Staghunt and (the) IRS

"IRS" stands for "Increasing Returns to Scale". It means if you double all the inputs, you more than double the output.

I first worried that US readers might think "IRS" stands for Internal Revenue Service (the US tax people). Then I realised that I want it to stand for that too.

Staghunt is a simple game with two players. Each player alone can catch one hare each; both players together can catch one stag, which they share. Staghunt has two Nash Equilibria: both hunt hare alone; both hunt stag together. (If the other hunts hare you want to hunt hare too, because you can't catch a stag alone; if the other hunts stag you want to hunt stag too, because half a stag is better than one hare.)

Staghunt exhibits increasing returns to scale technology, because two hunters producing one stag is more than double the meat of one hunter producing one hare.

Staghunt is also a model of society. We gain by cooperating, but it only works if we trust others to join in. It's different from Prisoners' Dilemma, the alternative model of society, because Prisoners' Dilemma has only one (bad) Nash Equilibrium, and so needs a Leviathan (which can be interpreted many ways) to enforce cooperation.

Staghunt is also a model of a "network externality". The individual who joins the network (hunts stag, buys a telephone) creates a positive externality for others already in the network, and increases the incentive for others to join the network. (That second effect should strictly be called "strategic complementarity", or "positive feedback", because it is not an "externality" in the strict sense. See my old post for the distinction.)

So Staghunt is a very good model with which to examine Brad DeLong's recent thought-provoking post.

In the "hunt stag" equilibrium of Staghunt, the two players earn less in aggregate than the sum of the values of their marginal products. If one player fails to show, output falls by one stag (or falls by one stag minus one hare if it's a sequential game and the second player can switch his move on observing the first's move). If each player insisted on getting paid the whole value of his contribution to the stag hunt, there wouldn't be enough stag to go round. If one player could make an irrevocable take-it-or-leave-it offer, he would say "I will join you in hunting stag if and only if you agree that my share will be the whole stag, minus meat equivalent to one hare, minus epsilon".

This is not an artifact of an integer constraint in a two person game. Consider Long Run zero-profit equilibrium in monopolistic competition. The demand curve is tangent to the ATC curve, and since the demand curve slopes down the ATC curve must slope down at that tangency point, which means (locally) Increasing Returns to Scale. Each input is paid its Marginal Revenue Product, which is less than its Value Marginal Product, because Marginal Revenue is less than Price. Holding inputs and outputs constant, if each input were paid its Value Marginal Product, the surplus would be negative. The Internal Revenue Service would need to subsidise Increasing Returns to Scale production to ensure each input is paid its Value Marginal Product (or to get output to the efficient point where Marginal Benefit=Price=Marginal Cost < Average Total cost).

It is nevertheless true that hunting stag together creates a surplus relative to each hunting hare alone. Starting from the hunt hare equilibrium, the Value Marginal Product of Labour of one person hunting stag is zero. It all depends on the starting point, whether we are worth less or worth more than we are paid.

Those of us in rich successful societies are in the hunt stag equilibrium. We have escaped the hunt hare equilibrium of the State of Nature (Rousseau rather than Hobbes, since it's Rousseau's Stag Hunt). To whom do we owe this good fortune?

You could say we owe it to our ancestors. In games with multiple equilibria, it is usually history that provides Schelling's focal point that leads us to pick one equilibrium rather than another. We drive on the right/left because that is what people did yesterday. But that is a debt we cannot repay; because they're dead. The institutions we live under are part of their legacy to their children and grandchildren. The hunt stag equilibrium is what they gave us; they did not give it to something called "the government". In fact, "the government" is not a thing; it is merely a name we give to one aspect of the hunt stag equilibrium.

But you could also say that it is not our ancestors who gave us the rules and institutions we live under in the same sense that they gave us houses and land that is cleared and drained. It is us who re-create those rules and institutions every minute of the day by acting the way we do and believing that others will act likewise. Like Tinkerbell the fairy, they don't exist except for our belief in them. Actually we did build that; and we re-build it every single minute of the day.

This post is not intended as a smackdown of Brad DeLong and his good post. (Though I would dearly love to smackdown any leftie who read it uncritically and who liked it simply because it seemed to confirm their "Value is a Social Product!" slogan used to validate any sort of government intervention they feel like.) He's onto something, and it's well worth thinking more about what precisely it is he's onto. That's what this post is trying to do.

And as for the so-called "Value Marginal Product Theory of Distribution": well, there isn't one. Because VMP is a function, as well as a point on that function, and which particular point gets chosen in equilibrium depends on...a lot of things. The example I tell my first-year students is that if Wayne Gretzky had been born where I was (England) when I was (....) his hockey playing skills would have been valuable only as a peculiar hobby, on par with being skilled at tiddlywinks. Which would be true whether hockey is IRS, CRS, or DRS. Guitar skills were valuable; bagpipe skills were not. It's all the luck of the draw, as well as your own effort and other choices. But we do need institutions that reward the right choices. And the human right (well, it's not an animal right, is it?) of free contract matters too, as in the Parable of the Vinyard.

Or, to say it more punfully, the world's IRSs should encourage, not discourage, the IRS hunting stag equilibrium.

Comments

[Henry]

"Those of us in rich successful societies are in the hunt stag equilibrium."

You have to be kidding, Nick.

The game is more like this. One player has a bow and arrow. He knows where there is a stand of trees the wood of which is strong enough to make a powerful bow. But the hunter with the bow needs one other to help him drag the stag out the the forest. The bowman convinces the other that the bow and arrow is the only way to get the stag. In return he kills a hare and gives it to his accomplice.

[Lord]

Convince others they are in a staghunt and hand over a jackaloupe. Clever.

[anon]

"... the two players earn less in aggregate than the sum of the values of their marginal products. If one player fails to show, output falls by one stag ... If each player insisted on getting paid the whole value of his contribution to the stag hunt, there wouldn't be enough stag to go round."

Good point. This also means that the distribution of payoffs in any IRS situation is critically influenced by bargaining power. If bargaining power (strictly speaking, elasticity of supply) varies a lot depending on the input and the increased-returns character of the market is especially high, we might get a distribution of payoffs that has little or nothing to do with even the average product. This is actually a desirable thing in abstract, since shifting price distortions onto the more inelastic factors is good for efficiency - but of course it makes equity considerations all the more important. (_Contra_ Brad De Long, I think this actually makes simple, no-strings-attached redistribution _more_ desirable rather than less, since that's by far the easiest way of ensuring a "fairer" market, where bargaining power itself is equitably distributed.)

[Majromax]

> This is actually a desirable thing in abstract, since shifting price distortions onto the more inelastic factors is good for efficiency - but of course it makes equity considerations all the more important.

In the short run yes, but there are also different short and long-run elasticities. If the commenters will forgive me for including 'society' as a factor of production, we can say that societal institutions (such as a culture of settling disputes in court rather than by personal, violent vengeance) are short-run inelastic. However, they're very elastic over the long-run precisely because they're reinforced (or not) on a day-to-day basis. If returns are distributed based on short-run elasticity only, then society can very efficiently undermine itself.

I think that this is getting at the Acemoglu/Robinson idea of extractive versus inclusive institutions in a roundabout way, from a purely economic rather than mixed economic/sociological perspective.

[rsj]

- The point about bargaining power is important because even for those jobs that weren't outsourced, the *threat* of outsourced hurt workers bargaining power. So things like NAFTA were devastating to the rust belt even across the board, and this can't just be measured in the number of jobs lost.

- For IRS more generally, this means a small (and even shrinking) proportion of the workforce accounts for the vast majority of productivity gains, and everyone else benefits via Baumol effects. This means that for a small town with just 5% of the workforce in one IRS company, when that company shuts down or moves, the entire town is devastated -- coffee shops, haircutters, lawyers, secretaries, landlords, landscape workers, construction workers, all of them see stagnant wages when previously they were enjoying increasing wages due to the productivity gains of the one IRS industry in that town. That also explains the widespread destruction and bombed out look to much of the central states even though only a small percentage of jobs were in the IRS industries.

- The problem with compensation as a solution

1. It's not stable. Universal benefits have broad popular support. Means tested benefits not only have people fall through the cracks but are more likely to be repealed.
2. It's not viable -- if the industries really had to compensate the workers for the lost income, they wouldn't have been advocating for these trade deals in the first place.

The above explain why compensation doesn't happen in practice. But even if we could overcome this, e.g. by limiting the political power of wealthier people and removing firms from the political process entirely, then still there would be issues

3. Who to compensate? Will you compensate the whole town that lost it's IRS industry or just the workers in that town? Will you continue to give them the same increases in wages that they would have gotten with the same level of bargaining power? For how long?

4. Humans need meaningful, well-regarded work. Without that you get a lot of social pathologies.

Bottom line, this is what politics is about. Decisions are drawn boundaries around acceptable forms of barganing, financial flows, workplace controls, and commerce in order to try to protect and encourage IRS industries as well as to share the gains enjoyed by the small IRS workforce with the rest of the workforce. These are messy decisions but they need to be made.

Simply retreating from this form of politics and letting the market sort this all out relies on stability guarantees that don't apply in the IRS setting. It's also bad policy. DeLong is himself gorging on stag by working in an industry dominated by Industrial policy -- namely the role of the large research university after WW2. Prior to WW2 UC was a small school next to some hills, with a budget in the hundreds of thousands of dollars. In 1950 it had a budget $11 Billion, with a national lab for nuclear research, and masses of students sent there by the GI Bill. It wasn't just the free market, just as the creation and destruction of Detroit, or Gary Indiana, or most of the rust belt as a whole was the result of political choices. We can't pretend that these choices don't belong to us, or that they shouldn't belong to us. There is a place for capital controls, anti-trust, legislation promoting worker bargaining power, and stabilizing fiscal policy in addition to redistribution.

[Majromax]

@rsj:

> The point about bargaining power is important because even for those jobs that weren't outsourced, the *threat* of outsourced hurt workers bargaining power. So things like NAFTA were devastating to the rust belt even across the board, and this can't just be measured in the number of jobs lost.

I don't buy it, at least for NAFTA. Take a look at the real earnings of goods-producing workers (PCE-adjusted). I see an increasing trend from the beginning of the series to 1973, a flat trend from there to 1993, and an increasing trend thereafter. It shows no particular correlation to manufacturing employment (red), which was at best flat from about 1966 onwards.

If anything, NAFTA (1994) marked a turnaround for manufacturing wages with the transition from steady to again-increasing earnings.

The best story I can make out of this is that wages stagnated as worker bargaining power receded, but it began increasing as trade openness improved the scale-improving returns.

[Nick Rowe]

Consider a game between one apple producer who wants to eat bananas, and one banana producer who wants to eat apples. Each individual has bargaining power against the other.

Now consider the same game with 100 apple producers and 100 banana producers. Each individual's bargaining power gets very small.

Some things are an artifact of having a small number of players in the game used as an example.

How much power does one individual telephone user have?

[rsj]

Well, I'd look at compensation of employees as a % of Gross Value Added, since this is closer to what I think of as the bargaining between labor and capital. You can find the data here: http://www.bea.gov/industry/gdpbyind_data.htm

Just look at 5 year buckets: 1989-1993, 1994-1998, 2000-2001

Share of compensation:
The economy as a whole staid relatively flat: 55.9, 55.0, 56.1

professional and business services went up: 67.9 69.6 71.7
educational services went up 80.2 81.4 82.7
Health care and social assistance went up 79.4 80.9 82.0

These in general are areas where you don't see big gains in productivity.

Manufacturing overall went down: 61.4 58.3 57.1
But the big plunge was automotive: 71.9 60.0 55.8

I don't point out automotive because I think it is special among IRS industires, but because much of the NAFTA debate centered around this industry, which is also funding much of the lobbying for NAFTA.

In terms of how much blame to ascribe to NAFTA versus WTO versus right-to-work or lack of labor law enforcement -- it's hard to say.

[rsj]

^^^ Sorry, the above said 5 year buckets 2000-2001, I meant 2000-2004 :)

I wanted to include the 5 years after the WTO ratified (and the Seattle protests, which both Krugman and deLong mocked).

[rsj]

Nick,

Exactly. With IRS you will tend to have natural monopolies and concentration of power, because a big firm can produce more cheaply on per unit cost than a small firm. The census bureau has concentration ratios, e.g. for manufacturing: https://factfinder.census.gov/faces/tableservices/jsf/pages/productview.xhtml?pid=ECN_EC0731SR13_FTP&prodType=document and for other industries.

It's fascinating. Household refrigerater and freezer market has 19 companies in all of the U.S with $2.5 Billion in value add. The 4 largest account for 92% of of this. Magnetic and Optical media recording has 121 companies, with only a half billion of value add, but 99% of it is captured by the top 4 companies. I recommend reading through the list, just to find the number and kind of different industries. Adam Smith will be interested to know that Fastener, button, needle, and pin manufacturing has 174 firms in the U.S., with the top 4 earning 52.6% of gross value add. This is all from 2007, so before the shakeout. I'm interested to know what the next report will contain.

I heard that in smart phone sales, apple has > 100% of the profits, in the sense that the rest of the industry is collectively losing money, but that figure is not from the census data.

Without IRS, you can pretend that the economy is scale invariant so that there may as well be an infinite # of small producers and no one has any pricing power.

[Tel]

Shorter Brad DeLong: "You didn't build that".

As I have said before, humans are, at a very deep and basic level, gift-exchange animals. We create and reinforce our social bonds by establishing patterns of "owing" other people and by "being owed". We want to enter into reciprocal gift-exchange relationships. We create and reinforce social bonds by giving each other presents. We like to give. We like to receive. We like neither to feel like cheaters nor to feel cheated. We like, instead, to feel embedded in networks of mutual reciprocal obligation.

Sorry but "gift exchange" ?!? It's muddle headed.

A gift is a gift, no exchange, no reciprocation.

An exchange is a trade, a transaction, not a gift at all, but a completely different concept.

[Henry]

"An exchange is a trade, a transaction, not a gift at all, but a completely different concept."

Tel,

I think he means an exchange of gifts. There is a net gain to the individual - and there is a social gain.

[Redwood Rhiadra]

@Tel

A gift exchange is like Christmas/Chanukah/other gift-giving holiday of your choice - folks give gifts out to family and friends, and they get gifts from family and friends - and they don't necessarily get gifts from the same people they give to, nor do they bargain to ensure the exchange is "equal". Society-wide there are a lot of exchanges going on during such a holiday - but these are NOT transactions.

Brad's thesis is that this is a deeply-ingrained behavior in humans - much more deeply-ingrained than the transactional exchange systems we have developed in recent millenia.

[Jacques René Giguère]

"Nor do they bargain": the rule is you don't do accounting for gifts. Then you if do accounting and if the accounting is not vaguely equal, allowing for temporary fluctuations, you stop the exchanges.
"Gifts" may go from the strong to the weak, showing power. They may go from the eak to the strong, showing submission. Gifts may be equal, showing equality.
Gifts are transactions that exclude money (except from a grown-up to a young child.) Money is excluded because it is too easy, too obvious and prevent the giver from showing how well it knows what is really pleasing to the recipient.
Gifts are a way of testing the level and strength of a relationship. They must be appropriate. A gift may be too small or too large. Both of those cases may hurt the relationship, even break it.

[Tel]

If DeLong is not talking about transactions, then why does he use the phrase "reciprocal gift-exchange relationships" that sure sounds transactional.

So what's the deal with "We like neither to feel like cheaters nor to feel cheated,"? That sounds like some sort of value accounting is happening here.

Then there's this:

And it is on top of this evopsych propensity to be gift-exchange animals--what Adam Smith called our "natural propensity to truck, barter, and exchange"--we have built our complex economic division of labor.

So apparently DeLong is attempting to build an entire economy on "gift-exchange", which surely must be transactional. How many factory workers think of themselves as gifting their labour? How many companies account for their payroll under "gifts".

Words have meanings, and "gift" cannot be equated to "exchange". The reason we use specific terms to refer to specific circumstances is because it does matter.

[Nick Rowe]

Tel's got a point. But maybe "gift exchange" should be understood as something in between pure gift (only care about other, expecting nothing in return) and pure exchange (only care about self). "Black white" means "grey".

[Nick Rowe]

IRS does not necessarily mean "big firm". The only variety store or gas station in a small village will probably have a downward-sloping demand curve, and downward-sloping ATC curve. The division of labour is limited by the extent of the (local) market.

[anon]

> If DeLong is not talking about transactions, then why does he use the phrase "reciprocal gift-exchange relationships" that sure sounds transactional.

It's a reciprocal exchange of _equity_. A valued gift/favor/etc. buys you a boost in equity (in both the financial and the everyday sense!) which the other person can redeem by giving you a gift/doing a favor of her own when (and if) the best opportunity arises. There is no "precise accounting" because that would lead to a _debt_-like relationship, where not returning a gift would amount to default (i.e. a severe violation of norms).

[Antti Jokinen]

Nick: Yes, ours is a "grey economy" :-)

You probably know that I'm writing a blog called (Impure) Gift Economics. What I'm arguing -- based on research I've made full-time for three years -- is this: our economy is a "multilateral gift economy with explicit record-keeping" (you might get an idea of it by reading this post. Alternatively, just read forward.)

What I mean by the horrible line "multilateral gift economy with explicit record-keeping" is this:

Take a gift economy and make it fully multilateral -- if it isn't already -- so that Andy can give a gift to Betty and get his "counter-gift" from Carol, while Betty will give a "counter-gift" (counter to the one she received from Andy) to Danny, and so on.

Imagine that the people in the economy rely on shared norms and (would really like to) trust each other to give and take in a balanced way (~lifetime budget balance). But behind the scenes, there is (implicit) record-keeping going on, because free-riders need to be spotted and corrective action needs to be taken. People have eyes, they have memory and they talk to each other.

For a smaller community this is a nice system which, among other benefits, solves the "double coincidence of wants" problem. But it obviously wouldn't work in a larger community.

Unless we made the record-keeping very explicit.

We could have the "gift-giver" (seller) and "gift-receiver" (buyer) agree on the value of the gift. They could price the gift in terms of an abstract unit of account (I have a post on this abstract UoA).

A third party (a book- and gate-keeper) could then keep records of the gifts people have given and received, by crediting the account of the giver (seller) and debiting the account of the receiver (buyer) with the price they have agreed on. The third party could be given certain rights, so that it could try to enforce a "lifetime budget balance" (say, by limiting negative balances).

What is especially good about this kind of system is that we don't need money to solve the "double coincidence of wants" problem. We only need people to trust the institutions involved -- not only the "third party" but the government, too.

This kind of system would allow economists to finally write a theory of exchange which would not only make more sense than the previous ones, but would actually describe the real world.

The good news is that we can adopt this system by adopting a new perspective on the existing monetary system. What might look like two different systems are actually one and the same system, seen from two different perspectives.

I don't know if DeLong has a similar system in mind?

[Jacques René Giguère]

Gifts must be reciprocal as they create a relationship..
When we render a small service, like holding a door, the beneficiary says "Thanks" to which we reply"That's nothing." It indicates that no relationship is created as there is no expectation of a future exchange.
Market transaction are not expected to create a relationship,though mutual satisafaction is nice and can induce the participants to favor one supplier-client over another in the long run.
A gentleman can invite a lady to a restaurant and theater (the first two part of a date being alimentation and distraction leading hopefully to the third part , affection.) We don't give money. You can't offer money. If you do, you're not a gentleman. If she accept, she's not a lady.
The non-money to money, the reciprocal to non-reciprocal spectrum is large and diverse and varies from society to society, including animal ones. (Male chimpanzees are known to kill vervet monkeys in front of female chimps and offering them the meat. It seems the killing is the distraction part of the event and eating is the alimentation. Affection follows.)

[Wonks Anonymous]

I believe Narayana Kokerlakota wrote a paper on money as an equivalent to memory, or Jokinen's record keeping.

[Andrew Lister]

How about this classification:

1. pure exchange: I'll give you this if you give me that.
2. pure gift: I give you this anonymously (so you can't give me anything in return, and can't even thank or recognize me)
3. instrumental reciprocity: I give you this, hoping that you will respond by giving me that.
4. strong reciprocity: I give you this even though you can't give me anything in return, but on the assumption that you would do likewise for me were our positions reversed.
5. indirect reciprocity: I give you this not expecting anything in return from you, but expecting you will do something similar for someone else, who will do something similar for someone else... and that in the long run I will also benefit from this general practice of mutual aid.
6. strategic gift in the context of generally accepted norms of mutual aid: I give you this big thing to exhibit my power and put you in my debt (or this small thing, to ingratiate myself with you).

3 involves purely self-interested motivation; 4 involves ethical motivation, and/or concern for relationships; 5 mixes the two; 6 can be purely self-interested but presupposes general recognition of social norms.

I think DeLong is supposing that we are wired by evolution for 4 as well as 3, leading to 5.

[Nick Rowe]

Andrew: yep, something like that. Lots of gradations between the extremes.

Maybe need to think about it from the recipient's perspective too.

[Antti Jokinen]

Wonks Anonymous: Yes, Kocherlakota's paper "Money is Memory" (1996/1998) is very much related to this. I refer to him in my second post. In my third post, I refer to Ostroy's much earlier paper (1973) which Kocherlakota built on.

The difference between Kocherlakota and me is that I make much more explicit claims about our actual monetary system, focusing on the actual accounting. I show how that system works as "memory". To see this in a clear way, one must get rid of the concept 'money' altoghether.

At least in the working paper version Kocherlakota seemed to be confused. He wrote:

"The money that John receives from Mary is merely a way of letting Paul know that John has fulfilled his societal obligations and given Mary her apples."

This is not how it works. Paul cannot know that John has fulfilled his societal obligations. Paul only cares for the record of him himself having given a gift of goods, and the cash he receives from John serves as that record. John could have got the cash by borrowing it.

(Of course, no one working within an "equilibrium framework" can get this right, because if they got it right, they would need to abandon that framework. Or so it seems. Money as properly understood cannot function as a numeraire, and this makes the price level indeterminate.)

[rsj]

Many different kinds of gift exchange:
http://www.zerohedge.com/news/2016-12-21/meet-ny-pension-head-portfolio-strategist-traded-hookers-and-blow-allocations

[Antti Jokinen]

Nick: I focused earlier mainly on my own agenda (thanks for turning a blind eye on my advertising it!). Now that I read your post for the third time, some new thoughts emerged. For instance, I got stuck on something which probably has little bearing on your main message:

"Those of us in rich successful societies are in the hunt stag equilibrium. We have escaped the hunt hare equilibrium of the State of Nature"

Even chimpanzees get together to hunt colobus monkeys, having no chance to catch one alone. Their debt is probably to the same ancestors as is our debt you mentioned?

Well, you probably meant that in rich societies we are closer to some kind of full "hunt stag" equilibrium, whereas societies where trust (in strangers, in government) is not as widespread see less stag-hunting and more hare-hunting. Did you?

Thanks for the posts and discussions in 2016! And thanks for giving me, in the form of your comments section, a platform to communicate my ideas and put them to a test :-)

[Antti Jokinen]

Btw, Nick: I'm planning a devastating attack on your distinction between "red money" (an overdraft) and a "bank loan" ;-) It would help if I knew what exactly I'm attacking.

Are you saying that one doesn't get rid of an overdraft (liability) by first acquiring "green money" and then using it to pay off the overdraft debt, while this is the case in a more traditional bank loan? Is this the decisive distinction? Or are there others?

[Nick Rowe]

Antti: thanks!

"Well, you probably meant that in rich societies we are closer to some kind of full "hunt stag" equilibrium, whereas societies where trust (in strangers, in government) is not as widespread see less stag-hunting and more hare-hunting. Did you?"

Yes. Something like that.

"Are you saying that one doesn't get rid of an overdraft (liability) by first acquiring "green money" and then using it to pay off the overdraft debt, while this is the case in a more traditional bank loan? Is this the decisive distinction?"

Yes. I think so. I can pay down my overdraft directly by selling (e.g.) my labour (if my paycheque gets directly deposited into my bank account). I can only pay down my mortgage indirectly by selling my labour.

[Antti Jokinen]

OK. Perhaps I could put your message about rich societies like this: In poorer societies, lack of inter-kin/inter-tribe trust limits the number, scale and sophistication level of feasible staghunt strategies. To establish that kind of trust one would need institutions (enforceable contracts, etc) that are either weak or missing.

So it's not that they hunt hare, but they hunt stag with bows and arrows, while in richer societies we have first of all someone feeding the stags throughout the year and then selling hunting licenses to strangers with rifles, rifle scopes, Gore-Tex gear and a quad to pull the carcass out of the woods and into the hunting camp where it can be butchered in full daylight even though it's night -- thanks to a diesel generator and some floodlights.

About overdrafts... I think I got it. So, a mortgage can be directly paid down ONLY by first acquiring "green money", while an overdraft CAN be paid down directly by selling goods or services. The latter can also be paid down directly with "green money" (for instance by depositing cash), but it's just one option. Correct?

I'll let you know when I have published my attack ;-)

[Nick Rowe]

Antti: yes and yes.

Except: "To establish that kind of trust one would need institutions (enforceable contracts, etc) that are either weak or missing."

Well, those institutions are themselves just a name for one part of the "hunt stag" equilibrium. Those institutions don't have any concrete existence (though they feel very concrete to us as individuals) apart from individuals (judges, cops, etc.) acting that way. They aren't mechanical things (unlike the Domesday Machine in Dr Strangelove).

[Jacques René Giguère]

I strongly recommend that book "Social traps and the problem of trust" by Bo Rothstein
It was inspired by a Russian asking Rothstein, a swedish politicologist:"How can we transform Moscow into Stockholm?" The answer is it took a long time to transform Stockhom into Stockholm. Few people realise how Scandinavia was a bacward,corrupt and badly governed,administered and managed till the early XXth century compared to more southern Europe (the hard-working, well governed Nordics is a recent myth according to Nordics themselves....)
The review by Elinor Olson is excellent.
https://www.amazon.ca/Social-Traps-Problem-Trust-Rothstein/dp/0521612829/ref=sr_1_18?s=books&ie=UTF8&qid=1483203930&sr=1-18&keywords=the+problem+of+trust

[Antti Jokinen]

Nick, I fully agree. I didn't mean (even if I said so!) that one can establish some agencies and those then bring the trust needed. It's a delicate process. Trust strengthens the institutions which strengthen trust, and so on.

Jacques: Thanks for the tip! As a Finn living in Norway I find that subject especially interesting ;-)