Full text
| |
| Author(s): |
Diana Lúcia Gonzaga da Silva
Total Authors: 1
|
| Document type: | Doctoral Thesis |
| Press: | São Paulo. |
| Institution: | Universidade de São Paulo (USP). Faculdade de Economia, Administração e Contabilidade (FEA/SBD) |
| Defense date: | 2017-04-07 |
| Examining board members: |
Carlos Roberto Azzoni;
André Luis Squarize Chagas;
Ricardo da Silva Freguglia;
Naercio Aquino Menezes Filho;
Leonardo Monteiro Monasterio
|
| Advisor: | Carlos Roberto Azzoni |
| Abstract | |
The goal of this study is to identify the contribution of agglomeration effects and spatial sorting for the determination of individual and local wages in Brazilian urban agglomerations. Administrative records from the Ministry of Labor (RAIS-MTE) show a spatial differential in wages, which can be explained by the different productive structures and compositions of workers across cities, and by differentials in cost-of-living. The longitudinal microdata allowed the inclusion of unobserved individual skills in the wage equation. Studies in Labor Economics show that skills are responsible for a large portion of the wage differential. However, the available studies on Brazil still find a significant differential, even after controlling for the individual components and the cost of living. This suggests the existence of specific effects associated with the location of firms and workers. The literature on Urban Economics considers the economies of agglomeration as a relevant wage determinant in dense labor markets. The higher productivity of dense areas can be attributed to the concentration of more productive workers and firms more productive, which became known in this literature as sorting. Studies in Urban Economics only control the sorting of unobserved individual attributes. This dissertation contributes to the literature by considering the spatial sorting of unobserved attributes of firms and of workers in the determination of wages and of the effects of agglomeration. The study uses a wage decomposition model to deal with multiple fixed effects in a matched panel of workers and firms. The pure effects of agglomeration (density) on local wages are estimated in a two-stage model. The first stage estimates a wage equation including the observed characteristics of workers and firms and the effects of location, with a microdata panel of RAIS (2002-2014). The second stage decomposes the location effects into components associated to local characteristics and to unobserved attributes of firms and workers. The identification strategy involves controlling for fixed effects of workers and firms, and using an instrumental variable to identifying the effects of agglomeration. Satellite data on illumination are used to estimate the proportion of the overall area occupied with population and firms in each local labour markets. The results indicate that the worker effects are more relevant to explain wage variation than the firm\'s effects. The model of preference indicates a density effect on wages of 4.9%, much higher than the literature lower bound (3%). This suggests that ignoring the variables included in this study can lead to an underestimation of the effects of agglomeration (AU) | |
| FAPESP's process: | 14/03077-7 - Agglomeration economies and sorting in determining of space wage workers of Brazil |
| Grantee: | Diana Lúcia Gonzaga da Silva |
| Support Opportunities: | Scholarships in Brazil - Doctorate |