Inorder to get trade data on imports and exports of federal units, Castilho, Mene´Ndez and Sztulman(2001) (p826), have used data colledcted by Secretaria de Come´rcio Exterior, Ministe´rio do Desenvolvimento, and Industria e Come´rcio Exterior (SECEX/ MDIC). Moreover, each state’s Gross Domestic Product (GDP) has been derived from the regional accounts in Brazil.
The authors’ contribution to the existing studies has been enhanced through meta-analysis whereby using units of observation that are sub-national, the study has been able to highlight the effects of trade liberalization on income distribution. Moreover, the existing data on household surveys have allowed the authors to measure inequality and household poverty in different states. Through the study, the authors have shown that the urban areas, which are more exposed to tariff cuts, experience small reductions in inequality and household poverty. However, the studies have also indicated that in rural states, trade liberalization has decreased inequalities and poverty. According to the authors, these results have been replicated in other studies (Castilho, Mene´Ndez and Sztulman pg826).
In this study, the authors have collected the data using a sample survey method. This is because the authors have used information from few states and not the entire population. Most of the data in this study was collected from a sample of 300,000 individuals, who came from urban and rural areas in the Brazilian Federation. Individual-level information from PNAD was important for the construction of summary variables on education, employment and income distribution. In order to come up with comprehensive data on poverty and inequality, the authors have applied indicators such as Theil and Gini indices.