An unsettling argument in favor of mandatory auditor rotation

In his post “Thinking, Fast and Slow” about Auditing, Tom Selling raises a very disturbing argument for auditor rotation. Most unsettling idea I’ve seen on the issue.

I believe that auditor rotation is a bad idea.  A secondary reason is that I believe it will raise costs. 

The bigger concern, I perceive, is the time it will take for a new audit firm to gain a deep understanding of the client.  As I have taken on new clients over the years I’ve learned it takes several audit cycles to get a deep grasp of what is going on at an organization that is even modestly complicated. I can’t imagine how long it would take the managers and partners to understand all the moving parts of a new Fortune 500 client.

I realize there’s a host of counterarguments to my points and many ideas in favor of auditor rotation that I haven’t (and won’t) address.

Ironically, my economic interest would be to argue for mandatory auditor rotation. As an intentionally one-person firm, I would not be able to do all the work that would come my way.

Imagine mandatory rotation was required after X years. That means I would lose 1/Xth of my clients every year. That also means that 1/Xth of all the NPOs in my market niche would have to change firms every year. I wouldn’t even be able to keep up with all the RFPs that would flood my e-mail!  Even a modest success rate in proposing for new audits would mean I could pick and choose who I want to accept as new clients and increase my rates at the same time.  I’d never have to do marketing again.  So, on one hand, let’s go!

Seriously though, I believe mandatory auditor rotation would be a lousy idea.

And then Dr. Selling discusses the book “Thinking, Fast and Slow”.

The author, Daniel Kahneman, discusses his upsetting experience in grading essays.  Dr. Selling quotes three paragraphs. Here is my exquisitely short summary.

The author found that when he graded a student’s essay book from front to back that the grade on the first essay produced a halo effect on how he graded the subsequent essays. The subsequent grades were close to that of the first essay. When he graded the first essay for all students, then went back to grade the second essay for all students, he found that the grades varied widely from one question to the next. When he jotted down the grade for an essay and it was divergent from the previous scores, he was surprised to see the differences and felt an internal pressure to change one grade or the other.

The issue?  Halo effect. Confirmatory bias.

Dr. Selling suggests those concepts apply to auditing.

Here’s the concept in one run-on sentence: If you firmly and sincerely and objectively believe that management has been honest and accurate and competent in every accounting issue for the last 10 or 15 years, it will be exquisitely difficult to maintain your skepticism when you come across something that an outsider would think is a problem because you are unintentionally and unconsciously biased to think everything management does is okay.

I am somewhat protected from that phenomenon (I hesitate to call that a bias, so I guess that’s one of my blind spots, huh?).  While providing audits to clients in my market niche I typically find a number of little things that need to be corrected during the course of an audit so I don’t see a halo effect that the accounting is right.

And yet the other hand, I’ve seen a small number of cases of dishonesty and lack of integrity in my career.  Ahhh, let’s just say I don’t work with those organizations anymore. 

Oh, so does that mean there is danger of a halo effect for the clients I do work with, because I’ve never seen dishonesty in them?

Oh dear.  This more complicated than I thought.

Like I said, a very unsettling discussion.  Might be fun to expand this discussion to journalism or academia, but that’s not my area.

By the way, just loaded the book to my Kindle.  From the first page:

Questioning what we believe and want is difficult at the best of times, and especially difficult when we most need to do it…

Gonna’ be a good read.

Update:  I’ve resolved my discomfort.  See a series of posts that start at If auditors won’t audit, mandatory rotation won’t help. See part 2, part 3, and part 4.