The New York Times has an article by Tina Rosenberg on the Bolsa Familia program in Brazil, which is modeled after the Oportunidades program in Mexico.
Both programs are conditional cash transfer (CCT) programs, which means that individuals or households receive a cash transfer if they statisfy fulfill requirements. In the case of Oportunidades, Rosenberg writes that “families must keep their children in school and go for regular medical checkups, and mom must attend workshops on subjects like nutrition or disease prevention.”
The empirical evidence on CCT programs is pretty solid. But as I tell the students in my development seminar, the problem with CCT programs is that they are very expensive: Duflo (2010) notes that it costs $1,000 to keep a child in school one more year under Oportunidades, whereas it costs $3.50 to do so under a deworming program.
During the first few weeks of the fall semester, I spend a lot of time teaching the students in my development seminar about the various ways in which we can assess the effectiveness of policy interventions.
To convince the most resilient of them — those who may be tempted to think that instrumental variables, randomized controlled trials, identification, etc. are eggheaded concerns — I tell them that knowing what works is useful because it allows policy makers — and thus presumably — taxpayers to get the best bang for the buck.
In the case of Bolsa Familia, I am for poverty reduction via policies that work, but I am also for poverty reduction via cost-minimizing (i.e., wealth-maximizing) policies that work. I am not familiar enough with the Brazilian evidence so as to be sure that Bolsa Familia is such a policy.
(HT: @poverty_action)