Monday, September 18, 2017

A couple of animations: demand curves

In my book, I illustrated Gary Becker's random agent model using a text diagram:


Here is an animation of that random agent model tracing out a demand curve when you change the price:


In the graph on the left, the black dot represents the average, and the black line represents the budget constraint. The graph on the right represents the demand curve traced out as we increase the price. A key thing to remember is that in order to achieve this result, we have to explore the full state space (the triangle under the black line). If we don't, then raising the price (or cutting it) doesn't necessarily change the quantity demanded:


Tuesday, September 12, 2017

Would economics exist without capitalism?


The first point listed, but the last point I'm addressing is this:
[In my opinion] characterisation of economics as the 'science of the state space' could be used to make [radical] political claims. E.g. does 'economics' only exist because of property rights/capitalism?
I somewhat re-wrote this based on the combining two tweets as I believe @unlearningecon intended.

The reason I'm addressing this one last is because it's the most intriguing and required the most time to sit and think about.

The question is basically whether the "economic state space" I talk about in the book (and which in economics jargon is referred to as the "opportunity set") is constructed by a particular system of laws, property rights, and institutions (e.g. capitalism, but also money in general) and therefore the study of that particular set of laws and institutions that we call economics is specific to those laws and institutions. Does economics as such cease to exist when those institutions change?

I know that Gary Becker and the Chicago school of economics thought that economics basically exists if humans ever make strategic decisions, and therefore economics should push into the study of sociology and politics.

I've always been a Star Trek fan, and indeed in Gene Roddenberry's future "economics" does cease to exist due to essentially the elimination of scarcity. This makes sense because institutions like money, capitalism, and property rights were all designed to deal with scarcity. You can read more about this in Manu Saadia's great book Trekonomics.

I'm going to have to say the real answer is: I don't know. I am ambivalent about how plausible the Star Trek thought experiment is. Can scarcity actually be eliminated? And if so, does the new institution that either mitigates or eliminates scarcity explore the state space? If the answer is yes, then "economics" will probably continue.

But I do think "science of the state space" will have lots of potential uses even if capitalism is crushed. When I say science of the state space, I am actually referring to what is essentially information theory and in particular the concept of information equilibrium which I have been exploring on my blog. In those explorations, I've already found a couple of examples: explaining a transistor using information theory and understanding Generative Adversarial Networks (GANs) as an analogy with information equilibrium. [Update: and traffic models (and see here).]

In particular, I think there is a deep connection between understanding the market algorithm and computer science (MIT recently started a new major combining the two).

I also think this science of the state space may be useful in neuroscience and understanding the brain. After I started my blog, Todd Zorick a neuroscience researcher and I wrote a paper on using information equilibrium to understand EEG measurements and distinguish between states of consciousness.

I have also speculated about the connection between the state space approach and evolution.

It's possible all of these disciplines may have a single framework based on information theory and the science of the state space, finally realizing Norbert Wiener's (who incidentally was a simultaneous progenitor of information theory along with Claude Shannon) desire for single field he called cybernetics.

...

PS That's a picture of a couple of my Star Trek models.

Update: Forgot to include traffic models above.


Tuesday, September 5, 2017

Name dropping

I put together a list of the people referenced in my book:
George Akerlof, Gary Becker, Jeremy Bentham, Ben Bernanke ,Ludwig Boltzmann, Guenter Borchardt, George Box, Sean Carroll, Keith Chen, Hillary Clinton, Arnaud Costinot, Bo Cowgill, Diane Coyle, Charles Darwin, Gerard Debreu, Lana del Rey, Dave Donaldson, David Dunning, Rochelle Edge, Martin Eichenbaum, Albert Einstein, Queen Elizabeth II, Peter Fielitz, Irving Fisher, James Forder, Cameron Freer, Milton Friedman, Galileo Galilei, Peter Ganong, Carl Friedrich Gauss, Nicholas Georgescu-Roegen, David Glasner, Alan Greenspan, Refet Gurkaynak, Robert Hall, Roy Harrod, Ralph Hartley, Friedrich Hayek, Cesar Hidalgo, Thomas Hobbes, Erik Hoel, Chris House, Nir Jaimovich, Edwin Jaynes, William Jevons, Lyndon Johnson, John Maynard Keynes, Israel Kirzner, Justin Kruger, Paul Krugman, Stanley Kubrick, James Kwak, Venkat Lakshminarayanan, Carl Linnaeus, John List, Robert Lucas, Rolf Mandel, Alfred Marshall, William McChesney Martin, Jason Matheny, Michael Mee, Benjamin Moll, Dale Mortensen, Tom Murphy, Isaac Newton, Pascal Noel, Emmy Noether, Harry Nyquist, Barack Obama, Karl Popper, Ed Prescott, Ronald Reagan, Sergio Rebelo, David Ricardo, Paul Romer, David Romer, Mitt Romney, Paul Samuelson, Laurie Santos, Claude Shannon, Adam Smith, Vernon Smith, Lee Smolin, Hugo Sonnenschein, Joseph Stiglitz, Scott Sumner, Joshua Tasoff, Paul Volcker, Harris Wang, Graeme Wheeler, Eugene Wigner, William of Occam, Alexander Wissner-Gross, Michael Woodford, Janet Yellen, and Eric Zitzwitz
I was curious about how under-represented women were in my book (especially since I was citing two fields that are over-represented with men — physics and economics).

A free version of the book



There was a great quote that I don't remember exactly about the Velvet Underground's first album. Paraphrasing, it said that very few people bought it, but everyone who did went on to start a band (looking it up, there are couple of versions). My greatest hope for my book would be something similar: only selling a few copies, but everyone who buys it goes on to help reform economic theory. The blueberry on the cover of my book is actually a reference to the album. It's definitely a bit of arrogance (hubris?) on my part.

Anyway, I think the ideas are more important than selling books so this represents something of a "free version" assembled from blog posts. It's really incomplete and the technical level varies wildly (however, nearly all are far more technical than the book). If these blog posts go over your head but seem interesting, then the book is for you!

*  *  *

Introduction

This semi-autobiographical chapter was largely written from scratch and represents a lot of new material. However, some of the basics are covered in a few posts:

https://informationtransfereconomics.blogspot.com/2015/05/about-me.html
https://informationtransfereconomics.blogspot.com/2015/01/is-market-intelligent.html
https://informationtransfereconomics.blogspot.com/2015/10/corporate-prediction-markets-aggregate.html

The critique

I actually excerpted an early version of this chapter after I wrote the first draft:

https://informationtransfereconomics.blogspot.com/2016/09/a-random-physicist-takes-on-economics.html

Physicists

Another chapter that is largely new, but the basic idea was captured in my post on Paul Romer and "mathiness":

https://informationtransfereconomics.blogspot.com/2015/05/the-irony-of-paul-romers-mathiness.html

Random people

The title was a reference to Pulp's Common People, but is probably totally lost on anyone else due to being way too subtle. Yet another chapter that is largely new. However, it can be considered an expansion on this post:

https://informationtransfereconomics.blogspot.com/2015/11/monkeys-and-markets.html

Another dimension

The chapter title is a reference to the Beastie Boys' Intergalactic. Part of this chapter is new, most of the main idea is presented here:

https://informationtransfereconomics.blogspot.com/2015/09/the-emergent-representative-agent-1.html

Advantage: E. coli

The title here is a weird reference to tennis, comparative advantage, and the idea that E. coli bacteria are better at trading than humans. This is a more technical version of the chapter that appears in my book:

https://informationtransfereconomics.blogspot.com/2015/08/obviously-e-coli-is-rational-utility.html
https://informationtransfereconomics.blogspot.com/2016/04/comparative-advantage-from-maximum.html

Great expectations

An obvious reference to Dickens, this is a far more technical version than appears in the book:

https://informationtransfereconomics.blogspot.com/2016/04/neo-fisherism-and-causality.html

Rigid like elastic

Title reference is supposed to look like a paradox, but then is explained: nominal rigidity is an entropic force like elasticity. This was completely re-written for a general audience. These posts are more technical versions:

https://informationtransfereconomics.blogspot.com/2014/10/wage-stickiness-is-entropic-force.html
https://informationtransfereconomics.blogspot.com/2015/03/nominal-rigidity-is-entropic-force.html

[There are actually several posts "X is an entropic force" on my blog. These are the most relevant two.]

SMDH

SMD theorem + H. This chapter's main premise is captured in this post, but it misses out on the blueberry pie metaphor of the SMD theorem that I'm particularly proud of:

https://informationtransfereconomics.blogspot.com/2015/10/the-smd-theorem-and-oh-no-not-another.html

The economic problem

This was the very terse starting point for this chapter:

https://informationtransfereconomics.blogspot.com/2015/03/the-price-system-as-communication.html

Economics versus sociology

This is another example where a post was greatly expanded:

https://informationtransfereconomics.blogspot.com/2015/10/economics-as-and-versus-social-science.html

Are we not agents?

The title is a reference to Devo's first album. I discovered a paper while the book was being written, and so this chapter was added based on it:

https://informationtransfereconomics.blogspot.com/2016/09/causal-entropic-forces-as-economic.html

Conclusions

Another chapter that is largely new in the book. One of the ideas I talk about was first presented here:

https://informationtransfereconomics.blogspot.com/2015/04/thinking-positive-is-thinking-different.html

*  *  *

But here's the whole thing in convenient Kindle form (and with less math):


Monday, September 4, 2017

Why no empirical work?

Another of @unlearningecon's suggestions was to showcase more of my "(successful) empirical work".

This was a conscious decision. I don't believe in self-publishing technical results without some direct mechanism for peer review. The journal submission process is one such avenue, but so are blogs with comments. Books don't have the same direct nexus with criticism (good/bad reviews on Amazon can function a bit like this, but are not quite the same thing as actual blogs). In fact, this mechanism is precisely why I wanted to have a book blog: so I could show and respond to both positive and negative criticism.

Given that reservation to putting non-peer-reviewed material in the book, it quickly became obvious that I should write a non-technical book aimed at a general audience. I scaled back the math, and that precluded inclusion of my empirical work.

However, if you are interested in exploring further, the empirical work is collected on my blog:


... especially at the aggregated forecast link where I track the performance of the forecasts I make (and comparisons to other models). There are results like this:


The green shaded region is the Information Equilibrium (IE) model forecast for the 10-year US treasury bond interest rate. The red line is a forecast from the "Blue Chip Economic Indicators" report from the end of 2014 (made up of a survey of expert). The purple dashed line is the CBO forecast from the end of 2016. The vertical lines indicate when the forecast was made.

The gray jagged line is the daily US interest rate data (from FRED) since the end of 2014. As you can see, the IE model was a much better forecast than the BCEI experts. I've been tracking this forecast for almost three years. Even the sudden rise in rates after the US presidential election hasn't thrown this forecast off (the bands are 90% confidence intervals for monthly data).

Friday, September 1, 2017

Another good point: economics vs sociology

Another of @unlearningecon's good points is this:
Concerned your separation of econ and sociology amounts to 'econ works except when it doesn't, which is what you criticise in mainstream
My response to this is that I've identified a particular mechanism (correlations in state space, which cause agents to not fully explore it), so it's not as vague. The general idea was speculative in the book (I explicitly said it was), and I also made the specific speculative claim that these correlations are caused by social factors. This last part may or may not be true: there may well be "economic" reasons for correlations that don't depend on our human nature.

But another reason I probably fell down on defending this particular claim (or making it more specific) is that it is based on the mathematics of information transfer and I haven't come up with a really good explanation that doesn't rely on math. That basically means I don't understand it very well, and that's true: hence the speculation.

The idea is basically that economics is "information equilibrium" (IE) and sociology explains "non-ideal information transfer" (NIIT) (definitions here). However, IE bounds the system dynamics even if you have NIIT (via some math). The result is that sociology should cause economics to fail in a specific way. I addressed this specific question in a FAQ on my blog:
But mindless atoms don't panic ... 
While information equilibrium treats agents effectively as random "mindless atoms" (but really treats them as so complex they look random), the information transfer framework is more general. If agents didn't spontaneously correlate in state space due to human behavior (e.g. panic, groupthink), then the information transfer framework reduces to something that looks like boring standard thermodynamics. However, they do in fact panic. In terms of thermodynamics, this means that the information transfer framework is like thermodynamics, but missing a second law of thermodynamics. The "mindless atoms" will occasionally panic and huddle in a corner of the room and you have non-ideal information transfer as opposed to information equilibrium. 
There is less the information transfer framework can say about scenarios where we have non-ideal information transfer, but it still could be used to put bounds on economic variables. 
Wait. Isn't this just saying sometimes your theory applies and sometimes it doesn't? 
Yes, but in a particular way. For example, the effect of correlations (panic, groupthink) is generally negative on prices. 
Additionally, empirical data appears to show that information equilibrium is a decent description of macroeconomic variables except for a sparse subset (i.e. most of the time). That sparse subset seems to correspond to recessions. Since human behavior is one of the ways the system can fail to be in information equilibrium, this is good evidence that information equilibrium fails in exactly the way the more general information transfer framework says it should. 
In a very deep way, one can think of information equilibrium being a good approximation in the same way the Efficient Market Hypothesis (EMH) is sometimes a good approximation. Failures of the EMH seem to be correlations due to human behavior.

Thursday, August 31, 2017

Good point about exploring the opportunity set

Via twitter, @unlearningecon makes a good point:
Potential problem with agents fully exploring the opportunity set is 'localised' optimising which could bias it in general
If I open a new piece of state space (or close one off), then it's likely agents "near" that piece of state space are the ones that are explore it (or leave for nearby open areas). This makes the process of evolution "localized". In a biological system, species in different ecosystems evolve in ways largely independent of species in other ecosystems, which may eventually result in some kind of conflict or catastrophe. More importantly, evolution happens from the current set of species (exploring parts of state space near existing species), creating strong path dependence.

In economic systems, e.g. firms will explore state space near existing firms, which may or may not find the best solutions.

This issue with "localized" optimization turns into a good argument in favor of government intervention in markets to both alleviate issues with path dependence (e.g. by providing workers with help if their industry needs to be reduced in favor of some other more socially optimal industry) and providing new seeds to grow in less explored areas of state space (e.g. funding science and innovation). I missed the opportunity to make this point in my book, but it is a good one.

In future posts, I will discuss unlearningecon's other points.