Labour market predictions for catalonia

Methodology The variables analyzed and forecast in the medium term are: Activity rate, employment rate, overall unemployment rate, absolute and relative variation of the active population/occupied/not active, real wages and productivity. At the short-term level, the models are based on quarterly statistical information, and they are univariate (ARIMA models of different variants, classic series) and multivariate (VAR, MRLM with mechanisms for error correction or not, SETAR, models based on Markov chains) to see which provides better predictions. It have also been implemented models to analyze whether there was a turning point, a change of trend in the labor market series. At medium-long term level, based on annual statistical information, it have been used multivariate models (regression, MVAR, multiequacionals, etc.) with major or minor complexity, depending on the number and types of endogenous variables that could eventually include the models.