The Enron Emails As Pseudoscience

It seems to make some intuitive sense, that scientists could look at the frequency, contacts, and content of emails sent through an organization and come up with a formula, a pattern, the hidden gem in all of it that would explain why Enron collapsed. Since there were no smoking guns in the emails, perhaps it was understandable that the search for one moved not to what was said, but who said it to whom, and how frequently. But that desire to know is not an excuse to be sloppy – which is exactly what happened with the Enron dataset.

The latest is this article, published on a site that calls itself Science Alert.

Changes in an organisation’s email traffic are a potential predictor of looming crises, an analysis of emails leading up to the infamous demise of Enron Corporation has found.

Researchers from the University of Sydney’s School of Civil Engineering analysed more than a quarter of a million emails from the now defunct energy, commodities and company, spanning three and a half years leading up to its 2001 bankruptcy.

The researchers found changes in emails patterns, a representative gauge of an organisation’s communication networks, started changing in the 12 months leading up to the company’s demise. Around this time, staff began interacting more frequently within small groups of peers more rather than relying on messages emanating from the company’s upper echelons.

I wonder if this might be the natural result of cliques becoming closer? Or the fact that Enron had done some consolidating — perhaps organizationally people were just closer together than they were before. And even if it was true, what does that indicate?

“We found it went from a highly centralised network to a largely decentralised and distributed system,” says Professor Liaquat Hossain, co-author of Understanding Communication Network Cohesiveness During Organisational Crisis.

“We saw structural changes in email communications in the 12 months before the crisis intensified as its CEO Jeff Skilling resigned and it started reporting earnings losses.

An earnings loss was reported in October 2001 (I am thinking of the write-down here, since Enron didn’t have an actual official quarterly earnings loss). It was also the time when Andy Fastow departed the company after having been unable to rally the confidence of Wall Street, and fudging the numbers of his compensation from LJM. So common sense says there would be plenty to talk about, and you probably wouldn’t go to your boss with your gossip.

But there are other reasons to increase conversation too. I am just confused why an increase in emails is associated with bad things.

Maybe I’m being too critical. I just think this is silly. Where is the science in this? What do we know from this study that we didn’t know before?

“As Enron approached disintegration in late 2001 more people communicated with their colleagues and at a higher frequency. During this peak crisis period, there was also a jump in the number of cliques forming within the company. This increase in communication is consistent with organisational theory that purports cohesiveness is greater under conditions of great anxiety.”

Professor Hossain says the paper’s findings suggest changes in communication patterns can predict potential crises in an organisation.

“If an entity’s email traffic is monitored – without breaching privacy – we think it’s possible to detect from changes in communications patterns perceptions of an emerging crisis. By identifying this early an organisation is better able to reshape its strategy in order to prevent a perceived crisis evolving into a debilitating one.”

Hm. Well who is going to be responsible making sure that emails are within “acceptable” boundaries? And what happens when they do increase? Do you demand an internal audit? Do you assume something has gone awry merely because coworkers are sending emails among each other?

I think this is a pseudoscience – a carnival trick, another way to indict Enron for something it didn’t do.

“Like The Enron Emails”

I’ve heard this statement for the last week or so in comparison to the Wikileaks documents and in speculation of what the Bank of America documents hold. But I think the general public has already forgotten what the Enron emails actually showed. Namely, nothing. Not one count in all the hundreds of aggregate counts was proved by email. Jeff Skilling didn’t even use email. His assistant wrote his emails for him. The DOJ would later view this practice as indicative of guilt (ie, that Skilling must be hiding something.)

The Enron emails did not prove particularly embarrassing. One mid-level Enron employee was sleeping with two women at Enron, and Ken Lay’s daughter asked her father to find a job for a friend, but that hardly rose to the level of scandal. Basically the emails were a peek into the day to day professional life of people who worked there. There were no huge surprises.

I’ve posted numerous emails that were not included in the dataset – mostly in an attempt to add some color to the personalities. But there are no amazing emails of two people conspiring or talking about their great conspiracy. They don’t exist. And nothing was ever written down that would imply there was anything illegal happening at Enron.

Whatever Julien Assange has on Bank of America, it has to be juicier than what we’ve seen with the Enron emails. The comparison is not apt, though I think that it’s funny that people still use Enron as a codeword for scandal. How wrong they are. How utterly normal Enron was.