Bayesian Econometrics | Bayesian Econometrics |
|
|
|
The major scientific goal of this project is to provide the researchers in the other working groups with elaborated, but user-friendly econometrics tools for answering their research questions. Existing econometric models will be extended to deal with non-standard features of their data like discrete outcome variables, non-linearity, and unobserved heterogeneity. Highly structured models will be considered where the distribution of the observed data is chosen from non-Gaussian and discrete distributions and is assumed to depend on latent variables. A typical example is to investigate the dynamics of individual transitions between discrete labor market states by dynamic transition modeling of discrete choice outcome variables under unobserved heterogeneity. Estimation and testing of these models is quite a challenge and Bayesian inference using Markov chain Monte Carlo (MCMC) algorithms is often the only approach that is feasible. Another scientific goal of this project is to improve existing and develop new MCMC algorithms for the efficient Bayesian estimation and model discrimination for a broad class of dynamic panel data models.
Project Leader
Project Members
|
|||||||||||||||||||||||||||||||||||||||
