In September 2002, the Financial Accounting Standards Board (FASB) of the United States published a discussion paper seeking views on whether U.S. standard setting should move from a principle based approach toward a principle based approach as sometimes associated with the International Accounting Standards Board (IASB). That paper was partly in response to the SarbanesOxley Act, which was itself a response to such accounting scandals as Enron and WorldCom. Schipper (2003) points out that the U.S. rules are often based on principles. That is, the standard setters use principles in order to produce the rules for the preparers of financial statements. Nelson (2003, 91) agrees, and suggests that a particular standard should rather be seen as more or less rules-based. He suggests that rules can increase the accuracy with which standard setters communicate their requirements and can reduce the sort of imprecision that leads to aggressive reporting choices by management. However, he notes that rules can also lead to excessive complexity and to the structuring of transactions. One of the reasons why standards on several topics need to contain rules is that the standards are inconsistent with the conceptual frameworks of the standard setters. For several topics, the use of the appropriate principle could lead to clearer communication and to more precision without the need for the current rules. That is, before asking how rules-based a particular standard should be, we should ask whether the standard is based on the most appropriate principle. I identify six topics on which the accounting standards have detailed technical rules. In each case, I suggest that part of the need for rules is caused by a lack of principle or by the use of an inappropriate principle (i.e., one that does not fit with higher-level principles). The lack of clear and appropriate principles can also lead to optional accounting methods in standards because no one policy is obviously the correct one; this leads to lack of comparability. I do not suggest that the use of appropriate principles would lead inexorably to standards with no optional methods but that, on some topics, optional methods could be eliminated. The six topics are examined one by one. In each case, I attempt to locate the principles being used, to assess the appropriateness of the principles, and then to identify any arbitrary rules or optional methods that result from the absence of appropriate principles. I start with the IASBs standards (hereafter, IFRSs), with frequent comparison with U.S. GAAP. One reason for examining

IFRSs in particular is that they are required for the financial reporting of listed companies throughout much of the world in 2005 onward,1 and the FASB has announced plans for convergence of its standards with IFRSs.2 The final section of the paper draws conclusions about how accounting might be improved by substituting principles (or better principles) for the existing requirements. PRIOR LITERATURE AND PURPOSE OF THIS PAPER Alexander (1999) investigates the nature of principles and rules in an accounting context. Below, I use the word principles to include Alexanders type A overall criteria (e.g., fair presentation, the definitions of elements of accounting and, in particular, the primacy of the asset and liability definitions) and his type B conventions (e.g., prudence). Such principles are contained in the standard setters conceptual frameworks. I contrast this to rules which are Alexanders type C rules (e.g., the requirement to measure inventories at the lower of cost and market). My definition of rules includes Nelsons (2003, 91) specific criteria, bright line thresholds, examples, scope restrictions, exceptions, subsequent precedents, implementation guidance, etc. The use of the terms principles and rules seems broadly consistent among Alexander (1999), Nelson (2003), Schipper (2003), and me. My purpose is not to investigate why the U.S. system tends toward the writing of rules (whether based on principles or not). Identifying the roles played by the existence since the 1930s of the Securities and Exchange Commission (SEC) as an enforcement agency and the perceived need of auditors to protect themselves from litigation by encouraging the setting of clear and detailed rules is left to Benston (1976), Zeff (1995), and future research. As discussed below, the IASB also frequently writes rules. Thus, my purpose is to evaluate how the failure to use the appropriate principles can lead any standard setter to rely too much on rules. As noted earlier, the imposition of rules has some potential advantages. Those identified by Schipper (2003) and Nelson (2003) include: increased comparability; increased verifiability for auditors and regulators (and a related reduction in litigation); reduced opportunities for earnings management through judgments (but increased opportunities through transaction structuring); and improved communication of standard setters intentions. Nelson (2003) and the American Accounting Associations Financial Accounting Standards Committee (FASC) (2003) review the literature related to these issues. FASC concludes: Concepts-based standards, if applied properly, better support the FASBs stated mission of improving the usefulness of financial reporting by focusing on the primary characteristics of relevance and reliability. (AAA FASC 2003, 74) (emphasis added) In addition to balancing the advantages and disadvantages of more detailed rules, the standard setters sometimes face competing principles. An obvious example is the difficulty of trading off relevance and reliability: for instance, estimates of current values or future cash flows might be potentially relevant data, but some such estimates have low reliability. Departure from one principle might be justified by the need to follow another one. Standard setters are also subject to political pressure, especially from the management of large companies (e.g., Hope and Gray 1982; Solomons 1978; Watts and Zimmerman 1978; Nobes 1992; Zeff 1997). Giving way to political pressure might be an explanation for departing from principles. However, a bad standard cannot be re-classified as a good one because issuing it enabled the standard setter to survive.

As noted earlier, my purpose is to identify several accounting topics for which the accounting standard could be improved by being based more closely on a principle from the conceptual frameworks. In some cases, merely removing a rogue principle that is not contained in the conceptual frameworks is sufficient. The improvements come in the form of increased clarity, decreased complexity, and decreased motivation for the structuring of transactions. That is, in some cases, increased clarity can be associated with a reduction in rules. This is not to say that principles-based standards are always clearer than rules-based standards. For example, development costs can represent an asset that meets reasonable recognition criteria; IAS No. 38 (para. 57) is based on this argument. In this context, the U.S. requirement (in SFAS No. 2) to expense development costs could be seen as an un-principled rule. However, in this case, the U.S. rule leads to a clearer instruction and to several resulting advantages (see above), although not necessarily to a better balance sheet. Because some accounting topics are not susceptible to solution by use of appropriate principles without rules, standard setters are then forced to choose, for example, between an unclear principle and a clear rule. However, I and most other authors quoted above do not welcome rules for their own sake. They should be kept to the minimum necessary to achieve the various advantages claimed for them, such as clarity. This warrants an examination of each accounting topic to see if a more appropriate principle could achieve the advantages of rules and yet reduce the amount of rules at the same time. As mentioned earlier, the use of appropriate principles can reduce optional accounting treatments, with a consequent increase in comparability. I am not talking here of judgments by preparers, but of overt optional methods in accounting standards. Optional methods are not prevalent in U.S. accounting standards, although some exist.3 However, several options continue to exist in IFRS even after the removal of many in December 2003. The options were needed to achieve a three-quarters majority on the IASC Board, but arguing for the options was easier in the absence of clear principles. Using appropriate principles does not guarantee a reduction in options, but the discussion below finds several instances where a focus on principles can reduce options.

WHAT IS THE MAIN POINT OF THIS ARTICLE? HOW TO SUMMARY THIS?