A paper read to the Cape Town Branch of the Economic Society of South Africa on 13 April 1945. First published in
Economica (February 1946).


The subjective theory of value and accounting ‘cost’


In a recent article
*21 in the
Economic Journal, Mr Harry Norris (an accountant) puts out certain ‘feelers’ into the overlap of the provinces of economists and accountants, hoping thereby to ‘stimulate economists into thinking about accounting procedures in the light of economic science’. Mr Norris tells us that a comparison of ideas is something of which accountants, and perhaps economists, stand in need, though economists may find difficulty in discovering exactly what accountants mean by certain terms which they use. Speaking of the subject in which he is particularly interested—income—he acknowledges the substantial truth of Professor Canning’s view that accountants have not developed, and probably have never put their minds to the task of developing, any complete philosophical system of thought about it. It is only fair to add that Mr Norris is ‘not able to find any great clarity of thought among economists as to what constitutes income’.

Some time ago, having suggested that ‘economic science has not yet become integrated into the philosophy of accounting teachers and writers’, I ventured to recommend that the results of such a study as Mr Norris desires should form part of the curriculum of university students of commerce, saying that ‘the ubiquity of accounting and the need for its reconciliation with economics rather suggests that part of a second course in accounting in the commerce curriculum should be called “Accounting in the Light of Economic Analysis”‘.
*22 It is natural therefore that
I should welcome, and even try to respond to, Mr Norris’s invitation.

My subject, however, is not income, but a term which Mr Norris uses incidentally, namely, cost. And I must confess that the main stimulus prompting me to discuss it was, not Mr Norris’s article, but one written by Professor C. S. Richards.
*23 In his article, in which he emphatically recommends the practice of cost accounting, Professor Richards himself deplores the use of ‘vague phrases and undefined terms… which lack clarity and the implications of which are seldom analysed nor their consequences appreciated’. But, although perhaps no term is used more loosely nowadays than ‘cost’, and although Professor Richards in his frequent uses of the term is traversing ground covered by both economists and accountants, he offers no discussion of the different meanings attached to the term in economics and accounting. The difference is fundamental.

The task that I have set myself is, not to deal exhaustively with all details of cost-accounting practice, but to suggest 1) the meaning of cost to a person—whom I shall refer to as the subjectivist—whose thought is conditioned or disciplined by the subjective theory of value; 2) the place and significance of cost in this sense in a philosophy of business administration; 3) the different meaning that the term has to an accountant; 4) the relationship of cost in this accounting sense to the subjectivist’s philosophy of business administration.


To the subjectivist cost would be understood to refer to the prospective opportunity displaced by the administrative decision to take one course of action rather than another.
*24 Cost is
inevitably related to the behaviour of a person. The person is faced with the possibility of taking one or other of (at least) two courses of action, but not both. He considers the relative significance to him of the two courses of action, and finds that one course is of higher significance than the other. He ‘prefers’ one course to the other. His prospective opportunity of taking the less-preferred course becomes the prospective cost of his taking the more-preferred course. By deciding to take the preferred course, he incurs the cost—he displaces the alternative opportunity. The cost is not the
which will flow along certain channels as a result of the decision; it is the loss, prospective or realized, to the person making the decision, of the opportunity of using those things in the alternative course of action.
A fortiori, this cost
cannot be discovered by another person who eventually watches and records the flow of those things along those channels. Cost is not something which is objectively discoverable in this manner; it is something which existed in the mind of the decision-maker before the flow began, and something which may quite likely have been but vaguely apprehended.

The alternatives between which the final selection is made are themselves a result of personal discovery and selection. The available alternatives cannot be said to exist unless the person making the decision is aware of them.
*25 If they could, their number would be infinite and their consideration by the decision-maker intractable. ‘Any number of potential applications “compete” for the use of the productive services.’
*26 But the human being is not omniscient. It is obvious that the very limitation upon human capacity necessitates the selection for consideration of only a few of the alternatives, and that the selection might easily be a
different one, either if the particular administrator (decision-maker) happened to notice different alternatives, and make a different selection, or if a different administrator made the selection.

The act of discovering cost, which really means discovering which of the considered alternatives is to be rejected, inevitably involves valuation. The decision-maker, in arranging the opportunities in order of preference or significance, is performing what is essentially an act of valuation, valuing the preferred opportunity higher than the alternative to be rejected.

This valuation necessarily involves estimates of happenings in the future about which the decision-maker can never be certain. The decision is based upon
ex ante reckonings, or advance calculations, which involve looking into the future, and consequently must, even for this reason, be matters of opinion. Yet ‘such advance calculations are made every day by scores of businessmen, either for themselves when they are making up their minds about the prospects of a contemplated business venture, or for potential partners or lenders when such are invited to consider participation in the enterprise’.
*27 This statement was intended to refer to plans for new industrial undertakings, but its reference can be legitimately extended to cover plans preceding all business decisions.

Cost is ephemeral. The cost involved in a particular decision loses its significance with the making of a decision because the decision displaces the alternative course of action.
If the accepted course of action
were completely planned at the time of the decision and
if the course of action
were taken and
actually carried out in accordance with the plan, no new decision—choice between alternatives—occurring in the interim, then no cost—no cost of ‘production’—could be said to occur in the interim, however many times money was converted into goods by purchase or hire, and however many times goods were converted into other goods and sold. In the meantime production would have been
proceeding according to plan—the plan accepted by the decision and put into
operation as a result of it—or, in other words, it would have been proceeding under
standing orders.

But usually new decisions will be made before the first one is completely implemented. And cost occurs every time a business decision is made, however large or small the matter under consideration, whether the decision is upon such a matter as to delay the execution of a small order for goods so that a previously unexpected rush order may be accepted, or whether it is to set up and carry on a large industrial enterprise.

The decision is the
primum mobile of production, without which nothing that occurs can be regarded as production. It is the logical starting point for any investigation which seeks an explanation of why production or the industrial structure is what it is.


The subjectivist sees ‘the whole direction of resources to ends as a continuous selection between alternatives, guided throughout by a weighing of the significance of the anticipated results, in which the “cost” of adopting any alternative is simply the relinquishing of some other alternative; reward and sacrifice alike being measured and determined by the ultimate significance of the respective products, as anticipated by the producers’.

By discussing an aspect of the functioning of an imaginary firm,
*29 I shall try to explain how this must be presumed to apply to the internal workings of a modern departmentalized firm, with divided administration. But first I must refer to the coordination process in a firm in which administration is
not divided—i.e. in a one-man business.

We can imagine a man in a small retail business deliberating upon the question of how much money to retain (or acquire) for
the purpose of investment in stock which is to be bought and sold over a (relatively short) forthcoming period. We will suppose that he is already in the middle of his deliberations. He has already considered a certain sum to be worth while investing in this manner. He has decided that he could do better with it there than elsewhere: the cost would be worth while incurring. If now the man is thinking of the advantage of using £50 more than that sum, he will be comparing 1) the significance of the alternative opportunity of using (or not acquiring) that increment of money, with 2) the significance of the result of investing it in stock and realizing the stock. And obviously he cannot consider 2) without considering 3) what extra stock he would buy at what extra price, and how much extra its sale would be likely to realize. Further, if he allows himself to consider different kinds of stock, he cannot consider 3) unless he considers 4) to which kind of stock to allot the £50, or in what proportion to allot it to different kinds. In other words,
there must be ex ante
coordination of 1)
the significance of the alternative opportunity of using (or not acquiring) the increment of money with 2)
the significance of the eventual returns from the investment of the increment in stock; and this coordination incidentally involves other acts of coordination, namely 3)
the coordination of the prospects of buying goods with the prospects of selling them, and 4)
the coordination of the relative significance of the prospective returns from investment in alternative kinds of goods.

We may now suppose that the man considers the retention (or acquisition) and investment of this extra £50 to be advantageous, that he then considers, in the same way, the advantage of using a further £50, and so on until eventually he thinks that the investment of another £50 would not be worth while—and that consequently he decides that the best sum to retain (or acquire) and invest is the total sum of money which does not include this last increment.

The description of the man’s deliberations up to this point is sufficient to illustrate the nature of the coordinated decision
ex ante which it is necessary to comprehend before a satisfactory approach can be made to the understanding of the conduct of
business under divided administration. But it will be convenient for my later discussion to assume that a contractual rent payment is made during ‘the forthcoming period’. So I am obliged to elaborate a little.

The deliberations cannot be said to be fully coordinated and completed at this point. The man’s calculations have led him to the conclusion that, in so far as he has calculated, it will be advantageous to him to continue business for ‘the forthcoming period’. But it is possible that this advantage would disappear if he considered the possibility of otherwise disposing of, not only the money, but also other factors (e.g. his premises or the lease of them) which he will use for the business in ‘the forthcoming period’ if he does decide to carry it on.

To avoid a long and complicated discussion concerning the extent to which these other factors can be varied in quantity, and the effect of varying them (e.g. the effect of letting off portions of the premises, or extending them, or allotting different portions to different portions of the business), I shall assume that the man does not at this time allow such variations to enter into his calculations. This seems to be a reasonable assumption to make, because, as the objective possibilities are infinite, a person must impose some rules (‘standing orders’) upon himself, intuitively or otherwise, to limit the number which he considers and the times at which he considers them. But I shall allow him to consider the
complete disposal of the business for a period or permanently.

I shall assume also that the man has a lease of the premises for a period longer than what I referred to as ‘the forthcoming period’. The coordination of the result of the calculations (or budgeting) already considered with the question of whether it would be advantageous to dispose of the business must therefore be considered to be a problem of coordinating the result of the calculations (or budgeting) already considered with a wider budgeting (‘wider’ here referring to a longer time period). It must be so regarded because the significance of the opportunities of disposing of the business for ‘the forthcoming [relatively short] period’ are not likely to be considered without taking into account what would happen in the more distant future. Why? For the
simple reason that to close down in the meantime would affect subsequent prospects—e.g. some contractual rent might be saved by subletting the premises for ‘the forthcoming period’, but some regular customers might not return after their enforced absence.

We may now suppose that the prospective advantage shown by the narrower calculation is either so great that the man is not prompted to consider this wider budgeting, or that, if it is small enough to prompt him to do so, his wider budgeting has led him to the conclusion that to close down temporarily or permanently would be to his disadvantage. Obviously he might come to this conclusion although his prospective net money returns for ‘the forthcoming period’ were lower than the contractual rent payment to be made for ‘the forthcoming period’, and although he might consider it possible to reduce the difference by subletting for ‘the forthcoming period’: his decision would depend partly upon what he thought of his more distant prospects. But whether the contractual rent payment is expected to be covered out of net money returns will not affect the issue to be discussed.

The language in which I have described this illustration of the coordination process of the small business man at the street-corner shop would perhaps be quite unintelligible to him; but the description is, I suggest, one that the subjectivist would give of a process that he supposes not only the small man at the street-corner shop, but also all firms, trading or manufacturing, to be continually performing.
*30 The subjectivist supposes the equilibration of which he speaks, and the functioning of industry and commerce to which this equilibration really refers, to depend upon the performance of the process, the decision based upon it, and the acceptance by the firm of the consequences of the decision.
The acceptance of the consequences includes the acceptance of the ‘automatic sanction’
*31 for error. The coordination process and the supervision of the execution of the decision may be loosely or negligently performed, or, on the other hand, they may be rigorously or carefully performed. They may be performed
according to any limiting rules that the firm chooses to impose upon itself. Different aspects of the process and execution may be delegated to different people. But, however this may be, the firm is supposed to accept the consequences of what it does. If, for example, a firm relapses into and works upon an unjustified assumption that the events of ‘yesterday’ will be repeated ‘today’, and tacitly issues standing orders based on such an assumption, it is presumed to do so on its own responsibility.

The process, decision and framing of orders for the execution of the decision constitute an act of business administration. In modern undertakings this act of business administration is often divided among a number of people (administrators). The lines of fracture of the act of business administration might be different in different cases. One man might be responsible for estimating the market for goods on the buying side and for the actual buying, another for estimating the market for goods on the selling side and for the actual selling. This could well be the arrangement in a firm in which purchases were made abroad and sales made locally. One or other of these men might be responsible for estimating the market for short-term funds and actually negotiating loans; or a third person might attend to this. The work on the buying side, or on the selling side, might be split, each of several men being responsible for the market for a particular type or range of goods. Or each of several men might be responsible for both the buying and the selling market for one of several types or ranges of goods. In one or other of these situations there might be a person who accepted responsibility for the estimates and actions of those amongst whom part of the work was so divided, and for the coordination of their estimates. This man, responsible to a higher authority, would be giving advice and criticism to those responsible to him, without usurping their initiative and discretion as administrators. Clearly he would be a man of broad knowledge of men and probably of the markets in which his subadministrators were operating: the judgement of people’s behaviour in advance is of the essence of administration—a matter which tends rather to be obscured when one speaks of judging
of what people are going to do as ‘estimating the future market conditions’.

Whatever the lines of fracture are,
the complementary activities of the people (administrators) amongst whom the act of business administration is divided must be presumed to be coordinated, for the purpose of making the ex ante
decision, by one, or by a committee,
*32of all or several, of the administrators. At the same time the very division of function gives rise to the danger of loose coordination, with the firm’s left hand not knowing what its right hand is doing.

The arrangements, or rules, laid down by itself, which the firm adopts to determine this division of function and coordination, together with other regulations, might be called the ‘standing orders’ of the firm; and a tree describing the division of the administrative function and coordination ‘the administration chart’ of the firm. The way in which administrative authority—authority to make decisions which
ex definitione involve cost—is divided and distributed through the organization, and how it is circumscribed, this arrangement is itself a matter for administrative decision. It is a matter of choice between this structure and some other. It involves subjective judgement; consequently no ‘right’ way can be objectively determined.

The organization that I have chosen to illustrate the coordinated decision
ex ante is one in which there are several (two) people, each responsible for the buying and selling market for a particular range of goods. Another man accepts responsibility for their activities. A third is responsible for estimating the market for short-term funds and actually obtaining them. This arrangement enables me to confine my discussion, in the main, to the aspect of the process of coordination relating to the linking of the market for short-term funds with the market for the goods into which the money is to be converted, without discussing in detail the coordination of the buying and selling markets for goods.

Let us suppose that, instead of being the small man at the street-corner shop, our firm is a mercantile firm, e.g. a department store, working under divided administration. Each of two department managers (buyers) A and B has discretion as to what varieties of goods he acquires for sale, and is responsible for making and coordinating the forecasts of the buying and selling markets for those goods. A higher authority, whom we will call the merchandise manager, is responsible for settling the
proportions in which money is alloted to the buyers for investment in stock. A still higher authority, whom we will call the highest authority, settles the total amount to be allotted to the merchandise manager for this purpose. All are planning their operations for ‘the forthcoming period’. The highest authority will carry out the process of determining the optimum amount of money to invest in the stock in the same way as the man at the street-corner shop did, except that its study of variations in anticipated results inside the business will not go further than considering the significance of variations of revenue which the merchandise manager offers to try to get from the buyers if one quantity of money or another is allotted to him. When it is eventually made, the decision of this highest authority, which will be the coordinated decision
ex ante, will finally settle the total amount of money, and incidentally any contractual payments for its use, planned to be invested in stock in the forthcoming period, and may be considered to be reserved to the highest authority by standing orders which require the merchandise manager to submit his offers to it.

But before the merchandise manager can do this, he will need to obtain offers from the buyers. He will require from A estimates of the variations in revenue which A expects to make with variations in the quantity of money allotted to him, and he will require from B estimates of the variations in revenue which B expects to make with variations in the quantity of money allotted to him, so that he—the merchandise manager—can choose whether to allot to A, or whether to allot to B, each successive increment of money which may be allotted to him by the highest authority, and so decide what increment of revenue to offer the highest authority for each increment of money which may be allotted. The decision
required of the merchandise manager is how to distribute the allotment of money, whatever it may be, between the buyers. This decision may be considered as being reserved to him by standing orders which require the buyers to submit their offers to him.

But before A (or B) can do what the merchandise manager requires of him, he will need to consider how to distribute his allotment of money from the merchandise manager, whatever that may turn out to be, amongst the purchases of the different kinds of goods that he contemplates buying,
a2… (
b2…). Just as the merchandise manager is conceived to be dosing prospective increments of money between A and B, so that he may decide what increment of revenue to offer the highest authority for each increment of money that may be allotted, so A (or B) is conceived to be dosing prospective increments of money between
a2… (or
b2…), so that he may decide what increment of revenue to offer to the merchandise manager for each increment of money that may be allotted. The decision required of A (or B) is how to invest the allotment of money, whatever it may be, in the various kinds of goods. This decision may be considered to be reserved to him by standing orders. (Obviously the decision requires simultaneous coordination of buying and selling prospects.)

So A (B) coordinates the prospects of investment in different channels in his own field (department); the merchandise manager coordinates A’s and B’s investment prospects; the highest authority coordinates the merchandise manager’s investment prospects with the advantages of using money outside the business (or of not acquiring money). After choosing the optimum sum for investment in the business, it makes the coordinated decision
ex ante. The coordinated decision
ex ante settles the amount of money to be acquired by (or retained in) the business and allotted to the merchandise manager, the proportions of it to be allotted by the merchandise manager to A and B, and the proportions which A (or B) intends to allot to the purchase of
a2… (or
b2…). At the same time, it settles any contractual obligations by the firm for the use of the money, the amount of revenue that the highest authority expects to receive eventually from the merchandise
manager, the amount of revenue that the merchandise manager expects to receive from A and from B, the amount A expects from the sale of
a1 and from the sale of
a2…, and the amount B expects from the sale of
b1 and from the sale of
b2…. The amount of money to be used by the firm, plus any contractual obligations for the use of the money, on the one side, and the revenue expected from the merchandise manager on the other side, might be referred to as the budget of the highest authority and be thought of as a wider budget than that of the merchandise manager. The allotments of money to be made to A and B, and the revenue expected from them by the merchandise manager, might be referred to as the merchandise manager’s budget and be thought of as a wider budget than that of A and B. Its details are not a matter concerning the highest authority directly. The amounts of money to be spent by, and the revenues expected by, A (or B) might be referred to as A’s (or B’s) budget and be thought of as a narrower budget than that of the merchandise manager. Its details are not a matter concerning the merchandise manager directly. The term ‘estimated profit calculation’
*34 might be used throughout as an alternative expression for ‘budget’. The contents of the budgets are anticipated results (of the coordinated decision
ex ante) which are expected to become objective. They do not disclose costs in the subjectivist’s sense of the word.

What costs do occur in this process? Cost to the highest administrator is the opportunity of disposing, outside the business, of money in its possession and money that it might acquire. This is not a cost to the merchandise manager. That is to say, the question of whether to allot money to the merchandise manager instead of using it outside the business (or instead of not acquiring it from outside the business) is excluded from consideration by the merchandise manager by standing orders which reserve the question for consideration by the highest authority. Cost to the merchandise manager is the sacrifice that he incurs, in deciding to allot (any particular increment in) the quantity of money to A, by displacing the opportunity of allotting it to B instead (or vice versa). But this is not a cost to A (or B). That is to say, the question
of allotting money to B instead of to A (or vice versa) is excluded from consideration by A (or B) by standing orders which reserve the question for the consideration of the merchandise manager. Cost to A (or B) is the sacrifice that he incurs, in deciding to allot (any particular increment in) the quantity of money to goods
a1 (or
b1), by displacing the opportunity of allotting it to
a2… (or
b2…) instead (or vice versa). Cost occurs whenever, and only when, an administrator makes a decision, choosing between prospective alternative courses which appear to be open to him, between which he has discretion to choose. Under divided administration, the action open to a particular administrator is dependent upon the action to be taken simultaneously by other administrators. Consequently coordination of his plans with those of the other administrators must occur before his final decision can be made.

We may assume, without elaborate discussion, that the highest administrator has coordinated the result of this narrower budgeting process with the wider question of whether the result justified the use of the premises for the forthcoming period—in the same way as the man at the street-corner shop did—and that the highest administrator has decided that the business shall continue for the forthcoming period. A contractual rent payment will fall due during that period.


‘Cost’ to the accountant means something quite different. What he refers to as cost would, but for a trick, or imaginary conversion, that he performs, be an objective result which emerges 1) after all the decision-making which has involved cost has been done; 2) as a result of the decision-making; 3)-which, of course, follows from 2)-as part of the
ex post events which are described or implied in the
ex ante plan to which the anticipated profit calculation belongs.

The trick, or imaginary conversion, which he performs is this. He assumes (implicitly) that when money has been spent or contracted
to be spent to acquire things, the money has not necessarily and inevitably been spent or contracted to be spent as it has, leaving the business with the things acquired, but that the things acquired carry the money with them, and that bits of the things flowing into different departments or products of the business carry bits of money with them, or that bits of the life period or assumed life period of the things acquired carry bits of the money with them; and that the money in question has not been wholly spent so long as any of the things acquired and still possessed has one of the bits of money attached to it. The bits of money are ‘costs’. These ‘costs’ are carefully distinguished from values: ‘It is costs we deal in, not values’. (Mr Norris, below).

This description of the accountant’s behaviour and attitude seems to be confirmed in the following statement appearing in Mr Norris’s article.

Earnings to an accountant are simply money revenues from operations minus the cost of performing those operations. There is an outflow of money costs to be classified (the labels used in the classification tree are wages, power, materials, components, finished articles, and so on) and linked up with the inflow of money revenues. The product flows out to the customer; a legal claim for money flows into the business as revenue. It is costs we deal in, not values. Some costs we attach to bits of material, writing them off when the material is sold, others we attach to the calendar and write off according to lapse of time. There are complications in this; and there are, in my view, some illogicalities and errors in common accounting practice; but what we
aim to do is simply what I have stated, to find the surplus of revenue over expired costs. To do this one may have occasion to
refer to the incidence of values—of raw materials for instance—but figures of value are not used
as such; they are merely an aid in cost apportionment.

Mr Norris refers to the ‘cost of performing… operations’. To the subjectivist the cost of performing an operation is the administrator’s alternative opportunity displaced by the administrator’s decision to have the operation performed. The displaced opportunity
might be the performance of the operation in some other way, or the following of some entirely different course of action. Not so to the accountant. The accountant thinks first of an observable (objective) ‘outflow of money costs’; something which can be computed objectively by observing and recording. It is clear that this ‘outflow of money costs’ is primarily understood to be the money flowing out of the business in exchange for things to be used in the business.

Subsequently, however, the accountant shifts his attention from the money outflow to the
inflow, of the things received in exchange for the money outflow, into the business and thence into the operation and the product. It might be thought that, if the accountant rigorously pursued his objective study of flows, he would record these inflows of factors into the product in quantities of things, without attaching figures of the money paid for them. It is perhaps not quite clear from Mr Norris’s statement that the accountant does attach the money figures; but it is well known that such is his practice. That is what is meant or implied by saying that ‘some costs we attach to bits of material…others we attach to the calendar…’. The ‘costs’ are then ‘expired’ by writing off in the manner indicated. Clearly the accountant is here tacitly assuming that, or behaving as if, the money which is spent, or contracted to be spent, on the purchase of factors is not spent, or contracted to be spent, at the time when it is actually spent, or contracted to be spent, but remains attached to the factors, to be spent subsequently according to whatever arbitrary or ‘conventional’ method of ‘expiring’ the money (‘cost’) is adopted by the accountant.


In section II, the discussion of the coordinated decision
ex ante had proceeded to the point at which the decision had been made
and the budgets or estimated profit calculations of the various administrators had come into existence. The place of accounting ‘cost’ can now be discovered by discussing subsequent events.

It follows from the opening quotation of section II that, in the subjectivist’s philosophy, everything that can be regarded as part of the firm’s business operations (‘production’) must be the result of one administrative decision or another.
*37 Some of these results, occurring subsequently to the coordinated decision
ex ante which is under discussion, will or may be the results of earlier decisions which had not yet been fully implemented: such, for example, as the contractual rent payment accruing due during ‘the forthcoming period’. Any other results must,
unless and until a further administrative decision is made, be results of the particular coordinated decision
ex ante which is under discussion. If no new decision
did occur, and if the coordinated decision
ex ante were completely implemented—a supposition which implies that all anticipations proved to be sufficiently correct to allow complete implementation—objective results would occur which would correspond exactly with the plans of the several administrators. Eventually
accounts could be produced, correctly recording results, which would correspond exactly with the budgets or estimated profit calculations.

None of these budgets or accounts would include cost in the subjectivist’s sense. To what extent would they include ‘cost’ in the accountant’s sense?

The contents of the several budgets (of which the subsequent accounts are replicas) have already been described. If the money used were borrowed money, the account of the highest administrator would include any objective payment (‘interest’) for the use of money. This appears to be an ‘outflow of money costs’, that is to say, ‘cost’ in the accountant’s sense before he shifts his attention from the money outflow to the
inflow of things acquired by the expenditure of money. The item does not appear in the account—as I have envisaged it—of the merchandise manager or of A or B: it is no concern of these people. If the accountant chose to
‘attach’ the item, or shares of it, to the money resources which its payment brought into the business—i.e. to the money flowing to the merchandise manager and thence to the buyers—and subsequently to the goods into which the money was converted, ‘expiring’ it as sales of goods were made, his doing so would have no apparent significance to any of the administrators. The item, as a
prospective payment in the original planning stage, appeared only in the budget (estimated profit calculation) of the
highest authority. It was the objective payment (at the time
prospective) necessary to achieve the optimum (prospective) revenue. The marginal increment of revenue having been considered worth the cost of the marginal increment of money to be invested, the item in question became one whose expenditure was expected to be justified by the whole activity of the business in the forthcoming period as planned by the coordinated decision
ex ante. The ‘efficiency’ of the subordinate administrators remains to be tested, not by whether they contribute the money allotted to them plus an ‘attachment’ of the item in question, but by whether they contribute the revenue which they offered.

Similar remarks apply to the contractual rent payment arising out of the earlier decision. To ‘attach’ this item to the flow of things through the business would appear to have no administrative significance.

But this does not exhaust the matter. The accountant would, I presume, say that so far I have referred only to ‘fixed cost’, or ‘overhead cost’, or ‘oncost’, or whatever he chooses to call the ‘interest’ and ‘rent’. There is still the outflow of money upon the purchase of stock to be dealt with. A, for example, will be spending his allotment of money. The accountant will see the outflow of money and the inflow of goods, and may wish to ‘attach’ bits of the money to bits of the goods, and ‘expire’ the money in the manner indicated by Mr Norris. His doing so appears again to have no administrative significance. A offered a certain revenue in return for being granted a certain allotment of money. It can be understood that an account should be kept of the actual allotment of money and the actual revenue, and used by the merchandise manager as a check upon A’s performance. And if A had offered to absorb the
allotment of money and return the revenue at certain rates during the period, it can be understood that interim accounts should be kept with the same object. But A was not asked to disclose what goods he would buy with the money, or what prices he proposed to pay and charge for the goods: such matters were left within his administrative discretion. A was not asked to supply the merchandise manager with a budget in respect of each line of goods, although he prepared one for himself. If an
account in respect of each line of goods were sent to the merchandise manager he would have no budget with which to compare it. The scope of accounting, as an administrative check upon A’s performance, appears to be limited to rendering an account in the same form as the budget approved by the merchandise manager. It is easy to construct simple cases to suggest the abortiveness for this purpose of further independent accounting.

Let us suppose, for example, that A’s mark-up on stock ranged between twenty and forty per cent on buying prices, and that he achieved his anticipations in all respects except one. In one line of goods he expected to make forty per cent, but, after ordering the goods and before making any sales, decided that he had over-estimated the selling market. In order to clear the stock, he put on a mark-up of only thirty per cent, and realized this. His failure will be shown in the account, by a shortfall in his actual revenue below the anticipated revenue in his original budget. But this comparison will not show wherein his failure lay. Neither apparently will the pursuance by the accountant of the practice of ‘attaching’ and ‘expiring’ and linking divisions of revenue to the divisions of ‘cost’, for obviously the thirty per cent result, in the achievement of which the failure occurred, appears to be a better result than others, in the achievement of which no failure occurred. Only if A had submitted his corresponding budget, showing that the result ought to have been forty per cent, would the account have significance. It is easy, too, to construct simple cases to suggest that, if formal budgets were submitted for the purpose of making such comparisons, the accountant would, in his accounts, have to accommodate his methods of ‘attaching’ and ‘expiring’ to the discretion allowed by the firm’s administrative
arrangements, and not proceed with his own independent methods of ‘attaching’ and ‘expiring’. Suppose, for example, that A, acting within his administrative discretion, planned to buy fifty homogeneous raincoats for £50 and to sell forty-eight of them upstairs for 30s (£1.50) each, and the other two in a bargain basement for 18s (90p) each, and actually achieved the results anticipated. Clearly there is here one piece of business which is indivisible: one venture which has to be read as a whole. To ‘attach’, for example, £2 to the two raincoats going to the bargain basement and £48 to the others, and to ‘expire’ £48 against the sales upstairs, leaving £2 to be ‘expired’ against the 36s (£1.80) revenue in the bargain basement, showing a ‘loss’ of 4s (20p) in a separate account, would be meaningless if not misleading. It would certainly be misleading to suppose that accounts incorporating ‘attachments’ and ‘expirings’ according to independent methods of the kind indicated could operate as a criterion of A’s efficiency in exercising his discretion to budget and act as he did.

Nothing that I have said should be regarded as suggesting that no separate accounts should be kept of sales of separate products. Obviously, if A issues goods to salesmen—whom I assume here to have no discretion to vary the prices which A puts on the goods—he is likely to want reports upon which products are producing his incoming revenue: he will want to know whether particular goods are being sold at the rate he expected. The collection of this information does not, however, require any ‘attachment’ and ‘expiring’ of bits of money. Invoice analysis, or some other method, could yield the required information either in physical units of stock or in resale prices. For A to receive reports as to how the raincoats were selling, it would not be necessary to ‘attach’ and ‘expire’ bits of the amount of money spent on them.

There is still another matter. I have suggested that the extended independent accounting could not in the circumstances be regarded as having the function of being a report to the merchandise manager on A’s performance. Could it have the function of informing A what he ought to charge for the goods in stock? Could the ‘unexpired’ bit of money ‘attached’ to the bit of
material be regarded as any criterion of what A ought to charge the public? The answer that the subjectivist must give is that it could not—emphatically not. To assume that it could would be to make an assumption which belongs to the category of ‘cost-of-production fallacies’.
*38 It must be added, again with emphasis, that the irrelevance of the ‘unexpired’ bits of money for price-fixing does not depend in any way upon the accountant’s method of ‘attaching’ and ‘expiring’. It is not a matter of petty illogicalities in particular methods. The irrelevance and the ‘cost-of-production fallacy’ lie in the very ‘attaching’ itself. The money ‘attached’ has already been spent. It appears only by the trick of ‘attaching’. A has the goods, not the money ‘attached’ to them.
*39 The money ‘attached’ is not a cost although the accountant gives it that name. The only cost which is significant for the purpose is the cost—in the subjectivist’s sense—which occurs if a new decision happens to be made.
*40 Under what circumstances
will a new decision be made? Selling prices are tentatively planned
ex ante, that is to say, before the goods are bought.
*41 But it is likely that often, as time passes and the relatively obscure future approaches nearer to the present, the administrator will revise his appreciation of the selling market conditions, and consequently revise the selling prices that he had in mind when he bought the goods. One of the simple examples that I gave suggested as much. In that case A decided that he had overestimated the selling market. He cut his expected selling price. The subjectivist argument is that the money spent on the goods has no relevance for fixing a limit to the extent of this cut. Is there then no limit
on the cost side to the extent of the cut? The answer
is that the limit has to be found in the contemporaneous and intertemporal opportunities which I have discussed elsewhere.

The problem before A here would be whether he would be better off eventually by cutting the price at once and realizing over a shorter period, or by hanging on for the higher price (and perhaps having to cut it eventually). This is not merely a question of choosing between two alternative total revenues; it involves also the question of money being available earlier or later—perhaps for reinvestment. The course of action (alternative opportunity) rejected by A would be his cost of taking the course which he chose.

In spite of this association of the practice of ‘attaching’ and ‘expiring’ with ‘cost-of-production fallacies’, it cannot be pretended that in the modern world firms do not adopt, as part of their standing orders, the convention of assuming that the ‘bit of material’ is to be regarded as having a cost equivalent in significance to the sum of money so ‘attached’ to it—in spite also of Mr Norris’s contention that the ‘costs’ are to be distinguished from values. It is well known that they do.
*43 Seeing this, the subjectivist, without questioning the business administrator’s freedom to do what he liked (providing that he accepted responsibility for what he did, and the ‘automatic sanction’ for error) would associate such firms with Wicksteed’s businessman whose ‘temper is expensive’.
*44 He might enlarge upon the dangers inherent in its practice to the
firm itself,
*45 and, where the practice was common to a large number of firms,
*46 or where the application of the ‘automatic sanction’ was modified,
*47 to society. But this is not the place to raise these discussions.


My discussion of ‘The subjective theory of value and accounting “cost” ‘ is intended to throw out some suggestions which may not be immediately apparent on the face of it.

1. If economics is to be useful to assist discussion of the problems of internal organization of the firm and the explanation of the industrial structure and its weaknesses, the aspect of economics that should be developed is that which deals with people’s behaviour when they are deciding what to do next with their resources. It is not sufficient, however, to assume that the decision-making is performed by individual people whose decisions are coordinated with those of other people
only through the medium of the market. Within the ‘large-scale undertaking’ decision-making is shared. Coordination occurs through other means than the market. The relationship between the buyer of factors and the seller of the product, for example, is
not a market relationship. What is required from economics is the presentation of models showing how the decision-making might be split up (shared or delegated) and coordinated, together with models of ‘standing orders’ determining the channels and timing of coordination. Energy might then be diverted from the impotent condemnation of monopolistic institutions to a critical examination of internal organization with a view to discovering its weaknesses, which incidentally lead to the formation of such monopolistic institutions.

2. If accountants studied a theory of administration, working from the Subjective Theory of Value, through the coordinated decision
ex ante, towards a set of theoretical models of administration charts, standing orders and budgets, they would discover
that both the orientation of accounts and the methods of accounting ought to be accommodated to the particular administrative arrangements of the particular firm or organization. Accounts would always be related to administrators’ budgets and would always be of a form corresponding to those budgets, instead of being prepared independently of them. There would be no pretence that money was attached to things when it had already been spent, or contracted to be spent, upon those things. All the pseudo-problems of ‘allocating’, ‘burdening’ or ‘charging’ would disappear.

The following comment is offered after a first reading of Mr Harry Norris’s article on ‘Profit: Accounting Theory and Economics’ in the August issue of

Mr Norris states: ‘We accountants grant the attribute of objectivity to “profit” if not to “income”…’ (p. 132). The difficulty of conceding that accountants are right in doing this will never be understood until it is recognized that the objective results upon which accountants work can be explained only by reference back to, and in the light of, the opinion of the decision-maker whose decision gave rise to those objective results. Out of his process of decision-making emerges the decision-maker’s budget relating to the course of action which he decides to take. This may or may not be recorded. Such a budget, without an analysis of the opinion of the decision-maker attached to it, would not disclose the subjective acts of valuation which determined that his planned course of action was in his opinion the most advantageous or ‘profitable’. For example, at a particular time and in a particular situation the budget might contain on its ‘expenditure’ side merely an enumeration of diverse non-monetary resources already in the ownership or control of the decision-maker, which he had decided to use for a particular job because he contemplated no better use for them. On the ‘revenue’ side might be a sum of money which he expected to achieve by selling the results of the job. The expected ‘profitability’ of the job would reside in his valuing his contemplated returns from this job higher than his contemplated returns from any (or the best) alternative use of his resources. This subjective valuation would not appear in the budget of anticipated objective
results. Neither would it, nor ought it to, apear in the subsequent account of actual objective results. Autonomous accounting which, without reference to the decision-maker’s opinion, but in order to make up an account of ‘profit’ in monetary terms, subsequently introduced an assesment of ‘cost’ into the record of objective results, would apparently be substituting (
ex post) a simulated objective result for the decision-maker’s subjective act of valuation (
ex ante).

Mr Norris perhaps makes his best approach to recognition of the link between the decision (and budget) and the objective results (and account) in his discussion of ‘fashion’ goods on p. 130. His remarks might be compared with my own reference to homogeneous raincoats (above).

Harry Norris. ‘Notes on the Relationship between Economists and Accountants’,
Economic Journal 54, Nos. 215-16 (December 1944).

G. F. Thirlby, ‘The University Commerce Curriculum’,
Sociological Review, 34, Nos. 3 and 4 (July-October 1942).

C. S. Richards, ‘The Task before Us: with special reference to Industry’,
South African Journal of Economics, 12, No. 3 (September 1944).

‘The conception of real costs as displaced alternatives is now accepted by the majority of theoretical economists’. L. Robbins, Introduction to Wicksteed,
The Common Sense of Political Economy (London, 1933), p. xviii. It is significant that Professor Robbins adds to these words ‘but…we are still a long way from making it part and parcel of our daily speculations on those problems to which it is most relevant’, and that on a previous page (p. xv) he has stated that ‘since the war [1914-18], there has appeared a great mass of literature on the cost question which, for all the awareness it displays of the essential problem at issue, might for the most part have been the same if Wicksteed had never written’.

Cf. ‘Resources and needs exist for practical purposes only through somebody knowing about them and there will always be infinitely more known to all the people together than can be known to the most competent authority’. Hayek, ‘Scientism and the Study of Society’,
Economica N. S., 11, No. 41 (February 1944), p. 37.

Fritz Machlup, ‘Competition, Pliopoly and Profit’, parts I and II,
Economica N. S., 9, Nos. 33 and 34 (February and May 1942), part I, p. 9.

Machlup, ‘Competition, Pliopoly and Profit’, part II, p. 156.

The Common-sense of Political Economy, p. 820.

I propose to confine my discussion in this section to a single coordinated decision
ex ante. It is my hope that this will be adequate to suggest that an understanding of this coordinated decision
ex ante is the appropriate starting point for the development of a philosophy of modern large-scale business organization.

The process would, I suppose, be commonly referred to as the planning of the acquisition (or retention) and use of short-term funds, or short-term capital, or working capital.

This term is used by Brutzkus in
Economic Planning in Soviet Russia.

On the limitations of committee management, see Hayek ‘Scientism and the Study of Society’, p. 31, footnote 2.

My abstract discussion is founded upon a section of a concrete discussion of Budgetary Control in Department Stores given some years ago by Professor Arnold Plant in his lectures on Business Administration.

A term used by Machlup, ‘Competition, Pliopoly and Profit’, part II.

Norris, ‘Notes on the Relationship between Economists and Accountants’ p. 376.

That economists sometimes tacitly adopt the same sort of assumption is apparent in a definition of depreciation cost by Mr. Hawtrey which is criticised in G. F. Thirlby, ‘Permanent Resources’,
Economica, N. S., 10, No. 39 (August 1943), pp. 247 ff.

This implies that results of breaches of standing orders issued to executives and other people, and results of ‘acts of God’, are excluded from ‘production’.

I presume that Marx would have ‘expired’ units of ‘labour’ instead of units of money.

‘The value of what you have got is not affected by the value of what you have relinquished or forgone in order to get it… You have the thing you bought, not the price you paid for it’. Wicksteed,
The Common-sense of Political Economy, pp. 88-9.

This cost might be of higher or lower significance to A than an amount of money—if he had it—equal to the amount ‘attached’ to the goods.

I find that students under the accounting influence sometimes find it a little difficult to understand this, particularly if the goods are bought in one country and sold in another. I have taken to asking them the question ‘If this is not so—if there is no
ex ante coordination of the buying and selling markets—how does the buyer of a commodity know how much to buy? Is he indifferent as to whether he buys a collar-box full, enough to fill a fleet of ships, or a quantity given by a number drawn out of a hat?’

See Thirlby, ‘Permanent Resources’.

Some evidence, if any is needed, is contained in the following extract from an article by Mr K. Lacey: ‘There are many…types of business (e.g. those producing proprietary lines), the selling prices of whose products lag very far behind the movement of raw material prices, and tend rather to be based upon the average cost of their stocks on hand. The profits of such businesses on the first-in-first-out basis do not vary quite so greatly over the Trade Cycle, and the adoption of the last-in-first-out basis might have the unusual effect in some instances of making their profits more unstable from year to year than they are at present. There is a fallacy here however, and it must not be assumed that the earning of a reasonably stable profit is evidence that no self-deception exists and that no alteration in method is desirable. The position here is that sales are made at too low a price relative to replacement costs when market values are rising, and at too high a price relative to replacement costs when market values are falling’ (‘Commodity Stocks and the Trade Cycle’,
Economica N. S., II), No 41 (February 1944), (Mr Norris joined issue with Mr Lacey in the following August issue of
Economica.) This article is further welcome evidence that accountants are becoming concerned about the effects of accounting practice.

The Common-sense of Political Economy, p. 387.

See twelve articles by R. H. Coase in the English
Accountant, 99.

See Lacey, ‘Commodity Stocks and the Trade Cycle’.

E.g. where there is compulsory cartellization. On the association of cost-accounting with cartellization,
see Burn,
Economic History of Steel Making (Cambridge 1940), pp. 494-5.

Essay 7, The Ruler