Hierarchy and wings

by Benquo2 min read6th May 201922 comments


PoliticsSocial StatusWorld Modeling

(Note for LessWrong: This is more overtly about partisan politics than the norm, but I think it's not more about that than The Two-Party Swindle, from the Sequences, and it proposes a structural model that doesn't require people to be as stupid.)

There are a few points I didn't make in my post on blame games because they seemed extraneous to the core point, which are still important enough to write down.


The Hierarchy game is a zero-sum game in which people closer to the center expropriate from people farther from the center, and use some of those resources to perpetuate the power imbalances that enable the expropriation. Players that fail to submit to expropriation by higher-level players are punished by those more-powerful players, often through intermediaries. Players that fail to help members of their class expropriate from those beneath them are excluded from their class, and often coordinated against more overtly.

This game isn't inherently majoritarian, - instead, it allows smaller groups to stably expropriate from larger ones, because every player has a short-run incentive to go along with the arrangement. Feudalism is a simple example of the hierarchy game. Modern states almost always have some hierarchical arrangements, such as the police and military, and (less formally) economic class. 

Political handedness

Around the time of the French revolution - a replacement of Feudal arrangements with Modern states - people started using terms like "left" and "right" to refer to political orientations. These terms are related to natural structural coalitions within a modern democratic state.

Political parties don't overtly promise to expropriate from 49% on behalf of an arbitrary 51%. This is probably in part because this would be correctly viewed as a proposal to massively increase expropriative activity relative to other activity, accelerating the rate of expropriation, which actually isn't in the majority's interests, and would quickly undermine the democratic paradigm without providing a replacement to enforce property claims. Instead, appropriation is opportunistic, and political coalitions seem to be oriented around natural power bases which could in principle replace deliberative democracy.


One natural sort of organization to orient around is the formal hierarchy with a monopoly on force - the military and police. The staffing needs of these organizations are substantial, especially in wartime (democracies perform well in wars in part because of their ability to mobilize a large share of the population without destabilizing their internal political arrangement) so they already form a natural constituency.

The obvious advantage of control over these organizations is in the event of a civil war, control over the army and police would be a massive advantage. So, building a group identity around those things is a pretty plausible way to expropriate the country from the other half.

The "right wing" is the part of the political spectrum that most resembles or is most naturally allied with this coordination strategy. Generally, if there's an identifiable majority group (ethnic, cultural, religious, etc.), the hierarchy of violence will perceive members of that group as more "central" and want to help them expropriate from minority groups more than vice versa, insofar as gaps in the rule of law allow this. People rewarded by the existing credit-allocation, the "upper classes," will also tend to favor and be favored by this arrangement.


The "left wing" is the natural complement to this strategy: a political "big tent" made up of all the noncentral groups. Such a coalition has a structural advantage as long as democratic institutions persist, since any new group (e.g. immigrants) that isn't part of the majority is a natural member of the "left wing" coalition. Such groups also tend to seek control of, and expropriate resources through, the parts of the state that are responsible for information processing, investment, and resource allocation rather than the administration of violence. In short, the bureaucracies and those who staff them.

Related: Nightmare of the Perfectly Principled, Arseholes, considered as a strategic resource


22 comments, sorted by Highlighting new comments since Today at 5:03 PM
New Comment
The "left wing" is the natural complement to this strategy: a political "big tent" made up of all the noncentral groups.
As before, both sides are winning this civil war, at the expense of the people least interested in expropriation.

While this appears to be true of conventional politics, it's worth noting that a very similar structure appears in less-expropriative contexts. For example, some technology markets naturally organize into a market leader vs. an alliance of everyone else; eg Microsoft (right) vs open source (left), or Apple (right) vs Android (left). In these contexts, overt force is replaced with soft power, and there is enough value created for everything to be positive-sum. Notice that people refer to an "Apple tax", and at the height of Microsoft's power referred to a "Microsoft tax".

This is a really helpful extension! I notice that in the two examples you gave, the split is qualitatively different from the political one in kind of the same way - the "left" is structurally libertarian / anarchistic, and the "right" is centrally planned. This suggests to me that "left" and "right" may not have an invariant meaning across instances of the broader sort of polarization, like they seem to in Enlightenment-influenced Euro-American democracies.

I wonder whether anyone here understands Japanese politics well enough to explain what's going on there.

FWIW this post made me want to tag it with a related question (a feature we just implemented for Q&A but haven't integrated with other posts), something to the effect of:

"Sampling randomly from a bunch of countries and cultures, how do political coalitions seem to be clustered?"

The abstraction you gave here made sense and I'd expect it to be one of the forces affecting coalition politics. But I'm less confident it maps cleanly to what I currently conceptualize as political-right-and-left.

(I think it might have been somewhat better to refer directly to the two strategies as the "big tent" strategy and "monopoly on force" strategy, or some such, since those are more concretely true, and don't [prematurely] frame the discussion around two dominant strategies, when it's possible there might be multiple strategies or they might map to different things in different contexts)

I do wish in hindsight I'd made that more explicit. I used those terms in an attempt to show (among many other things) how "left vs right" isn't a universally natural schema for political thought, but is instead the product of specific circumstances making a specific pair of coalitional strategies appealing, so that trying to categorize political thought outside the context of that particular coalitional battle as "left" and "right" is confused. But I don't seem to have communicated that clearly.

One relevant piece of evidence: the terms "left" and "right" were invented to describe physical position in the French parliament, where comparatively conservative groups aligned with the old regime sat on the right.

The "big tent" vs "hierarchy" dynamic might be somewhat specific to the US, but I don't think the left is ever more aligned with traditional hierarchy in a country that uses the left-right split.

Well, there are (at least) two ways to approach the question, one of which is explaining whatever happened to be going on with French politics at the time (or, in the US or other countries right now), and the other way is to be exploring the implications of the hierarchy game (which might or might not coincide with what was going on in France, the US or others).

Seems useful to explicitly note that the questions might have separate answers, and meanwhile this particular post seemed more to be doing the latter.

Compare: the political spectrum as Rentiers vs Apparatchiks

Thanks for linking this, it seems VERY closely related.

Could you give a reference for the Hierarchy Game? A quick google search did not turn up anything that sounded like game theory.

On a separate note, this post is IMO really toeing the line in terms of what's too political for LW. I'd consider it safely in-bounds if there were more explanation of the content of the hierarchy game, the conditions required for stability of expropriation in that game, and then discussion of evidence that those conditions actually do exist in current political systems. As it stands, the post has too many borderline-controversial claims and too little explicit evidence.

Could you give a reference for the Hierarchy Game? A quick google search did not turn up anything that sounded like game theory.

I think that was coined specifically for this post, and doesn't (yet?) have a corresponding formalism. I would be interested in seeing an attempt to formalize this, but there's enough subtlety that I'd worry about confusion arising from mismatches between the idea and the formalism.

On a separate note, this post is IMO really toeing the line in terms of what's too political for LW.

The way we currently handle this is with the Frontpage vs Personal Blog distinction; things that meet our frontpage guidelines, we promote to frontpage, everything else we leave on Personal Blog. We chose to front-page this, but I agree that it's borderline.

We considered giving a specific explanation of why we chose to frontpage this. I had decided not to at the time but will do so now.

In general, Benquo is writing a lot of stuff these days about group rationality and social dynamics. All of the topics are intrinsically a bit "hard mode" to write about – it's difficult to write about them without generating mindkilling dynamics. But, they're still a crucial part of rationality, and we need to try to deal with them at some point.

Basically each of the posts so far has been an edgecase of what I'd consider frontpage-appropriate (usually having at least one section that superficially violates a frontpage rule). But, we try not to be a slave to frontpage rules, instead treating them as guidelines to help us (and others) coordinate around an overall spirit.

In each case so far (including this post), despite superficially violating a guideline, my overall takeaway from the post felt very measured. I didn't feel incited to go join or fight a coalition, or to rally socially around an idea, or that Ben was employing rhetorical tricks to make it harder to come to my own conclusions.

We wrote explicitly about other posts here (and here). I also specifically didn't frontpage the North Korea post since it more overtly delved into object level politics of both North Korea and Effective Altruism (each separately are something we avoid frontpaging in most cases).

The borderline-nature of this post is different from Authoritarian Empiricism. Instead of worrying about LessWrong members getting mindkilled, the worry here is about attracting mainstream politics discourse into LW.

If that started happening (i.e. new users show up and start posting in this thread, or in the near future about controversial mainstream political stuff), I would change my mind about how easily to frontpage posts like this. (My current belief is that this is sufficiently abstracted to avoid that failure mode).

I don't think it makes sense to require more formalization of the Hierarchy game, since this post seemed like "the first post exploring it as an idea." Basically all game theory has some rooting in politics, and I don't think it's good to force new proto-formalisms to jump through more hoops than jumped through here.

(also note: I will be trying avoid commenting much more about the frontpage-decisions on the current series of Benquo posts. I think with this comment I've now covered most of concerns that the team has discussed. I'm hoping future posts in the series can just focus more on the object level)

Thanks, the decision makes sense given your reasoning.

I also agree more formalization shouldn't be required, if it's early exploration of an idea. I had read the post as saying that there was already a formal model out there, which cast the whole thing in a different light.

This doesn't seem to accurately describe contemporary politics, at least in the Western world. The left wing isn't just non-central groups, but the cultural/intellectual elites.

Meta: this topic is important, and general enough to fit within my views of what LessWrong can handle, but it's getting close to the edge, especially with the bias toward evaluating right-wing as "organized to expropriate" left-wing as "inclusive leftovers, who tend to expropriate as a side-effect".

I would like there to be a way to clearly label it as "politics-adjacent" or something to acknowledge that we don't actually want to move the window of what's appropriate for LW.

Nod. My current approach to that is to write comments whenever the mod team frontpages something that is an edgecase. I'm not sure whether tagging would add much to the equation but we have talked about that sometimes.

Especially for non-short posts, I often read and digest the post before looking at comments. Having a tag or other visible indicator that the mods have considered the question and I shouldn't be quite as distracted as I seem to be by the potential disruption of a political topic would help me.

I don't look at frontpage much (my bookmark is allPosts), so I don't distinguish much between promoted and just normal posts. Maybe that's a problem too - some indication of endorsement that shows on the post when something is promoted to frontpage.

I agree - I almost didn't post this on LW, and don't think I would have if The Two-Party Swindle hadn't been part of the Sequences.

I am happy you posted it here. It sounds reasonable, and from my perspective it doesn't feel mindkilling.

I have already assumed that state power is about taking resources (or defending from having more taken from you; but if you have the power to achieve that, most people won't stop there). Saying that the right and the left are two different strategies of creating coalitions to achieve this goal, that sounds quite impartial to me. Maybe I have mentally edited out something offensive, dunno. But I like the definitions of "the Schelling point of power" and "the natural opposition to the former" (this is how I abbreviate it for myself). Definitely more useful than "the guys who would hypothetically sit on the left/right chair in the 18-th century French parliament" or "it's just completely random coalitions".

Two things I would like to add:

1) This model seems to work for e.g. USA, but the situation e.g. in post-communist Eastern Europe is the other way round. The Schelling point of power is "let's bring back communism", its natural leaders being the former apparatchiks and secret service officers (many of them, or their sons, still active in current army and police). Yet, this is considered "left-wing". And "right-wing" is the hodgepodge of free market and religious fundamentalism and whoever's vision of the future does not include the return of communism.

More abstractly, the Schelling point of power depends on the recent historical events in given country. If the previous military power self-identified as "left-wing", the naming gets reverted.

2) It seems possible to go at least one level deeper in this analysis. You have the "natural Schelling point", and "its natural opposition" defined as people most likely to be oppressed by the former. But even the opposition oppresses someone -- there are "minorities within minorities" -- and thus we can sometimes get a second-order opposition, which may ally itself with the enemy of their enemy, despite not belonging there "naturally". Generally, "enemy of my enemy" strategy can create weird coalitions.

To give a real-life example from American politics, the left-wing coalition includes feminists, gays, and ethnic minorities. But what if you are an ethnic minority member who criticizes how given minority treats their own women or gays? You will get labeled as "right-wing". Even if you identify as left-wing, and your opinions and arguments are traditionally left-wing, picking the wrong target gets you thrown out of the coalition.

I think you're right about the Warsaw Pact vs NATO arrangements. Explicitly organizing the state around central economic planning altered the power arrangement so the info-processing bureaucracy (including secret police) ends up being the central coordinating point.

I agree the second-order effects you mention are there, and maybe important to model that so as to not get confused by it, but I don't think it has done very much so far except occasionally confuse people.

Maybe I have mentally edited out something offensive

I think the potential problem is that many participants in the discussion will strongly identify with one of the two factions I'm describing, so that what appears to be intellectual engagement might just be rooting for one's team, and arguments are more likely to be soldiers than usual. (I ended up deleting a comment on the North Korea post for this reason.)