December 27, 2003

A Blurb

Ever since the invention of agriculture, human beings have had only three social-engineering tools for organizing any large-scale division of labor: markets (and the carrots of material benefits they offer), hierarchies (and the sticks of punishment they impose), and charisma (and the promises of rapture they offer). Now there is the possibility of a fourth mode of effective social organization--one that we perhaps see in embryo in the creation and maintenance of open-source software. My Berkeley colleague Steve Weber's book is a brilliant exploration of this fascinating topic.

Steven Weber (2003), The Success of Open Source (Cambridge: Harvard University Press: ).

Posted by DeLong at December 27, 2003 06:45 PM | TrackBack


Well, Amazon says it is not yet released so I do not know what the book says about it. Would be interesting if Brad could post his view/summary on what is different about it. Personally, I am all for rapture :)

Posted by: Leopold on December 27, 2003 07:24 PM


Well, I've not read the book, but your claim that there have been "only three social-engineering tools for organizing any large-scale division of labor" cries out for counterexamples, so I've been trying to think of some. Here are some possibilities:

Scientific and other academic research -- it seems to me that this differs little from the open source software model. The incentives are surely similar.

What about large-scale communal efforts in times of crisis? (tragically topical, given the Iranian earthquake). The "spirit of the blitz" and mutual aid surely involves different incentives to the market, hierarchy, and charisma. Or does the idea of mutual aid fall under your definition of markets? That would be stretching things. Perhaps these efforts are not permanent enough to count?

Perhaps forced labour is distinct. My guess is that you would include forced labour under hierarchy, although much labour in "market" societies has been as near to forced as makes no difference (the Ragged Trousered Philanthropists and the inhabitants of Upton Sinclair's Jungle didn't see much free exchange).

In fact, the separation of market from hierarchy is an interesting one. I wonder what fits where?
My guess is that you are talking of organization at the national level, but surely it is disingenuous to classify working on the floor at Wal-Mart as market (which I guess you would) while working in Cuba is hierarchy. I would guess that Wal-Mart is in many ways a bigger organization than the nation of Cuba?

But to get back to the topic, I haven't read the book, but I suspect the novelty of the Open Source model is sometimes overstated by its adherents.

Posted by: Tom Slee on December 27, 2003 07:26 PM


I agree about the academic model! I don't know why people get all goofy when it comes to open source software. I mean, it's cool, and it's fun when people like your software and tell you, but let's not get carried away. we've been hearing about the open source revolution since 1998 or so, and it still just means linux, which itself is a high-quality implementation of an operating system that's over 20 years old! yeah, I know about apache, which is great, and I love the wikis, but really, nobody is using abiword or even mozilla. open source is great for programmers (all hail emacs!), but it's just a few baby steps further along to being acceptable for non-programmer general use.

of course, 20 years from now, who knows?

Posted by: c. on December 27, 2003 10:23 PM


The author has been reading his Weber. Those are his three principles of authority.

Posted by: Ian Welsh on December 27, 2003 11:46 PM


Ayyah! At long last Bradford DeLong the economist focusing on productivity, public policy, economic history and tomorrows economy takes up the open source process!

Folks, you got to know this, to appreciate the potential of open source:

Bill Gates thinks it is Unamerican!

But IBM doesn't agree with him.

Posted by: Bulent Sayin on December 27, 2003 11:58 PM


c: I'm using Mozilla to write this reply.

Posted by: Walt Pohl on December 28, 2003 12:09 AM


I tend to think that people are starting to drift away from empirical reality in their thinking on Open Source. The reality of software development on the Open Source model is pretty far from the romantic idea many people appear to have, viz. thousands of programmers hacking away at the code in a loosely-coupled way, with no central control structure. The reality is that Open Source projects are run very similarly to commercial projects (there are nearly always project managers, and there's quite frequently a very small group of people with veto power over all changes (think of Linus Torvalds)), and that even a relatively successful Open Source project has only a handful of active developers over the course of its life.

I can certainly understand why economists would feel enthusiastic about the romantic view of Open Source -- it feels an awful lot like a market, with a bunch of on-the-ground, decentralized, self-interested decisions driving the larger system ever closer to some optimal state. Unfortunately, the costs of entering a project as a contributor are very high, particularly given the low-level languages and idiosyncratic coding styles employed in many Open Source projects (coders: have you ever looked at the Apache code? the Mozilla code?). The theory of Open Source development many non-coders have latched onto appears to me to require the questionable assumptions that it's very easy to begin contributing to a software project, and that modern software development is the sort of undertaking that requires little central coordination.

Anyway, don't listen to me; listen instead to the much smarter Kieran Healy:

Posted by: Chad K on December 28, 2003 01:44 AM


Thank you for the link, I began reading it. The following sentence caught my attention:

"... We suggest that the gulf between active and inactive projects may be explained by social-structural features of the community which have received little attention in the existing literature...."

Posted by: Bulent Sayin on December 28, 2003 03:20 AM


By the way, what do advocates and participants of of OSS in America think about outsourcing of informatics jobs by IBM Oracle etc?

Posted by: Bulent Sayin on December 28, 2003 03:38 AM


From Healey and Schussmann, link by Chad above:

".. It (OSS) is a hybrid: part social movement, with idealistic principles and goals; part formal organization, with an intensive schedule and innovative products; part volunteer network, with time and energy to donate....."

By the way, I have a hunch that OSS people on one hand, and, on the other hand, the people who are intensively involved or interested in NIH National Health Institute process for placement of research funds, should pay attention to each other. Both communities may benefit from such attention.

Don't underestimate NIH. Currently it has a budget about twice as big as that of NASA.

Posted by: Bulent Sayin on December 28, 2003 03:51 AM


I will think carefully about this issue, for the wish is that technology transfer to developing countries will not be impossibly limited by increasing protection of intellectual property. The difficulty of developing a drug industry in Africa should be a matter of concern to us all.

Posted by: anne on December 28, 2003 04:52 AM


Bulent Sayin

In the strongest terms, I recommend reading the New York Times thoroughly and regularly. To compare the NYTimes with other papers can be useful, but there is no substitute for the NYTimes no matter a political bent.

Posted by: anne on December 28, 2003 05:07 AM


New models matter -- a lot.

But I'm wondering...the three social-engineering tools Brad cites (markets, hierarchies, and charisma) sound like a new twist on Weber's old categorization of forms of domination: patriarchal (taken, by an economist, as hierarchical/coercite), charismatic (a category that essentially tries to categorize the unpredictable and uncategorizable?), and rational (which an economist happily equates with markets?).

(I'm probably picking up on this by Brad's use of "charismatic" as one of the categories...).

Everyone probably agrees that (human) reality is more complicated than any three basic forms, so suggesting further models may be beside the point: but one might mention the "communal" -- seen especially in city-state contexts, particularly in post-1348 Europe, aggressively anti-hierarchical in ideals, also anti-market and deeply distrustful of charisma. It reached a kind of ideological high point in the rhetoric around the early Reformation and German Peasants' War before being sidelined (quite possibly for essentially economic reasons, tied to problems of scale that had always dogged city-states), and lost much of its utility as a social-engineering toolkit. But it had its day...

Posted by: PQuincy on December 28, 2003 05:18 AM


"I will think carefully about this issue, for the wish is that technology transfer to developing countries will not be impossibly limited by increasing protection of intellectual property. The difficulty of developing a drug industry in Africa should be a matter of concern to us all."


Left wingers in advanced industrial countries should (a) oppose exports to of old technology at high prices and (b) encourage export of advanced technology at low prices.

That policy would further spur productivity growth both in advanced and developing coutries and make the world a cleaner, better place to live -- I think.

OK, I'll overcome my laziness and fill out that form on NYT web page so that they'd let me read it.

About NIH funds placement process: To me it is what comes closest to the vague picture of direct democracy I seem to be developing in my mind.

Posted by: Bulent Sayin on December 28, 2003 05:41 AM



Please develop your "communal" metaphor more fully. I am quite interested. Be simple or story like for I am not a technology buff.


Posted by: anne on December 28, 2003 05:56 AM


"Advanced industrial countries should (a) oppose exports to of old technology at high prices and (b) encourage export of advanced technology at low prices.

"That policy would further spur productivity growth both in advanced and developing coutries and make the world a cleaner, better place to live -- I think."

Agreed completely. Trade that simply entails raw resouces or low skill fabrication for mere "cash," is not enough. By hook and crook, Americans gained hold of European technology and improved on it during the 1800s. Nigeria or Brazil or Vietnam must gain technology in trade. A problem can be a lack of incentive for a government, especially a non-democratic government to bargain over technology exchange.


Generally I begin the NYTimes with the arts section, then go anywhere from international to editorial to books. Science and health are always worth reading, and there are enitre sections on Tuesday. Arts is special on Friday and Sunday. Books on Sunday. Dining Tuesday, Technology thursday. But, the subjects are added to each day. Yes, I grumble about articles now and again. I have favorite critics. So what?

Posted by: anne on December 28, 2003 06:17 AM


Bulent Sayin

Thanks for the fine comments. I did not catch them till the screen refreshed. Perhaps there is a way to refresh I do not know of. I will ask. Another note about the NYTimes. Book reviews and some articles from all sections are kept open for years to months, but many stories are closed after 14 days. So, I cut important articles from my home copy or e-mail a full article to myself.


Posted by: anne on December 28, 2003 06:45 AM


Oh my.

There are four general categories of ways to create belief -- repetition, authority, apriority (it agrees with what you already believe), and rigour (you can test it.)

'The community of science' was the first successful approach to social mediation by rigour; you have to be able to demonstrate your assertion in a way independent of the contents of your head. It doesn't work ideally well, because people aren't evolutionarily hardwired for rigour, the way we are for repetition, authority, and apriority, but it does work.

Open Source, to the extent that it works, relies on rigour, and that's the fundamental difference between an Open Source project and a commercial one -- things which people do not care to work on, which people do not use, or which the maintainer does not like, are thrown away. You can put months of work into something and have it refused inclusion on a large project, or have it fail to achieve widespread use in independent release. Businesses can't afford to find utility like that, because (from a programmer hours standpoint) the process is horribly inefficient.

[The whole desktop usablility claim is heavily tied up with familiarity, on the one hand, and support effort, on the other. In environments where there's anything like a support structure for the open source desktop stuff, people use it just fine. As more such environments come to exist, the better the feedback into the projects on what sorts of things those users want.]

I'd strongly recommend Jane Jacobs' Systems of Survival and Stafford Beers' Platform for Change to anyone interested in issues of technologically enabling social change.

I'd also point out that one of the core causes of the current difficulties is the enormous resistance on the part of managerial classes to actually using the improved communications technology developed over the last thirty years to automate managerial functions like resource allocation and task identification. If that were being done, management would be a rapidly shrinking proportion of the workforce, instead of a stable-or-growing one.

Posted by: Graydon on December 28, 2003 06:47 AM


Well, I pressed refresh and this screen did refresh. Wow. I can learn.

Posted by: anne on December 28, 2003 06:57 AM


Graydon, last paragraph above:

That has already happened in auto industry in and after 1970s, I think.

Once top management in Detroit and AFL-CIO understood and agreed that the name of the game was "automate or evaporate", they found a new layer of resistance: Middle management.

But that obstacle was eliminated through top-down approaches, down-your-troath top management tactics.

I guess it is time now to develop entirely new, participative management processes and get rid of "top management too", this "star system" type of management processes that is giving us examples of greed and corruption on a daily basis.

Posted by: Bulent Sayin on December 28, 2003 07:13 AM


Lessening layers of management can as well mean fostering more general worker autonomy and control. No wonder there is reluctance even by top manages to encourage more technology use. Ever look at the costly absurdity of medical supply acquisition?

Posted by: anne on December 28, 2003 07:35 AM


There is a time when a difference in quantity creates a difference in kind, but there have long been human activities where folks coordinate their activities to create public or club goods. The maintainance of free markets, effective governance, creating open standards, nonprofits, rules for sports, etc. etc. are all examples of this.

Open source is a difference in quantity. We draw from a larger sample space of volunteers - increasing the chance of finding talented nonrival contributors. We have better coordination tools. We are less easily coop'd by distribution bottlenecks, i.e. we are harder to make excludible.

Posted by: Ben Hyde on December 28, 2003 08:29 AM


Open source in theory might reduce Transactions Costs of other tasks, making it a type of Institution (which are all supposed to be "Hierarchies", I think: but maybe hierarchies need not be of humans, only of chains of electronic digits). Long discussions might ensue upon (1) anarchy as a form of government, or counter-government, (2) the long dismissal by economists of matters of altruism, (3) what the open source writers have to do for a living, and (4) the proper length and breadth of intellectual protection.

Computers are hyped-up forms of simple language, and if open-source really gets rolling, it may take on some attributes. It would be ridiculous for me to try to invent a usable word like "elephant", and attempt to sell it to you all. The word (or phrase) "catch-22" --perhaps we've all used it--derived some income from a publisher, since it was first fixed in a novel.

Posted by: Lee A. on December 28, 2003 09:10 AM


Forgive the change but so very important -

Millions of AIDS Orphans Strain Southern Africa

CEMENTO, Mozambique — On the day last July that he and his wife died of AIDS, 36-year-old Samossoni Nhambo leaned up from a hospital bed a few miles from this dirt-road village of thatched huts and asked his preacher a despairing question: who would take care of his children?

Five months later the answer is glaringly obvious: no one.

Three-year-old Fátima died in early December, perhaps from AIDS, perhaps from malnutrition. Five-year-old João, infested with worms that have reduced his toes to red stumps, can walk only on his heels. His 7- and 9-year-old brothers, Ricardo and Samsoan, are covered with dime-size sores from scabies mites, which infect the entire family.

Sixteen-year-old Maria, who dropped out of school to care for her sick parents, was pregnant by a man whom she refuses to identify, and in early December she gave birth to a boy.

That leaves the eldest, José, a slim, short 17-year-old who just finished seventh grade, as the surrogate father. In their half-built shelter of stones and sticks on the bad side of a poor village, with no walls and a single cane chair for furniture, the Nhambo children reel from crisis to crisis.

"Life is very difficult," José said. "No food, no clothing, no bed covers. We have to struggle."

So do millions of others like them. Southern Africa is increasingly home to children like the Nhambos, robbed of their childhood by AIDS and staggering under adult-size hardships.

The United Nations Children's Fund estimates in a new report that 11 million children under 15 in sub-Saharan Africa have lost at least one parent to AIDS. About a third of them have lost both parents. By 2010, Unicef predicts, AIDS will have claimed at least one of the parents of 15 percent of the region's children — 20 million in all.

The social implications are enormous, Unicef and other relief organizations say. Orphans are more likely to drop out of school, to suffer from chronic malnutrition, to live on the street, to be exploited by adults, to turn to prostitution or other forms of crime and to themselves become infected with H.I.V., the virus that causes AIDS.

African social traditions dictate that relatives should take them in. But AIDS has pushed so many families to the brink that the surviving adults are beginning to turn away their young relatives. An aunt and a grandfather live down a dirt path from the Nhambos, but the grandfather says neither can help them.

So far, governments have done little either: of 40 sub-Saharan countries hit by the AIDS epidemic, only six have plans in place to deal with orphans, Unicef says. Their sheer numbers, plus the state of African bureaucracies, make even the simple act of registering orphans so that they can be exempted from school fees an enormous task. In countries like Zambia only 1 in 10 births is even documented.

For Mozambique, a nation nearly twice the size of California that stretches along the Indian Ocean coast, orphans are not a new phenomenon. When 17 years of civil war finally ended in 1992, hundreds of thousands of children were left without one or both parents.

But AIDS has sharply multiplied their ranks. Now, in a nation of 18 million, 16 percent of the children — more than 1.2 million — are already missing at least one parent. AIDS is responsible for the plight of a third of these children, according to Unicef.

Maria Cemedo, an official of an agency that serves women and children in Sofala, the region where the Nhambo children live, said an entire generation was being lost. "We may become a society of old people and children," she said.

Sofala, in Mozambique's narrow center, has been particularly hard hit because it has both a port and a major highway running to Zimbabwe. The combination of poverty-stricken women and lonely truckers spreads the AIDS virus all along the corridor. Now 1 in 4 adults in the province is infected.

Of the 46,000 registered orphans in the province, said Antônia Charre, the agency's director, few receive any government help. Fewer than 5 percent obtain food through the World Food Program, she said.

"It's a shocking situation," she said. "It is not clear how some of these children survive from one day to the next." ...

Posted by: anne on December 28, 2003 09:21 AM


>>But I'm wondering...the three social-engineering tools Brad cites (markets, hierarchies, and charisma) sound like a new twist on Weber's old categorization of forms of domination: patriarchal (taken, by an economist, as hierarchical/coercite), charismatic (a category that essentially tries to categorize the unpredictable and uncategorizable?), and rational (which an economist happily equates with markets?). (I'm probably picking up on this by Brad's use of "charismatic" as one of the categories...).<<

It's not supposed to be a new twist! It's a straightforward borrowing (except that I tend to see bureaucracy as a variant of patriarchal authority rather than as one of two branches of "rational" authority)!

Posted by: Brad DeLong on December 28, 2003 10:56 AM


We're all engaged here in some sort of open source process, by the way.

We too are coding programs -- though of a different kind.

Mein Gott there is so much to do!

For example, develop specs for a process of bureaucracy without the bureaucrats!

A code of program, by the way, is nothing but specs.

We are all spec writers!


A spec writers' civilization!

If we could only define how it is supposed to work!?

Posted by: Bulent Sayin on December 28, 2003 01:15 PM



It's odd that an economist should miss the fact that doing-more-with-less, mediated through trade, brings about division of labour.

If you have antelope shit in your yard and I have yams in my hummock, I will trade you yams for fertilizer, you will tend to collect more fertilizer, and get some of my yams as reward.

We have both tended in the direction of specialization -- but this is not an example of your market, it is something different: the physical reality and the exchange of information about the reality are both prior to the construction of the market.

Hence you are forbidden to use the phrase "the invention of agriculture" again until you have read your Jane Jacobs.

Posted by: David Lloyd-Jones on December 28, 2003 02:28 PM


Here is an example of an OSS project that appears to have succeeded:

"Stellarium renders 3D photo-realistic skies in real time with OpenGL. It displays stars, constellations, planets, nebulas and others things like ground, landscape, atmophere, etc."

Sourceforge Link:

Stellarium is the work of one sole person, as I understand it: Fabien Chéreau

Only one person developing an OSS project is the typical figure, according to Healy and Schussmann paper, link given by Chad above.

This is a free software, you can just download it and you have a planetarium on your computer screen.

An American company recently developed a portable dome projector for a price of about 15 grands and it is using the Stellarium software, again free of any charges.

Yahoo news link on this:

A conventional planetarium costs about two million dollars to build, at least.

So why does Chéreau do it?

Posted by: Bulent Sayin on December 28, 2003 03:29 PM


The feudal system is notorious for being hierarchical, but in actual fact hierarchy was emergent behaviour. On its own it could and did deploy efforts even in those cases where there was no hierarchy around, and even when there was a hierarchy around it was merely giving overall direction to efforts that were applied by the distributed processes of feudalism proper. Hierarchy wasn't mobilising the resources.

Posted by: P.M.Lawrence on December 28, 2003 05:04 PM


Open source is an improvement in LANGUAGE AND COMMUNICATION, as computers themselves are, at another level. In this sense, the fourth mode of effective social organization, although different from the three modes by human master and commander, has been concomitant with us forever.

To picture this, I wish to theorize that every form of organization is made up of at least one single-step hierarchy. (In this sense, the words "organization", "hierarchy", and "institution" will be nearly synonymous.)

Any such pattern will give division to labor, no matter what.

Blog is a good example, since each person adds her/his stuff to the round of ideas. But we are all using English, and we try to be polite, and to think clearly, i.e. there are central rules to which we loosely must ascribe. The differences of lifepath and experience give each one of us a special concentration of knowledge and insight--so an intellectual division of labor repays the whole. Moreover the good professor holds our court, pays money and time, and weeds out the nutshells. He is our institution, reducing our (non-monetized) transactions costs. It is (almost perfectly) a non-coercive single-step hierarchy with the center being “good rules + host”, and sometimes it goes on to build peripheral structure. That is to say, a discourse.

Speaking geometrically, I think "an hierarchy" could be redefined as “ANY single-step, ONE MANY connection": i.e. a CENTER, with logical struts radiating out to a multitudinous PERIPHERY. This can be found in machine-language, and of course in angel choirs.

The abstract visual would start from an asterisk, with a center point, and peripheral points. Since there can be transactions among members in the periphery, we must connect the outer points around, looking like a wagon wheel.

Of course these quickly stellate into sub-hierarchies of many such single wheels.

To be clear: this definition of “organization” covers not only social formations, but language itself (and so ritual, art, tool, ..., artifacts). Syntactic structures allow division into meaning (and redundancy), with increased returns to understanding-control.

The center may be powerful, or not. The members in the peripery may be powerful, or not.

The center’s existence tends to reduce the space-time between successive events at the periphery. This is an innovation by organization (as opposed to innovation by technology--which is a re-organization next level down, in the tool itself. Stone ax + haft = more force through space-time).

Defined this way, ALL organizations are hierarchies, but only SOME hierarchies are coercive, like patriarchies or (in some ways) bureacracies. Firms are hierarchies, coercive within their boundaries, only they must trade outwardly in the market.

And while free trading is the definition of rationality, the market system ITSELF is a single-step hierarchy, with all transactions peripherally around a huge CENTRAL mountain of laws, procedures and assurances, as well as the guarantee of bank currency. In this certain way the market system IS coercive, since you can get away with very LITTLE outside of these strictures--though of course we all think it's for our own good!

Charisma sets up a moving center (i.e. the leader)--and maybe not as many codified connections among the members of the charmed circle. Perhaps in this geometrical sense, blogs are an abstract charismatical form?

Advanced communication, by which I mean beyond the apes, is necessary for large-scale division of labor, and within our spoken and written languages there are requisite logical hierarchies of meaning. We do not tend to think of these as coercive, unless you are just over from Critical Studies (although sometimes, perhaps, we all should).

These “coercions” will be promulgated through open-source, too.

But of course, by "coercion" we WANT TO MEAN top-down, or center-out, control (particulary the despotic forms)--as opposed to bottom-up, or periphery-sideways (which I guess we call "democratic" or "rational"?).

It's not so much the existence of coercion, as WHO or WHAT is being coercive, and the degree of it you are willing to submit to.

But in non-despotic times, it is probably not a Newtonian-Einsteinian ENERGY/POWER metaphor, as often as a Darwinian-Batesonian coincidence-or-hindrance of PATTERN. You will want to GO ALONG with the “outside” organization to the extent that it MAPS ONTO the hierarchy or institution of YOU--you, at the CENTER of your own life, hoping to reduce the space-time among the important events arrayed round about you.

Posted by: Lee A. on December 28, 2003 05:07 PM


Post a comment