The Brazilian cash transfer programme, regional effects, and its impact on the labour market

Abstract

The aim of this paper is to discuss the effects of the Bolsa Família Programme on the Brazilian labour market. The data analysed derive from a sample from the Family Budget Survey, 2008–2009. The descriptive analysis highlights the importance of Bolsa Família in the composition of families’ income, especially in the case of those families that live in less economically developed regions of Brazil such as North and Northeast, as well as rural areas. The empirical analysis carries out econometric models (Seemingly Unrelated Bivariate Probit model and Instrumental Variables (IV)), and demonstrates both a strong and negative effect of the cash transfer provided by Bolsa Família on the Brazilian labour market.

Keywords:

• Bolsa Família Programme
• Brazil
• income
• region

JEL Classifications:

• I38
• N36
• R23

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1. For further details, see Lindert et al. (2007).

2. Considering an exchange rate of 2,00R$/US$ (July 2012).

3. For this study, it is considered that participation is linked to being employed in the labour market.

4. We took as granted that those enrolled in PBF are receiving the benefits.

5. Previously reported surveys that have addressed the same theme, albeit with different questionnaires took place in 1974–1975, 1987–1988, 1995–1996, 2002–2003 and 2008–2009. The POF survey began on 19 May 2008 and ended on 18 May 2009. Although the period is related to the arrival of the international crisis, the average unemployment rate in Brazil was 9.3 % (2007), 7.9% (2008), 8.1% (2009) and 6.7% (2010), indicating that the decision to participate in the labour market was more related to an individual decision rather than a decision conditioned by the international crisis (labour market conditions).

6. The Continuous Cash Benefit (BPC) is a monthly benefit granted to elderly people (65 years old or older) and disabled people in Brazil. The Eradication of Child Labour (PETI) is a Brazilian Programme for children and teenagers (up to 16 years old) exposed to child labour practices, except as apprentices, implying the direct transfer of income to the benefited families.

7. We drew on the same conceptual criteria applied in the POF questionnaire: electricity supply service and garbage collection service. The indicators of subjective evaluation were obtained according to the individual's perception, whose values range from 1 to 3 and in which higher values mean best indicators.

8. The test of Murphy was applied to verify the adjustment of the model due to excess kurtosis and asymmetry in the error distribution. The result is shown in the Appendix. For further details, see Cameron and Trivedi (2009).

9. The tests are in the Appendix. In addition to the exogenous test, the tests of instrument validation and strength were applied. In all IV regressions, the standard errors are robust to heteroscedasticity. The statistics Cragg-Donald Wald F test, Kleibergen-Paap Wald rk and Kleibergen-Paap rk LM, test, respectively, the strength of the instruments and identify the validity of the model. The results suggest that the equations are identified and the instruments are valid. In turn, the Hansen J statistic evaluates whether the instruments of the second equation (scholarship recipient family) are correlated with the errors of the first equation (equation of labour supply), that is, if the decision to enter into the labour market depends on the variables used to explain the receipt of family allowance. Evidence suggests the rejection of the null hypothesis, i.e. the decision to enter into the labour market does not seem to be influenced by variables that define the individual as a beneficiary of the family purse. Therefore, the results ratify the use of the IV model (Kleibergen-Paap rk LM and Cragg-Donald Wald F). For further details see Hansen (1982).

10. The command used in the estimation already corrects the variance-covariance matrix with regards to the efficiency loss of the estimator due to the limitation of the available information.