Duration analysis (RM/PhD)

In June 2019, I taught a condensed version of this course as PhD lectures at the Toulouse School of Economics. Earlier, I taught variants of it in the Danish Graduate Programme in Economics (Aarhus, Denmark) and at the University of Chicago (Econ 315+316), University College London (BP3), the ENSAE in Malakoff (Paris), Tinbergen Institute (721/I), the CIdE Summer School of Econometrics (Bertinoro, Italy), and Yale University.

The course provides an introduction to the econometric analysis of duration and more general event history data. It first reviews basic concepts and commonly used descriptive methods. Then, it discusses mixed hazard models and their relation to economic models in which heterogeneous agents take discrete decisions at random and discrete (Poisson) times. Finally, it covers mixed hitting-time models and their relation to optimal stopping problems. Throughout, it highlights the extent to which model primitives are uniquely determined by the data (identifiability), and studies computational and empirical methods. It focuses on decision problems, but may include recent developments in the analysis of stopping games. Theory and methods are illustrated with (potential) empirical applications to e.g. the evaluation of labor market policies, job search, and asymmetric information in insurance markets.

Joomla templates by a4joomla