Sunday, January 29, 2017

biggest unanswered question

I'd like to hear people's answers from other fields and from a variety of economic subfields as well. Go, mysterious meme powers...

My answer for economics generally and behavioral economics specifically is exactly Tyler's: "culture".

To try to be slightly more specific: I basically see cultures as collections of norms and methods of norm enforcement.  There are various types of norms (but note that particular norms do not often fall clearly into exactly one bucket):

  1. Coordination norms, like which side of the street to drive on, are fairly well understood both in origin and enforcement (i.e. self-enforcing).
  2. Social enforcement of norms that provide public goods (e.g. littering) is getting to be fairly well understood. But which public goods are provided by norms and how this comes to be is utterly mysterious and of utmost importance. There are lots of pieces of the puzzle identified but not put together at all.
  3. Norms for division of goods (splitting the pie) is a sort of subcase of #2 but it's a weird area of study because these are some of the simplest and most universal norms yet the theories (e.g. Binmore's) of them are some of the most complex. It's really really interesting, but almost overkill.
  4. There are also many norms that are better described as heuristics for dealing with uncertainty and limited information (don't pick up hitchhikers; eat breakfast), and I don't think the selection and enforcement of these norms is at all well understood, nor given serious attention. I'm sure most are written off as individual heuristics, but the degree of social enforcement and arbitariness suggest to me something more is going on that just judgmentally inferring bad things about other people from their stupid decisions. I could be wrong. At the very least, the individual motivations for adhering to these norms are closely related to individual motivations for adhering to cooperative norms, so the literatures are intertwined.
  5. Lastly there are norms that are completely arbitrary, don't fit in any of the above categories, and exist solely for signaling value, like wearing ties to weddings. These are not very mysterious, in the same way that sexually selected traits are arbitrary but evolutionarily well understood.
In summary I'd say the most important and mysterious unanswered question of economics is the point from #2: which cooperative norms are chosen to be enforced and how does this come about?


Anonymous said...

In the 90s and early 2000s there was a time when every Starbucks patio and airport lounge was ruined by people speaking loudly on their cell phones. A large portion of society hated these people, but somehow we never coordinated or arranged to consistently express our preference. Then very quickly after the iPhone came out everyone began messaging and emailing instead of speaking. Very quickly a new norm emerged in which it was agreed to be rude to have a phone conversation in a line at Starbucks. I still don't know how the consensus was formed or communicated.

Ray Lopez said...

Self-selection. If you are a geek who likes to invent just for fun, and I've met quite a number of these people working in Silicon Valley, you'll be befriended by a businessman who will pick your brain and profit from it. Same with a 'good student' who is befriended by a 'poor student', same as a person who tolerates X befriended by a person who performs X, same with a athlete befriended by an agent who takes a percentage of their earnings in exchange for setting them up with a team, and so on. Self-selection.

Vera said...

Anonymous: great example :)

Ray: wouldn't that mean most pareto improving norms would spread very quickly? I think the trouble is that adopting a coordination norm requires coordinated action, so the analogy is with group selection rather than individual selection.

Unknown said...

I think you need to look at the from the perspective of adaptation and the perspective of a single culture. The world we now inhabit is no longer mono cultural or even multicultural, it’s fractional cultured.

If you go back a few millennia, being part of a group was unquestionably key to your ability to survive long enough to procreate. Maintaining the integrity of the group was also key. An elevated status (authority, prestige, whatever) improved your odds for progeny. Key ways to maintain group cohesiveness would be to maintain fairness and provide clear cues in regard to achievement of status. (behavior, visual)
When hominids developed the requisite abilities for cultural evolution to take off, it was generally within the context of a single culture. Since individuals were always imbedded within a culture, many (most) of our cognitive abilities/behaviors are dependent on others for learning (mostly nonconscious) AND daily execution. Simplistically, our brains function under the ‘assumption’ that others will help fill in the blanks and keep us on track.

Note that none of these abilities would require much in the way of language nor perhaps even consciousness as we now know it. Individual behavior would be driven by feelings.

So to answer the question: some of the selection criteria for cooperative norms;

appear to threaten group cohesion if violated, unfair
have visible cues
easy to take action, e.g. shout
threatens the status of some at the expense of others
require little or no conscious thought to observe or react to.

Frederik Marain said...

Example of (4): If a phone call is interrupted, the person who made the call calls back.
I think I got this from Elinor Ostrom, who is obviously very relevant here.

starcourse said...

Check out the work of Martin Nowak and his colleagues at Harvard.

Roger Sweeny said...

Not sure how useful they'd be but at least on the periphery of your concernss are Joseph Henrich's The Secret of Our Success and Morris B. Hoffman's The Punisher's Brain.

Jd said...

Any observations about changing negative "norms" around domestic violence?

JohnRaymond said...

I might pick up on Briant's mention of the brain's role in norm building because it ties in with my own field: language and linguistics. In my research I work on the assumption that the brain is hardwired in certain ways (much of which is not yet well understood), mostly for language acquisition, but also for language use at all ages. Both language and the hard wiring of the brain play a significant role in norm formation: we do and say things in ways that can't be explained, that seem irrational sometimes, except from the standpoint of how the brain has been wired/programmed through evolutionary development. In language there are relations between parts of the sentence that follow language-specific parameters, which are obviously connected to the related culture in certain ways, but also to certain universals of language. So are there universals in the norms of human cultures? And can we state specific parameters of culture x vs. culture y? These probably have something to do with what norms are established in these cultures.

Vera said...

Briant: yes the fractional cultures we live in are really complicated compared to more traditional, isolated, and cohesive cultures - this is how I read Akerlof and Kranton's work on what they call "identity". Very interesting stuff.

Frederik: Interesting, I didn't even know about that norm! That sounds like a coordination norm to me (avoiding a collision in return calls in which neither can get through, or a lack of any callback).

Starcourse: You're absolutely right, I love the stuff Nowak works on but haven't read nearly enough of it :) Similar for Henrich actually, Roger, and I didn't know Hoffman so thanks for the heads up.

Jd: I wish!

John/dad: yeah it would be great if we could come up with some kind of universal grammar of norms. I would guess that the analogy with linguistics applies mostly to coordination norms (since language is a coordination problem) but I'd like to hear more detail sometime.

windwheel said...

I'd have thought 'mirror neurons' would be relevant for co-ordination games. Mimetic effects rapidly spread out such that, for example, people appear to spontaneously 'keep left' or 'right' as the case may be. The question is- why have a distinction between norm and action?
Perhaps the answer is that dis-coordination games have such great salience due to gender dimorphism wrt costs of reproduction in our species that 'pooling equilibria' have to be qualified and the system has to have some 'memory' or hysteresis.
'Identity'is a mix of cheap talk, featuring adherence or violation of norms, and some underlying structure of costly signals which establish a separating equilibrium. This permits a 'bourgeois strategy'- at least some people following the norm will enforce it rather than blink first in a game of chicken. Crucially, for this to occur, there has to be some notion of scarcity or imposed opportunity cost.

I think this is what happens when some 'subaltern' agents furiously insist their peers adhere to a norm costly to themselves and not group adaptive at all. Typically, the person violating the norm is accused of reducing the amount of 'Purity' or 'Honour'or some such bogus abstraction during a supposedly cataclysmic time when the commodity is scarce. Perhaps there is a link between this sort of 'highly correlated thinking' and the historical development of complex social choice mechanisms or abstract financial markets.
The enforcers of'bourgeois strategy'type norms could gain a rent as 'market makers' bridging coordination and discoordination games or they could prefer to keep a threat point open for themselves for future use. This would make cake cutting even more difficult with the result that cakes become scarce or mythical like 'Purity'.
No doubt mine is a jaundiced view of 'traditional' economies or cultural constellations. Still, it seems, young people's mirror neurons light up when they see modernity and they vote with their feet against normative bourgeois strategies and infinitely complex cake cutting problems.
The actual historical bourgeoisie was actually quite Coasian thus permitting our liberal conception of norms as flexible and 'ethical' rather than grim and punitive Norns set over us by a remorseless Destiny.

windwheel said...

Whoops! Did I just kill off an availability cascade which could run and run?

Vera said...

Which availability cascade?

Doug Jones said...

I come to this topic as an anthropologist interested in kinship. One of the frequently noted things about kinship is that many societies have norms of "generalized reciprocity" which call for sharing with your kin, even when the beneficiaries can't reciprocate. You do this not so much because you care deeply about the person you're helping (the standard evolutionary biology theory of kin section governed by Hamilton's rule), but because you care about your reputation among others of your kin (this lets you beat Hamilton's rule).

Here's an article where I do some of the math, including between-group selection of balanced reciprocity versus generalized reciprocity:

And a post from my blog:

Vera said...

Doug, yes absolutely, I also am working on indirect reciprocity a bit actually. This is definitely one of the enforcement mechanisms I was referencing when I said that the enforcement of cooperative norms is getting to be well understood :) The repeated game theory is very cool.

Nathan said...

Hi Vera,
When you say 'culture' as the big unanswered question, how do you see 'organisational culture' within that?
I wonder if there is a 'micro versus macro' thing here?
If you walk into an organisation (eg start a new job), you always find a specific culture. There will be norms around what time to leave on a Friday, how much (if at all) to dissent in front of the boss, etc. And there will be a lot of things half-way between norms and collective decisions -- 'To what extent do people who work harder get promoted faster?', or 'To what extent do people who come in early tend to get promoted ahead of those who stay late?', 'How open are you with your colleagues?', 'How risk averse with new projects?', etc.
And of course every organisation has its take on the wearing of ties.
Do you think if we understood organisational culture, we would understand culture generally?
Or do you think it goes the other way? Organisational cultures are just tidal pools buffeted by bigger forces, and you have to understand broad 'culture' first to explain them?

Vera said...

Nathan, yeah that interaction between society and subsociety norms is an interesting question that I think deserves more attention exactly because it can say so much about how norms form. I suspect the mechanisms of norm transmission/establishment/enforcement/evolution are similar within society and organizations (but obviously different in way those mechanisms manifest, since the group size and structure etc are different) and in that sense understanding organizational culture is really useful for understanding culture overall, as smaller scale and variable testbeds essentially. But organizations also inherit norms from the societies they're embedded in, so I don't think you can make complete sense of a particular organization's norms without understanding the broader culture: There's lots of variability in all of the norms you suggested but they can't get *too* far from some societal mean, and they start off by default at the societal mean. E.g. some U.S. businesses shun ties and some require them but none require kimonos...

Nathan said...

Hi Vera,
I thought more about this. Forgive me pushing it further, but I think there is a tendency in some of the above comments to think about norms like memes -- as if they are ideas that spread neutrally from a source based on some transmission mechanism.
But if you think about changes to norms within an organisation, they are typically the product of people intentionally trying to enforce their own preferences. A group of graphic designers might think "We need to stop swearing in front of clients" or "We need an affirmative action hiring policy". Others in the firm will think "Don't be so uptight" or "No, we don't". Everyone has the usual 'exit, voice or loyalty' options, and depending on who has influence and who is prepared to use it, things will either change or they won't. If a group of trainees think "Stop swearing" and the CEO thinks "Don't be so uptight", we can guess whose preferences will prevail. If the organisation is a family or a school we can guess that the collective decisions will be resolved in favour of the adults, and resistance by the kids will be costly.
Explaining the initial preferences seems difficult (eg the kimonos!) but explaining which preferences prevail seems like it might be more fruitful.
So a question for you...
You say the big question is "Which cooperative norms are chosen to be enforced and how does this come about?".
Someone else might pose it differently. They might say: "How do we make collective decisions within an organisation?" -- and propose that the answer (once we have it!) can be applied to organisational norms as a particular case, and then extended to societies, as very big organisations.
What do you feel is the advantage of your phrasing?
Re-reading your previous answer, I think it is because you think the mechanism that changes norms between strangers on the street is fundamentally different (and perhaps fundamentally more important?) to the one that changes norms between staff in a café?

Craig said...

It's interesting you described "don't pick up hitchhikers" as a heuristic for dealing with uncertainty and limited information, as that norm has changed. Before around 1960 or 1970 hitchhikers were more common and more likely to be picked up. I'm sure a part of the change has to do with fear, and stories about bad experiences with hitchhiking predominating over stories about good experiences. The change became permanent when any story about good hitchhiking experiences raised the question, why was the protagonist hitchhiking. Other norms change due to fear also. Intense passenger scrutiny at the airport came from fear of murderous hijackers, while earlier in commercial aviation hijacking was considered a rare inconvenience. Some norms change while overcoming fears. Attitudes about homosexuality changed despite the fear based heuristic "don't tell strangers about your homosexual tendencies"

King said...

When it come to such, the individual involve is let to make such decision by themselves.

Vera said...

Nathan, that's a really great point about trying to enforce preferences. I think that's the motive people have in general when trying to change norms (graffiti shouldn't be acceptable, sunbathing topless should be acceptable, etc) but in a huge society where a single person's preferences are negligible, the main channel where that has influence is in law. Laws and norms are obviously tightly intertwined and both influence each other, and there are some interesting studies showing that, but my personal interest is in non-legal, spontaneous-order-type mechanisms. But perhaps by limiting my attention to that I'm missing a bigger picture... power structures certainly seem extremely relevant even when they are not formalized.

But going back to organizations and smaller societies (which I guess are usually organizations with at least some degree of formality - homeowners associations, churches, clubs) a single person's preferences most definitely aren't negligible anymore, and then as you say influence will matter very much. I do think about norms with fundamentally different transmission mechanisms but I have to agree that thinking about norms as organizational decisions in an extreme case is very interesting...