Michael Perelman wrote: > Asset prices can convey information, but they can also convey misinformation. > These > prices can be correct. Even a stopped clock gives the right time a couple > times a > day. The problem is the inability to identify when the price is correct and > when it > is not.
I'll try again to have people think beyond financial markets. What's at issue here is not financial markets alone, but any mechanism that aggregates information from individual inputs that are not independent in the statistical sense. (Correlation implies statistical dependence, but the converse is not necessarily true.) Basically the same problem arises in, say, a political mechanism of choice aggregation (e.g. a democracy) in which individuals are not making independent choices, but instead seek to anticipate or adopt as their choice some function (say, an average) of the choices others may make. (The only society where I can imagine this curse being ruled out is ideal communism, because in such society the free development of each one is supposed to be the premise for the free development of all. Individual autonomy is premised on the full recognition, embodied in our institutions, of our social interdependence. But for now that is mostly an abstraction.) For a bit more concretion, think of a formal democratic society (e.g. a society trying to transition to socialism) in which large groups of people make their choices not on the basis of individual perspective and circumstance, but on their trying to look good to others or guess the direction of the political wind or follow blindly what opinion leaders offer them or out of fear of falling out of step with the crowd or the leader. The information range will be narrower, because most people will be free riding on information. "Errors" are likely to accumulate in one or the other direction (over- or undershooting) until some crisis happens. The free-riding problem is that of going to a potluck dinner without bringing food to it (in this case, informational food). If everybody does the same, there is no dinner. On the other hand, imagine that Keynes' prettiest-face contest in GT, 12 changes its rules. Instead of trying to guess the average preference, contestants are asked to use their own personal standards of beauty to judge the photos. To the extent, the personal standards of beauty are independent of one another (i.e. knowledge about the choices made by one of the players doesn't allow you to predict the choices made by another one), the self-referential character of the contest gets broken. This doesn't guarantee that the aggregated information will be "true" or "correct" or "good" in some absolute sense. But it'll definitely break the feedback loop that makes "errors" self-reinforcing. Again, the aggregated choice is simply a statistic or -- more generally -- a function of the choices made by the individuals. So it can only reflect the information that individuals have and use when they made their choices. That information may or may not be "true" or "correct" or "good" in reference to some absolute standard of truth or correctness or welfare. So, this doesn't exclude ideology, "false consciousness," or the "illusions of the epoch." The difference here is simply that the "errors" won't be self-reinforcing. By the way, as I wrote before on pen-l, you get the same garbage in, garbage out problem on these listservs to the extent the list members' views are more uniform. We contribute best to the common good by speaking our minds as we see things individually, and let the chips fall where they may. _______________________________________________ pen-l mailing list [email protected] https://lists.csuchico.edu/mailman/listinfo/pen-l
