The statistical mechanics of economic inequality

Economist Thomas Piketty has recently become a celebrity because of his new 700 page best selling book, Capital in the 21st century.

The latest issue of Science has a special section about "the science of inequality". It features a review by Piketty and his longtime Berkeley collaborator Emmanuel Saez. In the introduction the editors make an important point about an exciting future for economic research:

And in the past decade in developed capitalist nations, intensive effort and interdisciplinary collaborations have produced large data sets, including the compilation of a century of income data and two centuries of wealth data....
It is only a slight exaggeration to liken the potential usefulness of this and other big data sets to the enormous benefits of the Human Genome Project. 

Researchers now have larger sample sizes and more parameters to work with, and they are also better able to detect patterns in the flood of data. Collecting data, organizing it, developing methods of analysis, extracting causal inferences, formulating hypotheses—all of this is the stuff of science and is more possible with economic data than ever before. 

Hopefully, economics will move beyond its current situation where there are popular introductory textbooks that contain virtually no real data, just schematic curves. Previously I wrote about how different the book Poor Economics is.

Econophysics does not get an article but a one page story with the graph below. It is the work of Victor Yakovenko [who I know for his nice work on angle-dependent magnetoresistance!] and is nicely described in a Reviews of Modern Physics Colloquium article. The essential "physics" is that income distribution follows an exponential [Boltzmann] distribution which can be derived on the assumption that through random "collisions/exchanges" the total amount of money [energy] is distributed among all citizens.