This is a crosspost from https://250bpm.substack.com/p/controlled-creative-destruction.

Creative destruction is a process which lets the inefficient fail and the efficient survive. It's the economists' way of saying of "survival of the fittest".

In free markets, creative destruction reigns supreme. Firms build competing products. Consumers buy the best products. The firms that cannot deliver go bankrupt.

But it not just about money. Open source programmers work for prestige or joy, users choose the software to use based on what works best for them. Popular bloggers gain followers, others fade into obscurity.

Joseph Schumpeter, a Marxist economist who coined the term in 1940's, looked at creative destruction with a mix of awe and fear. Awe at its efficiency, fear that the destructive element may prevail in the long term.

The modern, colloquial usage is more in line with Daron Acemoglu, who believes that creative destruction is vital to economy and elites blocking creative destruction is a major force leading to economic stagnation.

So far, so good.

But sometimes it is the case that destruction of an institution would be extremely harmful. Think state. Nobody wants to live in a failed state and so the state is allowed to persist with no competition and no elimination of the underperforming. As time goes it accumulates dysfunction.

Once the situation becomes intolerable, revolution may occur. People rise up in arms and destroy the state. That's bad. No only is the disruption of the state often accompanied by bloodshed. Also, there's no telling how will the new institutions look like, whether they will be better or worse than the old ones. In history, there's no lack of revolutions that resulted in improvement, but also those that changed nothing or descended into chaos.

One possible way to deal with it is the libertarian one. The state is too big to fail, so let’s make it as lean as possible. Outsource as much stuff as possible to private enterprises, to NGOs, to voluntary associations. All of those are allowed to fail. Creative destruction naturally follows.

Yet, this approach only delays the inevitable. Even a minimal state will eventually accumulate dysfunction and face a potential revolution.

An alternative approach is to use what could be called "controlled destruction": The system is allowed to fail, but in a controlled, piecemeal manner, rather that collapsing all at once.

Some examples:

• Democracy: A change in government every four years wipes out the existing political patronage networks and starts the cycle of building them anew.
• Biological death: The genetically programmed death of an organism wipes out mutations accumulated in its lifetime, allowing only the germline to survive.
• IETF, the organization that creates standards for the Internet, forms working groups, where all the actual work is done, for a limited time. They dissolve automatically once the time runs out, preventing concentration of power and scope creep.
• Jubilee year: Some ancient societies dealt with the problem of exponentially growing indebtedness by cancelling the debts once every few decades.
• Different sunset clauses.

I don’t think we really have a theory of controlled destruction. Developing such a theory, however, would require understanding why the institutions, those little monarchies without internal competition creative destruction, exist at all. Why firms, for example? There's the Coasean theory that it's all about transaction costs. It is also often pointed out that big firms or monopolies do better at R&D (Bell Labs, Xerox PARC, etc.) But why would that be? Or, consider the bureaucracy of a state: The reason why we don't want it to fail may be that there's too much accumulated tacit knowledge that we don’t want to lose.

Once we know what the benefits are, maybe we can think of various systems that would periodically wipe out the dysfunction, while still preserving those benefits.

In any case, the mechanisms of controlled destruction are an intriguing and possibly important topic to think about.