In my previous post I linked to Mencius Moldbug’s argument that a sufficiently strong utterly rapacious state may be desirable to live in. The less explicitly unpleasant version of “Fnargocracy” is sometimes referred to as a “Vertically Integrated Proprietary Community”, and I’ve linked to Peter Leeson’s argument against Ed Stringham on their desirability a number of times before, along with Eric Crampton on the socialist calculation problem making non-benevolent autocrats more bearable. We don’t have to leave it at that though. Recently via the Sociological Imagination I came across Leeson & Stringham’s survey on the economics of anarchy. In it they referred to Mancur Olson’s argument on behalf of the “stationary bandit”, as well as a critique of it by some folks I had never heard of before, Boaz Moselle and Benjamin Polak. Their paper is A Model of a Predatory State, which points out that maximizing tax revenue is quite different from maximizing the welfare of subjects, and so a primitive society may be worse off in transitioning from organized banditry to statehood. The authors mention that in their working paper from 1999 they give similar results over the long-run with a dynamic learning model, but I haven’t been able to find that online. UPDATE: I think this might be it, though it is dated to 1997 (apparently developed from a 1994 paper with a different title).
On a slightly related note, I’ve gone around asking people what they think if Nick Szabo’s criticism of anarcho-capitalist usage of the Coase Theorem. His primary target, David Friedman, has not responded but Johnnie Lin did with a paper of his own.
January 31, 2011 at 5:39 am
The Lin paper mentions neither myself, Friedman, Olson, or Coase, so I wouldn’t characterize it as a response to any of the arguments by or about us. Rather it seems written from another, seemingly completely separate academic tradition. It thus takes some extra effort to decipher and after such a deciphering I fail to see where it refutes anything Olson or myself have said.
This tradition involves analogs from ecology which is a very useful approach. Predator/prey models seem to correspond to roving bandit models: many predators and many prey without territories. However, a good analog to a stationary bandit with many prey (or hosts) seems hard to find. Perhaps an ant colony (which can be treated as a single organism) domesticating numerous aphids. I don’t see that covered in the Lin paper though, albeit I didn’t try to decipher every sentence so I might have missed it.
As for Moselle and Polak, they seem to completely misunderstand Olson’s roving bandit model of anarchy. Rather, they just assume it away by making arbitary assumptions about anarchy that are contradicted by Olson’s theory as well as historical fact. Olson’s assumptions on the other hand involve straightforward reasoning about what bandits competing with each other for the same victims will do. They will leave hardly anything to the victim because that would just mean another bandit will get it, whereas a stationary bandit will leave the victim with sufficient capital so that the same victim can be harvested again in the future. So very straightforwardly the total Laffer maximum rate of the roving bandits is far higher than that of the single stationary bandit, contrary to the GIGO argument of the paper.
The authors choose to ignore Olson’s straightforward analysis of motivations in favor of launching directly into a blizzard of mathematical jargon about convexity, concavity, and so on which have everything to do with the mathematics of polynomials and nothing too with what actually motivates people. In other words, they simply make the mathematical assumptions that give the answers they want without taking the time to analyze whether those assumptions actually make any sense. GIGO.
Also missing are references to actual history, such as the history of the American Southwest I discussed, which amply confirm Olson’s model. We know from history that in large regions (e.g. most of Europe east of the Fulda gap) fixed farming was abandoned in favor of nomadism, despite very good agricultural soil in much of these areas (e.g. Ukraine). In other words, lacking large stationary bandits protected by geography, people quit farming entirely and reverted to nomadism and hunting-gathering. The roving bandit theory also explains the Dark Ages and much else of history, whereas these author’s models explain nothing.
BTW, I am currently reading Njal’s Saga, which contains much detail about the Icelandic legal system. To the extent the Saga reflects actual historical events, I’m afraid ddfr’s account is wearing rose-colored glasses.
In brief Iceland had a judicial and quasi-legislative branch (the Althing) but no executive branch — you had to collect your own posse to enforce a judgment or to prevent a judgment from being enforced on you. Imagine a culture in which you had to aggressively social network, not to get a job, but to protect yourself from violence, theft, and other crimes and torts. Furthermore, throughout the saga (and this may be a stylized simplification of the author) the wergild (payments for killing as a tort) fail to satisfy the sense of justice of sufficiently powerful members of the victim’s families that the cycle of violence escalates until the final _Gotterdamerung_.
Furthermore, while there’s no official tax collection, certain events do suggest that the more powerful leaned on the less powerful to “give” them things, e.g. food during times of famine. Furthermore, large amounts of wealth were transferred in legal settlements and judgments where a member (and usually a leader) of a powerful posse sued a member of a much weaker posse and coerced a settlement.
There is a ton of literature, from Njal’s Saga to the Greek tragedies to recent Westerns, about the evils of revenge cycles and the blessings of a legal system and stationary-bandit executive to quell these cycles. Perhaps this is all just a long tradition of stationary bandit propaganda, but precious little has been written for the counter-argument. Any references literature (near-)contemporaneous to the society described but written from an anti-stationary-bandit point of view appreciated.
In short, I recommend reading Njal’s Saga and other (quasi-)historical accounts directly and coming to your own conclusions instead of papers that confuse the mathematical properties of polynomials for human motivations.
January 31, 2011 at 12:47 pm
Nick, as I recall you’ve written elsewhere how much cheerleading there is for Roman and related ideas of law flowing from the directives of one arbitrary overlord. The sociology behind this is unclear, but seems clear to me that the enthusiasm for arbitrary central authority there is all out of proportion to how successful the idea has been. It seems very plausible to me that the sociology of cheerleading for central suppression of independent authority to deal with lawlessness could be similar, leading to similar disproportion.
Also, I never studied it much, and it has been many years since studied it at all, but I vaguely remember being impressed in some Icelandic primary source (probably Njal’s Saga) and in _Order Without Law_ the stability considerations that seemed to be built into some of the revenge traditions described there. From memory, the Icelanders had scary balls-to-the-wall customary requirements on what kind of secrecy and unaccountability were acceptable in revenge — my mental image after these years is something like having to find people nearby and declaim to everyone “I killed Joe for $such_and_such_offense” instead of just vanishing into the night like a prudent ancient Saxon, prehistoric New Guinean, or modern antiterrorist. And again from memory, the California farmers had taboos against enriching yourself in revenge: when your neighbors straying cattle damage you, you should discourage him from doing this e.g. by going to the trouble of driving the strays someplace where it will be inconvenient for him to get them back, rather than just shooting the strays and putting them in your freezer. So while I agree that history and indeed current events support the idea that revenge and bad faith can easily spiral out of control, there has been at least some attention paid to informal antispiral norms which can develop in the absence of a central authority with recognized moral authority to impose them.
Note also that we have a lot of recorded history of nations coexisting with each other with no recognized central authority to crush revenge cycles. Revenge and bad faith cycling out of control does happen internationally, but it doesn’t seem to me that it dominates international relations more than domestic relations. Thus any broad claim that revenge will necessarily spiral out of control without central authority seem to me akin to a broad claim that trade is an artificial blessing that follows from a central authority taming the state of nature — one really should address exhibit A, the historical record, before proceeding. And the truth may indeed look something like an emergent stability property of equations describing the situation — adjust the number (e.g.) of iterations of the prisoners’ dilemma and watch instabilities appear and disappear, even while human nature remains the same.
January 31, 2011 at 1:55 pm
how much cheerleading there is for Roman and related ideas of law flowing from the directives of one arbitrary overlord.
Stationary bandits don’t have to be, and usually weren’t, military dictators. “State” vs. “anarchy” is a false dichotomy. Most of history happened in between.
It seems very plausible to me that the sociology of cheerleading for central suppression of independent authority to deal with lawlessness could be similar, leading to similar disproportion.
But a disproportionate of many to zero? There is _no_ primary historical or quasi-historical literary source for anarchy working well. (For feudalism, federalism, and other such decentralized stationary bandit schemes yes. But for anarchy none). As opposed to many, many works from a variety of cultures portraying and decrying cycles of violence. And a bunch of modern literature about utopian (i.e. nonexistant dreamed-up) anarchy.
As a great believer in the wisdom of highly evolved tradition (as opposed to the conceits of modern utopians and cargo-cult “social scientists”), I must conclude that roving banditry has always or almost always turned out remarkably poorly.
the stability considerations that seemed to be built into some of the revenge traditions described there.
In Njal’s Saga it is portrayed as unstable, i.e. the cycle of violence escalates with ever-larger wergild payments that still fail to satisfy somebody in the opposing camp, leading to more revenge violence, etc. until Njal and his sons are destroyed in a conflagration.
Again I strongly urge reading the most primary historical sources possible and not the often highly distorted descriptions of modern utopians and cargo-cult scientists (as Feynman called social “scientists” who pride themselves on their physics-like mathematics).
California farmers had taboos against enriching yourself in revenge
The California farmers were operating against a backdrop of common law. Just the threat of lawsuit and criminal prosecution greatly changes behavior even if it is rarely carried out.
That said, I’d love to see somebody try anarcho-capitalist experiments on seasteads and watch them from a safe distance. Kind of how I’d like to see a one-way Mars mission as long as it’s somebody else going. :-)
January 31, 2011 at 9:28 pm
As I observed to a friend recently, the Icelandic sagas, although the society depicted has it’s good points, tend to go: “Axe murder, axe murder, lawsuit, lawsuit, lawsuit, axe murder, lawsuit, outlawry.”
January 31, 2011 at 9:36 pm
Sometimes the monotony is broken by “Go to Constantinople, join Varangian Guard, get paid big bux.”
(After reading the Sagas, I was glad that Bjork didn’t have an axe handy during that Thailand airport incident.)
How much weregild do you have to pay for chopping up annoying journalists?
January 31, 2011 at 6:13 am
Incidentally, it does turn out that per capita welfare — stated in purely economic terms, i.e. ignoring the greater violence of roving bandit societies — was a bit higher in roving bandit than in (pre-British-agricultural-revolution) stationary bandit societies. Which basically translates to saying per capita nutrition tended to be somewhat better in roving bandit societies. This was for straightforward Malthusian reasons, because lower welfare led to lower fertility, which freed up environmental resources equilibrating welfare at much lower population levels in the long term for roving bandits. Furthermore, even in a single genreration the greater violence of roving banditry freed up some extra environmental resources for the survivors. For example in an agricultural society the survivors of the increased violence could farm on better soil, thus extracting more nutrition with less effort.
These Malthusian effects are another crucial phenomenon ignored by Moselle and Polak in favor of exploring properties of polynomials which were discovered by mathematicians thousands of years ago and expecting that we will be all so impressed by such wizardry.
January 31, 2011 at 9:24 pm
http://books.google.com/books?id=EcFk5XKukwMC&pg=PA1146&lpg=PA1146&dq=%22Goodness+is+adorable%22+rebecca+west&source=bl&ots=huEMX3b-3Q&sig=rNgpWvRFgCz46MwITMhrhcyEJlk&hl=en&ei=B3xHTavDL4yusAPc7oikAg&sa=X&oi=book_result&ct=result&resnum=1&ved=0CBMQ6AEwAA#v=onepage&q&f=false
February 1, 2011 at 7:23 pm
When I said “response”, I meant to my comments. I doubt Lin had read your blog at the time he wrote that paper.
Moselle and Polak explicitly listed five assumptions (with some debatable justifications for why those assumptions are legitimate). Is one of them the issue, or an implicit assumption? They also maade some references to history, though it was mainly a theory paper.
As I read the paper, their references to sub-optimal measures taken to avoid appropriate reminded me of “The Art of Not Being Governed”. I don’t think their view on abandoning the best fields is that different from yours. The authors also claim that it explains why states arise in certain places (citing someone Scott says is completely debunked) and what they will invest in.
David Friedman cited Njal’s Saga here.
The Malthusian aspect does throw a monkey wrench in the works. In order to deal with you might have to grapple with the Repugnant Conclusion.
“Furthermore, even in a single genreration the greater violence of roving banditry”
Friedman claims Iceland was less violent than Europe at the time.
February 7, 2011 at 12:53 am
Pre-1991 Somalia was a stationary bandit society. Welfare improved when Somalia’s government collapsed, switching Somalia to a roving bandit society.
February 7, 2011 at 3:42 am
Nope, Somalia switched from larger-scale to smaller-scale stationary bandits. The warlords have territories every bit as well defined as a large county or small state. The only bandits one can characterize as roaming are the pirates, who have greatly degraded the welfare of shipping and sailors in the area.
February 8, 2011 at 5:13 am
I’ve long thought that the best system for Afghanistan over the next few centuries would be hereditary feudalism.
February 9, 2011 at 11:44 pm
I disagree, though I’m no expert. My hammer may be technocratophilia, but I think experts on Afghanistani governance would be better and even more stable for Afghanistan than hereditary feudalism.
February 10, 2011 at 12:13 am
Who are the experts on Afghanistani governance? The Taliban?
February 10, 2011 at 1:02 am
No, not the Taliban.
I like Azimi
http://www.chicagotribune.com/news/opinion/chi-0701210351jan21,0,7944559.story
and Ghani
http://en.wikipedia.org/wiki/Ashraf_Ghani_Ahmadzai
to a representative two.
February 10, 2011 at 1:02 am
should read “to name a representative two”.
February 10, 2011 at 10:45 pm
I had never heard of Ghani. I had heard of their banking ministers robbing the country blind, I was surprised to read they had the “best finance minister of Asia” at one time.
February 13, 2011 at 7:18 am
For epistemological transparency, I never heard of him either. After you snarked/asked, I went hunting for authentically afghanistani technocrats, and found those two guys.
February 13, 2011 at 1:21 pm
That makes me feel better in comparison. I often go googling in the midst of responding to someone, and sometimes note how I found the info or how recently I discovered it, but often I don’t.
February 13, 2011 at 3:55 pm
You shouldn’t feel that good about yourself. A population the size of the Afghanistani diaspora is going to members with technocratic credentials, just like any population that isn’t specifically nontechnocratically defined and numbers more than 100,000.
When you get to the scale of tens of millions, you’re going to have people with very strong technocratic credentials.
I don’t know the population of the Afghanistani diaspora off the top of my head, but I’m guessing 30 million Afghanistanis globally.
You replied to my good faith analysis not just with snark, but dumb snark.
You shouldn’t feel better about yourself because my rough bayesianism was buttressed by subsequent data-hunting.
February 13, 2011 at 3:59 pm
I pretty much nailed it:
http://en.wikipedia.org/wiki/Demography_of_Afghanistan
which shouldn’t be surprising, but I think it’s a bit nutty to think a population of 30 million isn’t going to have a decent share of people with solid technocratic credentials.
February 15, 2011 at 12:15 am
It’s not that surprising that there are some competent technocrats among the broader Afghan diaspora. It is surprising to find them at high places in the Afghan government (particularly finance), which I had been led to believe was irredeemably corrupt.
February 15, 2011 at 12:19 am
TGGP, I agree. It’s interesting to see the confusion of hegemonic proganda in our time (think of the schizophrenic mainstream narratives about what’s gone on recently in Egypt).
February 15, 2011 at 12:21 am
proganda should be propaganda
February 15, 2011 at 7:47 pm
There may be an element of schizophrenia with Egypt since we had been supporting one client and then turned on him when he seemed vulnerable, but for Afghanistan the narrative seems to have been more coherent (which is not to say accurate).
February 8, 2011 at 9:32 pm
nick, from what I heard Somali clan law is not based on territory (although clans restrict foreigners from owning land claimed by them). Wherever you travel, your clan membership and legal system go with you. “Polycentric law” is the term used. Taxation is supposedly prohibited under Xeer.
Not many people claim that something other than some universally best system of governance is appropriate for different places.