Arnold Kling  

Unable to Simplify

A Simple Reply to Krugman on A... The Mystery of Libertarian Pat...

Clay Shirky writes,

Complex societies collapse because, when some stress comes, those societies have become too inflexible to respond. In retrospect, this can seem mystifying. Why didn't these societies just re-tool in less complex ways? The answer Tainter gives is the simplest one: When societies fail to respond to reduced circumstances through orderly downsizing, it isn't because they don't want to, it's because they can't.

In such systems, there is no way to make things a little bit simpler - the whole edifice becomes a huge, interlocking system not readily amenable to change. Tainter doesn't regard the sudden decoherence of these societies as either a tragedy or a mistake--"[U]nder a situation of declining marginal returns collapse may be the most appropriate response", to use his pitiless phrase. Furthermore, even when moderate adjustments could be made, they tend to be resisted, because any simplification discomfits elites.

When the value of complexity turns negative, a society plagued by an inability to react remains as complex as ever, right up to the moment where it becomes suddenly and dramatically simpler, which is to say right up to the moment of collapse. Collapse is simply the last remaining method of simplification.

Thanks to Laura Freschi for the pointer. Naturally, when somebody mentions the possibility of a system becoming overly complex, maladapted, and unable to simplify itself, I think of contemporary government in the United States.

COMMENTS (19 to date)
Mike writes:

My instincts run in the same dark direction as yours. I should stop reading Econlog early in the morning.

Floccina writes:

I think of the tax code but then I think computers can handle a incredible level of complexity.

chipotle writes:

Eventually the laws become so arcane and byzantine that they are impossible to follow.

It's not that we live in a Kafka-esque nightmare. We will live in a place of constant low-grade chaos. Less tear gas and truncheons and riots but more daily humiliation and passive servility.

To coin an analogy, it's less like a fatal heart attack and more like incurable chronic fatigue syndrome.

Can one die from prolonged enervation?

Lee Kelly writes:

The issue is not complexity per se, but the absence of adequate feedback to calibrate the system for long term survival.

The inadequate feedback creates apparent "complexity" because antiquited, outmoded, mistaken methods and counterproductive rules are not discarded but added to the pile. Sometimes these methods and rules once solved genuine problems, but rarely are they discountinued when the old problems cease to arise. (It is a complexity of rules, rather than of stuff.)

Whatever feedback exists instead calibrates the system to serve those who are politically powerful, i.e. the rule makers, and tends to entrench interests so that feedback good for long term survival is obscured or ignored.

It's like the situation with unions. Sure, they might get their members ten years of artificially high wages, but eventually they'll kill the goose that lays the golden egg, so to speak. Without intervention by a more powerful organisation, they cannot persist.

The most powerful organisation (usually, a government) cannot be "bailed out," but must collapse, because eventually long term oriented feedback will make itself felt--in the long term, at least. Nature cannot be tricked by obscure accounting methods and eloborate jargon--in the long run the selection processes will favour those systems well-organised for long term survival, and punish those that aren't.

The meta-rules of the system must induce feedback to keep the rules of the game simple and adaptive, (think adaptive efficiency vs. static efficiency).

Lee Kelly writes:


These ideas of yours are very Popperian in spirit.

In Open Society Popper critiqued the traditional problem of political philsophy: who should lead? Popper proposed a different problem: what institutional arrangements can be established so that bad leaders may be removed?

The shift is from a static problem solvable by intellectuals, to a problem of establishing rules and procedures to discover solutions through trial and error--a static and designed sytem vs. an evolving and adaptive system

SydB writes:

Having studied Tainter and his ideas for years, I think Mr Kling distorts them for his own libertarian anti-government bias. The central issue associated with the collapse of complex societies is not government. It is the fact that ways of living become ingrained and people as a whole cannot climb out of the local extremum--instead they double down, investing more energy in a lost cause. It's not regulations. It's infrastructure. And not just infrastructure created by regulation.

Collapse is a natural outcome of complex processes. Feedback is not the always the cure. Because feedback will be channeled into ingrained ways of being and living. It often requires a higher mind, e.g. a collective look, to overcome the local extremum, a higher mind that can plan beyond it.

E.g. if the government had not intervened in the recent financial crises, the system may have collapsed. Often it is government that prevents the collapse, smoothing out oscillations in the system.

I'm certainly not proposing government as the solution, but it sure would be nice if thinkers such as Mr Kling could read and relate the phenomena of the world in a way less distorted by their need to persuade and convince a particular audience, or to confirm some existing bias.

Mr. Econotarian writes:

Clay's examples (Roman Empire, Hollywood) are both examples of systems destroyed by lack of economic freedom.

Rome faced massive problems due to bad monetary policy (coinage debasement) and taxes.

Hollywood is beset by unions, as Clay pointed out. Thus Hollywood has responded with the reality show, where there are no union actors or writers to pay.

Doc Merlin writes:

@Mr. Econotarian

Reality shows still have a lot of writers.

Loof writes:

Relative to societal collapse consider nature’s retribution for monopoly via over-specialization as the dark side of innovation. Thumb rule: in stable times specialists-complexity out compete generalists-simplicity; in turbulent times, opposite. Overspecialized species become extinct in times of terrible trouble.

Ella writes:

SydB, it'd be nice if you read what Arnold Kling wrote rather than distorting it for your pro-government bias.

He didn't say *Tainter* was talking about the government. He said, very clearly, that he personally thinks of the US government whenever *anyone* mentions an "overly complex, maladapted, and unable to simplify" system.

Second - what makes you think the financial system isn't in collapse, anyway? The entire crisis in 2008 was government engineered and resulted in a government power grab. In that six month period, August 2008 - February 2009, the Fed and Treasury, between them, printed or "promised" about $4 trillion dollars publicly to financial institutions (and God knows how much under the table), which has permanently and negatively distorted our monetary system. There is an insane amount of instability and uncertainty in the market - new business opening are down about 30%, housing is down almost 50% (in sales), real employment has lost about 15 million jobs. And this is still the first stage of the recession.

So, how, exactly, did the government intervention help?

Grant writes:

FYI, Tainter's book is short, to the point and a quick read; the way I wish more books on ideas were.

He cites some examples of organizations successfully reducing their complexity instead of collapsing. Its been a while since I've read his book, but I believe most or all were in the private sector.

Bill Drissel writes:

Shirky assumes complex societies must have over-arching govts. For a primitive society with 100 or so members, the govt (chief) might make the best decisions for everyone. The more numerous and complex a society, the greater need for small govt so that minute adjustment can be made for each person's circumstances. Small govts lead naturally to flexibility.

Why can't complex, rigid govts re-tool? Because the ruling class would rather fail than give up their perks.

Look at what's going on in climatology now! Look at Europe and the steady decline because the ruling class can't give up the "welfare" state.

Bill Drissel

Troy Camplin writes:

The problem isn't with complexity, but with complicatedness. There is a huge difference, but nobody seems to pay attention to the fact that there is a difference between the two. A complex system is in fact very robust, and can take several large hits, let alone small hits, without much of an effect. It is simple systems that are prone to catastrophic collapse. A complicated system is another thing entirely. A complicated system gives the appearance of complexity, but it is in fact simple. Complex systems are full of generative paradoxes; a complicated system is full of destructive contradictions.

The etymology of each is telling: complex means folded; complicated means knotted. If something is folded, it has layers and multiple dimensions hiding under the apparent 2-D surface. If something is knotted, all you have is a 1-D structure pretending to be 3-D. The solution to the Gordian Knot? Cut it.

Tracy W writes:

Troy, out of curiousity, what do you call a complicated system that both appears to be complex and is complex but is prone to catastrophic collapses?

Or are you arguing that such a system couldn't exist, and consequently all complex-appearing systems that collapse must have in reality been simple?

Scott writes:

This reminds me of the concept of irreducible complexity that is often discussed in evolutionary biology.

Yet another example of how economics is more like biology than physics.

Troy Camplin writes:


That would be an intermediate system. If there are simple subsystems within a larger complex system, though, it can cause local collapses that can lead to more general collapses. They subvert the robustness of otherwise complex systems.

Tracy W writes:

Troy - thanks for answering my question - it seems a rather undramatic name.
I'm now pondering, if a complex looking system collapses how do you know if the cause of the collapse was a complex bit or a simple bit? After all, if the system is really complex it is possible that you have misunderstood it and have mistakenly identified the cause of the collapse as something simple. I was talking with a friend from engineering school who has spent the last six months trying and failing to eliminate a stray frequency from a radio device, he can't track down the cause of the frequency even though he can replicate it. Sounds to me like a complex cause for a simple failure. And he can fit a radio onto his desk at work, which you can't do with a whole society.

Troy Camplin writes:

Actually, your example if a prefect example of what I mean by a simple system that is apparently complex. There is nothing an engineer does or studies in school which involves complexity. All mechanical systems are simple, though they are beginning to become more intermediate in structure. A complex system has the following features: they exist on the edge of order and disorder, they are interconnected, they are self-similar regardless of scale, they have feedback loops, they have what are called strange attractors, and they have emergent properties -- patterns of behavior not predictable from the constituent elements. They have distributed structures, meaning they are very robust, meaning breakdowns in the system get dampened out more often than not. When they are simplified, somehow, they "die." That is the definition of a living cell dying -- it is simplified into its constituent chemicals. And it is usually done so from external, simplifying processes. Digestion after being eaten by a predator, for example.

Tracy W writes:

There is nothing an engineer does or studies in school which involves complexity.

As an engineering graduate myself, who has studied system and control theory, I note this assertion is wrong. I may have forgotten a lot about complexity, but I did study it.

A complex system has the following features: they exist on the edge of order and disorder, they are interconnected, they are self-similar regardless of scale, they have feedback loops, they have what are called strange attractors, and they have emergent properties -- patterns of behavior not predictable from the constituent elements.

In the case of my friend's radio device:
1. Existing on the edge of order and disorder strikes me as a good, if poetic, description of engineering, or indeed any other human endeavour.
2. Radio devices are interconnected systems. If they're not interconnected the electricity doesn't flow and the device doesn't work. This is how you can turn your radio off by disconnecting the power supply.
3. Self-similiarity is not a requirement of a complex system. And if it is, then no real world system is complex.
4. Radio devices have feedback loops.
5. This radio device has emergent properties not predictable from the constituent elements. In particular, this stray frequency that is wrecking it.
6. Most consumer radio devices are very robust. Your car radio can work in a ski field and in the middle of summer in Death Valley, and if it doesn't take it back to the manufacturer. Of course, my friend's radio device isn't robust, because of said stray frequency.

All in all, your ignorance of complexity here makes me suspect that you had no basis for your earlier claim that complex systems are all very robust. Plus, of course, that you haven't actually presented any proofs of your earlier statement that complex systems can't fail catastrophically.

Comments for this entry have been closed
Return to top