I made one of those serendipitous discovery clicks on Twitter today that lead me to a short sponsored video exposition on the nature of social networks and their effect on the economy. The conversation is sponsored by Dow Jones, so it helps put the comments in the video and the reason for them being publicised into perspective, but all in all is a decent take and worthy of taking a look.

The video is by James H. Fowler, a behavioral economist, and starts out talking about a concept called Homo Economicus, which he explains as “a bunch of independent decisions leading to an connected outcome.” He attributes this behavior in modern day to the greased wheels of electronic communication in general, and in part to social networks in particular.

He then goes on to say that these decisions that affect economic swings controlled by social networks, because these networks “govern just about everything that humans see, think, feel or do.”

He goes on to say that from an economic perspective, this is a bad thing because herd behavior tends to amplify emotional and ultimately damaging decisions during times of crisis, both in terms of the general public as well as the key influencers and decision makers at the top of the economic food chain.

There are a number of interesting in valid points throughout Fowler’s analysis, but I want to quibble with one key point: I firmly do not believe that social networks are “governing” human decisions.  Yes, to a certain extent, there is herd behavior (a topic I explored in depth during the 2008 primary season).

Increasingly though, due to the mainstreaming of social networking, the concept of Seth Godin’s tribes concept we’re seeing decentralized monocultures or pockets of polarized opinions.

The Evolution of Digg in 2007 was a Precursor to the Current and Future Real Time Web

In my 2007 piece on Digg’s divergent path from their original vision of utilizing the wisdom of crowds, I explained what the original concept was and how Digg was pursuing something that mitigated the sociological precept:

I’m a firm believer in the concept of “wisdom in crowds,” one of the founding principals of Digg. It has very solid statistical roots, is a very old concept, and in most cases work.

The anecdote that is the genesis for the concept of wisdom of crowds I’ve heard many times over is the story of scientist and statistician Francis Galton from the late 1800’s, who was surprised that the crowd at a county fair accurately guessed the butchered weight of an ox. What made it interesting was not that any one individual came close to guessing the actual weight, but that the crowd did. When their individual guesses were calculated to the median, the resulting number was much closer to the ox’s true butchered weight than the estimates of most individual crowd members, and perhaps most surprisingly also closer than any of the estimates made by ‘cattle experts.’ Bringing it around to Digg, the idea is amongst the 21.5 million unique users to Digg per month, the wisdom of the crowds will tend to dictate that the most important, relevant, and interesting stories will rise to the top.

There’s a distinct difference, though, in what Digg tries to utilize the Wisdom for, and how the Wisdom was discovered in the first place. On Digg, what answers are being solved? They’re hypothetical question with only subjective answers. In Galton’s case, where the Wisdom was discovered, it was a very objective answer that was being sought. The answer wasn’t what most folks would like the butchered weight of the ox to be, it was what it actually is.

Here’s my point: if you pull your entire crowd from which you wish to derive Wisdom from the same monoculture, it will skew your median results toward a place that is very unwise. Based on the strategy Adelson divulges in the Guardian interview, everything Digg does to upgrade the system moves towards further niching and herding of it’s users into smaller and smaller monocultures, thus further diluting the system’s Wisdom.

The behavior that Digg was designing their system to enhance has naturally occurred to a certain extent on Facebook, but to a much greater degree on systems like LinkedIn, Twitter and Buzz. Early users of these systems set the trend on how they’re now being used, and tended to add people to their networks that are professional or ideological contacts, rather than friends, family, or geo-centric additions to their social graph.

As such, these smaller circles of “friends” have their own memes, herd-behavior and decisions that bleed over to one another, but rarely is there a top down and centralized herd behavior where the impetus can be directly attributed to the larger network as a whole. 

In other words, most trends spotted by those looking and analyzing the firehose can be more accurately described as correlation, not causation.