Brazil has come to be known as one of the world's most extreme examples of inequitable growth. Since 1973, when Albert Fishlow first brought the phenomenon to the attention of foreign scholars, Brazil's combination of low wages, fast growth, and rising inequality has become, for many, the symptom of all that is wrong with growth under capitalism.

The negative assessment of the critics intuitively struck me as misleading both as to the meaning and the causes of rising inequality. My experience in Brazil suggested that the whole issue needed a much closer, more thorough look. There were in my mind three key issues: First, what was the relationship between changes in the wage structure and widening inequality? Second, what caused the widening in wage structure? Third, what is the meaning of inequality measures in dynamic, growing economies?

These issues are controversial and important. Societies and their governments are judged on their ability to make the distribution more equal even as they are causing total income to grow. Although a combined focus on distribution and growth is definitely an improvement over the exclusive attention that economists used to pay to measures of growth, it is crucial to understand what these distribution measures mean, and better yet what they do not mean, and also how they are likely to change during the process of growth. That is the only way that we will ever be able to correctly evaluate growth and distribution performance.

To examine these three questions, I chose to focus on labor markets and what happened to them during the process of growth.