Here We Go Again and Again

Here We Go Again and Again

Here we go again…and again

When the IASB and the FASB were embarking on their joint project to revise the conceptual framework (CF) and published theirPreliminary Views on an improved

Conceptual Framework for Financial Reportingin 2006 I sent a comment letter entitled ‘Here we go again…’ suggesting that it was a pity that there was little consideration in the Discussion Paper of the kind of thing a CF actually is. In particular, I argued that the nature of the principles in the framework and how they were to be used in setting standards was not examined. I felt that this should have happened before the revision of the CF was undertaken if the project was to be successful. In the rush to show some results from the project this was ignored.In the 2006Discussion Paper the Boards ‘concluded that a comprehensive reconsideration of allconcepts would not be an efficient use of their resources’.Now, seven years later,the IASB are undertaking a revision of the CF, this time without the FASB. In the 2013 Discussion Paper A Review of the Conceptual Framework for Financial Reporting(IASB, 2013) the IASBhas decided to ‘build on the existing Conceptual Framework—updating,improving and filling in gaps rather than fundamentally reconsidering allaspects of the Conceptual Framework’. Here we go again…and again….

Why is it important to consider the nature of a CF and how it is to be used in making standard setting decisions?The IASB state that the purpose of the CF is ‘to assist the IASB in the development of future IFRSs and in its review ofexisting IFRSs’ (IASB, 2013, §1.25). This makes it clear that the CF is to be used to assist in deciding upon actions it will undertake in developing or revising accounting standards. When we decide upon actions in everyday life we start with a desire to bring about some end and consider beliefs about what actions will achieve these ends and then decide what actions we want to perform. Decision-making can be characterized as coming to want to do something as the conclusion of reasoning from statements that express these desires and beliefs. To take a simple example, I want to eat something and believe that going into a restaurant will result in my eating something so I decide to go into the restaurant. One way of thinking about the CF is to take it as expressing desiresfor financial reporting that are used in reasoning to standard setting decisions by considering what standards will achieve these ends. The objectives and qualitative characteristics chapters in the CF can be seen as statements expressing what is wanted from financial reporting, that is, a desire to meet user needs and a desire to produce information with certain characteristics that users want. This is part of the answer to the question about the nature of principles in the CF. Some principles in the CF express what is wanted from financial reporting.

Given what is wanted, the next question to answer is ‘what information will meet these needs and have these qualities?’. If those who construct CFs, conceptual frameworkers, believe that the information that will meet these needs and have these qualities is information about assets, liabilities and equity and about income and expenses and that information measured using certain measurement bases and recognized when certain characteristics are met then what is wanted is to produce such information in financial reports. It clearly important to understand what the elements represent if there is a belief that information should be about them. This gives rise to a need to define the elements which appears in the elements chapter of the CF. Another kind of principle in the CF is one that gives definitions of this kind. The recognition chapter and the measurement chapter are different. They express general desires to report elements when certain characteristics are met and to report the elements with monetary values in accordance with certain measurement bases. In effect, these general desires are in the nature of general rules. If one wants to report elements in certain circumstances and in certain ways then one wants something as a rule. Adopting a rule is to want to do something in general when the circumstances for following the rule exist. The recognition and measurement chapters thus express general rules that are used to derive more specific rules for accounting for particular elements. These chapters express a different kind of principle. It is interesting to note that the chapters on recognition and measurement are derived by reasoning from the statements in the chapters on objectives and qualitative characteristics.The statements in the CF are thus used in reasoning to other statements in the CF and in reasoning to decisions about accounting standards. The former is implied by the ‘pyramidal’ structure of the CF recognized in the FASB Discussion Memorandum No. 1 (FASB, 1974). The latter is implied by the stated purpose of the CF as expressed in the latest IASB Discussion Paper. What is not clear is the nature of this reasoning for this is not discussed in the paper or, indeed, very much in previous CF documents.

A number of writers on the CF have commented on the assumption that appears to be made by conceptual frameworkers that the reasoning involved in constructing and using a CF in standard setting is deductive (Power, 1993, p. 48; Davies et al., 1999, p, p. 81). This appears, initially, rather odd. After all, when we decide to do something in ordinary life, such as reasoning to a desire to go to a restaurant in the example above, the reasoning is not deductive. Given the desire and the belief that a certain action will fulfil the desire we do not deduce that we want to perform that action. There is no necessity in deciding to perform that action, the conclusion is not somehow implicit in the premises and the conclusion can be undermined by wanting something else, to conserve one’s money for example. This reasoning does not have the characteristics of deductive reasoning. The assumption that the reasoning is deductive in the context of a CF needs to be explained. One way of explaining it arises from the nature of the conclusions that are drawn in such a context and in the context of standard setting. The conclusions of such reasoning in these contexts are not desires to perform some particular action but desires to do something in general. The measurement and recognition chapters and the standards derived from the CF express something that is wanted generally, namely to perform actions as a rule. If one wants to do something generally then this must be because the reasons for wanting it also have a generality. One wants to adopt a measurement base because, on each occasion when it is used to report elements, doing so is believed to achieve a certain end or objective or desire. One must want to achieve what is wanted on a number of occasions, or generally. The belief that adopting a measurement base is also not a belief that adopting the base on a particular occasion will achieve what is wanted but a belief that adopting the base will generally achieve this end. Both the desires and the beliefs that are used in reasoning to something wanted generally are themselves general, that is, wanted on more than one occasion. The principles in the CF, whether they are expressions of desires or of something wanted as a rule, have generality.

It is a short step from thinking that the principles in a CF have generality to the thought that they are universal. This step is made in some kinds of systems of morality and law. If one always wants to achieve some end and believes that doing something will always achieve this end then one can deduce that one always wants to do that thing. From a universal desire and a universal belief that doing something will fulfil the desire one can deduce a universal desire to do that thing, that is a desire to do something as a rule. In the CF context, if one always wants to meet user needs and believes that using a fair value measurement base will always meet these needs then you can deduce that you always want to use fair value. If ‘you’ are a standard setter then you may reason from this universal desire to use fair value to a desire to measure financial instruments using fair value.You may then prescribe that fair value be used for particular elements in an accounting standard. If a standard setter wishes to use a CF to reason deductively to standard setting decisions then conceptual frameworkers will aim to construct universal desires and express them in a CF. This kind of objective is also present for those who construct moral systems and those who construct constitutions in legal systems. The problem they face, at least in the context of constructing moral systems, is that ‘few moral judges are equipped with an exhaustive set or exceptionless moral principles by reference to which all their moral judgments are made’ (Craig, 2005, p. 694). It is difficult to construct desires that are universally desired or to formulate beliefs that a certain kind of action will universally fulfil such desires and hence to deduce universal desires to act in a certain way as a rule. Why should conceptual frameworkers fare any better?

Solutions to this problem have been suggested in the moral sphere. One moral philosopher has suggested that the kind of universal desires or rules used in deciding on particular actions should be accepted as, in some way, provisional. The idea is to ‘suppose then that we find ourselves in certain circumstances which fall under the principle, but which have certain other peculiar features, not met before, which make us ask “Is the principle really intended to cover cases like this, or is it incompletely specified—is there here a case belonging to a class which should be treated as exceptions?” Our answer to this question will be a decision, but a decision of principle…. If we decide that this should be an exception, we thereby modify the principle by laying down an exception to it’ (Hare, 1952, p. 65).This provisionality can be seen to be required by the assumption that the reasoning using such universal principles is deductive. With deduction, if you do not like the conclusion of reasoning then you must reject at least one of the premises. Applied to the CF context, if you deduce a consequence for standard setting from a principle in the CF and you think, on other grounds, that it is unjustified then you must modify the principle in the CF. It is also the case with deductive reasoning that if incompatible conclusions are drawn from valid deductions then the premises must also be incompatible (Salmon, 1992, p. 26). An example of the latter might be the deduction of a rule requiring the use of a fair value measurement rule from the objectives statement and a belief about what kind of information the users want. If the qualitative characteristics require reliability and information based on fair values is not reliable then a rule requiring that fair value be not used would follow. The problem now is that principles in the CF are used in deductions to incompatible conclusions. At least one of the principles in the CF needs to be modified.

This might be described as an example of ‘reverse engineering’ (FRC, 2013). Given the pyramidal structure of the CF items higher up in the pyramid are supposed to be used to derive items lower down in the pyramid. If the items lower down in the pyramid are not acceptable, either on other grounds or because they are incompatible with other conclusions derived, then given that the derivation is deductive items higher up in the pyramid must be modified. That this kind of ‘reverse engineering’ has happened in the development of the CF has been maintained by several academics. O’Brien suggests that ‘as standard setters shift their focus toward momentary fair valuation and away from periodic measurement of performance and business activity, concepts like reliability, matching, earnings, stewardship, and prudence become inconvenient. Their presence in the framework encourages criticism of standards that promote fair value reporting and the primacy of the balance sheet. One solution, then, is to remove the troublesome terms from the framework’ (O’Brien, 2009, p. 264). The replacement of reliability by faithful representation may be one such example (Whittington, 2008, p. 147). Another example is arguably seen in the designation of primary users in the objectives chapter. The original desire to meet the needs of users creates a problem for standard setters where different users want different things. The debate between the use of financial statements for making economic decisions and for making stewardship decisions is symptomatic of this problem. Incompatible conclusions may be deduced from the original desire. This requires a modification of this desire of the kind that the IASB and the FASB proposed in the 2006 Discussion Paper. The quest for users and needs which ‘behave themselves’ and allow for deductions of consequences may lie behind the perception that users and their needs have been ‘made up’ (Young, 2006) to enable the requisite deductions to go through without the embarrassment of incompatibilities. Characterising a CF as a ‘constitution’, as the FASB have done (FASB, 1976, p. 2), encourages the conception of a CF as something from which deductions can be made. The point about a constitution is that it is meant to provide principles that cannot be overridden. The problem is that, as in the moral sphere, these exceptionless principles may not be found in the area of financial reporting. They seem to be difficult in the moral sphere which draws upon what it means to be human. How much more difficult will it be in the sphere of financial reporting where different human interests may conspire to frustrate the acceptance of universal principles or desires. The danger in conceptualizing a CF as something that expresses such universals is that it may become impossible to construct a framework that is acceptable without riding roughshod over the desires of some interested parties. Even if the hope of constructing such a framework still exists the likelihood that the one we develop will be the one we need is doubtful.The prospect of further revisions looks likely.

Is there an escape from this treadmill? One way out of the cycle of revision and ‘reverse engineering’ may be to accept that a CF is not something that can be used to deduce either statements within the framework or standards outside of it. What if the CF expressed general desires or general rules that are to be used in non-deductive reasoning to standard setting decisions? Just as reasoning to particular actions in everyday life is not deductive so reasoning from general desires to other general desires within the CF and to decisions about general rules in standard setting is not deductive. Instead, one starts with general desires, things one wants to achieve through actions in general, and then looks around to see what actions will fulfil these desires. It may be that one has other desires that one wants also in general that may result in a different conclusion being drawn about what else is wanted or what actions will fulfil them. The fact that one may also want something else and this may result in not fulfilling the first desire does not imply, as it would with deduction, that one does not want the first desire in general. Wanting something in general is compatible with not wanting it where one wants something else more. With reasoning of this kind one does not have to reject the first desire if it results in incompatible conclusions being drawn. If one wants, in general, to provide information that is useful for making economic decisions and one wants, in general, to provide information to assist in stewardship decisions, then if providing information that meets one of these desires but not the other one may still conclude that one wants to provide such information. This does not mean that one does not want to provide information to fulfil both desires in general. It merely means that in circumstances where both desires cannot be met then one has to decide between them. The reasoning may also involve beliefs about what actions, in general, will fulfil these desires, not necessarily about what actions will always fulfil these desires. Although one may conclude that one wants to perform some action as a rule this is a general rule that may not be followed on all occasions. If following a general rule may frustrate some other general rule this does not mean that one has to reject the first general rule. One may want to do something, in general, but this is compatible with not wanting to do it in certain circumstances. A general desire or a general rule is not a desire or a rule that has to be fulfilled or followed on every occasion in which it applies. For example, one may want, in general, to follow a rule to provide prudent information and also to want, in general, to apply the matching principle, but if in certain circumstances one cannot do both then one will follow one rule and not the other. The situation is similar to the UK situation where the desire to follow the rules in accounting standards, in general, is compatible with overriding them in specific circumstances. This does not mean that one cannot want to follow accounting standards, in general, and yet override them on occasion. The rules in accounting standards do not have to be conceived as rules which are always to be followed. Similarly, the rules in the CF, say of measurement were the CF to formulate such rules, may be followed generally but on occasions overridden by other general rules.