Abstract (Translated):
In 2006, the Bolsa Família Program completed its expansion, expecting to cover 11 million beneficiary families. The objective of this article is to evaluate whether this implied in deterioration in the program's efficient targeting and whether the target of 11 million families in fact covers the entirety of Bolsa Família's eligible target population. Our results are twofold. First, the increase in the coverage was followed by a slightly fall in the targeting. However, we do not find evidences that the coverage increase caused a worse targeting by itself. In addition, that coverage target is insufficient to cover the eligible population due to targeting errors and the size of the program. Taking into consideration the inevitable targeting errors, we estimate that the program should expand up to 15 millions of beneficiary families. Additionally, we show that a considerable part of the so-called inclusion error may be explained by the income volatility of poor families that makes the real target population of Bolsa Família (those families that are or soon will be below R$ 120 per capita) considerably larger that that calculated using a cross-section estimate of the income distribution in a given moment.