Shirky on Complexity

Clay Shirky has written another essay about the future of media. It's main point is that big established businesses will not adapt to the new media marketplace because it's actually impossible for them to reshape their large corporate structures to meet the new need. That's a good argument, but a familiar one - it's the essence of "disruptive innovation", and it's something Tim B Lee has been saying for a while.

What's new to me, though, is the analogy he introduces the idea with. He cites Joseph Tainter as arguing that a similar process applies to societies - they become more productive by being more complex, but when they become overcomplex, it's not possible for them to simplify instead, and they must eventually collapse.

"Tainter’s thesis is that when society’s elite members add one layer of bureaucracy or demand one tribute too many, they end up extracting all the value from their environment it is possible to extract and then some"

That's horribly persuasive. It needn't even be, as Shirky puts it, that the elites add more complexity beyond the point where it adds value. If circumstances change so that the previously optimal level of complexity is now excessive, the result is the same.

If there's one reason why libertarians tend to be in software, it's that software is more complex than other things humans design (since it doesn't have to be actually built), and that programmers are therefore more aware that complexity is a cost. The biggest cost of adding a feature to a piece of software is not the time you spend making it, it's the fact that your software is now more complex, and everything else you do with it in future is made more difficult by that complexity. Similarly, the biggest cost of adding a government program is not what you immediately spend on hiring people to do it, it is that you have made government bigger, in a way that is almost impossible to reverse when changes in the world or changes in what you want to do demand it.

Of course, saying that government cannot be simplified at the margin is just another way of saying that libertarian politics cannot be successful. The only people with any approach that can succeed in the face of Tainter's theory are these guys.