Fragility and Scale


There is an interesting interview over at Reason with Nassim Taleb, author of Black Swan, a book which presciently described the economic disorder preceding the financial collapse. I haven’t read the book, but there are some pretty arresting observations in the interview. Taleb argues that systems become too fragile if failure is not an integral part of them. Anti-fragility encourages disorder and a certain amount of failure because it allows systems to become stronger and more adaptive. A “too big too fail” banking policy essentially means that if a part fails, the whole system must come down with it.

To cite the great Yogi Berra, a good antifragile system is a system in which all mistakes are good mistakes. And the bad system is one, again to paraphrase Yogi Berra, where you tend to make the wrong mistakes. Let’s compare the banking system to, say, transportation. Every plane crash makes the next plane crash less likely and our transportation safer. Now, with the banking system, [a failure] leads to increased probability of failure of an entire system. That’s a bad system.

The overinflation of certain parts of a system occurs because of distorted goals for the system itself. A system committed to personal liberty will always emphasize that people are responsible for their actions, particularly if those actions are ill-considered. A more powerful system will insulate people against such consequences, but at the cost of their liberty.  For Taleb, all this relates to arguments for decentralization, for centralization distorts the otherwise socially-useful purposes of discrete organisms. Its attentiveness to liberty will help attenuate the free-rider problem (a problem so pervasive in our system it’s not even talked about):

What fragilizes an overall system? Three things: One, centralization. Decentralization spreads mistakes, makes smaller mistakes. Decentralization is where we converge with libertarians. A second one is low debt. The third is skin in the game.

Debt and centralization reinforce each other, and this reinforcement is driven by the defining purpose of centralized governments: war-making. The problem, he says, is one of size:

To make big mistakes and to be wiped out; this is the island effect at work. What we have had in this country is the progressive rise of central government. Particularly, deficits are the work of central government. [Scottish philosopher David] Hume figured it out. He said: Small states and city-states, they love commerce. And large governments love war. And that’s what justifies large government—war. There is no justification for large government other than war. And they’re not good at it.

It is increasingly evident that the key problems of our politics relate to problems of scale. Unfortunately, power acts in such a way as to distort such considerations. Its consolidating impulse requires that questions of size and scale – the bedrock questions of political theory going back to Plato and Aristotle – get disregarded altogether.



  • Share: