I’m possibly getting involved with someone doing some research on information markets. This is just a sounding board.
The main piece against which I should start is the Wisdom of Crowds. Do we even know the criteria to judge how to make these markets? Does it matter that they work experimentally even though there is no theory of why or how they work behind them – or vice versa, if there is a demonstrable benefit to information markets, who cares what the theory says. If they tap into the energy of the universe, does it really matter so long as they do a good job?
Anyhow, to task. There are three main problems that aggregation is meant to solve: (1) cognitive problems, where there is a solution, absolute or optimal; (2) coordination problems, getting disparate actors on the same page; and (3) cooperation problems, getting people with opposed interests to act in concert. Surowiecki’s independent variables are independence, diversity, and decentralization.
Beyond that, the theoretical justification for prices as more valuable sources of information than experts, individuals, or planning boards, is Hayak, and the “Use of Knowledge in Society” more specifically. Hayak is writing against the impulse towards central planning of economies, as opposed to using the price mechanism of markets. He argues that bits of knowledge “of the particular circumstances of time and place” are spread across a wide group of people, spatially separated, without much interaction (521). Those with good specific knowledge act in the marketplace when there is, in his example, a shortage of tin or some other commodity. When the tin prices go up, say because of a shortage or a new use for it in some industrial process, it does not matter than only a few people know about it. The knowledge is spread across the whole economic system via the price of tin. As he notes, “in a system where the knowledge of the relevant facts is dispersed among many people, prices can act to coordinate the separate actions of different people” (526).
What is Hayak saying and not saying here. He hedges a bit about scientific knowledge, at one point sidestepping the question of whether scientific knowledge can be ‘priced.’ At the same time he suggests that entrusting authority to a body of experts shifts the problem of knowledge to the choice of experts but does not make that knowledge more efficient. So he seems for the most part to be arguing pretty close to his case, for a market mechanism against planned economies.
The question I have, then, is whether other kinds of information markets, as opposed to price mechanisms in commodity markets, is as much a ‘metaphor’ as an actual marketplace. For my money, this is the central theoretical problem to be answered.
The idea in theory is two-fold: (1) to group information, under the assumption that many voices come closer to the truth than one voice; and (2) to use profit motive to induce people with private information to make it public. This is a derivation of Hayak’s theory of prices as information.
But, is the analogy to future markets just an analogy? In the Hahn and Tetlock primer on information markets, they group together TradeSports, the IEM, the Goldman Sachs future on economic indices, and the CME’s weather derivatives. But these are systemically different. Things like elections and movies and sports are ‘best guesses as to what the future will hold’; weather derivatives are based on the risk-offset idea, which actually underpins all actual derivatives.
The history there is that there is a risk that is too large to be borne by insurance (or that actually could be born by insurance), but is born by a market instead. So with weather, there are some organizations/firms which have systemic risks associated with weather – looking over to the California orange growers, for example. So to be able to offset weather-related risk would be helpful. Then speculators could be on the other side.
So in ‘real’ derivatives, there are hedgers and speculators. This may not hold in some real derivatives anymore, post-boom in black-box trading since the 2000s. But that’s the theoretical basis for derivatives.
But in information markets, where is the risk to offset? This is totally divorced from that model, using instead the information-gathering idea of markets. It seems to me that this appears as a Hayak-type markets-gather-information model glommed on to a futures-apart-from-underlying-spot market model. But not really one or the other.