Spatial Chow-Lin models for completing growth rates in cross-sections

Growth rate data that are collected incompletely in cross-sections is a quite frequent problem. Chow and Lin (1971) have developed a method for predicting unobserved disaggregated time series and we propose an extension of the procedure for completing cross-sectional growth rates similar to the spatial Chow-Lin method of Liano et al. (2009). Disaggregated growth rates cannot be predicted directly and requires a system estimation of two Chow-Lin prediction models, where we compare classical and Bayesian estimation and prediction methods. We demonstrate the procedure for Spanish regional GDP growth rates between 2000 and 2004 at a NUTS-3 level. We evaluate the growth rate forecasts by accuracy criteria, because for the Spanish data-set we can compare the predicted with the observed values. -- interpolation ; missing disaggregated values in spatial econometrics ; MCMC ; spatial Chow-Lin methods ; predicting growth rates data ; spatial autoregression (SAR) ; forecast evaluation ; outliers