• 2018-07
  • 2018-10
  • 2018-11
  • 2019-04
  • 2019-05
  • Introduction Brazil is a country known to have


    Introduction Brazil is a country known to have a high level of income inequality with a Gini coefficient of per capita household income of approximately 0.54 (IPEADATA, 2013). Aside from this factor, Brazil also has significant levels of both poverty and extreme poverty. In 2009, about 21.4% and 7.3% of the population was poor and extremely poor, respectively (IPEADATA, 2013). In order to help combat this poverty, a conditional cash transfer program – the Bolsa Família Program (PBF) – was created in 2004. The goal of the program was to increase the income level of economically underprivileged subpopulations, as well as generate direct incentives for children to improve their education and health outcomes. By 2013, there were around 14 million pak1 participating in the program, encompassing more than 50 million individuals. The total value of the cash transfers reached more than 0.5% of the Brazilian Gross Domestic Product (GDP). According to Foguel and Ulyssea (2007), public transfers have accounted for the major source of non-work related income in the past few years. Some studies, such as those by de Barros et al. (2007a,b,c), Hoffmann (2006), Soares et al. (2006), and Cury et al. (2009), find that the PBF has had a positive impact PBF on the reduction of income inequality and poverty in Brazil. Ferro and Kassouf (2003), Cardoso and Souza (2004), Bourguignon et al. (2004), and Glewwe and Kassouf (2012) note the positive effect that conditional cash transfer programs have on increasing the school attendance of children whose families are beneficiaries. However, there is no evidence of positive impact on vaccination (CEDEPLAR, 2007). Manipulation of income may shift the focus away from the subpopulation of interest, for which the benefit is badly needed. Data for 2012 show that there were 16.2 million people in Brazil with a monthly income per capita less than R$70.00, which is the extreme poverty line (World Bank, 2013). Hence, if non-eligible individuals benefit from the program, those most in need are likely to be excluded from the PBF as the budget becomes larger than expected. The aim of prey switching paper is to ascertain possible manipulation in the eligibility status for the PBF and verify if the manipulation is performed by behavioral changes on the time allocation of the individuals within the households. This manipulation is assessed by the test developed by McCrary (2008), which determines the presence of discontinuity in the density estimated by local linear regressions around the cutoff that defines the eligibility for the program, which was R$120.00 (US$60.00) in terms of monthly per capita household income in 2006. Next, we use a fuzzy regression discontinuity design in order to obtain evidences of the eligibility manipulation by changes on time allocation. The paper contributes to the public debate of social policy design. Particularly, there is a debate whether the targeting of social programs should be means tested or proxy-means tested. A means tested targeting strategy implies that household with income below certain thresholds qualify to the program. This is the case of Bolsa Familia program. On the other hand, a proxy-means tested strategy makes program eligibility to depend on a composite score of a household\'s characteristic such as asset holdings, demographic composition, and dwelling characteristics that are proxies for household income. This is the case of the Oportunidades program in Mexico. Although this is an important debate, there is scant evidence on the advantages or disadvantages of each strategy. This paper sheds new light to this debate by showing that means-tested targeting leads to inclusion error, because households change their behavior in order to become eligible to the program. Including this introduction, the paper is organized into six sections. Section 2 describes the main characteristics of the PBF and the incentives associated with participation. The construction of the database and the sample selected are noted in Section 3. Section 4 describes the methodology. Results are presented in Section 5, while Section 6 offers our conclusions.