2015 summer school student and LML visitor Yonatan Berman and LML Fellows Ole Peters and Alex Adamou have published an empirical study of wealth inequality in the United States. They add reallocation to a standard model of wealth growth (geometric Brownian motion) and ask at what rate the reallocation must occur to reproduce observed wealth inequality from 1917 to 2012. They find that this effective reallocation rate has been negative – from poorer to richer – over the last 30 years. In this regime, the wealth distribution cannot stabilise. Crucially, these dramatic conditions can only be observed by abandoning the assumption of ergodicity, standard in other analyses: in the model of Berman, Peters, Adamou (relative) wealth need not approach a stable limiting distribution. Lifting this constraint lets the data speak for themselves.
The paper is published at: http://arxiv.org/abs/1605.05631