Modelling brazilian regional formal labor market using global var approach

The assessment of economic variables is an important part of regional macroeconomic analyses. However, increasing integration of the markets has led to greater financial and economic interdependence between regions. Therefore, this paper uses the global vector autoregressive (GVAR) methodology, which can tackle the curse of dimensionality. Focusing in the Brazilian labor market, it has two main objectives: firstly, establishing a model accounting for the interdependencies between regions. Secondly, estimate the regional elasticity of employment in respect to the economic activity of the country. To this end, it is applied the so-called GVAR technique, which considers the interdependencies between several regions and their temporal dynamics in a multivariate framework. The model is estimated at the Brazilian mesoregion level, with 137 distinct mesoregions. The final model proved to be stable with 128 regions, 2 cointegration relationship and, and 9 regions having 1 cointegration relation. Focusing on the classical major Brazilian regions (North, Northeast, Southeast, South, Central) it is estimated that the most sensitive region is the South followed by the Northeast and the South region, while the Northern and Central regions are mostly unaffected. A long-run relationship is also estimated indicating a natural growth of 694 thousand jobs per year in Brazil.

See more details here



Escolas FGV

Acompanhe na rede