The answer to why groups of economists are no more often right than single economists is pretty easy: they're all using the same tools to generate their answers.

The "Wisdom of Crowds" functions not on the basis of experts following rational processes, but rather on the basis of people making irrational decisions and then aggregating the tiny bits of knowledge contained in those decisions to (hopefully) a more accurate picture of the whole.

If that's true, then the obvious corollary is that we would get better results by taking a general opinion poll on the economy then asking economists.

If you really want to understand why they are wrong, read "The Black Swan". But lets face it, economics is not a science at all, its an experiment in sociology mating with mathematics. The only thing that makes sense is microecomics, macro is a waste of time.

wayhey1:

If that's true, then the obvious corollary is that we would get better results by taking a general opinion poll on the economy then asking economists.

In a sense we do. If you ask a broad question such as "how will the economy be doing next year", Joe Six-Pack, Soccer Mom, and their thousand different types of cousins, properly polled, outperform the experts.

However, you can't realistically ask complex technical economic questions of such folks because they won't understand the question itself and the limited information you're trying to aggregate drops to no information at all.

What might illuminate this a bit is to realize that market capitalization has a minimal relationship to the actual 'value' of a company. It's really just a popularity poll - how confident investors are in a certain stock. That's why you have companies with virtually no material assets worth billions, but old-line Detroit automakers with billions invested in physical plants on the ropes. So polling the "experts" - who collectively comprise a trivial segment of the overall value estimations of an economy - is worse than polling the "rubes" - who collectively comprise the entirety of the value estimation.

Philosophically, the "wisdom of crowds" is really no different from something that we've known for a very long time: that certain questions are so complex that even the best possible expert is only able to at best formulate a process that might lead to an approximation of the answer.

I don't know about generally, but when related to economics and finance forecasting (as well as political forecasting), the wisdom of crowds isn't just from a survey of every geek of the street. Rather, it relies on people actually investing their money in something. Those with no opinion stay out, those with some inkling put some money in, and those with confident opinions and lots of money go long (that's the simplified model at least).

If you want to unleash the wisdom of crowds to forecast next year's GDP, you wouldn't ask Joe Six-Pack, who probably has no idea what normal, high or low GDP growth would look like. You'd instead set up a GDP derivative market and let it predict the level.

And for stocks, forgetaboutit. You don't need to ask the rubes what a particular stock, or the whole market will do in the future. The answer you would get would likely be nonsense. If you want to get an estimation of a companies future value, look at its stock price right now. That's exactly what a stock price is, the value set not by rubes, not by experts, but by investors. The same applies to any stock index.