The Econometrics of Imperfect Knowledge Economics

Conference paper By

Download paper

A core premise of contemporary economic models is that researchers can adequately specify in probabilistic terms how individuals alter the way they make decisions and how the processes underpinning market outcomes unfold over time.

Based on this core premise individual and aggregate outcomes at all points in time are represented with an overarching probability distribution. We refer to such models as determinate. To confront determinate models with empirical evidence a ‘theory-first approach’ to econometrics is typically used, see Hoover (2006b) and Spanos (2009). In the ‘theory-first approach’ the determinate theoretical model delivers a complete stochastic specification that relates aggregate outcomes to a set of explanatory variables, and the role of econometrics is solely to quantify the theoretical parameters of interest and test their statistical significance using regression or other statistical techniques (Spanos, 2006).