I know, I know: France and the Netherlands saying no, Deep Throat coming out of the closet, he-said-she-said about the Koran in the toilet, and here I am talking about accounting. Arthur Andersen, after many long years, has finally been acquitted. Am I the only one to remember this and care?

This is a funny one. Andersen no doubt played an important role in the Enron fraud, and the Feds desperately needed to take somebody down. This was a surprisingly difficult challenge, because the fraud was somewhere between 95% and 99% legal. But at the time, I was pretty well convinced by various commentators that Andersen's conviction wasn't legit. A few points:

• There was only one conviction of a person, the guilty plea by Dave Duncan, the chief auditor on the Enron account. Prosecutors can apply a tremendous amount of pressure to individuals to get a plea, and just because someone caves, does not mean that they necessarily did what they pled to.
  
• Despite the plea, and the protection against further prosecution, Mr. Ducan nevertheless testified under oath that his auditing was 100% correct and legal. That can't have made the Feds happy.
  
• Andersen was accused of illegally destroying documents when the SEC investigation was imminent but not yet announced. In order for the organization to have illegally destroyed documents, there must have been a person in authority in the organization, who knew it was wrong, who persuaded others to carry out the act. This is the "corrupt persuader". No corrupt persuader, no corrupt organization, as a matter of law. The Feds, however, never had a straight story on who, exactly, they believed this corrupt persuader to be.
  
• The question of the identity of the corrupt persuader was heavily argued in trial. Defense said the prosecution had to name somebody and prove it, prosecution said they could name a whole bunch of possibilities, and let each individual juror pick their own favorite. As long as everybody on the jury thought somebody-or-other did something bad, the organization could still be guilty, even though the jury could not reach a consensus on who it was that did the bad thing. The judge eventually allowed this final compromise: every juror could pick their own bad person, but they had to be reported to the judge, so that this question could get hashed out in appeals court.
  
• The jury couldn't reach a consensus on the corrupt persuader thing. They could, however, all agree that a particular person had edited a memo, and convicted based on that. Case closed, we can go home.
  
• Only one hitch: editing a memo, or at least that memo edit, was unquestionably legal. Not even the Feds had suggested otherwise. Prosecution: "Guilty means guilty, don't matter what they were thinking, nyah, nyah!" Defense: "WTF?" Thus the appeal.
  

Slate has a piece on the misunderstood business of pharmaceutical sales and promotion. It's actually not all bad. There are a few howlers, of course. It's scrip, not "script", naproxen, not "naproxyn", pharma marketing expenses are wildly overstated because free samples are booked at full retail price, not the pennies-per-pill cost to the company, and the COX-2 story is far more complicated than the MSM even seems to hint at. Still, it does a good job at discussing what pharmaceutical reps do, and more importantly, how their outcomes get measured. Database fans will note that the essential operation is a join, carried out on data acquired at great cost from disparate sources, none of which in itself seems to be particularly questionable, but the combination of which is highly useful. Intelligence gathering at its finest, and it tells you what tactics work on which doctors.

In other drug sales news, I learned today that Lipitor, the world's #1 drug, and the only blockbuster turning in more than $10B/yr in sales, is not in fact the world's best selling medicine. Erythropoeitin, mostly used to combat anemia after chemo, got almost $12B in 2004 sales total divided over all the companies which sell it. Biotech has normally been thought of as the poor stepsister to pharma, and overall it still is, but things are coming along. I remember when a mere billion dollars of sales counted as a blockbuster, but I don't think that's true anymore.