Third generations... were you thinking about Bismarck's "The first generation creates, the second administers, the third studies art history and the fourth degenerates completely"??
On self organizing processes and decentralization. The issue with the self organization you describe here is that most people, including political theorists and most certainly most countries' citizens and politicians, are deeply uncomfortable with the prospect of ever changing equilibria of indeterminate nature. People want some certainty. Even illusory certainty.
On federalism, and small, self contained experimental spaces that are safe to fail rather than fail safe: If you compare the US and the EU, you can see the trade offs. Centralization of presidential power in a federal system rather than remaining a decentralized confederation is what gave the US its unmatched ability to project its power globally. The EU meanwhile is often derided for its lack of single voice and inability of external power projection. On the other hand, yes the EU is a much better laboratory for diverse experiments.
I see a clear trade-off here between effectiveness and adaptability. The US has had the upper hand in many ways for the last good 100 years, comparatively speaking to Europe pre-EU and EU. The US' political stability goes back even longer of course, something we can't say about the EU yet due to lack of runway. And this in spite of the US civil war. That speaks for the US model. Then again as we now see, with its political makeup, the US may have been able to "win" comparatively speaking for a long time, but when it loses it may lose as a whole and very quickly. In this particular way, the US turned out surprisingly brittle. The EU loses smaller battles constantly but is too diffuse to be attacked, lose out, or decay completely as an entity. Its constituent parts may do all these things but elsewhere, something may clear up, fix itself, innovate, or grow. Interesting times.
The ideas on the use of randomness and the necessity of diversity are compelling. I think they are most characteristic not of an elected democratic system, but an allotted one. Replace elected representatives with randomly chosen representatives, who bring all the diversity of their varied lived experience to the table. Democratic lottery is not as radical as it sounds. It obtains in our jury system, in the original Athenian democracy, and in hundreds of citizens' assemblies around the globe, most recently in the Fort Collins, CO assembly in the US.
Indeed not. If we chose representatives by lot, it would not look or function like a legislature as we know it. More like the standing Paris or East Belgium assemblies.
Democracy is stable as long as elites believe in the system. When elites, as they are doing now, groom the voters that the system is bad, it becomes unstable.
This made me think of dictatorships are stable till the autocrat dies. Romania might be a classic, that schooled the elites that democracy might be a safer system.
Not even "belief" but merely their own self-interest in maintaining a stable system that put them at the top. It's the unspoken assumption throughout the American Constitution when it gets into specific requirements for holding office and voting.
So what are your thoughts on the stability (and adaptiveness) of the two main Democratic systems, coalition and two-party Democracies?
The Anglo-Saxon, mostly two-party world, had, until Trump, an almost flawless track record in coopting new ideas while remaining stably democratic.
Coalition Democracies now seem more stable to me (I must add that Hitler never got more than 33% of the popular German vote, so I don't regard his rise to power as a failure of the Democratic system in the Weimar Republic, but rather the failure of the leaders of the other parties to form a majority block against him).
And also, how do you evaluate Royalty? Being Dutch, I've always considered the Dutch Royal House, our Kings and Queens, who are our Heads of State, as a disgrace to the principle of Democracy, but the older I get, and the more turbulent the World, I now consider that folly a stabilizing factor, because it fulfills a need for exactly those people who are most prone to single man or party rule - to authoritarian rule.
For those people, I must add, who are least capable of understanding Democratic principles...
Hitler indeed never got anything close to a majority. He was able to grab state of emergency powers with the help of more standard conservatives (and violence against the left members of parliament). The nationalist/conservatives who helped him to power (financed by the then oligarchs) thought they could control him (and boy, were they wrong, because that 'control' assumed that the rules would remain the same). We might see this pattern repeat.
District based, especially the ones with 'first past the post' are more vulnerable than proportional representation systems. But they are by no means secure.
The idea of not "flattening the score" is not entirely original. Jazz musicians have been improvising on themes for decades. As Sun Ra used to say: "play what you don't know".
Electoral outcomes, when observed over time, display statistical regularities, yet these patterns are not temporally stable. While certain jurisdictions may appear structurally robust, the persistence of political equilibria is contingent on a range of economic and institutional factors. In this sense, stability is often illusory, a function of historical path dependence rather than an inherent characteristic of the system itself. It is particularly relevant to consider how shifts in political preferences can be conceptualized probabilistically, with different regimes corresponding to distinct probability distributions. Movements toward the tails of these distributions should be avoided, as they are typically dangerous and, more often than not, unidirectional.
1. Many of these descriptions of how democracy works (especially game theory) assume intelligent/rational players. But humans turn out to be not that intelligent, have convictions that steer their observations and reasonings far more than their observations and reasonings create their convictions. Instead, our convictions come much more from repetition and 'closeness'. In short: we are all potential flat-earthers.
2. Just as language comes from 'shared experiences' (Wittgenstein), culture comes from 'shared convictions'. Our culture is made up of those (vulnerable) (L)ego-blocks of individual convictions. If a society gets fragmented into parts that share little, society (which is based on 'shared convictions') fails. As less and less is shared between major f(r)actions, the foundation of 'togetherness' disappears. This is what we see happening, now. The information revolution plays a role here, but it already happened long before that got as influential as it is now — with positive-feedback from algorithms (frequency), in-your-face (closeness) 'influencers
3. Most people in western democracy still believe the security that a democratic rule-of-law fairness offers is worth a lot, this is an important shared conviction. But many also (and rightly so) see that this fairness is often more a paper thing than a real thing (e.g. outside influence of the rich). More and more of the (L)ego-blocks in our western societies lose the conviction that the system actually works in reality as well as it does on paper.
4. The big challenge seems to me how you create a system made up of vulnerable not-so-intelligent humans that is (a) open to good new ideas while (b) not so vulnerable to damaging or self-destructive ones. The simplistic idea that in a totally free exchange of speech the best outcome is the automatic result is very naive.
5. Societies that are increasingly built on purely individualistic principles will fail as these by definition undermine the 'shared' in 'shared convictions'. If individualism reigns supreme, autocracy or mafia/war lord/failed states are probably the outcome.
In the end it is about ethics and the question is: what convictions do we *share* on good versus evil, or right versus wrong? Add to that: the hedgehog-like messaging of autocrats ('we should protect society from terrorist gang members') is difficult to counter with a more fox-like messaging of plurality of ethical values. The solution? Maybe: find your inner hedgehog(s).
Henry, I hope you have sent a copy of this to Brian or one of his many layers, it would probably cheer him up, I could be wrong, but I detect a note of sadness in some of his most recent works.
There are two main kinds of system patterns, which old-school systems theorists called "feed-forward webs" and "directed attractors". I make a simple visual representation of each and juxtapose them, to compare the market system (a web) to various kinds of groups (attractors), including government. In this chapter-- https://www.youtube.com/watch?v=orP3HRJHufw --I illustrate the main failure modes of each (market failures, and organization failures of all kinds including government failures) and argue that good government will require a few dozen separate departments, each following "the rules for good rules." This all necessarily follows from the most basic fact that individual knowledge, time and attention are limited.
1) I don't fully understand where this idea comes from that democracy can be stable in the first place. Empirically, we don't have examples for it. Except maybe Switzerland, we don't have European democracies that are significantly older than a century, and while the US fits the definition, I wonder to what degree one can consider the first iterations democratic.
2) even if we assume stability modulo great shocks, I'd argue that the disappearance of the Soviet empire was a pretty big such shock that reverberates in all kinds of social policies.
This theory of democracy starts with the idea of the "loyal opposition". Wikipedia credits the term to Hobhouse in 1826, but I think the concept emerged in the late 18th century, once the association of the Tories with Jacobitism ended. It's only when the governing party is reasonably assured of personal and state survival after a change of government that they can be willing to relinquish power.
This is a very useful model. A relevant line of research by Henry Jenkins at USC found that a lot of Trump voters in 2016 had so completely lost hope, in a sense they could not imagine the future at all. This means they have no incentive to accept losses. The future didn't exist for them at all. https://www.civicimaginationproject.org/
I chatted with him about it circa 2019, and pointed out this other interesting paper on how PTSD manifests as a sense of foreshortened future. Time itself becomes unreal in a certain way.
Przeworski’s theory starts from a simple seeming claim: that “democracy is a system in which parties lose elections.” This is the start of trying to figure out and change things from where we are now This kind of thinking as a prelude to discussion can give us a different better foundation
Third generations... were you thinking about Bismarck's "The first generation creates, the second administers, the third studies art history and the fourth degenerates completely"??
On self organizing processes and decentralization. The issue with the self organization you describe here is that most people, including political theorists and most certainly most countries' citizens and politicians, are deeply uncomfortable with the prospect of ever changing equilibria of indeterminate nature. People want some certainty. Even illusory certainty.
On federalism, and small, self contained experimental spaces that are safe to fail rather than fail safe: If you compare the US and the EU, you can see the trade offs. Centralization of presidential power in a federal system rather than remaining a decentralized confederation is what gave the US its unmatched ability to project its power globally. The EU meanwhile is often derided for its lack of single voice and inability of external power projection. On the other hand, yes the EU is a much better laboratory for diverse experiments.
I see a clear trade-off here between effectiveness and adaptability. The US has had the upper hand in many ways for the last good 100 years, comparatively speaking to Europe pre-EU and EU. The US' political stability goes back even longer of course, something we can't say about the EU yet due to lack of runway. And this in spite of the US civil war. That speaks for the US model. Then again as we now see, with its political makeup, the US may have been able to "win" comparatively speaking for a long time, but when it loses it may lose as a whole and very quickly. In this particular way, the US turned out surprisingly brittle. The EU loses smaller battles constantly but is too diffuse to be attacked, lose out, or decay completely as an entity. Its constituent parts may do all these things but elsewhere, something may clear up, fix itself, innovate, or grow. Interesting times.
The ideas on the use of randomness and the necessity of diversity are compelling. I think they are most characteristic not of an elected democratic system, but an allotted one. Replace elected representatives with randomly chosen representatives, who bring all the diversity of their varied lived experience to the table. Democratic lottery is not as radical as it sounds. It obtains in our jury system, in the original Athenian democracy, and in hundreds of citizens' assemblies around the globe, most recently in the Fort Collins, CO assembly in the US.
It's a nice idea on the surface, but it seems not very far removed from anarchy.
Learn about it before dismissing:
https://www.youtube.com/watch?v=6UcFQ-eDhTk&t=3s
https://democracyrd.org
I can endorse the idea of citizens assemblies, but that's not the same thing as choosing random representatives.
Indeed not. If we chose representatives by lot, it would not look or function like a legislature as we know it. More like the standing Paris or East Belgium assemblies.
https://demnext.substack.com/p/how-a-permanent-citizens-assembly
https://www.publicdeliberation.net/the-ostbelgien-model-five-years-on/
Democracy is stable as long as elites believe in the system. When elites, as they are doing now, groom the voters that the system is bad, it becomes unstable.
This made me think of dictatorships are stable till the autocrat dies. Romania might be a classic, that schooled the elites that democracy might be a safer system.
Not even "belief" but merely their own self-interest in maintaining a stable system that put them at the top. It's the unspoken assumption throughout the American Constitution when it gets into specific requirements for holding office and voting.
So what are your thoughts on the stability (and adaptiveness) of the two main Democratic systems, coalition and two-party Democracies?
The Anglo-Saxon, mostly two-party world, had, until Trump, an almost flawless track record in coopting new ideas while remaining stably democratic.
Coalition Democracies now seem more stable to me (I must add that Hitler never got more than 33% of the popular German vote, so I don't regard his rise to power as a failure of the Democratic system in the Weimar Republic, but rather the failure of the leaders of the other parties to form a majority block against him).
And also, how do you evaluate Royalty? Being Dutch, I've always considered the Dutch Royal House, our Kings and Queens, who are our Heads of State, as a disgrace to the principle of Democracy, but the older I get, and the more turbulent the World, I now consider that folly a stabilizing factor, because it fulfills a need for exactly those people who are most prone to single man or party rule - to authoritarian rule.
For those people, I must add, who are least capable of understanding Democratic principles...
Hitler indeed never got anything close to a majority. He was able to grab state of emergency powers with the help of more standard conservatives (and violence against the left members of parliament). The nationalist/conservatives who helped him to power (financed by the then oligarchs) thought they could control him (and boy, were they wrong, because that 'control' assumed that the rules would remain the same). We might see this pattern repeat.
District based, especially the ones with 'first past the post' are more vulnerable than proportional representation systems. But they are by no means secure.
This is a nice post to have the day before the Australian federal election.
The idea of not "flattening the score" is not entirely original. Jazz musicians have been improvising on themes for decades. As Sun Ra used to say: "play what you don't know".
This is pure theory - make it a bit more applied: what should the US learn from the EU, Soviet Union or China?
Not to mention Denmark.
“‘It must have a responsive network of subsystems capable of autonomous behaviour’”
That’s SUPPOSED to be the Judiciary - hence the ongoing attacks against st it.
Electoral outcomes, when observed over time, display statistical regularities, yet these patterns are not temporally stable. While certain jurisdictions may appear structurally robust, the persistence of political equilibria is contingent on a range of economic and institutional factors. In this sense, stability is often illusory, a function of historical path dependence rather than an inherent characteristic of the system itself. It is particularly relevant to consider how shifts in political preferences can be conceptualized probabilistically, with different regimes corresponding to distinct probability distributions. Movements toward the tails of these distributions should be avoided, as they are typically dangerous and, more often than not, unidirectional.
By unidirectonial, do you mean unitopical, as in, for instance, solely concerned with immigration?
Hi, by “unidirectional” I mean that it’s hard to go back.
Interesting. A few remarks:
1. Many of these descriptions of how democracy works (especially game theory) assume intelligent/rational players. But humans turn out to be not that intelligent, have convictions that steer their observations and reasonings far more than their observations and reasonings create their convictions. Instead, our convictions come much more from repetition and 'closeness'. In short: we are all potential flat-earthers.
2. Just as language comes from 'shared experiences' (Wittgenstein), culture comes from 'shared convictions'. Our culture is made up of those (vulnerable) (L)ego-blocks of individual convictions. If a society gets fragmented into parts that share little, society (which is based on 'shared convictions') fails. As less and less is shared between major f(r)actions, the foundation of 'togetherness' disappears. This is what we see happening, now. The information revolution plays a role here, but it already happened long before that got as influential as it is now — with positive-feedback from algorithms (frequency), in-your-face (closeness) 'influencers
3. Most people in western democracy still believe the security that a democratic rule-of-law fairness offers is worth a lot, this is an important shared conviction. But many also (and rightly so) see that this fairness is often more a paper thing than a real thing (e.g. outside influence of the rich). More and more of the (L)ego-blocks in our western societies lose the conviction that the system actually works in reality as well as it does on paper.
4. The big challenge seems to me how you create a system made up of vulnerable not-so-intelligent humans that is (a) open to good new ideas while (b) not so vulnerable to damaging or self-destructive ones. The simplistic idea that in a totally free exchange of speech the best outcome is the automatic result is very naive.
5. Societies that are increasingly built on purely individualistic principles will fail as these by definition undermine the 'shared' in 'shared convictions'. If individualism reigns supreme, autocracy or mafia/war lord/failed states are probably the outcome.
In the end it is about ethics and the question is: what convictions do we *share* on good versus evil, or right versus wrong? Add to that: the hedgehog-like messaging of autocrats ('we should protect society from terrorist gang members') is difficult to counter with a more fox-like messaging of plurality of ethical values. The solution? Maybe: find your inner hedgehog(s).
Henry, I hope you have sent a copy of this to Brian or one of his many layers, it would probably cheer him up, I could be wrong, but I detect a note of sadness in some of his most recent works.
There are two main kinds of system patterns, which old-school systems theorists called "feed-forward webs" and "directed attractors". I make a simple visual representation of each and juxtapose them, to compare the market system (a web) to various kinds of groups (attractors), including government. In this chapter-- https://www.youtube.com/watch?v=orP3HRJHufw --I illustrate the main failure modes of each (market failures, and organization failures of all kinds including government failures) and argue that good government will require a few dozen separate departments, each following "the rules for good rules." This all necessarily follows from the most basic fact that individual knowledge, time and attention are limited.
Two quick remarks:
1) I don't fully understand where this idea comes from that democracy can be stable in the first place. Empirically, we don't have examples for it. Except maybe Switzerland, we don't have European democracies that are significantly older than a century, and while the US fits the definition, I wonder to what degree one can consider the first iterations democratic.
2) even if we assume stability modulo great shocks, I'd argue that the disappearance of the Soviet empire was a pretty big such shock that reverberates in all kinds of social policies.
This theory of democracy starts with the idea of the "loyal opposition". Wikipedia credits the term to Hobhouse in 1826, but I think the concept emerged in the late 18th century, once the association of the Tories with Jacobitism ended. It's only when the governing party is reasonably assured of personal and state survival after a change of government that they can be willing to relinquish power.
This is a very useful model. A relevant line of research by Henry Jenkins at USC found that a lot of Trump voters in 2016 had so completely lost hope, in a sense they could not imagine the future at all. This means they have no incentive to accept losses. The future didn't exist for them at all. https://www.civicimaginationproject.org/
I chatted with him about it circa 2019, and pointed out this other interesting paper on how PTSD manifests as a sense of foreshortened future. Time itself becomes unreal in a certain way.
https://pmc.ncbi.nlm.nih.gov/articles/PMC4166378/
The other piece that I had go up yesterday (a review of Klein/Thompson) is also about this - how to generate both variety and some reason for optimism about the future - https://www.combinationsmag.com/the-abundance-debate-were-not-having/. And I was really struck by this Liam Kofi Bright article from a couple of days ago, which looks to be bearing out in the UK local elections https://sootyempiric.blogspot.com/2025/04/more-stuff-and-fewer-people-to-share-it.html
Przeworski’s theory starts from a simple seeming claim: that “democracy is a system in which parties lose elections.” This is the start of trying to figure out and change things from where we are now This kind of thinking as a prelude to discussion can give us a different better foundation