Primarily PIL (Python Imaging Library). This is the guy who wrote 'Programming Collective Intelligence'. How he creates such graphs is well documented in that book. If you haven't read it already, I'd whole heartedly recommend it.
Fascinating! I wonder if this has predictive value, too -- if, for example, a board member who was on the board of a poorly-performing company predicts poor performance in the future.
Unfortunately, boards are pretty toothless. Lately, it looks like a single hedge fund with 5% of the stock and no board seats can cause more change within the firm than board insiders can -- on the other hand, the changes that happen internally may just be more subtle (e.g. if the audit committee gradually relaxes standards, it doesn't show up in the headlines until there's a disaster, and then it's hard to trace it to who is actually responsible).
"I generated a graph from the 400 largest companies by market capitalization. What’s shown here is the largest strongly connected component, which has 212 nodes."
Perhaps technology companies have more interlocked boards because many of them were funded by VCs. 100 years ago, GE would have been connected to all the other Morgan-backed companies.
I don't read Fortune because it seems to be written for middle-school students. I do know that lots of former GE managers are now running their own companies. However, it seems very likely that since most large tech companies were founded in the last 30 years or so, and many of them were funded by a relatively small number of VCs, that those VCs might form a very well-connected network, even if other companies also 'spun off' lots of good managers. I would be surprised if the number of Fortune 500 companies run by GE alumni exceeded the number of Fortune 500 companies funded at some point by Sequoia, Kleiner Perkins, Benchmark, etc.
You and I must be seeing two different things when we look at that graph. I see a pretty standard list of Fortune 500 companies, like Proctor & Gamble or Exxon. Some tech in there to be sure too, but no reason why GE would be excluded. And I'm sure they'd have more nodes than almost any non-tech company, even if your assertion is true.
It's interesting how Microsoft is far away from the other tech companies.
Does anyone know what was used to create the plot given the data?