I presented FASRI’s report on Financial Statement Presentation to the IASB (last Thursday) and to the FASB (yesterday).  The report is based on an experiment conducted with Frank Hodge, Pat Hopkins and Kristi Rennekamp.  Several Board members seemed pleasantly surprised by the ability of the research to address the issues being deliberated, and are clearly hoping FASRI will generate a lot more of it.  FASB’s Larry Smith seemed particularly surprised to have been converted from his rather skeptical position on the benefits of academic research; Leslie Seidman also had nice things to say about both this project and the Staff’s field test, conducted by Regenia Cafini.

But there was a bigger surprise for me, on a point that was tangential to the ones I was making.  (See here for a summary of the report).  Members of both Boards seemed very interested in my suggestion that it is reasonable for investors to infer that information on the face of financial statements is more reliable than information that is merely disclosed in footnotes.  This belief comes partly from the Conceptual Framework, which emphasizes that events are not recognized on the face of statement sunless they are sufficiently relevant and reliable (or representationally faithful, if you prefer).  But the show-stopper, which resulted in questions during both presentations, came when I pointed out that the belief is also supported by a study by Libby, Nelson and Hunton (JAR, 2006. 44(3):533-560), who showed that auditors are more likely to waive known errors when they affect a number presented in a footnote rather than on the face of a statement. Members of the IASB interrupted to point out to me that such a result is clearly inconsistent with a variety of U.S regulations, and wanted more clarification. Leslie Seidman of the FASB also wanted to hear more.

Well, here is a little more:  While flying out of Ithaca this morning to a conference in San Jose, I happened to run into Laureen Maines, who had just presented a paper at Cornell on the same issue (co-authored with Shana Clor-Proell).  In the context of a larger experiment, which I won’t go into here, Laureen and Shana asked a number of CFOs and others with corporate reporting responsibilities the following questions, assuming that the FASB is debating between two accounting standards.  Standard R requires that a contingent liability like the one the firm faces be recognized on the face of financial statements, while Standard D requires that it merely be disclosed in footnotes.

  1. How would the amount of effort that you exert to estimate a contingent liability differ under the two standards?
  2. The FASB defines reliable information as information that is reasonably free from error and bias and faithfully represents what it purports to represent.  If the FASB concludes that contingent liability estimates are reliable, then how does this conclusion affect the FASB’s choice between Standard R and Standard D?
  3. How do you think the FASB’s choice of standards will affect financial statement users?

All three results show statistically significant indications that the participants would exert more effort under standard R (Q1), would expect the FASB to be more likely to choose Standard R (Q2), and would expect investors to attend more to the information under standard R.

Despite whatever US regulations may say about the reliability of footnotes and financial statements, I am not surprised.  Are you?