‘Partial identification in non ergodic agent based models’
Marie Curie Research Fellow, INET Oxford, Associate Member, Nuffield College, Carlo Alberto Affiliate, Collegio Carlo Alberto, Moncalieri
Assistant Professor, Department of Economics and Statistics, University of Torino
Tuesday 11th November, 12.30-14.00 Seminar Room 1, Oxford Martin School
The ergodic axiom lies at the core of economics. The fact that ergodicity is seldom testable in the data favors assuming it from the onset. However, non ergodic, multiple equilibria models might offer valuable insights into the functioning of the real world and be more appropriate in many modelling circumstances. Far from being intractable, we show that non ergodic models are amenable to estimation, the condition for identication being similar to those for ergodic models. We show that non ergodicity can be usefully thought of as an additional source of uncertainty which can be quantied in empirical applications. We test our method by means of a Montecarlo experiment on an agent based model of the labor market.