Maybe echo chambers are evolving because they are efficient?

Our analysis reveals that segregation into small, homogeneous groups can be a rational choice that maximizes the amount of information available to an individual. In fact, homophilic segregation can be efficient and even Pareto-optimal for society. Why is that? Our argument builds on the idea that people have not only different information, but also different preferences. These differences in preferences can prevent successful communication, because people do not want to reveal their information to those who are different, and distrust the motives of those who speak to them. It then becomes easier to exchange information in segregated, homogeneous cliques than in large crowds. Echo chambers, though they may cut off potential communication with a great number of people, make actual communication possible, and are hence useful for society.

That is from a new paper by Ole Jann and Christoph Schottmüller.  I believe Jann is currently on the job market from Oxford this year.  Here is their other paper on the economics of privacy.  And from Schottmüller: “The quality of
advice can be highest if the adviser’s competence is uncertain.”


Sure. Works fine if one echo chamber doesn't insist on imposing it's narrow view on other people, or even better if there is no contact at all with people outside the echo chamber. It keeps life simple and predictable, suitable for the exquisitely educated.

Of course you could falsify the whole thing by mentioning Nicolae Ceaușescu, who lost his head because of an echo chamber.

Maybe the paper writer should have used the term Green Zone instead of echo chamber to clear their minds of priors.

He said, in what he hopes is his kind of echo zone.

Simply put, if everyone in the community looks like you, they're more likely to honestly tell you what the local schools are like--rather than hide the bad parts that make minorities look bad.

I'm not sure to understand your point about minorities. But the authors assume that somehow the members in an echo chamber have different pieces of a complete view and, within their echo chamber, they're more likely to get a complete view by exchanging with those who will not cause excessive transaction costs of their exchange of knowledge. There are many assumptions built into this. First, that members have some explicit goal of deepening their understanding (I didn't use broaden because that might imply exchanges outside their echo chamber). My assumption is that this is exactly backwards - there is no explicit or implicit desire to deepen knowledge, but simply a desire to maximize their emotional tribalism. Second, that each member possesses an incomplete view of the whole (again, limited to the echo chamber's implicit holistic view) and those incomplete views are uniformly distributed across the entire member set. I suspect the "ding" of getting upvotes is a stronger motivator for membership than some vague expectation of a more complete knowledge set.

Interesting, considering how some think the greatest advantage that capitalism possesses is its ability to allow anyone to interact with anyone else in a market context. And some of the proponents of this view definitely believe it is the communication of information that is the fundamental advantage of a market economy. Meaning that to work effectively, it is directly opposed to segregated, homogeneous cliques not participating in the swirling information market around them.

It is how one reaches the point of enjoying a symphony orchestra, enjoying the fruits of the vast unplanned coordination that led to the music filling the concert hall.

But notice: the symphony is not usually as enjoyable if a punk-rock band is playing at the same time. Coordination means alignment of operational goals, or at least an intersection of them. Arbitrarily forcing collaboration with parties that do not share your goals will devolve into factions vying for control of the group rather than just each working towards their independent goals; a nasty, inefficient system in which basically no one is happy with the process or outcomes... In a word: "politics".

'a punk-rock band is playing at the same time'

Or if the symphony hall is next to an airport.

Nonetheless, this point is derived from a recent Prof. Cowen post - 'I heard Mozart’s 39th symphony in concert last night, and it occurred to me (once again) that I also was witnessing one of mankind’s greatest technological achievements. Think about what went into the activity: each instrument, developed eventually to perfection and coordinated with the other instruments. The system of tuning and the underlying principles of the music. The acoustics of the music hall. The sheet music on paper and the musical notation. All of those features extremely well coordinated with the kind of compositional talent being produced in Central and Western Europe from say 1710 to 1920. And by the mid-18th century most of the key features of this system were in place and by the early 19th century they were more or less perfected.

Sometimes I think of the Industrial Revolution as fundamentally a Cultural Revolution. The first instantiation of this Cultural Revolution maybe was the rise of early Renaissance Art in Italy and in the Low Countries. That too was based on a series of technological developments, including improved quality tempera paint, the development of oil painting, the resumption of bronze and marble techniques for sculpture, and the reintroduction of paper into Europe, which enabled artists’ sketches and drawings.'

'Arbitrarily forcing collaboration with parties that do not share your goals will devolve into factions vying for control of the group'

The market does not force collaboration, it rewards it. And factions vying for control of the market is a good thing, at least if one believes that competition leads to progress.

"The market does not force collaboration, it rewards it."

Agreed, I'm responding to your conception of coordination that implies every person in direct coordination with every other person: "Meaning that to work effectively, it is directly opposed to segregated, homogeneous cliques not participating in the swirling information market around them."

The symphony works only if there is sufficient diversity to not all be french-horns, but enough segregation that there aren't punk-rock players or airplane pilots contributing too much to the experience.

"factions vying for control of the market is a good thing, at least if one believes that competition leads to progress."

I think this is a misunderstanding of how I was using the word 'control'. Vying for customers within a market is a very separate entrepreneurship activity than vying for control of the market in a meta sense (e.g., regulations).

I think a definition of echo chamber that is not so distinguishable from "an open conference on a subject".

However in the "Straussian" sense, this is useful material for right wing arguments against diversity and migration (social segregation of people with different cultural preferences is efficient!) so can deserve one cheer.

Occupations can be echo chambers, at least law enforcement:

So Diversity is NOT Our Strength?

I'm floored in shock.

Diversity may be our strength, but is not our laziest destination.

I only skimmed, but the conclusion makes sense to me.

Perhaps these researchers know of others who are testing social media designs for better outcomes.

(Contrary to many Twitter stars who want to keep their Likes, I suspect dropping it reduce this effect.)

This paper mostly reveals the narrowness and limitations of economic analysis. I can well believe that echo chambers are welfare enhancing - but I don't like that measure and can probably think of dozens of horrendous actions that might also be welfare enhancing. I truly believe that if we lived in a world which operated solely to enhance human welfare (as defined by economists), the quality of our lives would be much diminished. Fortunately, we have numerous institutions whose purpose is to restrain the worst impulses of economic welfare.

Before you attack my comment, let me clearly say that I do not know of a better system and that more reliance on markets will probably result in net benefits compared with anything else. I just don't really care whether echo chambers are Pareto optimal (and, I suppose, they really mean potential Pareto optimal, not truly Pareto optimal).

Right, obviously one could become a vegan and taylor a silo to bring happy feedback for that lifestyle, just as one could do the same as a big game hunter.

While that is fun in the short run, it could lead to a somewhat shattered Society, if bridges and some level of interchange are not maintained.

Quick, is the Caravan mostly small children far away, or is it the killers on our doorstep?

At the risk of "piling on" in our little message thread echo chamber within the larger message thread, I agree with Dale ;-).

I think the authors' assumptions that there are shared goals for communication are naive (optimizing info sharing balanced against costs of possible rejection). Echo Chambers are certainly "efficient" but it's not about communication towards some sort of aggregate, maximized understanding of a possible preferences. I would rewrite the paper using this headline: "Echo Chambers enable members to maximize the benefits of their emotional tribalism."

Plus it's naive to assume that within a group the individuals each own a "partial view of the truth" that is uniformly distributed and that they simply lack a complete vision.

reductio ad absurdum: The person you are (one hopes) most similar to and share the most in common with is yourself. So, we can assume (mental illness aside) that the most efficient communication is via internal dialog. Efficient yes, effective? hmmm. There was a talk (some years ago (highly recommended) or similar) about communication. I think the phrase the speaker used was "shorthand". He pointed out that much of what we say is "short" for more thorough and detailed meaning. That is, it is pretty obvious that the more understood/shared the context is, the less time/effort it takes to communicate. So, given our limited time and attention span it would seem obvious that more can be communicated when the context is shared (understood) than otherwise. But who said that diversity was efficient? It's about the difference between finding a local minimum (extremum) and a global (or at least regional) one. The less activation energy available, the less likely the global extremum will be found. I don't at all like the question posed here. Sloppy use of the word "evolving" (meaning proliferating, I think) when all you need to understand is the way human nature will cause most of us most of the time to speak to people who think like us and who confirm our own beliefs. This is also naturally compounded by the effect repetition has on our beliefs. I'm a bit puzzled by this post since the military seems to be the poster child for "echo chamber" efficiency. Arguably, there's more non-echo chamber communication in today's more effective military than ever. Efficient is to tactics as effective is to strategy.

So, monasteries.

Chris Follow at stumbling and mumbling had a post on this recently. You should all get together and agree about it.... whey!

A TCR platform that allows diverse arbitration panels could help.

If I have information that I think could persuade people to my point of view, I could write two (or more) versions of the same article.

One would be geared for people who already agree with me. I would assign friendly arbitration panels as eligible to take my tokens if this article is ruled inaccurate.

The second version would omit the weakest claims. Because it would be more difficult to claim this version is false, I would risk my tokens for this version with neutral/antagonistic arbitration panels. This would allow me to share the strongest subset of my claims with the readers who likely disagree with me (because they search for articles subject to arbitration by panels the reader finds agreeable). To maintain credibility, even a hostile arbitration panel might not rule against me if I stick to facts.

I'm liberal, yet a room (or forum) full of other liberals is where I'm most afraid to express opinions or share information, due to the possibility of group-think shaming. Similarly, I'm sure the RNC was not the best place for conservatives to share their concerns about Trump.

I'm skeptical. When you have two distinct groups with different preferences and information, one of those groups is almost always significantly better adapted to existing economic conditions than the other. So, any equilibrium is likely to be unstable with one eventually replacing the other. The only time when this does not hold is when their environments are very different from each other and one set of preferences is better adapted to one niche, and the other is better adapted to another. Historically, good farming lands were occupied by farmers who had one set of preferences and less arable land was occupied by herders who had a different set of preferences. But, when climate events caused historically arable farming land to be less productive, the herder "barbarians" would swoop in because they were better adapted to the new ecological conditions and would overwhelm the farmers. Then, when farming conditions were good, civilization would rebuild and the invading herder elites would be culturally converted to a farmer set of preferences.

All of these models kind of bore me right now. What I want to see are the most 2 similar models they can produce that result in different behaviors. Remove a link, change the relative weight of a value signal, etc.

Comments for this post are closed