Hard to go past last week's piece Trusting Bosses Not to Cheat by Floyd Norris in the New York Times. You can read all of it here.

Seems that for all the fear and loathing about Sarbanes-Oxley, the costs are actually harder to measure than we first thought.

Norris casts a critical eye on the research by Ivy Zhang, an assistant professor of accounting at the University of Minnesota which has been used as ammunition by the critics of Sarbanes-Oxley who claim it's had a trillion-dollar negative impact on the US economy.



"Ms. Zhang did estimate that consideration of the law cost shareholders almost that much. But now she has backed down. The latest version of her paper, she told me, indicates a far smaller effect, although she declined to put a number on it.
"Still, a few hundred billion is a lot. But to me, her argument seems a little disingenuous: if the market went down when Sarbanes-Oxley was being considered, that was the cause. If it went up, something else happened.
"Consider the day the House-Senate conference committee reached agreement on the law, with Republicans backing down from efforts -- reaffirmed the previous evening -- to weaken some provisions. The market soared that day, but she figures that is irrelevant because everyone knew a tough bill would pass.
"But she notes that the market opened down slightly -- and attributes that to the arrests that morning of executives from Adelphia, the cable company.
"How does she deal with the fact that the market bottomed in the fall of 2002, about the time efforts to enforce Sarbanes-Oxley got under way, and has had a sustained rise since? That is probably irrelevant, she told me, since the market would have already discounted the effects of the law, and whatever happened later was caused by something else.
"Explaining market moves is not so easy, and her paper would be unworthy of much attention but for its use by opponents of regulation.
Section 404 of Sarbanes-Oxley, requiring audits of internal controls, has cost more than anticipated. That could mean internal controls were in bad shape, or it could indicate that auditors are doing too much work. Both are probably correct.

David Weidner makes a similar point, that trust costs a lot to verify and it's hard to measure, in his piece in MarketWatch.

"U.S. investors deserve to know that companies, boards and senior executives stand behind their results.
"Would Sarbanes-Oxley have prevented Enron? No one knows. But come the day when Lay and Skilling receive years in jail for building the Enron fraud, you can bet that somewhere in the back of their minds they wish something or someone with more teeth than Sherron Watkins had tried to stop them."

Tom Peters has pretty much the same approach in his take.