I've always had doubts about the whole “small world” thesis, feeling that Milgram’s 1967 study didn’t quite warrant the conclusion that any two people in the world are six degrees apart. After seeing Milgram cited once again today, I got curious about whether anyone have attempted to replicate the study recently and on an international scale. After a Google search, it sounds like a team at Columbia tried to replicate the study a in 2003, asking over 60,000 people to forward messages to 15 targets around the world. They got the same average of around 6 for completed chains, though – just as in Milgram’s study, most messages never got delivered. (In fact, apparently, only a bit over 300 messages got delivered – a success rate of just 0.5%) The study thus suffers from the same problems as Milgram’s original study.

While googling, I also came across an interesting article by Judith Kleinfeld which seemed to capture some of my doubts about Milgram’s experiment and support it with some evidence from Yale archives. Most interestingly, Kleinfeld cites a less known study from the 1970s that found that while middle-class networks may form somewhat of a small-world, low-income people are largely cut out of those networks.