Samuel Bowles is giving a Berkman lunchtime talk called: “Kudunomics: Property rights for the information based economy.” He wants to look at how institutions are likely to evolve in the “weightless economy.”
NOTE: Live-blogging. Getting things wrong. Missing points. Omitting key information. Introducing artificial choppiness. Over-emphasizing small matters. Paraphrasing badly. Not running a spellpchecker. Mangling other people’s ideas and words. THIS TALK WAS ESPECIALLY DIFFICULT for me and certainly contains howlingly wrong misrepresentations of SB’s ideas. You are warned, people.
“In an economy based primarily on embodied and relational wealth, individual property rights are difficult and socially harmful to enforce.” Adam Smith’s invisible hand fails in important ways. SB says that that’s not a new idea. The new idea is that we should be able to gain insight about the evolution of institutions by studying the reverse transition from the Late Pleistocene forager economy to the agrarian economy. So, SB thought he should run that history backwards, which he may get to talking about in today’s session. The forager economy may provide clues for the weightless economy of the future.
SB puts up an equation explaining wealth, which I could not follow or capture, a cobb-douglas production function. [I hear Ethanz typing. He's certainly doing a far better job liveblogging this than I.] One point: Once we domesticated animals, we turned wealth into something we could own. Network wealth = the value your connections bring you. The number of people who will help you in your field, share food, etc. Embodied wealth = the value of what’s in your head that’s actionable by your body. [I'm not sure I got that, and I'm certainly paraphrasing.]
The basic idea of the invisible hand theorem is that good fences make good neighbors. Arrow and Debreu showed in 1953 that competitive market allocations will be optimal (in the Pareto sense), but only if the markets are complete (“the effects of the actions of economic actors on one another take the form of contractual exchanges”) and increasing returns to scale are absent or small [I don't know what that means]. “Under these assumptions, goods will be priced at their marginal cost which will equal their true scarcity (social marginal cost): p=M =SMC” SB is going to show that that is not true in a weightless economy.
Much of the economy – the grain and steel economy — fits this invisible hand theorem. It works best if the goods are tangible, easily measurable in standardized ways. In this classic economy, there was sufficient competition.
But, it’s different in weightless economies, where there’s high first-copy costs, and low marginal costs. E.g., it costs a lot to produce the first copy of a CD but very little for the rest of the copies. E.g., the first copy of Windows 97 cost maybe $50M, but the second copy cost $3.
In the weightless economy, enforcing property rights paradoxically force a violation of the invisible hand theorem: You let someone charge $20 for a cd the marginal cost of which is $0.85.
In the economy of grain and steel, market structure was a mix of competition and stable oligopoly (“competition restricted to a handful of firms”). The info economy may exhibit a serial monopoly structure, but that’s not what he wants to talk about.
SB gives a summary of what he’s said so far: Dilemmas of the weightless economy: Increasing returns on both the demand and supply side make competition difficult to sustain. This winner-take-all dynamic generates lots of inequality. The critical thing: Private firms cannot conform to the p=MC rule, and property rights are both ambiguous and difficult to enforce. The institutions that have worked well for the past 200 yrs are likely to work less well in the future.
Kudu = An antelope of some sort hunted in Tanzania for its massive caloric value. When one is killed, it’s widely shared (perhaps 2/3 outside of the nuclear family). The culture of the foraging band: generosity, modesty about one’s success, sharing. Christopher Boehm (1982) wrote that group sanction is “the most powerful instrument for regulation of individually assertive behaviors.” But mobile foraging bands “and its collectivist and egalitarian norms and properties was eventually displaced by agricultural production.” The critical fact is that that increased land productivity so that a small plot of band was productive enough to live on, which provided an incentive for putting up fences and defending it. These prop rights were not enforced by states but by some form of mutual consent.
Just as agricultural facilitated unambiguous prop rights, the info economy is reversing this process. We’re returning to the early Pleistocene economy. Most of the animals could not be domesticated. Some became more valuable when domesticated. Is an online song more like a cow or like a kudu? “Will the attempt to domesticate the modern day kudu’s prove costly and ineffective?”
Arrow: “Information is a fugitive resource.” It runs away. “We are just beginning to face the contradictions between the systems of private prop and of info acquisition and dissemination.” “If Arrow is correct, how would we expect our economic institutions to evolve under these new conditions?” Institutional change is very hard to study. There aren’t that many French Revolutions to study. He is doing Markov chain models with others at the Santa Fe Institute.
“Could between-group competition and technological advance combine to induce a new property rights revolution?” Darwin explained change via in-group revolution, while Marx looked at between-group. This is complex between there are both individual and group selection processes, so they’re almost impossible to predict using math. But you can use models. There are many quilibria. Initial conditions do not matter.
He talks about his agent-based model of institutional persistence and innovation. (You can play with his “artificial history” models here: http://www.santafe.edu/~bowles It looks like a Windows executable you can download.) He describes three strategies in the model: bourgeois (own prop and defend it), civic (share and penalize those who do not), share. [See Ethan! Or watch the webcast when it's posted in a day or too. Sorry.]
If prop rights are stable, then an all-bourgeois society (protect what they have) is in equilibrium. Likewise if all civics. If all civics (share and punish for non-sharing), you can drift toward all sharers because they are behaviorally indistinguishable if there are not B who are trying to protect what they have. Using these parameters (which I am expressing totally inadequately and probably inaccurately), he and Jung-Kyoo Choi have run simulations. If prop rights are stable, the system tends towards equilibrium. If they are not — a bourgeois contests ownership — there is no equilibrium, although there is some moving clustering. Summary: “Evolutionary success of the ‘bourgeois equilibrium’ depends on prop rights being unambiguous.
But this is not the right way to understand the future because we don’t know how ambiguous prop rights will be, which depends on technological advances and the legal system.
Diff institutions have diff advantages. States are good at coercing, Markets allocate well. Communities handle the ambiguity of prop rights but fail where inequalities among members are very large. The problem of the info economy is that information creates both substantial ambiguity or prop rights and a lot of inequality (winner-take-all). The ambiguity makes it hard for the state to adjudicate. The inequality makes it hard for the communitarian values to succeed.
He ends by quoting Hayek: Whether central planning or competition works depends on whether you put all the pricing info in the hands of a central authority or adjust the prices by giving the pricing info to individuals. But now we have a third player: Markets and states, but also communities. Fifty years ago, people speculated that computers would solve this problem. SB says that we need a high level of info creation as well as making it available at its marginal cost. This is the question asked for hunters in hunter/gathering societies: Why should hunters hunt if they give it all away? Understanding this activity — mirrored in today’s collaborative environment — may help solve the problem.
Q: What do we know about the scalability of communities? The ambiguity seems to grow as groups get bigger.
A: How many people work on Wikipedia?
Q: The ambiguity there occurs in small groups.
A: Hunter-gatherers can’t take advantage of economies of scale or of diversity. Can moral sanctioning be done in on-face-to-face environments? We’re finding out.
Q: Can you talk about common pool resources (Ostrom)? [and two more questions]
A: The value of the network is the number of possible connections. There are therefore huge economies of scale. That’s where you get the winner-take-all from. Ostrom took some insights of Ronale Coase and extend them beyond firms, to include things such as communities. Are the motivations for sw engineers the same for hunters? Reputation. Fun.
Q: [me] What’s a community?
A: The non-state, non-market ways that humans connect and interact. [Hugely paraphrased!]
Q: [me] Is there enough in common among all those ways to enable it to be used as a factor in your model?
A: Communities have in common that they have a public thing, they have to figure how to share the benefits of this, and they;re not doing this primarily through enforceable contracts. But I don’t want to pin it down too much. Read “Against Parsimony” by Albert Hirschman.
Q: One of the child’s first words is “mine” because that it eanables it to differentiate itself from its environment. I think your theory would change if you asked if that’s a universal.
A: It’s not. Children differentiate themselves from their mother, but they don’t universally claim physical objects as their own. Private property is incredibly recent.
Q: In your agent-based model, could you drill down to see which types of prop rights are likely to be stable?
A: Yes, but not with agent-based models. Our theory lets us address this. We just haven’t done it. You should be able to look at the nature of the project — first copy costs, e.g. — and develop a typology of the sorts of things that are hard to solve, although changes in tech or law would change this.
Q: The gov’t role has be quite diff if you an economy of cows or kudus. How does this affect gov’t regulation?
A: My preliminary ideas: I don’t think it leads to more or less gov’t. It leads into different kinds of gov’t interventions. The aim is to take seriously when designing incentives you have to take into account that people have their own motivations. And if you introduce monetary incentives, you may get worse outcomes; I’ve recently written about this for Science. The solution to problems is always some combination of incentives designed by economists et al. and the moral incentives of most humans. These two are inseparable; addressing one without recognizing this can be disastrous. Some problem are solved not just by financial incentives but by some combination of people’s incentives and motivations.
[NOTE: Samuel Bowles is way more coherent than this livebloggery makes him sound. I lack the background to follow much of what he says. Much for me was like typing in the dark. So, I apologize to him and to you. And here's Ethan Zuckerman's far superior bloggage.]