Friday, September 5, 2014


300 posts! I thought I'd celebrate with a list of the top 3 posts since the blog started back in April of 2013.

1. How money transfers information 
This is still one of my favorite posts to introduce people the information transfer model. It's likely here because I link to it so frequently.

2. If physics blogs were like economics blogs 
This post was my crossover hit! It was a comedy piece inspired by what happened to my inbox after subscribing to the comments at this post by John Quiggin. A large fraction of the hits came from the post being tweeted by Unlearning Economics (who is giving up the blog for greener pastures and post-graduate studies -- and was a good sport). 
So many of my posts begin with a reference to Scott Sumner and some claim he's making. This post discusses how the interpretations of economic data are model dependent. Of course the latest inflation data is consistent with your model and inconsistent with your interpretation of another model -- why wouldn't it be?


  1. I notice I can now stop at "information tr" when typing into Google, and that puts you right at the top of the list. It wasn't that long ago that I think I had to type "information transfer e."

    1. At ~ 1 bit per English character, that represents a gain of 7 bits of information!


    2. Why ~ 1 bit? It takes at least 5 to encode them, no? Is that the compression ratio of English text? 5:1?

      Your posts are too short, so I went over to Miike Freimuth's blog, copied all his posts on the front page into a text file, and compressed it. A 3:1 savings. :D

      (This one only compressed to about 70% it's uncompressed size: too much "header" info I guess).

    3. Of course it's highly dependent on the type of text too I think... I just put three large "English" text files into a directory (two timing reports and one synthesis report from a Xilinx FPGA build), and here's the figures:

      uncompressed/compressed = 35297609/1178072 or about 30:1

      Kind of a crude measure of "information content." Mike should feel good about that I suppose (in comparison). You should feel even better! Lol.

    4. ... that would be fun to do: leave comments on every one's blogs: "I calculate that this post is about 300% more wordy than it needs to be! Let's try to get to the point in the future, shall we?" Lol.

    5. Ha!

      Regarding the 1 bit, that's the average information per letter in English words. For example, the letters "transfe" could only be followed by "r" so that is actually zero bits of information (it has probability one). The letters "transf" could be followed by "i", "e" or "o" for transfer, transfix or transform ... So that is a bit more than one bit. On average there are about two letters given the earlier ones, so on average each letter reveals one bit.

  2. Well done on reaching your 300th post. I had a look at the first link here and I had a further thought about information transfer.

    Suppose you are right that information flows from demand to supply and that some is lost in transit. For sake of argument, suppose that each customer has two units of information and that each supplier picks up one of these units of information per successful transaction and none if no transaction takes place.

    Now suppose that there is a market with 1,000,000 customers and 10 suppliers. Suppose that, in a particular period, 80% of customers engage in a transaction with one of the suppliers.

    The average customer will then have two units of information about the market i.e. 2 units times 1 customer.

    The average supplier will have 80,000 units of information about the market i.e. 1 unit times 80,000 customers (equal to 80% (active customers) of 10% (market share) of 1,000,000 customers).

    The dynamics of the market are set by the total information available to each participant in the market. In total, the demand side has 2,000,000 units of information compared with the 1,000,000 units for the supply side. However, it’s the concentration of the supply information that favours the supply side in contrast to the diffuse nature of the information on the demand side.

    I think that is one reason why I’m suspicious of your gas analogy as that makes it appear that all participants are equal.

    1. Thanks, Jamie.

      Regarding your example, it is important to realize that those 80,000 bits are information about the demand -- they are only useful if you want to be a supplier. A member of the demand doesn't care what price someone else pays (in an auction or where there is price discrimination things are different ... But those are not the kinds of markets we're concerned with here).

      Also, there really isn't equality between supply and demand ... In the ideal gas analogy, supply is volume and demand is energy -- apples and oranges.

      You can think of those 80,000 bits as an approximation to the maxwell Boltzmann energy distribution of molecules in the ideal gas. The molecules only know how much energy they are carrying. Using the pressure (price), the volume knows the full distribution.

      (Sorry for typos etc ... Using my mobile as I'm traveling for work again)


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.