Real effects of direct cash transfers shocks: evidence from one of the largest social welfare programs in the world

ABSTRACT

We examine the socio-economic consequences in Brazilian municipalities of one of the world’s largest social welfare programmes – the Bolsa Família. Since the Bolsa Família endogenously targets municipalities with weaker local socio-economic conditions, it becomes challenging to reason about its causal effects on the local economy. We address this endogeneity by exploiting a change in the Bolsa Família regulation in 2012 that established a minimum per capita income of 70 reais for recipient families. Previously, additional financial benefits were available based on the number of children in the family, up to five. Following the change, large families effectively received a positive income shock. We investigate this discontinuity by comparing recipient families of different sizes with income nearing the 70 reais per capita threshold shortly before the shift at the municipality level. Municipalities’ employment, income, and health metrics show consistent and positive results. However, we find no effect on education, which could be explained by the lack of robust education-related conditionalities that Bolsa Família recipients are required to report continuously to enjoy the programme’s benefits.

KEYWORDS:

• Bolsa familia
• direct cash transfer
• public policy
• causal effect
• economic growth

JEL CLASSIFICATION:

• I30
• I38
• C20
• C50
• C54

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

¹ See also a discussion on the impact of cash transfers on female schooling in Punjab, Pakistan (Chaudhury and Parajuli 2010) and also Sodokin (2021) for an empirical evaluation of international cash transfers and their favourable impact on poverty and inequality in Togo, while also increasing household investments..

² There is a substantial selection problem when analysing direct cash transfer programs. The PBF analysis is no different: the values transferred under the PBF are larger precisely in municipalities with lower socio-economic conditions. Thus, many unobservable factors could drive our empirical conclusions if we compare groups of municipalities that receive a large volume of PBF transfers (concerning their GDP) against those that receive little PBF transfers. For example, municipalities that do not receive PBF could be wealthier than those that do because of a more robust and dynamic local economy.

³ The use of the difference-in-differences approach to extract causal estimates has become widespread in the empirical literature (Silva, Hasan, and Tabak 2021; Silva, Wilhelm, and Tabak 2022). We follow this standard.

⁴ Other papers have discussed several aspects of the PBF such as the mitigation of school dropouts due to maternity (Santos and Corseuil 2022), fraud detection (da Silva Azevedo et al. 2021), the effect on intimate partner violence (Litwin, Perova, and Reynolds 2019), and that PBF alone is not enough to reduce inequality and poverty (Campoli et al. 2020)..

⁵ Equation (4) is a difference-in-differences specification. However, it differs from the traditional equation because it does not contain the marginal terms $Pos{t}_t$ and $Trea{t}_{i,g}$. We do not add them because they are collinear with the fixed effects ${\alpha }_{g,t}$ and ${\alpha }_i$, respectively.

⁶ Simões and Sabates (2014) conducted a more restricted analysis of the results of PBF in children’s education. They discover evidence for a positive effect of participation in PBF and the amount of money received on academic achievement, but with substitute effects. Additionally, their research indicates the importance of determining how much support to provide and how long to ensure academic progress.

Additional information

Funding

This work was supported by the CNPQ (302703/2022-5, 305485/2022-9) and CAPES (88881.610990/2021-01)