Skip to main content
SpringerLink
Log in

Brazilian Wage Curve: Further Evidence Based on Spatial Interactions in Times of Recession, 2012–2019

  • Article
  • Published:
The Indian Journal of Labour Economics Aims and scope Submit manuscript

Abstract

This paper verifies the existence of a spatial Brazilian wage curve based on individual hourly real wages using quarterly data from the Brazilian National Household Sample Survey for the period 2012 to 2019, probably the largest dataset ever used in this field. Our spatial regression model enables us to estimate both the magnitude and dispersion of the local wage’s rigidity and its response to variations in the level of unemployment in neighbouring states—the spatial spillover. We find strong evidence for negative and significant spatial spillovers affecting local real hourly wages in the whole sample and 13 out of 20 different worker categories. For the entire sample, a 100% increase in local unemployment reduces the individual real wage in Brazil by 2.61% while the same increase in unemployment in contiguous states leads to an additional 1.00% reduction in wages. The findings indicate an overestimation of the real wage elasticity when the regression model neglects significant spatial autocorrelation. The results are robust to spatial effects present in the data, the weak instruments problem, and endogeneity of the regressors.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Log in via an institution

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1

Similar content being viewed by others

Data Availability

The data that support the findings of this study are openly available in the Instituto Brasileiro de Geografia e Estatística at http://www.ibge.gov.br/.

Code Availability

This paper uses R, a free software environment for statistical computing and graphics. R is freely available at http://www.r-project.org/, and the codes used are available upon request.

Notes

  1. Instituto de Pesquisa Economica Aplicada in Portuguese.

  2. In the same year, the IPEA estimated a decrease of 3.55% in the real GDP, which was the most profound slack in decades. Holland (2019) discusses the adopted economic policies which led to the most substantial slowdown in decades and other political and economic issues associated with the abrupt fiscal crisis in the Brazilian economy from 2014 to 2017.

  3. We put all the specification test results in the Appendix. See Table 8.

  4. Card (1995) and Baltagi and Blien (1998) observe that the relevant number of observations for our model specification is the number of individuals multiplied by the number of regions (\(4,045,664 \times 27 \ states = 109,232,928\)). In their study for Brazil, Baltagi et al. (2017) use a sample of only \(739,490 \times 27 \ states = 19,966,230\) that corresponds only to 18.27% of our sample data. In the case of Poland, Baltagi and Rokicki (2014) use only \(102,924 \times 16 \ states = 1,646,784\) observations, which corresponds to just 1.51% of our sample data.

  5. Baltagi and Blien (1998, p. 135–6) observed that “The wage curve is simply a standard wage equation normally used to estimate the returns to education of the male-female wage gap but with the addition of the local unemployment variable to the set of regressor”.

  6. https://www.ibge.gov.br/.

  7. We do not include all spatially lagged values of the explanatory variables X because we do not have the point coordinates of the individuals. Halleck Vega and Elhorst (2015, p. 346), when referring to the SLX model, observe: “a strong aspect is that there are no prior restrictions imposed on the ratio between the direct effects and spillover effects, which was a limitation of the SAR and SAC models”. In LeSage (2014) and Halleck Vega and Elhorst (2015), the acronym SAC refers to the spatial autoregressive combination model, while the SAR model refers to the spatial autoregressive model. The SAR model accounts for endogenous interaction effects, and the SAC model accounts for both endogenous interaction effects in combination with the interactions among the error terms.

  8. In relation to the SAR model chosen by Barufi et al. (2016), LeSage observed that “Global spillover specifications are more difficult to estimate and correct interpretation of estimates from these specifications is more difficult" (LeSage 2014, p. 15).

  9. For example, the direct effect in the SAR model is given by \(\frac{(3-\rho ^{2})}{3(1-\rho ^{2})}\beta _{k}\), and the indirect effect (spillover) is given by \(\frac{(3\rho +\rho ^{2})}{3(1-\rho ^{2})}\beta _{k}\), for k explanatory variables and \(N=3\). After obtaining these values, it is necessary to test their significance to assess their validity in the population, obtaining their dispersion measurements through Monte Carlo simulations. The valid method for inference with SAR/SAC and spatial Durbin spatial models involves obtaining the effects matrices and their corresponding dispersion measurements [see LeSage and Pace (2009) and also Elhorst (2010)]. The spatial Durbin model accounts for both endogenous and exogenous interaction effects but not for interaction effects among the error terms.

  10. For the sake of brevity, we only report \(\beta\), but the full results are available upon request.

  11. For brevity, we only report \(\beta\) and \(\chi\), but the full results are available upon request.

References

Download references

Funding

This research article has not received any funding support.

Author information

Authors and Affiliations

  1. Department of Economics, Federal University of Paraíba, Cidade Universitária, s/n, Castelo Branco, João Pessoa, PB, 58051-900, Brazil

    Guilherme Cemin de Paula & André M. Marques

Authors
  1. Guilherme Cemin de Paula
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. André M. Marques
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to André M. Marques.

Ethics declarations

Conflict of interest

Guilherme Cemin de Paula and André M. Marques declare that they have no relevant or material financial interests that relate to the research described in this paper.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Appendix

Appendix

See Tables 5, 6, 7 and 8.

Table 5 Variables used in the regression models
Full size table
Table 6 Grouping of occupations
Full size table
Table 7 Grouping of activities
Full size table
Table 8 Tests for weak instruments and the exogeneity of regressors: Results
Full size table

Rights and permissions

Springer Nature or its licensor holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

de Paula, G.C., Marques, A.M. Brazilian Wage Curve: Further Evidence Based on Spatial Interactions in Times of Recession, 2012–2019. Ind. J. Labour Econ. 65, 689–708 (2022). https://doi.org/10.1007/s41027-022-00390-w

Download citation

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s41027-022-00390-w

Keywords

JEL Classification

Search

Navigation