A1324
Title: A state-space frame for modelling time-varying parameters in panels: An applicacion to the Okun's law
Authors: Juan Sapena - Catholic University of Valencia (Spain) [presenting]
Mariam Camarero - University Jaume I (Spain)
Cecilio Tamarit - University of Valencia (Spain)
Abstract: A fully-fledged State-Space frame is developed for modeling panel time series that extends the simple framework generally employed into a panel-data time-varying parameters framework combining both fixed and varying components. Fixed parameters can be modeled either as country-specific or as a common parameter for the whole sample. Under determinate circumstances, this setting can be interpreted as a mean-reverting panel time-series model. Additionally, the mean fixed parameter can eventually include a deterministic trend. Regarding the equations governing the transition of the unobserved components, our structure allows for the estimation of different autoregressive alternatives that eventually include control instruments. The coefficients for the control instruments can be set up either as common for the panel, or as country-specific. The latter is particularly interesting for detecting asymmetries among individuals (countries) to common shocks. The code, that has been performed by the authors in Aptech Gauss, also allows for imposing restrictions regarding the relative size of variances of the error terms of the transition and measurement equations. Finally, we perform an empirical application of the proposed frame to estimate the estimation of a time-varying specification of the Okun's Law for a panel including both European and non-European industrialized countries on the support of its usefulness in solving complexities in macroeconomic empirical research.
