Tacit Knowledge and Prediction Markets

Lynne Kiesling

OK, so I’m a little slow … and Mike’s gonna give me the big eyeroll I deserve when I say this … it wasn’t until I saw the title of this Burgh Diaspora post (I’m a definitive member of the Pittsburgher diaspora) that I consciously thought about this question: do prediction markets discover and aggregate tacit knowledge?

I’m not sure they do. Tacit knowledge is stuff we don’t know that we know, or that we can’t recall having learned; it’s just there. But how much of that is really what prediction markets are getting at?

I think it’s pretty apparent that prediction markets are institutions that enable agents to discover and aggregate diffuse private knowledge. But do you decide on where to put your money in prediction markets based on your tacit knowledge? And are the kinds of questions that are amenable to prediction markets those in which tacit knowledge is relevant and/or useful?

Discuss.

4 thoughts on “Tacit Knowledge and Prediction Markets”

  1. This is an interesting question. I think it’s good to remember that prediction markets are only one piece of the wisdom of the crowds concept (the aggregation piece) … the other pieces being, of course, different types of knowledge. Surowiecki spells them out in the book. They are: diversity of knowledge (what you know about the question at hand); diversity of beliefs (Muslim, Democrat, Free Marketer, etc.); and diversity of expertise (doctor, banker, ditch digger, etc.).

    A prediction market which has traders diversified in their knowledge will probably work better than one that is made up entirely of, say, professional football players from the Pittsburgh Steelers.

    So, I don’t know exactly, but I would imagine that tacit knowledge would have some role in there. While perhaps not the insider knowledge that would be most effective in a market, it would certainly help shape our thinking when insider information is not available.

    ~alex

  2. I’ll have a thing or two to say in response. I thought the Google prediction market paper — discussed at the Freakonomics and Midas Oracle blogs, among other placed — was pretty interesting for a number of reasons.

    Hadn’t thought about the tacit knowledge issue. In general I sort of assumed that what office-neighbors shared was articulated information — reasons to buy X and sell Y — but maybe there is a tacit knowledge angle. When it comes down to making a choice — buying X or not, selling Y or not — each person will inherently act on reasons to do with both articulate and tacit knowledge. (And therefore, like other markets, prediction markets will reflect tacit knowledge.)

    The interesting question is how prediction markets interact with tacit knowledge — does it lead to the articulation of previously unarticulated knowledge, does it lead to corrections of inaccurate tacit knowledge, does it foster the spread of tacit knowledge within an organization. In each case the question is a relative one: how do prediction markets do compared to other approaches for drawing upon the organization’s dispersed tacit knowledge?

    I addressed a kind of related issue, without explicitly thinking of the tacit knowledge issue, in my post on prediction markets and the flow of information within organizations. Maybe I’ll go back and think about it some more.

  3. I’ll have a thing or two to say in response. I thought the Google prediction market paper — discussed at the Freakonomics and Midas Oracle blogs, among other placed — was pretty interesting for a number of reasons.

    Hadn’t thought about the tacit knowledge issue. In general I sort of assumed that what office-neighbors shared was articulated information — reasons to buy X and sell Y — but maybe there is a tacit knowledge angle. When it comes down to making a choice — buying X or not, selling Y or not — each person will inherently act on reasons to do with both articulate and tacit knowledge. (And therefore, like other markets, prediction markets will reflect tacit knowledge.)

    The interesting question is how prediction markets interact with tacit knowledge — does it lead to the articulation of previously unarticulated knowledge, does it lead to corrections of inaccurate tacit knowledge, does it foster the spread of tacit knowledge within an organization. In each case the question is a relative one: how do prediction markets do compared to other approaches for drawing upon the organization’s dispersed tacit knowledge?

    I addressed a kind of related issue, without explicitly thinking of the tacit knowledge issue, in my post on prediction markets and the flow of information within organizations. Maybe I’ll go back and think about it some more.

  4. Well, whaddya know?

    Greetings from Pittsburgh! From Oakland, specifically, stone’s throw distance from the Cathedral, and only about two miles from my new home in Point Breeze.

    Anyway…I’d venture an opinion that says yes, tacit information is aggregated in prediction markets. My point would essentially be a retelling of the Polanyi arguments about tacit information — the “know more than we can say” view — so I won’t hash it up by trying to restate it. I would think the mental process for choosing some position to take in a prediction market, the calculation of risk and expression of some belief (including strategic behavior) would have to rely in some part on the unmeasurables like experience.

    The harder thing to point to, I think, is discovery. Tacit information may well be included as part of the decision-making process that induced some position in the market, but the price that comes out is too rough to at all see what it is that drove people to that position. I suppose there might be more that could be gleaned if demographic details could be pinned to traders. “63% of middle aged, black, upper class women bought shares at…” and so on. The best you might hope for is correlating some personal characteristics with trading behavior. Is that tacit knowledge, though? I dunno.

Comments are closed.