Back in 1967, the social psychologist Stanley Milgrim published a fascinating article in the very first issue of Psychology Today called “The Small World Problem.” Milgrim was already famous for his shocking (literally) “obedience to authority” experiment, in which regular people zapped strangers with increasingly powerful electrical shocks, all in the name of “following orders.”
But Milgrim’s 1967 article came to an equally startling conclusion: Every person in the world was connected by an average of six people. A random person in Iowa could be connected to a complete stranger in Belarus with a chain of six or fewer acquaintances. It was a small world, after all!
Milgrim’s study made waves in the ’60s, but it really launched into mainstream culture with the 1990 play “Six Degrees of Separation” and subsequent 1993 film starring a young Will Smith. Since then, the idea of “six degrees of separation” has been accepted as a hard-and-fast rule of social networks. But how accurate is it?
The Flaws in the Original Experiment
Milgrim’s original 1967 experiment worked like this. He recruited “starters” in Kansas and Nebraska and instructed them to deliver an envelope to a “target” individual in Massachusetts. They were told to mail the envelope to a close acquaintance who had a better chance of knowing the target. The acquaintance would do the same thing until the envelope reached the target.
Milgrim reported that “chains varied from two to 10 intermediate acquaintances, with the median at five.” If there were an average of five intermediaries between strangers, then they were connected by six degrees.
But when you take a close look at Milgrim’s data from that first envelope experiment, the results were less than stellar. Of the 60 envelopes he mailed to the “starters” in Kansas, only three ever made it to the intended target. That’s a 5 percent success rate. And worse, they passed through an average of eight people, which is nine degrees of separation.
In the follow-up Nebraska study, Milgrim was able to achieve a 30 percent success rate and an average of six degrees. But Judith Kleinfeld, a psychology professor at the University of Alaska, wrote that Milgrim used a “passport” made of thick blue cardboard with the words “Harvard University” embossed in gold on it as the document that needed mailing, which might have increased the odds that the senders would try harder to find someone to take it. She also noted that most of the senders and targets were upper-income and therefore more likely to have a wider network of acquaintances than lower-income people.
Kleinfeld called the evidence from Milgrim’s landmark 1967 study “scanty” and wondered if six degrees of separation was just an “academic urban myth.”
Math Shows It’s a Small World After All
Decades before Milgrim posed his experiment, mathematicians were wrestling with the “small world problem.” They wanted to figure out the mathematical odds that two complete strangers would have an acquaintance in common. “You also know Brianna from Boise? What a small world!”
Milgrim’s social experiment seemed to prove that “small world moments” like those weren’t so rare or unexpected, but mathematicians still couldn’t explain how it worked.
It wasn’t until the late 1990s that two mathematicians built a computer model that successfully replicated Milgrim’s results. The trick, they discovered, is that most people have one or more “long-range connections” in addition to lots of local connections. One of the mathematicians, for example, was a professor in New York, but played chess online with a friend in Holland. Therefore, all the mathematician’s New York friends were just two steps away from everyone his Dutch buddy knew.
Modern experiments using email message chains (who sends letters anymore?) have confirmed that any two people in the world are connected by a chain of six to seven acquaintances. And in 2011, Facebook found that its billions of users were connected by an average of 4.57 “friends” or 3.57 intermediaries. Facebook called this “3.5 degrees of separation.”
Milgrim’s early experiments might have been flawed, but they hinted at mathematical truths that now form the basis of “network theory.” The surprisingly tight interconnectedness of our social networks — both in the real world and online — explains how COVID-19 exploded into a pandemic, or how misinformation can go viral and influence presidential elections.