Tuesday, March 26, 2002

Dope

Is it only Limited Inc's imagination, or should economists take more of an interest in "small-world" theory, associated with Duncan Watts and Stephen Strogatz?

Yes, my readers roar, in numbers too big to ignore. Read your Watts and Strogantz and sin no more! or something like that.

Well, to explain...

These two wrote a paper a few years ago, in which they tried to find the minimum path length for an undeterminately large network. They called these networks, with their improbably small dimensions, small worlds, after the Milgram experiment that supposedly showed that there are six degrees of separation or less between any two randomly selected people in the world (well, the experiment didn't make a claim that vast, but it has been made since then). The problem, from the perspective of networks, was that most individuals are connected to a cluster of individuals, in which each individual has a high chance of sharing acquaintances. So how do you break out of the cluster to connect to random, unfamiliar individuals? Here's a quote about the system set up by Watts and Duncan from a September, 1998 Physics Today article

"This result is actually quite general," says Watts (who will shortly be moving to the Santa Fe Institute in Santa Fe, New Mexico), "and does not depend on the choice of a ring substrate used in the model. All that is required to generate the small-world phenomenon is a network that is locally ordered (which means simply that two nodes with a mutual 'friend' are significantly more likely to be connected than two randomly selected nodes), and which has a small fraction of long-range shortcuts. The effect also does not depend on the specific nature of the network nodes or connections--only their topology--so the small-world phenomenon ought to arise in all sorts of large, sparse networks."

To check this, Watts and Strogatz examine the length and clustering properties of three real networks: the collaboration graph of movie actors (including approximately 225 000 actors of all nationalities since the start of motion pictures); the power-transmission grid of the western US; and the neural network of the worm Caenorhabditis elegans (the only organism whose neural network is completely known).

As Watts explains, they show that, in each case, the characteristic path length of the network is close to its theoretical minimum (that of an equivalent random graph), yet the clustering coefficient is far from minimal, indicating the presence of significant local order. So all three networks exhibit the small-world phenomenon. "

There is a book coming out in May from Mark Buchanan, Nexus, that not only explains Watts and Strogatz' work, but expands on it, explaining that Watts and Strogatz had stumbled on one form of small world network, and that there is at least one other possible form of small world network. This other form is related to the principle of "the rich get richer" -- that certain individuals are more connected, and by that very fact will become more connected. There is a network form for the fact that wealth is unequally distributed. Among network people, this is known as preferential attachment.

Now here's the question. One of the big rightwing pushes right now is to promote the idea that poverty in the third world is rather a mirage. Or, if not a mirage, caused by ... as you might have guessed, big government. The idea comes from Hernando de Soto, and it isn't quite as silly as it sounds. In two books he has promoted the idea that small, informal vendors and makers and homeowners need a system that recognizes them as free economic agents with capital. That is, if we strip away the onerous bureaucracy and government thievery, we could unleash, in the third world, value that is already there. This, after all, is partly what happened in the French Revolution. Anybody who reads Le Rouge et le Noir is going to have some sympathy with de Soto's point, because Stendhal is very conscious of the effect of liberalism, ie stripping away big guv and its thievish attachments, on the French landscape.

Is the rightwing idea going against the rule of preferential attachment, or seeding it?

Hernando de Soto is being presented to the American public as some kind of third world guru. The NYTimes magazine, last year, presented him as the answer to our dreams (insofar as our dreams involve giving up none of our stuff and not feeling guilty about it). But the earlier image of de Soto wasn't so heroic. Tina Rosenberg wrote a review of The Other Path in the New Republic, in 1991, that pretty much blasted de Soto as an egomaniac and a crony of Fujimori. Here is what she said about de Soto's grand vision:

"To Reaganites, however, the most marketable aspect of The Other Path is what it does not say. It does not talk about helping small businessmen acquire the infrastructure, technical assistance, or capital they need. It does not propose improving education, health care, or other programs that could get Peru's poor off to a better start in life. It does not address discrimination against Indians, which has shut Peru's poor out of many opportunities. Most informals are one rung above beggars. Redefining them as entrepreneurs doesn't cure what made them poor, especially in an economy that has experienced one of the worst declines in modern history. (The informal sector exploded in part because traditional jobs dried up; only 9 percent of workers in Lima earn a salary they can live on.) Not even legal businesses can get credit. But the book asserts that legal reforms alone will suffice to unlock the informal sector's engine of growth. De Soto compared the state to a dying emergency-room patient and told me, "I want to burn down the hospital."

The burning down the hospital phrase has been toned down, lately, and there is a little bit more heed being paid to infusing capital. There is an organization, Trickle Up, which just announced its association with the ubiquitous De Soto. Trickle Up is dedicated to making micro grants to the third world street neediest. Because it promotes the solid virtues of entrepeneurship and self reliance, Trickle Up has become a favorite for conservatives trying to summon up a little chic compassion.


"Grants are made by TUP to selected groups of five or more people after a business plan is reviewed for them by unpaid TUP project coordinators. The maximum grant is $100, and recipients must pledge to reinvest at least 20 percent of their profits in their businesses. In the past ten years, more than 90,000 individuals have participated, 15,000 businesses have been started in 86 countries, and over $7.5 million in profits have been generated from TUP-funded businesses. All of this has been achieved without the involvement of governments, large staffs, or social researchers. By now, you probably see why it's called the "Trickle Up Program." Funds aren't lavished upon government entities in poor countries with the hope that a small portion will somehow "trickle down" to the very poor. The grants go directly to the cagey entrepreneurs of the streets, including those in Port-au--Prince, Haiti.

Now, LI is fascinated with the project here: can beggars become choosers? It all looks very much like... like the 1966 War on Poverty project. One of the oddities of contemporary conservativism is this adoption of sixties forms, from classical rock to agit-prop. Hmm. In any case, LI is going to go further into this issue, this grassroots wealth issue, in another post soon.

No comments:

Asking

Yesterday, I watched a very sparkly Biden official, who looked like he had just come from the Ken-at-High-School-UN box, answer questions fr...