EDITORIAL
Recent proposals for the adoption and implementation of the IFRSs, discussion papers about the joint IASB/FASB Conceptual Framework project (IASB/FASB, 2008a, 2008b), and several other professional and regulatory initiatives come in the aftermath of perceived global financial disturbances—such as unexpected company failures, concerns over private equity takeovers, and more recently numerous matters related to the sub-prime loan crisis. Over many decades, similar crisis settings have presaged re-examination of accounting and auditing practices (see previous Editorials by Dean, June 2001, February 2002, March 2006, March 2007; Dean and Clarke, June 2004, June 2005; and Dean, Clarke and Wolnizer, June 2002).
A recurring element in many of the current professional initiatives is reference to the need for a principles-based rather than rules-based reporting regime—a regime that considers the issue of measurement as fundamental to an effective and efficient reporting and disclosure system. ‘Measurement Issues in Financial Reporting’ was the subject of a symposium at the April 2008 EAA.1 The proceedings (including questions and answers) suggested the need, as part of the conceptual framework exercise, for a better understanding of fundamental measurement concepts and the link to accounting's role in decision making.2
A previous Editorial (June 2008) outlined the potential for ‘a persistence of the “constitution” approach’ to developing a conceptual framework.3 Further, it was noted:
there is need to recognize that accounting is a service activity—its outputs must be useful for the uses ordinarily made of accounting data in cognate disciplines. It is not an end in itself. Accounting concepts need to be forged with a view to satisfying simultaneously the needs of both internal accounting uses—taking specific account of the requirements of the data to be useful for the uses ordinarily made of the data by disciplines like operations research, finance, management and (say) law . . . If this aspect is ignored the latest CF exercise is destined to become another example of Sterling's‘recycling of ideas’—without ‘resolving problems’ (1975, 1979).
Interestingly, such concerns have a long history. For example, they were well canvassed during the current cost accounting debates of the 1970s. The present Editorial places the current expressions of concern against the backdrop of some selected 1970s correspondence between the respected U.K. post-Keynesian economist G. L. S. Shackle and the founding editor of this journal, Ray Chambers. Chambers is well known to the readership of this journal, Shackle perhaps less so. Shackle was a highly respected and influential economist. He wrote several well-known books, including Expectations, Investment and Income (1938), Expectation, Enterprise and Profit (1970), Keynesian Kaleidics (1974) and Imagination and the Nature of Choice (1979). It is instructive to see what implications his understanding of measurement processes have for the practice of accounting.
The correspondence drawn on below occurred late 1975 and early 1976 and is sourced from the University of Sydney Archives Unit.4 Essentially, it reinforces the importance, in the context of decision making, of the differences between facts and expectations (especially ‘income’ and ‘expected income’), and the need therefore for accounting, in providing factual data about the past and present, to satisfy the canons of measurement. Those canons were clearly identified, for example, by Chambers in his March 1998 Abacus article, ‘Wanted: Foundations of Accounting Measurement’, which is essentially a previously unpublished review, written in 1972, of a 1971 Report by the AAA Committee on the Foundations of Accounting Measurement:
INTRODUCTION: 1997
From time to time for over 150 years pervasive flaws in accounting thought and practice have been brought to notice. The principal focus has been the almost limitless array of valuation (or quantification) rules that have emerged [in 1994/5 the FASB averred five different valuation bases, called ‘attributes’] under the widely fostered view that each mercantile firm may choose (and vary as it pleases) the combination of rules that its accounts will follow . . .
CRITIQUE: 1972
It might have been hoped that the Committee on Foundations of Accounting Measurement of the American Accounting Association (AAA, 1971) would have produced some conclusions in the nature of foundations; something of an elementary kind which might provide the groundwork for more elaborate or more sophisticated exercises. Its report is disappointing. It ranges over an almost limitless territory. But it finishes almost exactly where it started: ‘New guidelines are desperately needed . . . The lack of new guidelines for accounting is the most serious problem to be faced’ (p. 48). But worse than the failure of the Report to give direction is its potential for misdirection.
The Report begins with an ‘adopted’ definition: ‘accounting measurement is an assignment of numerals to an entity's past, present and future economic phenomena, on the basis of observation and according to rules’ (p. 3).5 This, as we shall show, is complicated and vague enough. But it is made even more vague by the following sentence: ‘Under this definition, it should be pointed out, the rules employed need not be good ones and observations made need not be correct to qualify as accounting measurement’. Freed from the need to employ ‘good’ rules for assigning numerals and from the need to make ‘correct’ observations, apparently any numbers that masquerade as (look like) measured magnitudes will qualify as accounting measurement. That, of course, is the standard recipe for the success of fraudulent misrepresentation, as any reader of the history of commercial fraud will know.
But return to the definition. It allows that accounting measurement includes the assignment of numerals to an entity's past, present and future labour turnover, market share, unfilled orders, total labour force and many similar things; for, as well as I can judge, these are all economic phenomena of an entity. Yet none of them falls within the only two types of accounting system, ‘equity accounting’ and ‘operational accounting’ (as they are commonly understood), contemplated by the Committee. The definition allows also that accounting measurement includes the assignment of any kinds of numerals. But the knotty problems which confront accountants concern themselves with the assignment of numbers of dollars, not numerals at all. As I understand it, the object of a definition is to limit the scope of usage of a term so that one can come to close grips with it and make good use of it, as of a fine tool, for the purpose in hand. But the Committee's definition is so open, it provides so many things to talk about, that we should not be surprised if it comes to close grips with none of them.
Properties, Scales and Units
It would have been instructive if the Committee [1971 AAA Committee on the Foundations of Accounting Measurement] had begun with a discussion of measurement simpliciter. It is true that most of the discussions of measurement occur in the context of particular branches of science. But they have sufficient elements in common to justify some general specifications of the object and process of measuring.
In the first place, every particular measurement scheme requires the specification of the property of a class of objects which it is of use and interest to measure. The Committee, almost as an aside, alludes to what might have provided such a specification: ‘accounting is indispensable in measuring and reporting organizational wealth and its changes’ (p. 7). But the point is not developed. What wealth is, how it is measured and how changes in it are measured are not discussed.
In the second place, every measurement scheme requires the specification of a scale of some kind which makes it possible to distinguish the extent to which every object in the class possesses the specified property. At least some important contributions to the literature of measurement distinguish ordinal scales, interval scales and ratio scales. These scales (or rather measurements taken in them) have different mathematical characteristics. As the addition of measures is so common in accounting processes, it is necessary to consider the conditions under which addition (and other forms of relation) are mathematically permissible—and in fact to stipulate the kind of scale which is appropriate to accounting. Nowhere in the whole Report is there any discussion of scales of measurement.
In the third place, every measurement scheme requires the specification of a unit in the scale, and the conditions under which unit measurements shall be deemed to be of equal significance. In brief, this requires specification of the meaning of the ‘standard’ unit. This is necessary since measurements may be taken in a variety of non-standard situations, such that the raw or crude measurements are not comparable or addable. Nowhere in the Committee report is there any discussion of the unit of measurement, either in general terms or in the terms appropriate to accounting measurement. (Chambers, 1998, pp. 36–8)
To put context into the previous and ongoing measurement debates, the mid-1970s correspondence between Chambers and Shackle is reproduced. It reveals insights that there is a divide in the literature and practice of accounting between what might be described as practice-driven and more technical approaches to measurement. This divide is agreed to be critical in evaluating the serviceability of accounting for business decision making.
Chambers’ Typewritten Letter to Shackle, 7 November 1975
Dear Professor Shackle,
I am taking the liberty of sending under separate cover a paper I have submitted for publication in Accounting and Business Research [‘The Functions of Published Financial Statements’, Spring 1976, pp. 83–94]. Workers in one specialism do not commonly cross their boundaries. But it is of more than passing interest to find one's views corroborated by another's. I have long been uneasy about the glib adoption of the apparatus of mathematical probability as a model of the processes of choice. Most of those who exploit it seem to me to fail to consider that choice is constrained by present states; that it is not so wide and free that a decision-maker must fall back on statistical probabilities; and that the world is not so well-structured that the use of statistical probability is relevant. These questionable, and in many respects mistaken, presumptions have influenced many of the recent generation in my own field.
For over 20 years my own disposition to establish the significance of knowledge of states or positions from time to time has met with little sympathy, notwithstanding that it lies at the core of the accounting process in my opinion. It has long been comforting and some encouragement, therefore, to find in your work a sustained development of the decision making theme with which my own conclusions would be consistent. Being concerned with a singular aspect of a particular class of choices, my paper will not be of more than limited interest to you. But it cuts across much that is confused and irrelevant in the literature and practice of accounting; much that must be cleared away before the accounts, balance sheets and so on of business firms can be serviceable in the ways in which so many people, including economists, believe them already to be.
Sincerely,
R. J. Chambers
Shackle's handwritten response to Chambers’ letter, 15 November 1975
Dear Dr Chambers,
It was a very great happiness to me to have your letter. I have realized for many years how fascinating a study the meaning, purposes and principles of accounting could have been for me, if only my early education as an economist had been conducted under less extreme pressure of time [for example, to finance his education Shackle worked as a bank clerk whilst studying for his undergraduate degree, as well as when he started a PhD, initially under Hayek]. The two disciplines are concerned with the same field, but see it in a different light and discuss it in different languages. Can your thesis be summed up by saying that a set of accounts should confine itself to statements to which the test of a factual truth can be applied, and that it should, of course strive to adhere to the truth? I am in perfect agreement with this. We can be eye-witnesses only to the present, including its records and traces of the past. We cannot be eye-witnesses of the future. In recent works I have avoided references to the ‘future’, which suggests by its derivation there is something objective to be discerned if only we were more clever or more strenuous. I prefer to say ‘time-to-come’. It seems to me plain that, as you insist, a business man's choices are bounded by his list of assets and obligations as these exist when he is making his choice. I think his list needs (in principle) to be expressed in terms of technological and physical equipment and not only in terms of dated market value. I wonder whether you would feel that (as it seems to me) a business man is entitled to value his physical plant's equipment, buildings, et cetera, in the light of plans which he may have for exploiting them, and that these plans may sometimes, with some justificatory basis, assign them a higher value than the market currently does? One of the things which most of all astonishes me is the speed at which huge fortunes can be made, well within a man's lifetime. Must not such a thing be the result of his being able to buy for a low market price, materials and tools which he can envisage as forming a system capable of uses which would give them a far greater value than their purchase price? Their high value in his thought would arise from his personal orientation of this system, an orientation which those who compete in the market to buy the elements of this system do not have in mind or vision. We may be entitled to entertain such ideas and act upon them. We would not be entitled to include assets in his balance-sheet on such a basis. The basis may later be secured, if the performance of his enterprise in practice gives the market a vision like his own. The prospectus and the balance-sheet are different things with different purposes and different meanings, they must not, as I am sure you are right in insisting, be entangled with each other.
I have read your article with very great interest and I very much hope it will be read by many businessmen. It was most kind of you to write to me. If only I had more time I should be eager to study the engrossing fields of problems on which you are engaged.
With kind regards,
Yours Sincerely,
G. L. S. Shackle
Chambers’ Reply to Shackle, 6 February 1976
Dear Professor Shackle,
It was very kind of you to write to me at length. Your view of the relation between information about the present status and the imaginative construction of future states tallies exactly with my own. I recall that in some of your work you mention the confusion which arises when economists write of ‘income’ when they mean ‘expected income’. Many people in my own field quote from economists without noting that there is a difference between the attained and imagined.
Certainly, as you suggest, a businessman may value his assets ‘more highly’ than the market in some cases. But, as it seems to me, the two valuations are different in kind and cannot be compared (as in the ‘more highly’). The market assigns an exchange value, the businessman, a user value (i.e. an expected net yield, to some time horizon, discounted to the present). Whether he chooses to sell or to hold an asset depends on a comparison of the net present values of (a) using the asset and (b) using otherwise the present proceeds of selling the asset. This is a comparison of two imagined consequences—which is quite proper; not a comparison of an actual magnitude with an imagined magnitude, which seems to be a fallacious comparison. It is the capacity to imagine profitable uses of goods which distinguishes the entrepreneur from the administrative manager. His imagined ‘valuer’ does not subsist in the goods represented in the balance sheet, but in his belief in the fruitfulness of their use in the way he envisages. Of course, he does not always succeed.
Thank you most cordially for sending me a copy of Keynesian Kaleidics, which I read with considerable pleasure—and profit. It prompted me to dip again into some of your other work.
Sincerely,
R. J. Chambers
We might well contemplate what is to be gleaned from those letters. And for whose benefit.
They contain a critical message for standard setters: the need to consider the commercial domain of accounting as the basis of the foundations of accounting (Chambers, 1991), especially as it relates to measurement. Contemplate further some other critical matters noted there by Shackle and Chambers:
- 1
The need to contrast the ‘attained’ from the ‘imagined’, the ‘past and present’ from, as Shackles observes, the ‘time-to-come’.
- 2
Hence the importance, as Chambers noted, of the need to be informed of one's ‘financial state’ (one's wealth and changes in it from the last observed state), whenever a decision is considered.
- 3
Actual and expected income are different concepts.
- 4
A balance sheet and a prospectus are, as Shackle notes and Chambers would agree: ‘different things with different purposes and different meanings, they must not, as I am sure you [Chambers] are right in insisting, be entangled with each other’.
- 5
The general canons of measurement should be the cornerstone of any application to an accounting conceptual framework—as Chambers (1998) noted, especially those pertaining to measurement, namely properties, scales and units.
Chambers aptly titled his 1998 piece, ‘Wanted: Foundations of Accounting Measurement’. Ten years on they are still wanted! ‘Measurement’ in accounting as elsewhere, now as then, is a technical not a haphazard affair. It is not something subject to the dictates of political whims. The canons of measurement have been well documented and are understood and accepted in many disciplines. Attempts by accounting standards setters to specify measurement methods will de doomed to repeat the mistakes of the past if they ignore that earlier discussion and if they do not adhere to the technical measurement canons noted above by Chambers (and by others, such as Canning many years earlier), namely the need to specify ‘the property of a class of objects which it is of use and interest to measure’, ‘the specification of a scale of some kind which makes it possible to distinguish the extent to which every object in the class possesses the specified property’ and ‘the specification of a unit in the scale, and the conditions under which unit measurements shall be deemed to be of equal significance’.
