Schumpeter’s Demon

For a while now, I’ve been dissatisfied with our shared mental models around the creative destruction being unleashed by the Internet.

On the one hand, we have coarse-grained and abstract models based on long-term historical cycles and precedents. This is the sort of thing I’ve explored quite a bit in previous posts. It involves careful analogies to previous technological revolutions. It involves debates around whether or not technological progress is stalling and whether a return to growth is possible.

On the other hand, we have detailed situational models, full of incomprehensible minutiae, that seem to develop around specific important decisions. An example is the  set of mental models that drove the “fiscal cliff” farce, which just played out in the US Congress.  Another is the set of mental models in evidence around the SOPA/PIPA debate last year.

The first kind of mental model is so large-scale in its concerns, it is effectively a fatalistic level of analysis. The other kind is ineffectually preoccupied with each immediate situation in turn. It quickly drives itself into a dead-end each time, and defaults to buy-more-time decisions.

I’ve thought of an allegory for understanding economic creative destruction, that I’ll call Schumpeter’s Demon. It just might be capable of informing meaningful action.

In thermodynamics, the allegory of Maxwell’s Demon is often invoked to illustrate and explore the subtleties of the second law. That sort of thing, I suspect, is the level of abstraction needed here. Especially since the operation of economic creative destruction is very similar to a thermodynamic process.

The Allegory

Imagine a town of a thousand people, with some mix of rich, middle-class and poor people.  The town has been ravaged by uncertain economic times.  The people live and work in a few hundred buildings: homes and workplaces. There is a general gloomy consensus that the town can only hold out for another year before nameless, inchoate horrors descend. But nobody is quite sure what those horrors are.

On this gloomy scene, a schizoid malevolent-benevolent demon appears.

The demon declares that after exactly one week, he will destroy half the buildings and kill half the people. The townsfolk can decide which buildings and people to sacrifice by marking doors and foreheads with red X’s. If they fail to do so, he will choose randomly.

But the demon also promises to leave behind a huge treasure as compensation, once he’s had his fun. He does not specify the nature of the treasure, beyond dropping a few hints about where he’s hidden it. But he promises that it will be enough to rebuild the town and its economy thrice over, put it back on the path to increasing prosperity, and raise more than enough children to replace the adults lost.

To make things more confusing, the demon throws in an exchange clause: the townsfolk can choose to trade three lives for one building, in either direction.

And to build in time pressure, he offers to trade time for either people or homes, at the exchange rate of an extra day for every additional home or every additional three people marked for sacrifice.

So what happens next?

A town meeting is convened. The rich generously supply cheap beer. Three basic conversations get underway.

  • The Futurists: One group ignores the immediate situation and furiously sets about debating what to do with the treasure, based on the little that is known about it through the demon’s hints, and the priorities suggested by the town’s woes.
  • The Situationists: The second group ignores the promise of treasure and furiously debates the question of which buildings and lives to sacrifice, based entirely on notions of fairness, values, rights and responsibilities as understood within the existing social order.
  • The Pragmatists: And the third group, the smallest, frantically tries to merge the conversations and talk about how to distribute the impending destruction in order to leave behind the social order best able to exploit the promised treasure.

Id-superego-ego in short.

While deliberations are in progress, a few start to despair of the the debate getting anywhere.

Some of them simply sneak off and camp outside the town limits. They pray that the demon’s random malevolence will not cross those limits, but that its promised benevolence eventually will. 

These are the wannabe freeloaders looking for a heads-I-win-tails-you-lose solution.

Others bravely decide to go seek the treasure itself, based on the demon’s tantalizing hints, and steal enough for themselves outside of the bargain with the town. They meet with some success, but can never be quite sure whether they’ve found a piece of the promised treasure, or something else. Some return to town with what they’ve found, and rejoin the deliberations.

The ones who return are the entrepreneurs. The ones who never return are the anarchists.

And still others assume that the townsfolk who choose to stay will be unable to engineer anything other than a worst-case outcome. They leave, and make plans to return and rebuild after the demon is done. They find themselves waiting a lot longer than they expected to.

These are survivalists of various sorts, permanently waiting for an apocalypse that seems to be taking its time.

No conclusion is reached by the end of the week, so the townsfolk hurriedly use the time extension clause, and buy another day by sacrificing the three most drunk people, who are too drunk to notice the X being painted on their heads. When they sober up and look in the mirror, despair descends.

The cycle repeats itself, a few days at a time, depending on the number of clueless drunks around. Occasionally if there are enough drunks from the same part of town, the more sober ones mark both the drunks and their buildings with X’s, buying a lot more time at once.

And all the while, people are also being born or dying in the natural course of events. Buildings are falling down and new ones being built, implicitly changing the terms of the deal with the demon.

How will the debate end?

I have concluded that the outcome has very little to do with the relative merits of the different arguments in play. It is driven, instead, by how much of the cheap beer different people drink, and how rapidly people leave the town following each failure to make meaningful decisions.

And perhaps most interesting, the demon does not need to exist for this drama to play out as described. People merely have to believe that the demon and the deal exist, and that the situation without the demon in the picture is heading towards an unspecified disaster.

Happy New Year!

About Venkatesh Rao

Venkat is the founder and editor-in-chief of ribbonfarm. Follow him on Twitter


  1. Bravo! This is a well-smithed and, if we are to be honest, painfully discomforting allegory for contemporary meta-observers.

  2. Given the intro, I expected the demon to be a very different sort of entity: a mere agent, without malevolence of benevolence or any overtones of supernaturality, used to illustrate how exceptional microeconomic circumstances might be argued to violate macroeconomic principles if they were to occur often enough.

    If you haven’t already read it, I recommend the book that used to be called Maxwell’s Demon (it’s now called Warmth Disperses and Time Passes) for a narrative of how macro-effects of thermodynamics were eventually reconciled with notions of individual atoms. Spoiler: information about individual atoms comes at a cost out of all proportion to those atoms’ influence on macrocscopic circumstances, and thus Shannon entropy has more than a cosmetic relationship to traditional ideas of entropy.

    The sort of demon I expected would, for example, accomplish unrealistically deft and rapid feats of HR, realty, and banking so as to eliminate any delay between destruction and creation. Your allegory might be good, but it was too nearly orthogonal to my expectations for me to notice one way or another.

  3. This post is exactly of the kind why I follow this site. Thnaks for the ‘hintsight’…

    And as for future times, may your reality be up to your dreams.

  4. I wonder why not everyone leaves the town, waits for the home destruction of the demon and returns with a 50/50 chance that ones own home hasn’t been destroyed or waits until most people have left and then marks someone else home with an X ( or as many other homes as possible )? There is little incentive to actually cooperate other than possibly building a gang, which commits violence before getting violated. No need for the beer to catch the fools.

    In actuality the destruction of the town is a collateral damage of the energy released by exploiting a new found treasure. Events happen in reversed order. SOPA will always be late and reactive. It attempts to remove the revolutionary disorder by re-establishing the old one which is already lost. The revolutionary moment persists until the new town is understood conceptually and then framed legally.

    The demon has been here and passed by. Its former malevolence/benevolence is the cause for retroactive debates, which attempt to justify its past actions and regain control. But be aware about the narrative distortion – a concept bought up by Taleb – which turns the demon into a historical necessity post fact.

    You too, a happy new year!

  5. Jay Chilcote says:

    I like this allegory, and where you seem to be going with it, that the demon is perhaps an invention to maintain a kind of low boil equilibrium of decision making (or delaying). I’m not clear enough on how you see the internet measurably changing the dynamics of creative destruction.

    It might be interesting to take the allegory all the way to an almost fairy tale extreme, showing the transformational difference the internet could make. In your scenario, some of those (the entrepreneurs) who’ve gone in search of the treasure could return with a curious magical device, a black plastic-and-metal hand-held object. The entrepreneurs set to work to reverse engineer it, quantities are made, long lines form and soon most of the sober citizens have their own portable magical device that allows for instantaneous inter-connectivity. How might the presence of such devices speed the creative destruction cycles in your allegory? How might that affect the decision making cycles at the town hall level? (Once upon a time, Time was relatively cheap…)

    You could also run a parallel analysis with a thought experiment based more on thermodynamics, as it relates to indirect agency? (Admittedly I’m much better at fairy tales than physics!). In your allegory there’s the town hall scenario, but at the macro level in our world such as for SOPA or the fiscal cliff, part of the frustration is seeing the imperfect and indirect influence linkage between willful agents. Could you play with the idea of a container divided into graduated chambers for poor, middle class, rich and then politicians? Perhaps the demons are more like arbiters, Editorial Daemons and Academic Daemons operating trap doors in the walls dividing the chambers, regulating the molecular movement of information between chambers and thereby influencing decision making. Suppressed, non-voting zones are essentially inert; more plugged-in ones tend to have higher temperatures. Perhaps the Editorial Daemons can only spot fast frequency molecules/information bits and allow them to pass bi directionally and individually (narrow but fast aperture). Same for the Academic Daemons, though they can only spot slow cycle information bits, and can only open the gate to broad groupings or summarizations of these bits (wide but slow aperture). The Academic gate is engineered to have a much weaker effect than that of the Editorial gate; in fact, for certain chambers, the Academic summarization signals barely signify. In addition, could there be Free Radical Filters capable of suppressing or even deleting information bits? These Filters might be Religion, Apathy, Alcohol and so on.

    The disruption represented by the internet in this scenario might again be the idea of disintermediation; the internet could enable a new form of P2P energy that pierces the barriers between chambers in new ways… going from a distributed model to more of a circulated model. What effect could that have on the influence equilibrium? This model might offer some extra dimensions to work into your original allegory.

    I really like your allegory and would be curious to see you develop it more on the generational, long-term level!

  6. I vote this for the best short story (business fable) of the year award.

    One vague fear I have is: if you churn out more of these you may tip over to the side of being comprehensible to tens of thousands of readers.

  7. I’m surprised there isn’t another group in your allegory: a few people who immediately leave the meeting and start marking other people’s houses and businesses with big, red Xs.

  8. Great story. Reminds me of something we used to say at one of my former employers: In a meeting , decisions get made either by whoever shouts the loudest or, failing that, whoever has the bigger bladder. (The ones with smaller bladders have to eventually leave the meeting room.)