While I'm not one to endorse Cass Sunstein in general, I recently read a good portion of his _Infotopia_ for part of a dissertation chapter. I should say that I read him against the grain. I don't know if this will read well outside of the chapter, but I'm basically discussing the Austrian theory of prices (especially the Hayek article on "Use of Knowledge in society" which is the reified position of the neo-liberal cheerleading for markets.) In this section, I discuss Sunstein's argument that markets are better at discovery and information agglomeration than deliberative democracy--a central claim of the above. He supports this by looking at various prediction markets and other kinds of statistical instruments that register fairly simple questions. However, nearly all of them work because they are parallel markets--i.e. the results and actions "the market" aren't directly engaged in producing the result they are supposed to predict. E.g. they aren't stock traders who--though they are directly engaged in producing the value of the stock--are partially trying to predict the value of the stock, but are more trying to predict what other people will predict the value of the stock will be (I cite that famous passage from Keynes on investment, beauty pageants http://tinyurl.com/keynespretty ) The inducement to affect the results of the market, using the market itself, make this third level of prediction quite inefficient. I guess I said that in a much tighter fashion than the passage below, but if anyone is interested in a longer elaboration of Hayek, Sunstein, etc., I would be grateful for any feedback.
In any case, I think we actually live in a time when the technology and practices would allow for this more (though, it should be said, the problems Woj mentions about irrational kinds of demand would be easily manipulated by even a system of parallel information/prediction markets.)
thanks, s
<OPEN QUOTE> Like the concept of value, the concept of rationality is relative (a point that Sunstein would likely make when wearing his Behavioral Economist hat): if money can be made off speculative trading—and people can be found to take the other end of the bet—then the rational way to make short term gains is to speculate[1]. These “predictions” have no necessary connection to measuring reality—to the fundamentals of the firm in question—but they have the potential to affect outcomes. Pulling out of a currency (as in the Russian, Mexican, and East Asian crises of the 1990s), company (Enron, World Com) or type of investment (mortgage backed securities) after inflating its value has a resounding effect on the rest of the economy, even if most of the speculation was on future market transactions rather than actual supplies. If the purpose of the market is to accurately predict and/or value (e.g. price) the actual commodity, it may not create quite the success Sunstein claims. On the other hand, if the goal is to predict what other people will think the value is at a given time—or what they may think other people will think—then this is a different story. In fact, here is where people like Stiglitz point to the problems of the market itself introducing information—and especially faulty information-into the feedback loop. Though tales of “pump and dump” may be wildly exaggerated[2], when the betting and hedging is taking place on commodities with real prices in the real economy, the upstream activity of these “prediction markets” can have effects that appear adversely in the actual commodity markets.[3]
This is the other dimension of the way markets normally work. In Sunstein’s examples, the incentive to be right is separate from the incentive to make money: in those above, the two are paired, often in inverse proportion to what he would find to be the “correct” answer. When the incentive is just to guess, for instance, the outcome of an election rather than affect that outcome directly, it stands to reason that the numbers will likely drift closer to an accurate prediction. It was fairly clear to anyone watching—including most of the Republican establishment—that Obama’s win was much more likely than the 53/46% split national polls were showing (even if the 85/15% IMTrade prediction was a bit overzealous).[4] However, polls don’t just aggregate the information about the likely result: they also demonstrate, in a highly abstracted way, people’s real world interest in the result. Electing one candidate over another is the closest we get to deliberation on a national level and people have an interest in this. The desire to see one set of policies over another (or in some cases, one person’s TV personality) for the next four years drives people’s answers in polling data: polls therefore register voters’ desires. And, in terms of predicting the final split in the electoral votes, they do a decent job. Prediction markets, though linked deep down with these desires on the part of the traders, are more interested in picking the “right” winner according to Hayek’s calculus: the winner who wins is right.
These two indexes show two different kinds of information agglomeration—one of which barely appears in Sunstein’s studies. He focuses on the way the incentives of markets help to produce accurate predictions. Here the interest in winning helps drive the accumulation of information. But in the deliberation studies, the controlled, hypothetical experiments take interest and motivation out of the equation: supposedly this measures some intrinsic psychological reaction. But taken out of a social and political context—the context in which deliberation actually takes place—these results merely show unconscious ticks that can appear in human perception and interaction, ticks which are also largely contingent on cultural assumptions about human interaction and the world and which are present whether people are participating in deliberation, collaboration, or prediction markets. Sunstein’s insights here are interesting, but he processes his data to reach a conclusion that would be very different than Hayek.
If the effects of speculative bubbles were limited to the stock market or to the actors in a private betting pool, the informational aspect might be useful in and of itself—even when it was incorrect. But since they are connected to the real economy, and have real effects—effects which, when possible, the predictors themselves have an interest in manipulating—it becomes an issue of public policy. For Keynes it is less a question of how to reshape the rationality of these actors. The “animal spirits” of the investor class as essential to any investment to ever occur: for him, as for David Harvey, speculation is the heart of market capitalism. The question is how central this process should be to deciding and distributing the use and distribution of basic resources.
Sunstein’s somewhat naïve reading of Hayek sees this “informational” aspect as similar to a sort of democratic voting process. He lends credence to a position that, at least in his earlier book, _Free Markets and Social Justice_, he would not agree to. This is the notion that the market is somehow more democratic than democracy. Hayek tends in this direction as does the position that Sunstein takes on deliberation
[. . . .]
When participation in decisions about distribution is limited to those who already possess land or money, there should be little surprise when the results of that “plebiscite” favor a different range of interests. Just who was participating in the market varies according to how it is set up. For the internal predictive markets at HP, Microsoft and Google, everyone in the company began with a certain number of shares and the “economic incentive” Sunstein finds essential to their functioning were arranged in terms of prizes for correctness.
For others, like the IEM, participation is based on payment, thus, “some people might refuse to participate not because they lack information, but because they lack the money to gamble.”[8] Sunstein’s may be correct in his assumption that these may demonstrate the power of this particular kind of market institution—that, if separate from the actual effects, it helps to collect and consider the information that everyone individually possesses.
But Hayek doesn’t just want them to be predictive: he wants them to discipline people, to effect their decisions in one way or another no matter if they are willing or able to play the game. It is a Hobbesian world of “live and let die” which he sees as functioning best when no one can opt out of it and no force can come to anyone’s rescue. In other words, as the pinnacle of civilization, Hayek pitches us the world of Leviathan without a Leviathan. The latter—as observers from Marx to von Mises, MacPerson to Marcus Rediker and Peter Linebaugh have noted—was a society and culture that, in so far as it ever flourished, expressly depended on what Bourdieu calls “the enforced conversions”[9] to the market modality. Yet the libertarian virtue depends on bracketing this process, seeing, instead, the “extended moral order” to be at once completely natural and the pinnacle of civilization. From this Hayek concludes that we are not only communicating across space, but across time, making the function of tradition supposedly the same thing as that of the free market
[. . . .]
Here, as observed above, Sunstein’s version of communication through prediction markets is quite interesting in and of itself. If he were to de-link it from the libertarian ideology he finds so central to these discussions, he would likely find a much more productive conversation. The “incentives” he talks about in relation to prediction markets are very different than those that von Mises or Hayek describe. Though they all use deceptively similar language, the purpose of Mises and Hayek (as well as Coase) is to limit the ability of states to provide directly for the welfare of their citizens. This appears as a principled claim in relation to fears of fascism, but the calculation debate rose first in relation to socialism and communism, which had very different goals in terms of economic production and industrialization. The ironic upshot of Sunstein’s prediction markets is that they effectively provide an alternative solution to the calculation debates.
As mentioned throughout this chapter, the reason planning was presumed to be ineffective was that it had no effective “communication” mechanism, like the price system, to sort “heterogeneous goods.” Overlooked was the fact that creating this price system, and having it actually function, would require the wholesale reordering of society, privatizing every aspect of the social order that would need to be allocated and subjecting every person to the discipline of the market—theoretically on an equal basis, though unless the conservative assumptions about the past were expunged, there would always be significant inefficiencies and inequities in its operation. All of this was excused, as mentioned in chapter three, because the utilitarian functionality of the system in sorting values appropriately was, in the long run, good for everyone. This is more efficient, produces more stuff, which is good for everyone (except that distributing it outside the market undermines the market's disciplinary power and therefore it can only be distributed on the basis of effective demand, thus reinforcing class based society.)
Sunstein potentially solves this by having the market in information simply run parallel to the production process, rather than determining it completely. In this, the answer that late, nineteenth century socialists gave would be improved by separate incentive structures outside of the market. The earlier answer was that, since goods would still be distributed based on demand—though in this case it would be absolute demand rather than merely effective demand—monitoring inventories would provide the information necessary to plan production. This idea of a parallel price system would seem to work right alongside Coase’s description of why the monopoly firm was able to subsist; Sunstein’s idea of prediction markets with separate incentives would likely be the kind of push the model into an even more effective productivity.
In any case, this market would not be for every sort of human value and it would not be the equivalent of all culture or society. Moreover, the idea of “communication” taking place here would be limited to the “communication” of actual production issues, rather than assuming the “sorting of preferences” taking place represented the whole of all possible preferences. Finally, since this system—like the free market utopia of Smith, Hayek, and Mises and the corporatist reality of the “railroad economists”—had to be invented and imposed, it would be essential that there be some mechanisms for democratic decision making—perhaps even deliberation—that took place alongside these processes. I hope to explore more of these in the final chapter,
<END QUOTE> .
________________________________
[1] Though this still depends on which side of the transaction you’re sitting: In the one year Andrew Lahde manager ran his hedge fund—from about Sept 2007-Sept 2008—he achieved returns of 866% betting against the subprime collapse. After closing up shop he sent a snarky letter to Portfolio.com in which he summed up his strategy for success thusly:
I was in this game for the money. The low hanging fruit, i.e. idiots whose parents paid for prep school, Yale, and then the Harvard MBA, was there for the taking. These people who were (often) truly not worthy of the education they received (or supposedly received) rose to the top of companies such as AIG, Bear Stearns and Lehman Brothers and all levels of our government. All of this behavior supporting the Aristocracy, only ended up making it easier for me to find people stupid enough to take the other side of my trades. God bless America.
Matthew Malone and Andrew Lahde, Hedge Fund Manager: Goodbye and F--- You (Portfolio.com, Oct. 17 2008 [cited Dec. 17 2008]); available from http://www.portfolio.com/views/blogs/daily-brief/2008/10/17/hedge-fund-manager-goodbye-and-f-you.
[2] Though likely less so than people who deny it happens. http://www.sec.gov/answers/pumpdump.htm
[3] Though his analysis remains contested by many orthodox, neoclassical financers—who assume that this connection is always bilateral and prices, even during what are obviously speculative bubbles, in some way reflect upon actual demand—this was the upshot of Michael Masters’ May 20, 2008 testimony before the Senate Committee on Homeland Security and Governmental Affairs. Here he claimed that,
Commodities prices have increased more in the aggregate over the last five years than at any other time in U.S. history. We have seen commodity price spikes occur in the past as a result of supply crises, such as during the 1973 Arab Oil Embargo. But today, unlike previous episodes, supply is ample: there are no lines at the gas pump and there is plenty of food on the shelves. If supply is adequate - as has been shown by others who have testified before this committee - and prices are still rising, then demand must be increasing. But how do you explain a continuing increase in demand when commodity prices have doubled or tripled in the last 5 years? What we are experiencing is a demand shock coming from a new category of participant in the commodities futures markets: Institutional Investors. Specifically, these are Corporate and Government Pension Funds, Sovereign Wealth Funds, University Endowments and other Institutional Investors. Collectively, these investors now account on average for a larger share of outstanding commodities futures contracts than any other market participant.
The testimony was made during a period of very high oil prices—nearing $150 barrel—and food prices that rose so rapidly on a world scale that it led to food riots. However, this was not a case of there being less supply or a sudden spike in demand: it was that the value of the commodities was being pumped up by speculative trading. Though I have yet to find a comprehensive analysis, some confirmation can be found for Masters’ hypothesis in the fact that the collapse (or at least freeze) in world finance has corresponded to the price of oil falling by almost half. The closure of factories in China (some reports calculate these at close to 75,000) could have some effect on this price, this cannot account for the full drop in world oil prices—any more than the demand from China was the factor driving the earlier rise. Committee on Homeland Security and Governmental Affairs Testimony of Michael W. Masters Managing Member / Portfolio Manager Masters Capital Management, Llc, 110th Congress, 2nd session, May 20 2008.
[4] realclearpolitics.com, General Election: Mccain Vs. Obama (realclearpolitics.com, November 3 2008 [cited December 16 2008]); available from http://www.realclearpolitics.com/epolls/2008/president/us/general_election_mccain_vs_obama-225.html.
[5] Ludwig Von Mises, "Profit and Loss," in Mont Pèlerin Society 4th General Meeting: Planning for Freedom (Beauvallon, France: Ludwig von Mises Institute: Auburn, Alabama [2008], 1951). p. 13.
[6] Ibid. p. 18-19.
[7] It is also a patently pseudo history, which refracts its market system into the past. Thus the sentence before the second quote is, “The popularity of motor cars, television sets, and nylon stockings may be criticized from a “higher” point of view. But these are the things that people are asking for. They cast their ballots for those entrepreneurs who offer them this merchandise of the best quality at the cheapest price.” This proposition ignores the monopoly position of the producers of these products and, in the case of cars and TV, the government assistance in helping to launch them and sustain consumer demand. Nylon stockings might have an inverse relation to this at that moment since they were rationed during the war and would have only recently been available again.
[8] Sunstein, Infotopia : How Many Minds Produce Knowledge. p. 105.
[9] Bourdieu, The Social Structures of the Economy. p. 6.