Replication is one of the fundamental tenets of scientific inquiry, and the absence of replication in accounting is a non-trivial issue for initiatives such as FASRI.  Understanding the need for replication involves understanding the consequences of not replicating.  To illustrate, I was recently speaking with a faculty member about a NOVA episode (Intelligence Overload) on the use of classifiers by the National Security Agency to extract information based on context and meaning, and thus identify threats to the United States in the mountains of data they collect (think petabytes per month).  As I began to describe this NOVA episode, the faculty member I was speaking with immediately responded, “Wow, the consequences of a false positive or negative are pretty steep.”  While we were not talking about replication per se, I think the comment sheds light on the implications of accounting scientists seeming indifference to replication.  Either (i) we don’t believe our work has consequences sufficient to warrant replication or worse (ii) our work truly is sufficiently unimportant to warrant replication.

Considering the case of the FASRI which anticipates the use of academic research to inform policy, I think there is enormous downside risk to not replicating our results.  Given the economic consequences of accounting policy, basing a new policy on faulty or fabricated research could have disastrous consequences.  While I have heard numerous potential explanations for not replicating, I pose the following (possibly related) questions.  First, why do we not replicate accounting research?   Second, can our work be taken seriously by standard setters without replication?  Finally, if the lack of replication is a problem, does anyone have a solution?