[Excerpt] The purpose of this paper is to reexamine one of these two challenges, namely, the distributional impact of Brazilian economic growth during the 1960's. My results lead to a quite different interpretation from the conventional one. I will show that the poor in Brazil did participate in the rapid economic growth of the decade. Estimates presented below indicate that average real incomes among families defined as poor by Brazilian standards increased by as much as 60 percent while the comparable figure for nonpoor families is around 25 percent. However, since nonpoor families receive incomes which are much greater than those of poor families, the bulk of the growth of national income over the decade was received by families whose incomes placed them above the official poverty standard. Thus, it would be incorrect to say either that 1) in achieving a high rate of economic growth in Brazil the rich got absolutely richer while the poor got absolutely poorer, or 2) the incomes of poor families increased more slowly (percentagewise) than those of nonpoor families. These and other findings are presented below in Section II, and some of the reasons for the observed changes are discussed in Section III.
In assessing the distributional consequences of Brazilian economic growth, this study explicitly adopts an absolute poverty approach. In so doing, it is at odds with the bulk of the economic development literature, which while urging a poverty focus, has long relied on measures of relative income inequality and Lorenz curves. Thus, this paper does not merely offer "one more measure"; it is, rather, the use of a different type of measure that causes the divergent results. The paper concludes in Section IV by reviewing the principal findings and exploring some further questions of more general applicability raised by the Brazilian debate.