Labour market analysis through transformations and robust multivariate models

Abstract The work presents a robust approach to labor share analysis. The estimate of labour share presents various complexities related to the nature of the data sets to be analyzed. Typically, labour share is evaluated by using discriminant analysis and linear or generalized linear models, that do not take into account the presence of possible outliers. Moreover, the variables to be considered are often characterized by a high dimensional structure. The proposed approach has the objective of improving the estimation of the model using robust multivariate regression techniques and data transformation.