How often do we get to use big esoteric words from philosophy when we are talking about accounting standards?  Not often enough!

Usually, the only philosophical terms we use are ‘normative’ and ‘positive,’ and I think many of us are pretty comfortable with the notion that standard setters are trying to answer normative questions (what should we do), and that academic researchers can help out by providing positive evidence on what happens, how it happens and why.  (My thoughts on that in this post, called Unabashedly Normative).

Thinking more about the issue leads me to note a different but related distinction between two ways of thinking about the what standard setters should do.  In the philosophy of ethics (the mother of normative questions), people distinguish between deontological and consequentialist methods of answer the question of what people should do.

Deontology (the science of duty) looks to intrinsic definitions of good and bad actions without reference to the consequences of those actions.  For example, arguing that lying is bad because it is forbidden by the 10 commandments would be a deontological argument.  In contrast, consequentialists would argue that lying is bad because it causes bad outcomes (a lack of trust, which leads to the breakdown of society, etc.).

Discussions at the Financial Reporting Issues Conference make me think that standard setters start with a largely deontological view of financial reporting, with the duty imposed by the conceptual framework.  (I understand that IASB members sign a contract agreeing to pass standards that are consistent with the framework.)  The Framework’s asset-liability view imposes quite a bit of discipline on the process of determining standards.  Specifically, it provides a straightforward four-step approach to determine an appropriate accounting treatment for a particular transaction:

  1. Determine when assets and liabilities should be recognized and derecognized, according to the Framework’s definitions.
  2. Determine how to measure those assets and liabilities when they are originally recognized, and when and how to remeasure them.
  3. Determine how to assign net balance-sheet changes to income statement line items.  Does a change result in revenue or some other gain, or cost of goods sold, operating expense or loss?
  4. Finally, determine how to present the balance sheet and income statement.  On the balance sheet this often becomes a question of net vs. gross presentation of items that include both a debit and a credit (such as a forward contract, property with accumulated depreciation).  On the income statement, this is often a question of the degree of disaggregation, and the placement of the information.

This process is deontological, because it looks to the intrinsic nature of the transactions to determine what the standard setters should do, without reference to the consequences of the ultimate standard.  To some extent adherence to the deontological argument seems likely to result in good consequences — or at least one can say that deviations from this process might result in very bad consequences.  However, it does not directly address some of the most important consequences of standards.  For example:

  1. Will the standard result in better investor decisions?
  2. Will the standard distort commercial arrangements as preparers rewrite contracts to get better accounting treatments?
  3. Will the standard allow firms too much ability to manage earnings through distortion of unverifiable estimates and judgments?
  4. Will the standard result in such complex standards that users won’t understand them?
  5. Will similar transactions be treated so differently that users won’t be able to compare across firms?

As Jeffrey Hales pointed out in today’s roundtable discussion, debates over steps 1 and 3 tend to be most deontological because we have reasonably precise definitions of assets, liabilities and many income statement accounts.  However, 2 and 4 don’t currently benefit from definitions.  But that leaves the question:  are the current deontological arguments also leading to good consequences?

That is a job for positivists!