exchange for those things.


It follows from what I have just said that the 'long period' at the end of which 'fixed' equipment cost reappears does not necessarily run until the time comes for renewal of equipment.*66 It is not admissible to assume that the consideration of equipment cost is restricted to occasions when decisions are taken to instal equipment, even if these occasions include those upon which renewals or extensions are contemplated: the consideration of the cost of using equipment occurs at other times. The process of settling what this cost is consists in an administrative judgement upon whether it is better to use the equipment for this or for that product, to use it immediately or to reserve its use for the more distant future. Obviously this is a judgement upon what is going to happen in the selling market, and one which might be expected to be made by somebody who is operating in the market and who probably spends his life there. It is a judgement upon what people (the consuming public) are likely to be willing to do with their future incomes. There seems to be quite inadequate allowance in academic discussion for the fact that this sort of thing has to be done—an inadequate allowance for the fact that the selling prices of the product and the quantities saleable are not known, but have to be judged,*67 at the earlier point of time when the decision to produce is made, and that the judgement has to go on continuously, or at least constantly, if appropriate rebudgeting and readjustment of production is to occur subsequently.*68

The conjuncture out of which the cost-revenue relationship emerges. Its indefiniteness and fallibility. Competent administration more important than the rule?


The cost-revenue relationship will vary over a period of time as often as acts of administration occur. It is as indefinite and fallible as the opinion of the administration about the forthcoming market conditions and their significance, and as the laxity of the administrative coordination allows it to be. This is still more apparent when it is seen that at any point of time at which one of the acts of administration occurs, that particular act which determines the input and output of the firm as a whole, and incidentally the marginal unit of output,*69 is a coordinated decision dependent upon a number of simultaneous subordinate acts of administration (decisions) which are equally indefinite and fallible in the sense of being personal opinions about future conditions. In an abstract model which I have used elsewhere*70 the subordinate decisions in the particular undertaking and in reference to the particular coordinated decision ex ante appear in as many different places as spikes on a porcupine. Each of these subordinate decisions, and its fallibility, has its effect upon total output, and upon the total and marginal*71 cost of that output, settled by the final coordinated decision ex ante. It seems to be seriously open to question whether the result of this highly subjective and constantly modified conjuncture could be externally controlled by enforcement of the rule; and, from the point of view of getting output appropriate to the market environment, it might seem to be more important to secure that administrative positions were occupied by competent administrators than to try to doctor the cost-revenue relationship by issuing a rule to the effect that it must be this and not that.


An attempt to apply the rule might conceivably lead to the transformation of the Ruler into the administrator. This being so, it is as well to emphasize what his becoming administrator would mean.

The Ruler turned administrator


I suggested earlier that the advocates of the enforcement of the rule might deny that they were concerned with planning stages. Alternatively and oppositely they might conceivably say that the Ruler would be responsible for all planning, not only at each point of time at which planning occurred, but at each administrative point in the organization at which planning was occurring at any point of time. This could mean either of two things. It might mean that the Ruler would make all the subordinate, as well as all the coordinating, planning or budgeting (cost, revenue and output) calculations, leaving no administrative responsibility to any other person in the undertaking, all of whom would in the execution of the plan become executives or operatives functioning under standing orders (the plan) laid down by the Ruler (the sole administrator). In this case, even if the Ruler (as a sworn member of the Rule Party) could be relied upon always to apply the rule to his subordinate and coordinative calculations, it is obvious that (unless membership of the party also implied a particularly high degree of omniscience) the sphere of activity (size of undertaking) of any particular Ruler would have to be fairly restricted, to avoid the undertaking's becoming top-heavy. In this case all the personal (administrative) judgement would reside in the Ruler, but that personal judgement with all its limitations and weaknesses would not be escaped.


Alternatively responsibility for all planning might mean responsibility for coordination and overriding responsibility for subordinate acts performed by others—in the manner indicated in the abstract model of the mercantile organization.*72 Here his immediately subordinate administrators (subordinate Rulers) would be telling the Ruler what results they intended to achieve, but he would be relying largely upon his judgement of men to tell him how far they were likely to be right. (In the mercantile organization the highest authority might well have been relying largely upon an incentive to efficiency given to the subordinates by the promise of a commission varying with results.)


In considering the application of the rule I hope by implication to suggest why an attempt to apply it might lead to the transformation of the Ruler into the administrator.

The application of the rule, on the assumption that what is required is that total cost and total revenue should be equal. The nature of the cost-revenue relationship


From now onwards I must for the sake of clarity distinguish the different rules from one another, and for the sake of brevity confine my discussion to one of them: the rule that requires total cost and total revenue to be equal to each other.*73


What is to be the nature of the cost-revenue relationship that is to be the subject of examination by the Ruler? Is the expression to refer to

1. cost and revenue in the sense implied in my discussion so far,*74 or
2. the anticipated objective results in the budget, or 'estimated-profit calculation', which inevitably depend for what they are upon the deliberations leading to 1, and which consequently may be different according to who is the administrator, or
3. the realized objective results in the account, which ought, subject to certain exceptions, to agree with the (revised) budget?

1 The application when the relationship refers to cost and revenue in the sense so far implied


In this sense revenue must refer to the revenue expected by the administrator to accrue from the contemplated investment of certain resources in a particular way (say the investment of liquid resources and the administrator's own services in a railway undertaking); cost must refer to the revenue*75 that he would expect to accrue if the same resources were invested otherwise (in what he thought to be the best alternative way—say in a chain of saloon bars). I should say at once that it seems to me that to require that these two figures should be equal requires that the administrator should discover two avenues of investment between which he cannot choose without resorting to a toss-up. There is also the other difficulty that the cost figure will never become objective, i.e. it will never be possible to check whether the forecast of the alternative revenue was correct, because the alternative undertaking will never come into existence to produce the actual alternative revenue. Both these difficulties occur not only at this highest level of deliberation—i.e. where the total revenue from the contemplated undertaking is being compared with the total revenue of the potential undertaking proposed to be rejected—-but at all other levels of deliberation where a choice has to be made, e.g. where two alternative variations in structure are being compared with each other. I pass over the two difficulties as such, but ask the further question whether the Ruler would be expected to be content with nothing but two bare alternative revenue figures relating to the highest level of deliberation, or whether he would be required to look deeper into the two conjunctures throwing up the two figures. Going to the extreme, let us suppose that the administrator is to make a full statement concerning all the alternative outputs (and their variations over time) and technical structures that he has thought fit to consider as possibly being suitable to meet the demand conditions that he thinks likely to mature in the particular situation of the enterprise contemplated, and an appropriately incorporated statement (required to determine the cost of his proposed activities) relating to the alternative opportunities, for investment in other situations, to be displaced by investment in this particular one, and also a full confession of his doubts and fears concerning the accuracy of his judgements and the appropriateness of his tentative decisions.*76 Let us suppose further that both the administrator and the Ruler are agreed that as many relevant factual data (obtained e.g. by market surveys) as can be procured have been procured, and have been equally available to and considered by both, but that on some significant question which such data cannot completely answer, but which is a question for administrative judgement and decision (e.g. one relating to the probable strength and continuity of future demand), the administrator and the Ruler disagree. For example, suppose the case were similar to the railway case which I cited earlier, the administrator being of the opinion that a horsetramway from the trunk line would, in view of his low estimate of the strength and continuity of demand for service and his doubt about it, be the most appropriate way of serving Stellenbosch, the Ruler being more optimistic about demand for service and of the opinion that a railway line with heavy works and locomotives was justified. In these circumstances whose view would prevail? The point that I wish to make here is that to the extent that the Ruler overrides the administrator's opinion in such matters, and makes the decisions, the administrator ceases to be the administrator, and ceases to be responsible for covering costs.*77 The Ruler himself becomes responsible, the other person to that extent—if he is told to carry on in accordance with the Ruler's decision and does so—-his executive operating under standing orders.


This is not to say that there cannot be more than one administrator contributing to the coordinated decision ex ante. The Ruler or somebody else might, for example, be the medium for coordinating one administrator's anticipated results from investment in one undertaking (e.g. a railway undertaking) with another administrator's anticipated results from investing the same resources elsewhere (e.g. in a chain of saloon bars) instead, i.e. with cost in the proper sense. But it is intended to suggest that the terms and method of application of the rule should be explicitly defined, to avoid amongst other things the usurping of administrative authority without accepting responsibility*78 for its results—to say nothing of sanctions for error.*79 And I would add that it is extremely difficult to understand why the Ruler's stamp should be placed upon a proposition simply because the administrator thought that its yield would be equal to the yield of the same resources in some other proposition. It would be much easier to understand that the Ruler should be expected to choose, or confirm a choice, between two alternative propositions, whose expected yields were different—-and accept responsibility for doing so.


It is conceivable then that the advocates of the rule intend that the Ruler should be concerned with revenue and cost in the sense of accepted and displaced opportunities; but I suspect that they have not thought rigorously in these terms, but have thought instead of revenue and cost as the two sides of a budget or 'estimated profit calculation'.

2 The application when the relationship refers to the 'estimated profit calculation'


We now become interested, not in two alternative revenues expected to be achievable by the investment of the same resources, but primarily in the money outlays upon factors (as cost), and (as revenue) the revenue expected to be achievable in one of the alternative applications. These two things, the money outlays upon factors expected to be brought into the undertaking, and the revenue expected to be achieved, which are to be the subject of scrutiny, and which—subject to the adjustments for 'opening stock' and 'closing stock' or 'residual assets' which will be mentioned eventually—have to be equal to each other, are the anticipated objective—results in the budget or 'estimated profit calculation' relating to the proposed undertaking (e.g. a railway undertaking). The 'costs' which interest the Ruler are primarily anticipated objective money outlays.*80 That is to say, the administrator is an administrator in that his displacement of alternative opportunities, at any level of deliberation, is not to be queried. All that he has to show is that his expected objective outlays—subject to the same adjustments—are equal to his expected objective revenue. He shows this in a budget which he submits to the Ruler.


I do not propose to raise the issue whether the adoption of this procedure of working upon anticipated objective results (instead of accepted and displaced opportunities) does or does not mean that the rule has lost touch with its purpose of securing the direction of resources into the best social uses. Apart from this question, the most significant aspect of the procedure is the dependence of the Ruler upon the administrator for the accuracy of the administrator's forecasts of the objective results. If the budgetary period were anything like the physical life period of the longest-lived of the physically depreciating assets of a railway concern, the figures for the remoter years would be extremely tentative; and even if the budgetary period for an enterprise which was to possess such assets were shortened, the tentativeness would only be transferred to the administrator's calculation of the money value of the residual assets; but at the same time it would be necessary to produce such figures if the budget was to (pretend to) show that outlays and revenues were to be equal to each other in the end.


The budget that I envisage as being submitted would be of the kind referred to by Machlup as an 'estimated profit calculation', and illustrated in his article.*81 If, however, it had to show the anticipated time dispersion of the outlays and revenues over the budgetary period, its outlays and revenues would not be—or ought not to be—telescoped into annual figures. The statement would show separately the anticipated outlays and revenues as they were expected to occur. A particular outlay (e.g. upon a new set of locomotives) would appear (undivided) at the date at which it was expected to occur. If it were anticipated that there would be residual assets at the end of the budgetary period, the statement would presumably conclude with the anticipated money value of residual assets, in order to show that outlays and revenues would ultimately balance. This money value of residual assets would represent either what the administrator assumed the assets would sell for at the end of the budgetary period, or (what is not strictly an anticipated objective revenue) what he assumed they would be worth in terms of net revenue in subsequent budgetary periods, according to whether it was to be assumed that the enterprise would cease at, or continue after, the end of the budgetary period. If there were non-monetary assets at the beginning of the first budgetary period—and there would be at least the administrator's own services—a similar calculation would have to be made in respect of this 'opening stock'—otherwise the Ruler could not strike the balance (which ought to be nil) in money terms. These calculations would be made by the administrator as such.


If it were deemed to be necessary to break down the budgetary period into 'years', each 'year's' figures purporting to show whether or not a 'profit' was anticipated in respect of that 'year',*82 what would be the appropriate procedure? Clearly the method of charging an outlay undivided at the date at which it was expected to occur would itself affect the 'fluctuations' in the difference between anticipated outlays and revenues in any 'year', perhaps giving rise to apparent 'deficits' in 'years' of heavy outlays, and apparent 'profits' in others: the outlay might be in respect of a factor (e.g. a new set of locomotives) which was intended to be used over a series of subsequent 'years'. How would this have to be dealt with? The answer is that it would be necessary to insert the administrator's calculation of the money value of the residual assets*83 for the end of each 'year' instead of merely for the end of the budgetary period.


Clearly then to shift the application of the rule to the figures of the 'estimated profit calculation' does not obviate the Ruler's dependence upon the administrator's opinion about what is going to happen in the future—unless the Ruler usurps the administrator's function. But there is still another point. If the Ruler confines his scrutiny to the budget, he will have no check upon either the administrator's good faith, or his competence to achieve his ostensibly expected results. To this matter I shall return after considering the application of the rule when the relationship refers to the account.

3 The application when the relationship refers to the account


I now change my assumption, and suppose that it is not the budget at all, but the account only, that the Ruler has to scrutinize. Beyond inspiring fear ex ante, and applying any sanctions ex post, the Ruler would apparently have no function but to see whether cost and revenue in the sense of realized money outlays and revenues—subject to adjustments for 'opening stock' and 'closing stock' or 'residual assets'—were equal. What are the important issues in this case? One of them is this same difficulty with regard to the treatment of opening and residual assets as was encountered in the budget case. An account submitted with opening assets or while there were still significant residual assets would presumably have to have added to it the administrator's calculation for opening or residual assets. It may be asked whether this requirement could not be obviated by the employment of a professional accountant. A professional accountant—if the Ruler were not himself behaving as such—might, and probably would, be employed to ascertain whether the objective results were in fact those which had been reported in the accounts by the administrator to the Ruler: such is the accountant's function. But further than this he cannot go. Without usurping the administrator's function he cannot assess the (net) revenue still to be yielded by the use of residual assets.*84


So this new shift in the application of the rule—-to the account—does not, so far as interim accounts are concerned, obviate the Ruler's dependence upon the administrator's opinion about the future. Again there is another point. Knowing that his realized 'costs' and revenues would be required to be equal, and knowing that whatever his original estimates were they would not be likely to be exactly realized, and because it is probably easier to get rid of an emerging surplus than an emerging deficit, the administrator would probably be inclined to embark only upon undertakings in which, for any period in which he had to show equality of 'costs' and 'revenue', he felt fairly sure of being able to make a surplus or 'profit'; that is, he would be disinclined to embark upon undertakings unless, for each such period, he could budget for a surplus, and could rely upon being able to avoid the surplus somehow if he subsequently found that he was in danger of realizing it—e.g. by not responding to an upswing of demand to the extent provided for in his undisclosed budget.

A possible addition to the rule: the stipulation that the account should agree with the 'estimated profit calculation'


When the application of the rule shifted to the 'estimated profit calculation', a certain impotence resulted from the lack of the account; when it shifted to the account, a certain impotence resulted from the lack of the budget. Could potence be restored by requiring that the account should correspond with the budget, the Ruler scrutinizing both?*85 The new requirement—that both the budget and the eventual account should be submitted—would seemingly provide at least a reflection of the accuracy of forecasting—a check or report on efficiency in this sense. But it cannot be left at that. There is something significant to be said arising out of the point that nobody would expect a first budget, constructed with any definiteness of detail, to be exactly realized in the event. I have already pointed out that the figures relating to the remoter future would be extremely tentative. During the course of the budgetary period the administrator would be constantly revising his estimates as he started and continued his 'voyage into the unknown',*86 and corrected his earlier judgements*87 by his fuller realization of the circumstances, which at the first budgeting point of time were circumstances of the relatively remote future. His first estimate would not be likely to be correct. If he had to operate with his own resources, he would not be likely to embark on the venture at all unless he thought that however wrong his estimates were likely to be, he would still realize a profit—or at least as good a living, besides the return of his resources, as he could get elsewhere. It seems reasonable to suppose therefore that, under this new arrangement which requires both budget and account to be submitted, it would have to be understood at the outset that much scope would be allowed to the administrator for subsequent variation of his original figures, revised budgets being constantly submitted to the Ruler.*88 If the administrator were not allowed much scope, it seems likely that he would introduce undisclosed safety margins*89 into his estimates amounting in effect to very much the same thing as budgeting for a surplus—in much the same way as, I believe, Ministers of Finance do, only in our case the administrator would take care to avoid the surplus if it tended to emerge. In other words, the administrator might be very unwilling to embark upon an undertaking unless behind the disclosed budget showing no profit he had a secret budget with a profit which he intended to lose if his anticipation proved correct. The item 'money value of residual assets' might often, I suppose, be a very useful medium for introducing his safety margins. It might also be used for temporarily hiding an impending deficit.



On the application of the rule when it requires the equation of marginal cost with something else. If we are contemplating the investment of liquid resources in a railway undertaking, the alternative opportunity to be displaced being a chain of saloon bars, what*90 is the marginal cost of producing, say, the final ton-mile of railway service which is expected to yield, say, a penny-halfpenny as marginal revenue, following the investment of an additional penny of liquid resources?*91 The marginal cost must be regarded as the displaced marginal revenue,*92 say a pennyfarthing, which might be expected to accrue from the investment of the same penny in the production of, say, an extra pony of beer, instead of the final ton-mile of railway service.


If the rule is intended to be applied to marginal cost in this sense, the marginal cost will escape the Ruler's observation in the same way as total cost in this sense escaped the Ruler's observation: the prospective alternative revenue is displaced by the decision, and will never be realized. From this point of view the same remarks apply to marginal cost as were applied to total cost.


If the rule is intended to be applied to 'cost' in the sense of anticipated objective outlays, the marginal outlay in the example will be a penny, against marginal revenue a penny-halfpenny. To disclose this to the Ruler, it would be necessary to submit, or at least to prepare, two budgets or 'estimated profit calculations', the one relating to a higher, the other to a lower, level of contemplated output. The penny (marginal outlay) would be the difference between total outlays in the two budgets; the penny-halfpenny the difference between total revenues in the two budgets. In other words, the 'estimated profit calculation' relating to the margin would be 'marginal outlay, a penny; marginal revenue, a penny-halfpenny'. So apparently marginal 'cost' in this sense might be budgeted.


If the rule is intended to be applied to 'cost' in the sense of realized objective outlays (in the account) it cannot be so applied, because only one level of output is, or, in the particular situation of time and place, can be, produced: no variation from that level occurs to show the required difference between the realized objective outlays of one level of output and another.*93


It will be remembered that, when the application of the rule shifted to the 'estimated profit calculation', where this referred to total outlays and revenue, a certain impotence resulted from the lack of the account. Now that the calculation refers to marginal outlays, potence certainly cannot be restored by requiring that the account should correspond with the budget, because the account cannot exist.


So the objection to the proposed application of the rule to the total apply to its proposed application to the margin a fortiori.



1 It has been suggested*94 that the principle of covering total costs cannot be enforced 'every year if there are general fluctuations in trade', and that it 'should rather be applied over a period with surpluses in good years and deficits in bad'. Ignoring for the moment the doubt about the possibility of satisfactorily identifying and measuring the 'total costs' and 'total revenues' that are to be equalized over a period during which they fluctuate in relation to each other, would it be stretching analogy too far to suggest that the concession to allow 'surpluses in good years' to compensate for 'deficits in bad' might be extended to allow surpluses in good transactions and in good markets to compensate for deficits on bad transactions and in bad markets? It is difficult to see why, if allowance should be made for these 'general fluctuations in trade' allowance should not be made for particular fluctuations affecting the particular undertaking too.*95 Silence on the matter gives ground to suspect an implicit assumption that (apart from these 'general fluctuations') the administrator operates in conditions akin to stationary equilibrium (supporting itself by its own bootstraps!), or that he has little or nothing of difficulty to do in the way of trying to discover how market conditions are going to change, of being prepared for and adjusting to the changes in advance, and of making allowance—particularly in the choice of technical structure—for the inevitability of uncertainty and error.


But if concessions for particular 'fluctuations' are to be made, how does the rule apply? When ends the long run, and where ends the group of transactions (or markets) to which the rule is to apply? Appropriate extension of the idea of permitting 'surpluses in good years' to compensate for 'deficits in bad', to allow for the inevitability of uncertainty and error, seems to render quite indefinite the limits of time and range of operations within which equality of 'surpluses' and 'deficits' was to be achieved.


2 The quandary that one is led into by the inadequate definition of the terms and method of application of the rule can be shown by examining closely Mr Wilson's suggestion,*96 following the one for an allowance for 'general fluctuations in trade', for dealing with a 'permanent decline in the demand' for a socialized industry's product. In order not to be misled by the telescoping of the ex ante with the ex post, I shall translate some of his statements into terms which I have used before. The industry is 'faced' with this permanent decline in demand. I take this to mean that the administrator is rebudgeting and working on the assumption that demand for the product will for the remaining future be permanently lower than he expected it would be when he originally budgeted. 'In these circumstances', Mr Wilson tells us, 'private firms will go on producing so long as the excess over prime costs is greater than the interest on the scrap value of the equipment.' This I adjust to mean that the administrator would, if he were a private firm, decide to continue production providing that he valued his anticipated revenues higher than cost in the proper sense, i.e. his reckoning of the best yield, from all the resources which he contemplated using (including, of course, the 'equipment') if he transferred them to his best contemplated alternative opportunities of use—which might be to the production, by himself, of a different product. The cost of using existing equipment cannot be deemed to be adequately described by the expression 'interest on its scrap value'. Mr Wilson then remarks that the socialized industry might at this time have 'a large fixed-interest obligation', and refers to this as (being included in?) 'total costs' (thus telescoping past outlays—i.e. money 'capital' invested—and money payments already contracted to be made—i.e. 'interest' on the 'capital'—with outlays neither made nor contracted to be made). Having done this, he suggests that 'it would be better [than to adopt Mr Lerner's principles] to write down the capital of the industry...and then tell the manager to cover total costs at the reduced level'. But what are these 'total costs at the reduced level'? I can only interpret 'total costs' as being intended to mean cost in the proper sense plus either nil (if the 'capital of the industry' were written down to the level of the excess of cost in the proper sense over anticipated objective outlays)*97 or a positive amount (if the 'capital of the industry' were written down to a smaller extent). But what could this positive amount be other than a surplus of anticipated revenue over cost in the proper sense? And how much surplus would be allowed?*98 That from the output promising the maximum surplus? Or that from some other output, and if so, which? And according to what principle would it be chosen? If it is the output with the maximum surplus that should be chosen, so long as even this choice still involves writing down the 'capital of the industry', why obscure the issue by referring to the choice as telling 'the manager to cover total costs at the reduced level', without defining the principle for determining what the 'reduced level' is?


Mr Wilson follows up with*99 the statement than 'in general, undertakings which can cover total costs should be made to do so', thus 1) obviously telescoping the ex post-the 'fixed interest obligation' or 'capital of the industry'-with the ex ante, and 2) implicitly asserting that at this rebudgeting point of time the principle to be observed, at least to the extent that this is necessary to meet the 'fixed interest obligation', is that of maximizing the surplus of revenue over cost in the proper sense.


But he concedes the possibility of an exception to this rule being made 'in the case of some old industries where capital is a very large proportion of total costs and where there is little likelihood of new investment', so as to allow the expansion of output beyond the point at which surplus is maximized to the point, presumably, at which marginal cost is supposed to be equal to 'price' ('...output could be expanded till it approximated more closely to the theoretical optimum'). The subsidy of 'fixed amount' which for this purpose, he states, 'might sometimes be justifiable' would, I presume, though this is not explicitly stated, be the amount by which the expansion of output caused a deficit in the amount required to meet the 'fixed interest obligation' (i.e. in anticipated revenue minus anticipated objective outlays). The subsidy having been given, 'the management should then be told to cover total costs less the subsidy'. But 'in no circumstances...should the State adopt Mr Lerner's policy, and offer to make good whatever deficit emerged as long as marginal costs and price were equated'. Upon this I must make two comments. First, if it is Mr Wilson's point that his desired output differs from Mr Lerner's, I am at a loss to know in what respect. Secondly, if his point is that Mr Lerner's policy would be conducive to carelessness about covering 'total costs', whereas Mr Wilson's would not, I would point out that, unless the Ruler, and not 'the management', were the administrator, the Ruler would, under Mr Wilson's policy, be dependent upon 'the management's' estimates for fixing the amount of the subsidy. Possibly Mr Wilson has this idea at the back of his mind when, in his concluding paragraph, he remarks that 'this policy implies a degree of understanding and discrimination which may well be absent in practice'. It would be interesting to know whether, if he developed his views, explaining precisely what was to be understood by the terms 'costs' and 'revenue', untelescoping time so as to distinguish the ex ante from the ex post and elaborating the administrative arrangements and their functioning, he would discover that in this second respect he was, after all, in the same boat with Mr Lerner.


By pursuing this untelescoping and substitution of terminology I have endeavoured to show that the proposal to write down a 'fixed interest obligation', and instruct a 'manager' to cover 'total costs' as so reduced, avoids the issue as to the extent to which the 'fixed interest obligation' is to be written down or, what amounts to the same thing, it avoids the issue whether 'the manager' (or whoever is the administrator in the circumstances) is or is not to have a free hand to try to maximize his net returns ('fixed interest obligation' excluded). The proposal fails at this stage to define the principle of operation, although a subsequent reference to it seems to imply the principle of maximizing net returns.


Two other difficulties emerge when it is suggested that this proposal might in some cases be modified to allow subsidized expansion of output. The first is some obscurity on the matter of whether the proposed output is or is not that which would equate marginal cost and price. The second, and critical, difficulty lies in failure, at this point, to notice (or at least to make clear) that the fixing of the subsidy depends upon administrative estimates, and that if administration resides in the management, the Ruler is in the management's hands—a situation which, presumably, it is sought to avoid.



In considering the application of the rule when the cost-revenue relationship referred to the account, I suggested that, without usurping the administrator's function, a professional accountant could not independently assess the (net) revenue still to be yielded by the use of residual assets. But cannot he avoid having to do this by independently apportioning to an interim accounting period part of any outlay upon an asset which has an unexpired (residual) use at the end of the interim period? The answer is that all such apportionments must be unsatisfactory. To understand why this is so it is necessary to revert to the decision to acquire the asset.


Suppose the administrator is faced with the problem of deciding whether to buy an asset for £5,000. Let us assume for simplicity that 1) the asset would be completely specific to the production of a particular product of the particular undertaking, and that if it were to be used at an unvarying rate (considered by the administrator to be the optimum rate) over a period of four years it would still at the end of that period have a residual physical 'life' (reckoned on the basis of the same unvarying rate of use) of one year; 2) the administrator expects that the demand for the product will cease suddenly at the end of the fourth year, but by that time the use of the asset and the sale of the product will, in terms of net revenue, yielded at an unvarying rate, have recovered the £5,000, and he thinks the best thing to do*101 is to acquire the asset; 3)the administrator acquires the asset, produces at the expected rate over the first year—during which his expectations about demand have so far proved correct and, for the future, have not changed.


Now let us suppose that the first interim accounting period ends with the first year, and that the professional accountant attempts to apportion part of the outlay of £5,000 to that interim accounting period. Observing, presumably with the help of an engineer—that the asset has during the year been used up at a rate which if maintained will consume its whole physical 'life' by the end of the fifth year, he 'charges' against revenues one fifth of £5,000=£1,000 as 'depreciation' of the asset.*102 Clearly, because one fourth of £5,000=£1,250 has been recovered as revenue (net, before charging 'depreciation') the interim account will show a 'profit', and get the administrator called over the coals.*103


Possibly the accountant would reply that the remaining £250 would be 'charged' as 'obsolescence'. But the point is that to ascertain what the figure for obsolescence is amounts to exactly the same thing as obtaining from the administrator his calculation of the money value of the residual assets.

Notes for this chapter

Read, in part, to the Cape Town branch of the Economic Society of South Africa on 11 October 1946. The final stimulus prompting me to write the paper was Mr. T. Wilson's 'Price and Outlay Policy of State Enterprise' Economic Journal, 55, No. 220 (December 1945). Originally it was a running commentary upon that note, elaborating criticism which Mr Wilson himself put forward, but eventually joining issue with him upon his own proposals.
Accordingly, the paper has relevance to the prevalent idea that the efficiency of industries can somehow be judged by measuring their costs.
Because his function is to measure.
The fragments of information about the Cape Town and Wellington Railway are derived from minutes of evidence taken before Select Committees and other official documents. (See Votes and Proceedings of Parliament of the Cape of Good Hope from 1854 onwards, particularly two Reports of 1859, Appendix 2 [C.1 and A.6).]
And the liability to be dismissed—by whom?
Cf. Mr Wilson's suggestion that Mr Lerner has no time to spare, in his Economics of Control, for the 'other problem, which is also partly economic, of ensuring technical efficiency in the use of resources'. (T. Wilson, 'Price and Outlay Policy of State Enterprise'.)
This statement is prompted by my own experience.
It may be argued that the same, or a similar, lack of coordination can occur in a single mind. But I prefer to treat the matter as it appears in the form of the elusive 'diseconomy' of divided administration which 'budgetary' control may to some extent offset. Cf. G.F. Thirlby, 'The Subjective Theory of Value and Accounting "Cost"', reprinted here, pages 135-61.
I suppose every spouse can recall some occasion on which he (or she) would have done something different had the other spouse told him (or her) something which she (or he) ought to have told him (or her). I doubt whether Wicksteed adequately treated this aspect of human affairs.
It is difficult to believe that the railways could not have been more administratively effective in forecasting and appropriately responding to the effects of road motor competition.
Fritz Machlup, 'Competition, Pliopoly and Profit', part II, Economica, N. S., 9, No. 34, (May 1942).
See Appendix I.
See Appendix I.
See Appendix I.
The 'long run' is precisely as long as a piece of string which is longer than a shorter piece.
Mr Wilson queries this behaviour. In referring to the propensity to consider 'the operation of increasing returns industries... from a short-period point of view when a certain amount of fixed equipment is in existence...' he says that, 'in discussing this situation economists are able to make full use of their intriguing discovery that "bygones are bygones"...' 'Price and Outlay Policy of State Enterprise', (third section).
For circumstances under which it will not have such a cost, see Hayek, The Pure Theory of Capital, (London 1941), and G. F. Thirlby, 'Permanent Resources', Economica, N. S., 10, No. 39 (August 1943).
P. T. Bauer, 'Notes on Cost', Economica, N. S., 12 No. 46, (May 1945), p. 96. All uses, including the use for the unit of output under consideration, that are contemplated at any planning stage must, to a greater or leaser extent, be future uses. Some of them—uses for alternative purposes—may be contemporaneous with the use for the unit of output under consideration. Later uses also may be for alternative purposes. With these adjustments the expression quoted seems to be in order.
It appeared to be necessary to make this matter quite clear because Mr Wilson ('Price and Outlay Policy of State Enterprise'), does not seem to have done so. Distinguishing it from the 'short-period' reckoning in which he more or less concedes that 'fixed' equipment cost is an irrelevant 'bygone', he conceives of the longer-period reckoning, that is to say, a reckoning in which 'fixed' equipment cost appears, as occurring when a decision is taken to install equipment.
It is quite irrelevant to the present issue to point out that many businessmen do in fact judge that they will be the same tomorrow as they have been today. The relevant point is that somebody has to be presumed to be responsible for forecasting or for not bothering to forecast.
There is implicit recognition of this fact in the discussion following Mr Wilson's suggestion that 'probably the only way to get at any sort of approximate answer [to the question whether a large "widening" of the capital structure should occur] would be to consider whether it would be possible to cover the total costs of the undertaking if it were run monopolistically'. ('Price and Outlay Policy of State Enterprise', p. 458). The force of his subsequent remarks upon the relative incompetence of 'socialist managers' of an 'undertaking run according to the rule' to pronounce upon 'what consumers would give for the product', and to determine investment policy, really depends upon the fact that the monopolist of whom he is thinking is, unlike the 'socialist managers', a person operating in the market with a constantly developing and revised understanding of the behaviour of the consuming public in the sphere in which he is operating. (I do not wish this remark to convey the impression that I think that 'socialist managers' could not be such persons—much less that I think that it would not be their function to employ this competence. The indispensability of business administration does not of course depend upon whether the industrial organization is 'socialistic' or 'capitalistic'. That this truth was discovered in Russia very soon after the 1917 revolution is apparent from the literature (e.g. Hubbard's Sovist Labour and Industry.)
See Appendix I.
'The Subjective Theory of Value and Accounting "Cost"', reprinted here, pages 135-61, passim. If the fact that this model relates to a mercantile rather than to an industrial undertaking is disliked, the term 'buyer' may be changed to 'product-line superintendent', and the term 'merchandise manager' to 'general superintendent'. The administrative structure is then similar in the two cases. But in the industrial case it would be hard, with a clear conscience, to abstract from discussion of the disposal of the use of plant—e.g. the use of a common foundry disposed of by the general superintendent between the two product-line superintendents.
See Appendix I.
See my 'The Subjective Theory of Value and Accounting "Cost" '.
In Appendix I, however, there is some deliberation on the margin.
I should perhaps say here, what is implicit in the earlier part of my paper, that if the cost-revenue relationship that is to be scrutinized is the cost-revenue relationship of economics (or ought I to say of the Subjective Theory of Value?), it is the ex ante deliberations which have to be investigated, for only in them is there any indication of the displaced opportunities, the effect of the marginal variations from the proposed level of operations (see Appendix I) and the safety margins (allowances for uncertainty) which have to be understood if the decision is to be understood.
Strictly, to the significance of the revenue. Revenue from saloon bars may not be as significant as an equal amount from railways.
The difficulty or impossibility of extracting this evidence is understressed, perhaps by speaking of 'the administrator' instead of a number of people contributing to the conjuncture, certainly by speaking of a first-budgeting process and decision instead of to the conjuncture changing over time as anticipations change and rebudgeting occurs. What seems to be required for a perfect scrutiny is a perpetual record of the budgetary thoughts of every person with discretionary responsibility in the undertaking—and capacity in the Ruler to deal with it with less fallibility than that of the administrator! But to what extent and how often administrators would have their thoughts scrutinized, it is for the advocates of the rule, and not for me, to say.
Or, what amounts to the same thing, achieving the additional revenue (equal to the displaced alternative revenue thought likely to accrue from investing the extra resources—required to make the works 'heavier'—elsewhere, in or outside the railway undertaking.)
To say that 'the state' is responsible is to avoid the whole issue: 'the state' has to be given content by an exegesis of the devolution of administrative responsibility upon persons.
In his deliberations on the Administration Chart in Some Modern Business Problems (ed. Plant (London 1937)) Mr Paul Wilson suggests that an administrative officer should be subordinated to that higher administrative officer who will stand to suffer most by his likely errors. It is ideas of this kind that should be taken into account in considering the relationship between the Ruler and the administrator and the incidence of sanctions for error.
It should be clear that this budget would not include any outlay which would necessarily measure the value of resources in their best alternative use. It might include, e.g. anticipated objective 'interest' payments on money to be borrowed for a railway undertaking, but these 'interest' payments could not be presumed to measure the yield that would have been expected from borrowing the same money and investing it in a chain of saloon bars.
Machlup, 'Competition, Pliopoly and Profit', part II.
Cf. Mr T. Wilson's suggestion 'Price and Outlay Policy of State Enterprise', that the principle of covering total costs could not be enforced every year, which seems to imply that some other people think it could be. (This suggestion is specially treated in Appendix II.)
This matter of calculating the money value of residual assets is much the same as that of enterprise valuation, which has received much attention by Bonheight and others. A subjectivist, after working round to the attitude that to get at the value of the enterprise it is necessary to discover what the future revenues (less outlays) are expected to be, must, I think, conclude that 1) to get at these in turn it is necessary to ask the particular administrator of the enterprise what he thinks they will be; 2) that what they are likely to be may depend very much upon the enterprise being run by the particular administrator; 3) that the appropriate 'rate' at which the future figures should be 'discounted' depends, too, upon his views.
On 'charging depreciation', see Appendix III.
This addendum to the rule may possibly be implied in the idea that if 'the manager' disobeys the order to observe the rule he should be 'dismissed': I do not know. But it certainly is not implied in the mere injunction that cost and revenue should be equal.
I believe I am indebted to Mr Shackle for this phrase.
Subject to the limitations (e.g. by the acquisition of specificities) that he has already imposed upon himself. His later knowledge might suggest that he ought to have adopted a different technical structure (e.g. a different route for a railway), but it does not follow that he ought then to change it.
If perpetual submission is conceived of, so that all alterations in plans are approved before being executed, the eventual account must, I think, be identical to the approved revised budgets—if we abstract from the results of 'acts of God', breaches of standing orders, and any variations allowed to sub-administrators, executives and operatives by standing orders (e.g. a standing order to buy a material at a certain rate per unit of time so long as the price did not vary outside certain defined limits). Cf. my 'The Subjective Theory of Value and Accounting "Cost"', reprinted here, pages, 135-61.
On the introduction of safety margins into estimated profit calculations, see Machlup, 'Competition, Pliopoly and Profit', part II.
Abstracting from the time at which the marginal unit will emerge.
Since decimalization 5p = 12d.
Strictly, the relative significance of the displaced marginal revenue.
It might, of course, be retorted that at this stage it would be a fairly simple matter for the ruler to calculate what extra 'costs' and what extra revenue would have been involved in the production of an additional 'unit'. But this retort is unsound. Calculations of this kind could not be accepted as satisfactory without allowing for imponderable subjective or administrative elements. An alternative retort might be that the ruler would get his results by experimenting: changing the output and watching the results. But obviously this method is not generally satisfactory either, not only because the administrative imponderables might be overridden, but also because the different outputs would be produced at different times. It cannot be assumed that the circumstances affecting 'costs' and revenue are the same at different times. The overriding of the administrative imponderables is easily illustrated. Illustration. In response to a request from the ruler, who wishes to observe the results of increasing the output of a particular product of a particular firm, the general manager of the firm issues a peremptory order to increase output by twenty-five per cent. (He would not be likely to behave in this way in the ordinary course of business: it is for the purpose of demonstration that I assume that he does so.) The Ruler, subsequently passing through the firm's premises, encounters the sales manager and the production manager. A discussion ensues in which the sales manager informs the ruler that he could have told him in advance what the approximate effect of the increased output on revenue would be: it was his job to be able to do that. The production manager tells him that the stepping-up of output involved delaying the execution of a remunerative order, which required the use of the same machines. The sales manager tells him that this order has now been cancelled. (Surely the loss of this job is a cost element—but it will not appear in objective outlays).
T. Wilson, 'Price and Outlay Policy of State Enterprise', section 2 (d).
Incidentally it would often be difficult to distinguish with any exactness the effect of the general fluctuation from the effect of a particular fluctuation.
'Price and Outlay Policy of State Enterprise', in the remainder of his section 2 (d).
It should be clear that these anticipated objective outlays are different from cost in the proper sense, which would include, for example, the value of existing equipment in an alternative use. That Mr Wilson has in mind (as 'costs', excluding the 'fixed interest obligation') something over and above anticipated objective outlays is apparent from his analogy with private enterprise in which he includes 'interest on the scrap value of the equipment' in what a private firm will require to be covered by revenue. The 'capital of the industry' would not be entirely written off unless it were anticipated objective outlays which were to be (only just) covered by revenue.
Let revenue (r) minus cost in the proper sense (c) be x, and let c minus anticipated objective outlays (a) on the same budget (excluding, of course, those outlays which are already contractual obligations) be y, so that anticipated objective surplus on the budget is (r-c)+(c-a)=r-a=y+x.

Then Mr Wilson's 'total costs' appears to mean either c, if the 'capital of the industry' is written down to y, or c+x, if the 'capital of the industry' is written down to y+x.

But x (and, of course y, a, c and r) will be different according to what output is planned.

'Price and Outlay Policy of State Enterprise', in section 2 (e).
Cf. a discussion of 'depreciation' in section III of my 'Permanent Resources'.
In order, we may assume, not to maximize net revenue, but to make outlays and revenue equal.
I do not of course wish to suggest that this method exhausts accounting methods of 'charging depreciation'. I wish only to suggest that all 'objective' methods—all methods which do not involve appeal to the administrator's calculation of the money value of the residue of the asset—are unsatisfactory.
I might well have added that if such accounting apportionments of outlays (or charges for 'depreciation') were conventionally adopted, the administrator knowing in advance that they would be, and knowing that he would be required to have made his 'costs' (including such apportionments) and revenue equal in the interim account, would be likely to try to adjust his ex ante plans accordingly. (In the above example he probably would not have bought the asset at all, because, in order to come out in the end, he would, on the assumptions made, have been faced with having to show a 'profit' in the first four years and a 'loss' in the fifth.) Production would be distorted by the adoption of an accounting convention: technical specialism would be cutting across, or imposing a rigidity upon, administration. (If it were urged that the imposition of this convention amounted to the laying down of a standing order by a higher authority, and that 'the accountant' was therefore taking part in administration. I should probably accept the correction. But it would still be open to economists to consider the effect of such a standing order upon production.)

Essay 8, The economist's description of business behaviour

End of Notes

11 of 15

Return to top