September 2022

Sun Mon Tue Wed Thu Fri Sat
        1 2 3
4 5 6 7 8 9 10
11 12 13 14 15 16 17
18 19 20 21 22 23 24
25 26 27 28 29 30  
Blog powered by Typepad

« Information Processing, Knowledge Acquisition, and Decision Making | Main | Mises, Hayek and the Austrian School -- CCA Lectures at Hillsdale College November 2016 »


Feed You can follow this conversation by subscribing to the comment feed for this post.

Is that a proof that Jason Brennan's theory of epistocracy wouldn't work?

This comes back to the local knowledge problem; that experts may indeed have general knowledge about class of problem abstracted from its setting, but that only works for problems that are truly able to be abstracted. As a result, effective experts usually need to embed, or 'condescend' to understand local conditions when addressing a problem in the specific. However, when community problems are fundamentally about the 'community,' the experts are likely to favor being 'objective' and 'distant' rather than 'involved' and perhaps compromised. As a result, there is a conundrum. They cannot sit on high in judgement on the community and still understand it; but if they become involved, the problem will not appear the same. It's a kind of relativity, particularly well known in families.

The answer is not to have contests among the experts to see who is more frequently right. This favors cherry picking and all sorts of bad strategies. The answer is to have experts as local as feasible; and keep them local, not giving them broad authorities. They can learn from each other but not subsume each other. There are costs to this approach, but it will be more robust than the current brittle strategy.

Well it's much less academic but I recommend "SuperCrunchers." It contains the results of a lot of research on experts and shows that their overconfidence makes them biased. Simple regressions do consistently better in almost all cases.

For this election, I bypassed the "experts" and paid attention only to the economic models, such as Ray Fair's Yale model predicting elections. Fair had Trump winning back in July. It's pretty obvious if you take voters at their word: the economy is the most important issue for them.

As to how Trump won the primary the answer is equally simple. Trump was the handiest Molotov cocktail for blowing up the Republican establishment.

I think experts who serve an 'elite' aren't going to be objective because an elite, by definition, believes that it's values and preferences are 'hegemonic' in the Gramscian sense- i.e. they are prescriptive because of some obvious virtue which everybody recognizes as attaching itself to the 'elite'.
In other words, the elite has an incentive to employ an expert who predicts that which is in their narrow interest and tries to pass it off as a 'Muth Rational' solution.
If Elites are insecure or subject to rent-contestation, sure, they may consult 'expert cognition' mavens so as to hedge their bets but they still have an interest in supporting official 'experts' who either predict what they want them to predict or who make a policy space multidimensional in a manner that gives the Elite 'agenda control'and thus the ability to rig the outcome in their favor.

If experts fall short at at what they are really good at will result in their failure. If a writer is not able to convey what he is expected to convey, it will be his failure and none of the readers will give interest to go through his work. So, experts should be good at what they are really good at constantly in order to maintain their standing in what they do. if you would like to know why experts are good at what they do, get through so that you will be able to know what makes experts the real experts.

I was not surprised by the result of the presidential election, because my principal source of news is the Drudge Report. Days and weeks before the election Drudge headlined polls telling Trump had the lead.

The comments to this entry are closed.

Our Books