Zusammenfassung (übersetzt):
This text describes in detail how to simulate the poverty impacts of changes in the entitlement rules and benefit values of the Bolsa Família Program. This simulation is especially challenging due to: i) the absence of a specific question on Bolsa Família income in the PNAD household survey and, more importantly; ii) the 2.4 million family disparity between administrative records and household survey data. The first problem we solve using a typical value identification methodology for the other income variable in which lies buried the program's income. The second problem we solve by using various matching methods to make the household survey data coherent with the administrative data. In addition to describing the simulation mechanics, we simulate various changes in the Bolsa Família program and find that: i) the most cost efficient way of achieving a modest poverty reduction is to eliminate the limits on the number of child benefits per family; and ii) linear increases in benefit values lead to linear reductions in poverty, which means that the present benefit values are still far from offering decreasing returns.