Friday, August 7, 2015

Definitions: information and effective

There are two words (maybe more added to this post in the future) that I use on this blog that have different meanings from how they are typically used. One way to look at the information transfer model is as an effective theory of information, but I'm not saying it is a "successful theory of knowledge" with that phrase. That's because the words mean something a bit different.


On this blog, "information" is not meaningful knowledge (like how to do a Fourier transform or knowing the capital of Texas), but rather a measure of (reduction of) uncertainty. Flipping a fair coin once gives you one bit of information about the heads/tails state of the coin. You've reduced the uncertainty from not knowing if it was going to be heads or tails to knowing that it is exactly one of them. In economics, we use information as a measure of e.g. who buys a widget or not. Information equilibrium as used here means that the uncertainty reduced by figuring out the distribution of widgets supplied is equal to the uncertainty reduced by figuring out the distribution of demands for widgets.


On this blog, "effective" is not a synonym of "successful" or "useful" (though those definitions sometimes also apply), but rather an adjective applied to an explanation or theory focused on an effect rather than a cause. You could have an "unsuccessful effective theory". A good example is the use in the phrase "effective tax rate" where the details of a progressive tax structure, tax deductions and tax credits are combined (in physics we might say "integrated" or "integrated out" since the details disappear), leaving you with a single number. Aggregate has a similar meaning in economics. For example, the price level is an effective price that comes from combining all (or just many) of the prices in an economy.


So effective theory of information means a theory of the reduction of uncertainty associated with aggregated constituents.


  1. Good to know, especially the "effective" one.

  2. O/T: this guy's been posting links in (long and combative) comments at Nick Rowe's site. What do you make of it? The word "bizarre" comes to mind for me.

    1. As a physicist with a university email, you tend to get links to stuff like this:

      As for axec, there doesn't seem to be any connection to data and the axioms don't specify how the various variables are allocated ... It's a series of accounting identities.

      But I think that is taking it more seriously than it deserves :)

    2. Lol, I'm glad to hear that: I reached almost exactly the same conclusion after my very brief analysis... but I decided to scale it back to "bizarre" (did you see the bit about the copyrighted ideas?). I don't get the motivation: a built up to a Peter Schiff style hedge fund marketing campaign perhaps?

    3. That "Yun Qi Kingdom" post is hilarious. I've spent some time on yahoo answers, and I've seen some pretty bizarre stuff there too. One guy in particular I played along with... he claimed he had two PhD's in physiology and something else and his brother was a PhD in physics. He casually mentioned that Einstein's general relativity had been disproved some time ago and that gravitational lensing was bogus. I told him how excited I was to know him before his Nobel Prize award (I was laying it on pretty thick, but he didn't get the sarcasm). Anyway, he couldn't take credit for it and pointed me to this guy instead. Fascinated, I spent a good hour or so looking into this crackpot... the good folks at some website that tracks physics cranks told me he was on their radar, but they hadn't found the time to do a piece on him yet... in fact I got the impression that crackpots are popping up faster than they can deal with.


Comments are welcome. Please see the Moderation and comment policy.

Also, try to avoid the use of dollar signs as they interfere with my setup of mathjax. I left it set up that way because I think this is funny for an economics blog. You can use € or £ instead.

Note: Only a member of this blog may post a comment.